Page 1

JMRE

Journal of Medical Research and Education Volume 1 info@jmre.org

Number 1

2012

www.jmre.org


Cover photo by Viranuj Sueblinvong, M.D.


From the Editor Please be welcomed to Journal of Medical Research and Education (JMRE), a new peer reviewed English based journal aims to contribute the novel knowledge regarding various fields of medicine and medical education. The journal was initiated by the Medical Advancement Fund under the support of Thai Ministry of Public Health through the Collaborative Project to Increase Production of Rural Doctor (CPIRD). It was conceived out of a dream to provide a forum for the vastly potential and creative biomedical researches and medical education for both medical students and teachers especially in the emerging economics area of Southeast Asia which will be soon merged into the Asian Community in 2015. However, the first issue of JMRE is launched. Hope you like it and find it useful. Enjoy!

Thammasorn Piriyasupong, M.D., Ph.D. Editor-in-Chief of JMRE


Own & Published by The Medical Advancement Fund & Khon Kaen Medical Education Center Under the Patronage of Collaborative Project to Increase Production of Rural Doctor Thai Ministry of Public Health

T H E A D VI S O R Y B OAR D ON PU BLI CATI ON Varavudh Sumavong, Professor, M.D. Prasert Boongird, Professor, M.D. Waraporn Eoaskoon, Assoc. Professor, Ph.D. Kitpramuk Tantayaporn, Assoc. Professor, M.D. Somchai Nichpanit, M.D. Suwat Lertsukprasert, M.D. Sirijit Vasanawathana, M.D. Surachai Saranrittichai, M.D.

ED I TOR- IN-C HI EF Thammasorn Piriyasupong, M.D., Ph.D.

ED I T ORI AL B OARD Hiroshi Nishicori, M.D., Ph.D. Madawa Chandratilake, MBBS, MMEd. Maria-Paz Loscertales, M.D., M.Trop.Paeds. Bothaina Attal, MBBS., Ph.D. Duangjai Lexomboon, M.A., D.D.S., Ph.D. Kanokwan Sriruksa, M.D. Benjaporn Silaruks, B.Pharm., Ph.D.

M A N US C RI P T EDI TI NG Thammasorn Piriyasupong, M.D., Ph.D.

GRAP HI C ART Kwanroedee Sawangnuk Material printed in the Journal is covered by copyright. No copyright is claimed to the Thai government. No part of this publication may be reproduced without written permission. The Journal does not hold itself responsible for statements made by any contributor. Statements or opinions express in the Journal reflect the views of the author(s) and do not represent the official policy of the Journal unless stated. w w w. j m r e . o r g

info@jmre.org


In this issue PE R SP E CTI V E in Medical Education ๏ The Journey of Imparting the Morality of Medicine ๏ Community-Oriented Medical Education

O R I G IN AL AR T IC L ES by Medical Students ๏ Effects of Breast Feeding on Infantile Diarrhea: A Systematic Review ๏ Effects of Antibiotics on Length of Hospital Stay in Children with Acute Diarrhea ๏ Effects of High Dose Vitamin C Supplements on Urinary Oxalate Excretion: A Systematic

Review ๏ Effectiveness and Safety of Influenza A (H1NI) Vaccine: A Systematic Review ๏ Weight Gain during Pregnancy and Risks of Caesarian Section ๏ Validity of Symphysis-Fundal Height Measurement for Estimating Infant Birth Weight ๏ The Association Among Capillary Blood Glucose Levels, Sepsis and Death in Patients with

Melioidosis and Diabetes Mellitus: A Retrospective Cohort Study ๏ Maternal Urinary Tract Infection is Independently Associated with Preterm Delivery


Perspective in Medical Education

Photo by Thammasorn Piriyasupong, M.D., Ph.D.


JMRE

Journal of Medical Research and Education

PERSPECTIVE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

The Journey of Imparting the Morality of Medicine

Madawa Chandratilake (MBBS, MMEd) Research Officer, Centre for Medical education, University of Dundee, UK, and Lecturer in Medical Education, Faculty of medicine, University of Kelaniya, Sri Lanka Medical professionalism has become one of the central foci of medical education especially in the we s t e r n wo l d . T h e re n ewe d e m p h a s i s o n professionalism education is triggered by several factors. Regulatory and professional bodies of the practice of medicine in different countries faced public outcries about the way their doctors practice medicine. For example, isolated but significant incidents like the infamous case of Harold Shipman, a general practitioner who killed hundreds of patients under his care, triggered substantial reforms to the professional practice guidelines of the General Medical Council in the UK; ‘professionalism’ became a formal requirement to practice medicine.1 The selfregulation of the professional by doctors themselves has been replaced with enhanced public scrutiny.2 In USA, there were serious concerns about the way doctors manage conflict between self-interests and the interests of patients; altruism emerged as one of the corner stones of medical professionalism.3 In addition, evidence emerged to suggest that students or trainees who demonstrated lapses of professionalism might be likely to face fitness to practice issues in the future as independent practitioners.4 As a result professionalism emerged as an explicit competence or outcome of modern medical curricula.5,6 There is evidence suggesting that professionalism is becoming increasingly important to the medical education community in other parts of the world as well.7,8 Therefore, this account intends to generate an insight into what we already know about professionalism education.

changes and expectations, and has characterised medicine as a profession.9 For example, in the early days of medicine, where doctors with their limited knowledge on disease processes, were not much different to Soothsayers, ‘professionalism’ formed the identity of medicine as a profession by adopting very basic elements of an ‘evidence-based approach’.9 Centuries later, in the era of industrial revolution, ‘professionalism’ was focused on protecting patients from being the mere victims of commercial and research interests of booming scientific and technological advancements.9 During the last few decades, with the demand for patient-centred medical practice, ‘professionalism’ has become a process of scrutinising the professional autonomy of doctors from the perspective of patients. 2 Therefore, professionalism has increasingly been recognised as a social contract between doctors and society.10 Because of its nature, defining professionalism is challenging. The efforts of doing so have also been further complicated by its sensitivity to cultural backgrounds. 11,12 What is defined as professionalism or unprofessionalism by the western cultures may or may not make the same sense in eastern cultures. Therefore, currently medical educationalists and researchers attempt to ‘understand’ professionalism in relation to their own cultural and society context. However, there is general agreement that it encompasses a set of values, attitudes and behaviours. These attributes represent not only the relationship of doctors with patients, but their relationship with colleagues and other healthcare professionals in the team, and their behaviour in society.13 The attributes of professionalism appear to range from personal qualities (e.g. honesty and integrity, managing conflicts of interests) to regulatory requirements (e.g., ethical

Dynamic of professionalism Professionalism is a dynamic concept. Throughout the history of allopathic medicine, since the days of Hippocrates, it has been responsive to societal

8


Journal of Medical Research and Education

JMRE

conduct, accountability) to skills (e.g. effective communication, reflective practice, teamwork).14

professional identities of our students. Therefore, the professional culture of an institution plays a vital role in fostering professionalism among students and trainees. Professional culture can be explored through attitude surveys, student narratives, observations and critical incident analysis, and may be manipulated accordingly to achieve the desired educational o u t c o m e s . T h e p r i m a r y go a l o f ‘ t e a c h i n g ’ professionalism, however, should be to open up the informal and hidden components of the professionalism curriculum through student reflections and discussions, and to provide conducive environments and appropriate platforms for the purpose in the formal curriculum. Medical educationalists have taken several approaches to assessing professionalism. However, the l a c k o f u n i v e r s a l l y a g r e e d d e fi n i t i o n f o r professionalism has made assessment challenging.22 The assessment of students or trainees awareness of professional expectations and requirements, and demonstrations of this awareness appears to be the most basic approach. The commonly used assessment methods include written test such as scenario-based multiple choice questions, Extended Matching Items, and OSCEs.23 Although this step is important, students can ‘fake’ professional attitudes and behaviours in such assessments, i.e. demonstrating professionalism rather than becoming professionals.24 Therefore, the assessment of professionalism should essentially be extended beyond the assessment of the cognitive base. The assessment of reflection in and on practice on professionalism issues23 may be a step higher than the assessment of the cognitive base. However, it should most appropriately be assessed in educational or work-place environments using observations and multisource feedback (MSF).23 The observation of students / trainees for their professional approach to day-to-day clinical practice is an essential component of the assessment process. However, observations warrant considerable amount of staff time which may be a limiting factor of using t h i s m e t h o d i n re s o u rc e - p re s s e d c l i n i c a l environments. In addition, collated perceptions of different members of the healthcare delivery team (e.g. peers and nurses) or different stakeholder groups (e.g. managers, patients, etc. In addition to the team members) could contribute substantially to understand professionalism of individuals. The role of assessing ‘doctors’ by other healthcare professionals or patients in situations such as MSF, however, may not be acceptable in certain cultures. Although this is a professionalism issue itself, this attitude may t h re a t e n t h e u t i l i t y o f s u c h a s s e s s m e n t s . Observations and MSF should be used repeatedly and in multiple settings to obtain credible insight into the professionalism of a student or trainees.23 Unlike clinical competence and incompetence, determining professionalism and unprofessionalism may not be straight forward or ‘black and white’; assessors may feel ‘uncomfortable’ of failing their students or

Educability of professionalism The nature of the concept naturally generates scepticism on the educability of professionalism. Is professionalism teachable, learnable or assessable? There is a growing body of evidence to suggest that it is an educable concept. Skills like communication,15 reflection16 or teamwork17 have been known as teachable constructs. Even certain personal qualities like honesty and integrity could be fostered.18 However, it is important to understand that these diverse groups of attributes cannot be inculcated in our students using a single approach. In any educational programme, there are three facets of curriculum operating in parallel: formal curriculum, i.e. what is planned and documented; informal, i.e. what is not planned and documented but what is known to be happening in the educational settings; and hidden curriculum, i.e. what is unknowingly transmitted to learners through the institutional culture.19 It is not difficult to include skills components of professionalism in the formal curriculum. In fact, these have already been incorporated to many modern curricula. For examples, most of us as clinicians take part in explicit teaching and assessment of communication skills. However, encompassing ethical values and conduct, and personal qualities within the limits the formal curriculum alone would not entirely be successful. For example, we could teach students the correct technique of washing hand and ethics behind doing it properly as a moral obligation towards patient safety, and assess these aspects in simulated environments (e.g. OSCE). However, in wards, clinics or theatres, hand washing may not take place in the same vein as we taught to students. In a rare situation, there may not be proper hand washing practice at all in certain settings. When students are exposed to clinical environments they face a dilemma between what is learned and what is practised, and often they succumbed to the influence of the latter, i.e. informal or hidden aspects of curriculum. Therefore, the informal and hidden aspects of curricula play a greater role in fostering some aspects of professionalism.19 This does not mean that the formal curriculum should not explicitly address these aspects. Rather, students should be aware of morality and moral conduct by providing knowledge , experiential learning opportunities, reflection on practice or observations. 20 Most of such attributes, however, are socially learned and are sustainable only in conducive institutional environments.19 Even at the early stages of medical education, teachers and clinicians become role models for students, and their behaviours, expressions and conducts become the sources and drivers of the informal and hidden curricula.19 21 Our actions or inactions and commission or omissions as clinicians ultimately influence the formation of

9


JMRE

Journal of Medical Research and Education

trainees solely on professionalism ground and such decisions may not been entirely ‘acceptable’ to assessees.25 This issue is further complicated by the fact that most of the professional lapses are less ‘dramatic’ but vital, and there may not be consensus on how to respond to them.26 Therefore, although the summative assessments are necessary to ensure patient safety and professional accountability, professionalism should be assessed primarily for providing feedback and facilitating the development of right professional persona, i.e. assess formatively.14 The main emphasis of professionalism assessment, therefore, should be assessment for learning and such assessments should be carried out with a sound mutual understanding between the assessor and assess about the constructive nature of the process.

professionalism, there is much room for generating new evidence or supporting or questioning the existing evidence. There are also gaps in research on assessing professionalism; the evidence on how to assess certain areas (e.g. reflectiveness, lifelong learning) is inconclusive and there is no literature evidence at all on the assessment of certain aspects (e.g. advocacy).23 In general the available evidence to address all three questions appear to be methodologically less robust and atheoretical.27 Therefore, the scope for researching medical professionalism with a robust and theoretical approach is vast.

Researching professionalism The research into medical professionalism has been centred on four fundamental questions; what is professionalism, how is it learned, how is it best taught, and how is it best assessed? As mentioned above, the research on ‘defining’ professionalism has shifted to ‘understanding’ professionalism in different cultural contexts over the past two decades. However, the multi-centre comparative studies addressing this question are very limited. Although there is some evidence on learning and teaching

In summary professionalism should be understood in the backdrop of culture and context. It is learned by students primarily from role models. The professional culture of institutions also plays a major role in fostering professionalism. It should be assessed mainly to facilitate further learning, and multisource feedback and observations are the most useful methods of assessment. All areas of professionalism, i.e. definition to learning to assessment, need to be researched fur ther with sound theoretical underpinning and methodological robustness to advance the existing evidence base. Being aware of what we already know will help us face the future challenges of professionalism education.

REFERENCES 1. Jewell D. After shipman: reforming the GMC - again. Brit J Gen Pract 2005;55(511): 83-84. 2. Irvine D. Doctors in the UK: their new professionalism and its regulatory framework. Lancet 2001;358(9295):1807-10. 3. ABIM. Project Professionalism Philadelphia: American Board of Internal Medicine, 2001:1 31. 4. Papadakis MA, Hodgson CS, Teherani A, Kohatsu ND. Unprofessional behavior in medical school is associated with subsequent disciplinary action by a state medical board. Academic Medicine 2004;79(3):244-49. 5. Harden RM. AMEE Guide No. 14: Outcome-based education: Par t 1-An introduction to outcome-based education. Medical Teacher 1999;21(1):7-14 6. Frank JR, Danoff D. The CanMEDS initiative: implementing an outcomes-based framework of physician competencies. Medical Teacher 2007;29(7):642-47. 7. Zaini RG, Bin Abdulrahman KA, Al-Khotani AA, Al-Hayani AM, Al-Alwan IA, Jastaniah SD. Saudi Meds: a competence specification for S a u d i m e d i c a l g r a d u a t e s . M e d Te a c h 2011;33(7):582-4. 8. Tsugawa Y, Tokuda Y, Ohbu S, Okubo T, Cruess R, Cruess S, et al. Professionalism MiniEvaluation Exercise for medical residents in Japan: a pilot study. Medical Education 2009;43(10):968-78. 9. Sox CH. The ethical foundations of professionalism; a sociologic history. Chest 2007;131(5):1532 - 40. 10. Cruess SR, Cruess RL. Professionalism: a contract between medicine and society JAMC 2000;162(5):668 - 69.

11. Chandratilake M, McAleer S, Gibson J. Cultural similarities and differences in medical professionalism: a multi-region study. Medical Education 2012;46(3):257-66. 12. Al-Eraky MM, Chandratilake M. How medical professionalism is conceptualised in Arabian context: A validation study. Medical Teacher 2012;34:S90-S95. 13. Chandratilake M, McAleer S, Gibson J, Roff S. Medical professionalism: what does the public think? Clinical Medicine 2010;10( 4): 364-9. 14. Hodges BD, Ginsburg S, Cruess R, Cruess S, Delport R, Hafferty F, et al. Assessment of professionalism: recommendations from the Ottawa 2010 Conference. Medical Teacher 2011;33(5):354-63. 15. Aspegren K. BEME Guide No. 2: Teaching and learning communication skills in medicine - a review with quality grading of articles. Medical Teacher 1999;21(6):563-70. 16. Henderson P, Johnson MH. An innovative approach to developing the reflective skills of medical students. BMC medical education 2002;2:4. 17. Aarnio M, Nieminen J, Pyorala E, Lindblom-Ylanne S. Motivating medical students to learn teamwork skills. Medical Teacher 2010;32(4):E199-E204. 18. Bryan CS, Babelay AM. Building character: a model for reflective practice. Acad Med 2009;84(9):1283-8. 19. Hafferty FW. Beyond curriculum reform: confronting medicine's hidden curriculum. Academic Medicine 1998;73(403 - 407). 2 0 . C r u e s s R L , C r u e s s S . Te a c h i n g professionalism: general principles. Medical Teacher 2006;28(3):205-08.

10

21. Cruess SR, Cruess RL, Steinert Y. Teaching rounds - Role modelling - making the most of a powerful teaching strategy. British Medical Journal 2008;336(7646):718-21. 22. van Mook W, van Luijk S, O"Sullivan H, Wass V, Schuwirth L, van der Vleuten C. General considerations regarding assessment of professional behaviour. European Journal of Internal Medicine 2009;20:e90 - e95. 23. Wilkinson TJ, Wade WB, Knock LD. A blueprint to assess professionalism: results of a systematic review. Academic Medicine 2009;84(5):551-8. 24. Rees CE, Knight LV. The trouble with assessing students' professionalism: theoretical insights from sociocognitive psychology. Acad Med 2007;82(1):46-50. 25. Rees CE, Knight LV, Cleland JA. Medical educators' metaphoric talk about their assessment relationships with students: 'you don't want to sort of be the one who sticks the knife in them'. Assess Eval High Edu 2009;34(4):455-67. 26. Roff S, Chandratilake M, McAleer S, Gibson J. Preliminar y benchmarking of appropr iate sanctions for lapses in undergraduate professionalism in the health professions. Med Teach 2011;33(3):234-8. 27. Rees CE, Monrouxe LV. Medical students learning intimate examinations without valid consent: a multicentre study. Med Educ 2011;45(3):261-72.


Journal of Medical Research and Education

JMRE

Photo by Viranuj Sueblinvong, M.D.

11


JMRE

Journal of Medical Research and Education

PERSPECTIVE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Community-Oriented Medical Education

Kanokwan Sriruksa, MD, DiploMedEd Medical Education Center, Khon Kaen Hospital Khon Kaen, Thailand Many medical schools currently adopt SPICES models their education strategy. The model focuses on student-centered, problem-based, integrated, community-based, elective and systematic approach. However the term "community-based" (CBME) or "community-oriented" medical education (COME) is sometimes confusing among educators. This article aims to explore the importance and relevance of CBME and COME in medical education, how COME curriculum can be developed by providing an example from a medical school in Australia.

the patients. The CBME has a slightly narrower scope. CBME emphasizes on learning that takes place in a community for a certain time throughout the training period. It can be said that CBME is a subset of COME. Why COME? The agreement from the 30th World Health Assembly in 19774 was that all member countries have to adopt the "Health For All by the Year 2000" as their common health goal. The concept is that all individual are to achieve a level of health that will permit them to lead a socially and economically productive life. The main strategy is based on the concept of primary health care as described in The Report of the International Conference on Primary Health Care, Alma-Ata in 1978. The key strategy is the development of the health system infrastructure to ensure that primary health care can be delivered to the whole population in the country. In 1986 the First International Conference on Health Promotion was held in Ottawa, Canada and The Ottawa Charter for Health Promotion5 was declared to ensure that all member countries can be able to firmly deploy the strategies to achieve the Global Strategy "Health For All by the Year 2000". However the main focus is still on the primary healthcare delivery. Medical school takes full responsibility in producing medical doctors to serve the global health goal. The most important obstacle, however, is that most medical schools around the world are tertiary care hospitals. Most teaching and learning are diseaseoriented rather than health-oriented. Medical students were exposed to rare, complicated cases rather than common health problems in community where they will serve. The community medicine or preventive medicine, in the past, was just a small area of the whole curriculum and usually ignored by teachers and students. By adopting SPICES model, medical school currently give more attention to

What is COME? In order to understand the concept of COME, the meaning of "community" should be clarified because community is a key word in COME. The meaning of "community" defined by a dictionary is "a social, religious, occupational or other groups sharing common characteristics or common interests and perceive or perceiving itself as distinct in some respect from the larger society within which it exists".2 Taken into account the meaning of community, COME in the context of medical education can be defined as the curriculum that takes into account community health problems throughout the process of development. As Hammad B. defined in his article that COME is a wider term than CBME. COME is "relevant medical education, which takes into consideration…the priority health problems of the country in which it is conveyed...It aims to produce community-oriented doctors who are able and willing to serve their communities and deal effectively with health problems at primary, secondary and tertiary levels…" 3 Therefore COME is not to produce the second class doctors. On the other hand, doctors graduated from COME curriculum should be able to use communityoriented approach in solving the major health problems in the community they work. It can be a family doctor, a surgeon or a pediatrician who keep community-oriented approach in mind while caring

12


Journal of Medical Research and Education

JMRE

COME. Many new medical schools use CBME as their main educational strategies. As mentioned above that COME aims to produce doctors who will be able and willing to work in community. If this goal could be achieved, this would have resulted in having more doctors who are willing to work in remote communities. Hopefully, it can also solve the problem of uneven distribution of doctors which is the common problem in many countries either developed or developing countries.

3. Government and community 4. Personal principles and professional expectations Medical students are the center of the 4 axises connected between the 4 key components. Each axis represents perspectives from each components including; (i) the clinical axis: clinician (medical student) and patient, (ii) the institutional axis: health service (medical student) and university research, (iii) the social axis: community (medical student) and government and (iv) the personal axis: personal principles (medical student) and professional expectations

Do the current medical curriculum apply the concept of COME? Hammad B. proposed a set of questions for medical school to explore how much the school adopt COME concept in the curriculum development. 1. How much the goals, objectives or principles of the curriculum take into consideration the needs or major health problems of the communities they are serving? 2. Does the curriculum cover all health dimensions including health promotion, prevention and rehabilitation in designing learning activities? 3. What is the proportion of community-based and primary healthcare-based activities comparing to the secondary and tertiary care-based? When do community-based and primary healthcare-based activities start in the curriculum? 4. Do teaching and learning make use of the resources or technology in the community? This to help medical student aware of the situation in community where resources or technology may be limited. 5. How much the healthcare system components such as administration and human resource are integrated into the curriculum planning? Are there any healthcare personnel involve in the whole process of curriculum development including planning learning activities, monitoring and evaluation process. 6. Are there any measures to determine medical student competency in community-oriented health care such as community leadership in health supervision, teamworking, and caring for the amjor community health problems. This set of questions aims to be used as a guide for medical school to consider when planning a curriculum. How much to apply depends on the context of each school.

Figure 1. 4Rs model represents the relationships between the 4 components; clinicians and patients, university research and health service, government and community, personal priniciples and professional expectations

The development of curriculum was based on the 4Rs model. As a result, teaching and learning activities that involve or take place in community were integrated as important parts of the curriculum from the first year until the fourth year of training. For example, the first year medical students will learn through the activity called Cultural Safety Workshop. Medical students learn about history, social and culture of the Aborigine who is the native population in Australia. This workshop was designed to increase student awareness of cultural, ethnic and beliefs that may affect healthcare service in community. In the second year medical students will spend one week working with healthcare personnel as well as participate in an activity in community such as disaster preparedness demonstration. They have to conduct a research in community health. In the third year students can choose to spend a year in a program called PRCC which student will work under supervision of general practitioner in community. The fourth year medical students will be able to choose a rotation in different

An example of medical school that adopted the concept of COME Flinders University Rural Clinical School 6 adopted the CBME concept in developing the curriculum called Parallel Rural Community Curriculum (PRCC). This is a 4-year medical training program that enroll graduates in basic sciences into the medical school. The conceptual framework is a 4Rs model which emphasizes on relationships between 4 components. 1. Clinicians and patients 2. Health service and university research

13


JMRE

Journal of Medical Research and Education

specialties such as medicine, surgery, anesthesiology in a community-based hospital. The example from Flinders University, Australia demonstrates how a medical school apply COME concept and adapt to individual context in designing CBME curriculum. People in community including healthcare personnel (Flinders University Rural Health Society) play major role in curriculum development as well as teaching and learning activities, assessment and curriculum evaluation. In summary, to achieve to global goal; Health For All by the Year 2000 and Beyond, and to produce

community-oriented doctors, medical schools need to put effort and resources to integrate COME into the curriculum. Although modern educational trend is moving towards more community-oriented in designing curriculum, translation the concept into reality is still challenging especially for old medical schools where traditional curriculum have been embedded for a long time. Collaboration from all stake holders including people in community as well as local healthcare personnel in development of the curriculum may be the key success.

REFERENCES 1. Harden RM, Sowden S, Dunn WR. Educational strategies in curriculum development: the SPICES model. Med Educ 1984; 18(4): 284-97. 2. D i c t i o n a r y. c o m http:// dictionar y.reference.com/browse/community access 08/07/2012: 11.00 AM

3. Hammad B. Community-oriented medical education: what is it? Med Educ 1991; 25: 16-22. 4. World Health Organization. Global Strategie for Health for All by the Year 2000. Health For All series 3. 1981. 5. World Health Organization. The Ottawa Charter for Health Promotion. http://

14

www.who.int/healthpromotion/conferences/ previous/ottawa/en/ access 26/05/2012: 7.52 PM 6. Worley P. A New Way to Analyse Community-based Medical Education? (Part1). Education for Health 2002;15(2): 117-128. http://furcs.flinder s.edu.au/clinical/cbme/ default.htm access 26/05/2012: 9.51 PM


Journal of Medical Research and Education

JMRE

Photo by Thammasorn Piriyasupong, M.D., Ph.D.

15


Original Articles by Medical Students

Photo by Rapeepun Sangsomwong, M.D.


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Effects of Breast Feeding on Infantile Diarrhea: A Systematic Review Khwanjai Kamhoong1 Suttida Wijitpan1 Teerasak Ouncharoen1 Thammasorn Piriyasupong2, M.D., Ph.D. 1Medical

student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen, Thailand 2Medical Education Center, Khon Kaen Hospital, Khon Kaen, Thailand

ABSTRACT BACKGROUND It has been well-known that human milk has many protective effects for infants. However, effectiveness of breast feeding on infantile diarrhea is still unknown. METHODS A systematic review was conducted by including randomized controlled trials and cohort studies investigating the effectiveness of breast feeding on infantile diarrhea from four online databases. Outcomes were collected in relation to prevalence of diarrhea at three and six months. RESULTS Three studies with 17,265 patients were included in the present review. Comparison between exclusive breast feeding and non-breast feeding, there was no significant difference in relation to prevalence of diarrhea at three months between the two groups. However, at six months, exclusive breast feeding seemed to be able to reduce the prevalence of the diarrhea (relative risk (RR) 0.87; 95% confidence interval (CI), 0.77 to 0.98). These findings were also consistent for the subgroup analysis of the studies from urban areas (RR 0.89; 95% CI, 0.82 to 0.96). However, moderate heterogeneities were observed from both analyses (I2 50% and 30%, respectively) CONCLUSIONS Exclusive breast feeding infants up to six months was likely to be effective on reducing the prevalence of infantile diarrhea.

18


Journal of Medical Research and Education

JMRE

Acute diarrhea is the second most afflicted conditions in children worldwide.1 It is a leading cause of morbidity and mortality in children especially in developing countries with 15 to 30% of deaths in children younger than five years of age.2,3 Moreover, its global incidence in published reports from 1990 to 2000 slightly increased compared with previous reports.4 However, this can be preventable by proper hand washing, few feces dispose and exclusive breast feeding.5 Breast milk is the considered as a good source of nutrition for infants recommended up to six months of age as it can reduce inflammatory response to stimuli in the newborn intestine.6,7 Breast feeding can decrease diarrheal mortality by 24-27% among infants aged 0-5 months, that demonstrate a protective effect of breast feeding on diarrhea in both developed and developing countries.8-13 However, it remains unclear whether breast feeding protects infants against diarrhea. Thus, we designed this review to identify the effect of breast feeding on infantile diarrhea.

sought and retrieved. Additional required data for abstraction into the meta-analysis were requested directly with the authors if necessary. Quality assessment and data extraction The included studies were assessed for their quality using Jadad score with the acceptant level of three or more points for randomized controlled trials. For cohort studies, JBI-SUMARI assessment was used with the cut point of the number of "Yes" counted six or more. After that data of relevant studies were abstracted according to the prevalence of diarrhea in infant who received different breast feeding (exclusive and non-breast feeding) in three and six months. Data synthesis Data were analyzed using the Review Manager 5 software. For dichotomous data as prevalence of infantile diarrhea, the analysis was reported as risk ratio compared between exclusive and non-breast feeding. Heterogeneity of the combined studies was examined using chi-square and I2. The meta-analysis was performed using random effect model

METHODS

RESULTS

Study design This is a systematic review included randomized controlled trials and cohort studies investigating the effectiveness of breast feeding on infantile diarrhea.

We searched 1224 studies from four databases (Figure 1). Of 1224 relevant studies, 11 studies met the inclusion criteria. We excluded five studies; one was case-control study, one was cross-sectional study and three studies reported outcome did not meet inclusion criteria, one study was excluded because of its outcomes showed only incidence rate of death among infant, one showed incidence rate of diarrheal episodes per child-year and another showed incidence rate of gastro-intestine tract infection. Six studies were assessed using Jadad score for RCTs and JBI SUMARI for cohort studiesand this excluded three studies as Jadad score were less than three points or JBI SUMARI assessed that the number of "yes" response from the evaluation was less than six (Table 2 and Table 3) Finally three studies were included in the meta-analysis, one was a randomized controlled trial and the others were cohort studies. The included randomized controlled trials included 1115 infants in eight communities, 464 mothers (intervention group) were randomly assigned to receive exclusive breast feeding counseling and 411 mothers (control group) were randomly assigned to receive routine counseling (Table 4). Its outcome was intervention effects on exclusive breast feeding diarrheal prevalence at three and six months of age. For two cohort studies, one study included 15 890 healthy, singleton and term infants. Their information was collected and assessed by the questions about breast feeding and hospital admission for diarrhea that outcome measures were parental report of

Criteria of included studies Participants Studies of the infants age between 0 to 24 months Intervention Breast feeding that individual infant received that were categorized into exclusive and non-breast feeding. Outcomes Prevalence of infantile diarrhea at three and six months Exclusion criteria Study with following characteristics were excluded; study with participants with underlying disease such as congenital anomaly and infection disease who used antibiotics, mother that breastfed more than one infant, study that resulted the outcome in episodes, breastfed with formula or not observed in three or six months. Search strategies Included studies were identified and screened for title and abstracts through four databases including the Cochrane library, SCOPUS, PubMed and Ovid. Initial keywords to be used for the review components and their respective combinations used for each search are illustrated in Table1. Full relevant articles were

19


JMRE

Journal of Medical Research and Education

Screening

Records identified through four database searching (n = 1224)

Excluded on title and abstract (n = 1213)

Records screened (n = 11)

Full-text articles assessed for eligibility (n = 11) Excluded on content (n = 5)

Eligibility

Research design not met (n = 2) Outcome measures didn’t meet the criteria of outcome measure for this review (n=3)

Studies include in qualitative synthesis (n = 6)

Included

Excluded on quality assessment(JBI-SUMARI) from cohort study <6 (n = 3)

Studies include in quantitative synthesis (meta-analysis) (n=3) cohort=2, RCTs=1

Figure 1. Flow Diagram of Selection Process of the Included Studies

20


Journal of Medical Research and Education

JMRE

Table 1. Combination of Keywords Used for Search Strategy Number of studies

Terms used for search in each database PubMed "Diarrhea, Infantile/prevention and control"[Mesh] AND "Breast Feeding"[Mesh]

88

"Diarrhea, Infantile/prevention and control"[Mesh] AND "human milk"[Mesh] "Diarrhea, Infantile"[Mesh] OR "Diarrhea, Infantile/prevention and control"[Mesh] AND "Breast Feeding"[Mesh] AND "Milk, Human"[Mesh] "Breast feeding"[Mesh] AND "Diarrhea"[Mesh] AND "Infant"[Mesh] AND "prevention and control "[Subheading]

0

156

"Milk, Human"[Mesh] AND "Diarrhea"[Mesh] AND "Infant"[Mesh] AND "prevention and control "[Subheading]

49

46

Cochrane registry Infantile diarrhea and breast feeding and prevention

2

Infantile diarrhea and breast feeding

25

Scopus ALL(infantile diarrhea) AND ALL(breast feeding)

659

ALL(infantile diarrhea) AND ALL(breast feeding) AND prevention

179

Ovid (Infantile diarrhea AND breast feeding).

3

(Infantile diarrhea and breast feeding).

17

Total

1,224

hospitalization for diarrhea at three months. Another study included 170 healthy, singletons, birth weight more than 2500 g and term infants were followed for six months. Its outcomes were incidence of diarrhea by category of breast feeding (exclusive breast feeding and non-breastfeeding) at three and six months of age. We analyzed data and reported as risk ratio and 95% confidence interval (CI) comparing between exclusive breast feeding and non-breast feeding. We assessed statistical heterogeneity of the combined studies using I2and chi-square test (Table 5). In the present review, high heterogeneity was observed (I²=95%; chi-square=64.69; P=0.47).

1.49, P=0.97) and there was high heterogeneity (I²=95%; chi-sqaure=20.88; P<0.00001). However, the effect was significant at six months (RR = 0.89, 95% CI=0.82-0.96, P=0.002) with moderate heterogeneity (I²=30%; chi-square=1.43; P<0.23). The effect of breast feeding on diarrheal prevalence in rural (Lopez et al.) is presented in Figure 2. In relation to time on breast feeding, the protective effect was not shown for exclusive breast feeding for three months (RR=0.28, 95% CI, 0.07 to 1.16, P=0.08). Addition to the similar effect was found for exclusive breast feeding for six months (RR=0.36, 95% CI, 0.09 to 1.37, P=0.13). Regarding the three combined studies, exclusive breast feeding showed no significant protective effect on prevalence of infantile diarrhea breast milk (RR=0.92, 95% CI, 0.61 to 1.37, P=0.67) with high heterogeneity (I²=92%; chisquare=23.76; P<0.00001) at three months. Nevertheless, there was significant protective effect of

Outcomes The effect of breast feeding on diarrheal prevalence in urban is presented in Figure 2. In relation to time, the protective effect was not significant on prevalence of diarrhea at three months (RR=0.99, 95% CI, 0.66 to

Table 2. Quality Assessment using Jadad Score for Randomized Controlled Trial Jadad Score

Bhandariet al.

Was the study described as randomized?

1

The method of randomization was described in the paper,and that method was appropriate

1

Was the study described as double blind?

0

The method of blinding was described, and it was appropriate.

0

Was there a description of withdrawals and dropouts?

1

Total

3

21


JMRE

Journal of Medical Research and Education

exclusive breast feeding on infantile diarrhea at six months (RR=0.87, 95% CI, 0.77 to 0.98, P =0.02) and there was moderate heterogeneity for the study combination (I²=50%; chi-square=4.01; P<0.13).

birth weight; 2.7 kilogram for Bhandari and 3.45 kilogram for Maria A study. Contrary to reports from other populations, birth weight did not influence breastfeeding significantly at any age. Explanation is that because low birth weight is common, small infants are accepted as the norm and size does not influence feeding decisions.20 Our study has many limitations, we pooled the results of studies that were not originally intended to explore the outcomes. Much of the studies have depended on socioeconomic communities, self-report and recall data, memory is not precise and biases are often occurred.21 Bias can be present in cohort studies are selection, information, confounding and follow-up bias, the way measures are taken, or data analysis. There is evidence that mother who breastfed exhibit more nurturing behavior than mother who feed their infants with artificial breast milk.22 Our study also has another bias and confounder. In relation to selection bias, limited study design, published and language of study make it narrow down to collect the papers. Our data suggest exclusive breast feeding until six months of age associated with reduce rate of diarrhea. In Bangladesh, comparison between partial or no breastfeeding with exclusive breastfeeding in the first few months of life showed risk of infant deaths from diarrhea is 3.94 to fold higher.23 Thus infants who are exclusively breastfed for six months have less morbidity from gastrointestinal infection than those who are mixed breastfed of three or four months.24 However, our study shows high heterogeneity, providers should carefully consider because in practice there is difference among women who give breast feeding such as feeding practice, duration of feeding, pattern of dietary practices evident in first six months of life, salutary effects of human colostrum and cultural tradition.25 Incidence rate of infantile

DISCUSSION This meta-analysis reviews the effect of breast feeding on infantile diarrhea. As the result, our study showed significant effect of exclusive breast feeding compares with non-breast feeding on prevalence of infantile diarrhea at six months but not at three months. The diverse outcomes of the two-time period might be due to biological effect of human milk that provides protective effects on infants.14 This might be due to the time requirement for development of immune system of the infants after three months of birth. Before having its own defense system, transfer of protective components from mother such as immunoglobulin in breast milk will help the infants fight against exogenous pathogens, coat the intestinal mucosa and prevent bacteria from entering the cells. 15,1617

In the present review, high heterogeneities were observed in the meta-analyses. The differences of characteristics such as socioeconomic status as it is the important confounders in communities-based studies as well as the various maternal education, feeding practice, collected data that was gather from questionnaire, sanitation, trained field nurse who gave counsel to mothers and dissimilarity of maternal age can affect the homogeneity of the combined outcomes. For instance, in Bhandariâ&#x20AC;&#x2122;s study, the average age of mother was about 23 years while nearly 30 years old in Maria A.18,19 Moreover this might be due to various birth weight (kilogram) in relation to two studies included difference of mean Table 3. Quality Assessment using JBI-SUMARI JBI-SUMARI assessment Is the sample representative of patients in the population as a whole?

Lopaz-Alarcon et al. Maria A. Quigley et al. Y Y

Are the patients at a similar point in the course of their condition/ illness?

Y

Y

Has bias been minimized in relation to selection of cases and of controls?

Y

Y

Are confounding factors identified and strategies to deal with them stated?

Y

Y

Are outcomes assessed using objective criteria?

N

N

Was follow up carried out over sufficient time period?

Y

Y

Were the outcomes of people who withdrew described and included in the analysis?

Y

Y

Were outcomes measured in a reliable way?

N

N

Was appropriate statistical analysis used?

Y

Y

Y=7, N=2

Y=7, N=2

Total Y = Yes, N = No, U = Unclear

22


Millennium Maria A. Quigley et al. 15980 Cohort Study singleton infants (natural mother), age between 6 and 12 months (mean and median: 9 months) who did not have major problems at birth

Bhandari et al.

2

3

23 1115 normal infants from 8 communities of Haryana, India, age between 0 and 6 months

- Episode, odd ratio, relative risk, survival analyses plot, multiple regression - Observe at age 1, 2, 3, 4, 5 and 6 months

Outcome measure

Incidence, prevalence, and duration of individual episodes of diarrhea were also lower in breast-fed infants.

Result

Control group: Infants who were not gave beast-fed in first 3 months after birth.

The 7-day diarrhea prevalence was lower in the intervention than in the control communities at 3 months (0.64, 0.44â&#x20AC;&#x201C;0.95, p=0.028) and 6 months (0.85, 0.72â&#x20AC;&#x201C;0.99, p=0.04).

Experimental group: - Hospital admission for Reduced prevalence of diarrhea was Infants who were gave partially diarrhea in first 8 month significant in experimental group breast feeding or exclusively breast - Association between feeding in first 8 months after birth. breastfeeding and hospital admission per month in first 8 Control group: months. Infants who were not gave beast-fed - Association between in first 8 months after birth. months since breast feeding Cessation (in 1-4 months and 5-7 months) and hospitalization for diarrhea in first 8 months - No. of infants who were gave exclusively breast - fed at 4th and 6th months

Control group: Non-breast feeding-formula

Experimental group: Exclusive breast feed, Partial breast feed.

Intervention

Community- Experimental group: -Diarrhea prevalence based Infants who were gave exclusive at age 3 months and 6 months randomized breast feeding or partially breast control trial feeding in first 3 months after birth.

170 normal Cohort infant that study delivery in hospital, age between 0 and 6 months

Lopaz-Alarcon et al.

Study design

1

Participant

Author & Year

Reference number

Table 4. Summaries of the Included Studies

Journal of Medical Research and Education JMRE


JMRE

Journal of Medical Research and Education

Table 5. Heterogeneity of the Included Studies Outcome In urban, at 3 months

Chi-Square

P Value

I2

20.88

<0.00001

95

In urban, at 6 months

1.43

0.23

30

Comparison of three studies, at 3 months.

23.76

<0.00001

92

Comparison of three studies, at 6 months.

4.01

0.13

50

!A b.#

*!

**!

!B a.#

Figure 2. Forest Plots of Study Outcomes Panel A: Comparison of the three studies; Panel B: Sub group comparison in only in urban area *EBF exclusive breast feeding **NBF non-breast feeding

24


Journal of Medical Research and Education

JMRE

diarrhea in rural is more morbidity than in urban more over sanitary is associated with risk of diarrhea thus breast feeding in this communities should be promoted.For proper results, we need more studies with high quality and low bias. In conclusion, this meta-analysis shows that exclusive breast feeding provide significant protective

effect on infantile diarrhea at six months, despite it has high heterogeneity because of difference between birth weight, maternal age, collection of data and feeding practices. We consider that our results are in l i n e w i t h t h e Wo r l d H e a l t h O r g a n i z a t i o n recommendation of exclusive breast feeding until infants are six months.

REFERENCES 1. The United Nations Children’s Fund (UNICEF)/Wor ld Health Or ganization (WHO), 2009.Diarrhoea: Why children are still dying and what can be done,2009:1(http:// whqlibdoc.who.int/publications/ 2009/9789241598415_eng.pdf) 2. Measurement Issues in Trials of Pediatric Acute Diarrheal Diseases: A Systematic Review Bradley C. Johnston, PhD et al. PEDIATRICS Vol. 126 No. 1 July 2010, pp. e222-e231 3. Kosek M, Bern C, Guerrant RL. The global burden of diarrhoeal disease as estimated from studies published between 1992 and 2000.Bull World Health Org.2003;81:197-204 4. Stefano Guandalini,Diarrhea, Emedicine.medscape.com,2010. 5. Family and community practicesthat promote child sur vival,growth and developmentA REVIEW OF THE EVIDENCEZelee Hill, Betty Kirkwood and Karen Edmond, Public Health Intervention Research Unit, Department of Epidemiology and Population HealthLondon School of Hygiene and Tropical Medicine 6. Newton ER: Breastmilk: The gold standard.  ClinObstetGynecol  2004; 47:632. 7. For cause-specific mortality: The World Health Report 2003, WHO, Geneva. For malnutrition: Pelletier, D. L., E. A. Frongillo, and J. P. Habicht, ‘Epidemiologic evidence for a potentiating effect of malnutrition on child mortality’, American Journal of Public Health, vol. 83, no. 8, August 1993, pp. 1130-1133. 8. Family and community practices that promote child sur vival,growth and development ,A REVIEW OF THE EVIDENCE , Zelee Hill, Betty Kirkwood and Karen Edmond ,Public Health Intervention Research Unit Department of Epidemiology and Population Health London School of Hygiene and Tropical Medicine 9. Adair L et al. (1993).Growth dynamics during the first two years of life: a prospective study in the Philippines. European Journal of Clinical Nutrition, 47:42–51; Cohen RJ et al. (1994). Effects of age of introduction of complementary foods on infant breast milk intake, total energy intake, and growth: a randomised intervention study in Honduras.

Lancet, 344(8918):288–293; Simondon KB, Simondon F (1997). Age at introduction of complementary food and physical growth from 2 to 9. months in rural Senegal.European Journal of Clinical Nutrition, 51(10): 703–707; Dewey KG et al. (1999). Age of introduction of complementary foods and growth of term, low-bir th-weight, breast-fed infants: a randomized intervention study in Honduras. AmericanJournal of Clinical Nutrition, 69(4): 679–686; Kramer MS et al. (2001). Promotion of Breastfeeding Intervention Trial (PROBIT): a randomized trial in the Republic of Belarus. Journal of the American Medical Association, 285(4):413–420. 10. Feachem RG, Koblinsky MA (1984). Interventions for the control of diarrhoeal diseases among young children: promotion of breastfeeding. Bulletin of the World Health Organization, 62:271–291. 11. Leon-Cava N et al. (2002). Quantifying the benefits of breastfeeding: a summary of the evidence.Washington, DC, Pan American Health Organization (ISBN 92-75-12397-7). 12. Howie PW et al. (1990). Protective effect of breast feeding against infection.British Medical Journal, 300(6716):11–16; Dewey KG, Heinig MJ, Nommsen-Rivers LA (1995).Differences in morbidity between breast-fed and formula-fed infants.Journal of Pediatrics, 126(5):696–702; Scariati PD, Grummer-Strawn LM, Fein SB (1997).A longitudinal analysis of infant morbidity and the extent of breastfeeding in the United States.Pediatrics, 99(6):E5; Van Derslice J, Popkin B, Briscoe J (1994). Drinking water quality, sanitation and breastfeeding: their interactive effects on infant health. Bulletin of the World Health Organization, 72,589–601; Ahmed F et al. (1992). Community-based evaluation of the effect of breast-feeding on the risk of microbiologically confirmed or clinically presumptive shigellosis in Bangladeshi children. Pediatrics, 90(3):406–411. 13. Breast-Feeding Lowers the Frequency and Duration of Acute Respiratory Infection and Diarrhea in Infants under Six Months of Age, Mardya Lo´ pez-Alarco´ n, Salvador Villalpando2 and Arturo Fajardo.

25

14. WHO Library Cataloguing-in-Publication Data(2009) , Infant and young child feeding : model chapter for textbooks for medical students and allied health professionals:p9 15. Simister NE: Placental transpor t of i m m u n o g l o b u l i n G . Va c c i n e 2 0 0 3 ; 21:3365-3369. 16. Newburg DS, Peterson JA, Ruiz-Palacios GM, et al: Role of human-milk lactadherinin protection against symptomatic rotavirus infection. Lancet 1998; 351:1160-4. 17. Michael S Kramer, RitsukoKakuma, Optimal duration of exclusive breastfeeding (2009) : 2 18. WHO Library Cataloging-in-Publication Data, Diarrhoea: Why children are still dying and what can be done,2009: page11, 19 1 9 . Wo r l d H e a l t h O r g a n i z a t i o n . Contemporar ypatterns of breastfeeding. Report on the WHOCollaborative Study on B r e a s t fe e d i n g . G e n e v a , Wo r l d H e a l t h Organization, 1981. 20. (Barros FC, Victora CG, Vaughan JP, Smith PG. Birth weight and duration of breastfeeding: are the beneficial effects of human milk being overestimated? Pediatr ics. 1986;78:656–661 ;;;;; Adair L, Popkin BM. Low bir th weight reduces the likelihood of breastfeeding among Filipino infants. J Nutr. 1996;126:103–112 21. Gabbe: Obstetrics: Normal and Problem Pregnancies, 5th ed.(2007) 22. Crow RA, Fawcett JN, Wright P.: Maternal behavior during breast- and bottle-feeding.    J Behav Med  1980; 3:259. 23. R2Shams Arifeen Exclusive Breastfeeding Reduces Acute Respiratory Infection and Diarrhea Deaths Among Infants in Dhaka SlumsPediatrics2001;108;e67). 24. UNICEF, WHO, UNESCO, UNFPA, UNDP, UNAIDS, WFP and the World Bank,Facts for Life Fourth Edition: 2010 : page 47 (http://www.factsforlifeglobal.org/resources/ factsforlife-en-full.pdf) 25. Bernardo L. Horta et al.Evidence on the long-ter m effects of breastfeeding SYSTEMATIC REVIEWS AND METAANALYSES,Wor ld Health Organization 2007:p4


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Effects of Antibiotics on Length of Hospital Stay in Children with Acute Diarrhea Kiriya Jitanapakarn1 Wansiri Inhangwa1 Warujpong Boonkum1 Weeraphat Sabphoke1 Thammasorn Piriyasupong2, M.D., Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen, Thailand 2Medical Education Center, Khon Kaen Hospital, Khon Kaen, Thailand

ABSTRACT INTRODUCTION The role of antibiotics for treating diarrhea is still debatable. This study aims to identify the effect of antibiotics on length of hospital stay. METHODS This was a retrospective cohort study analyzing the medical records of children age five years old or younger with the diagnosis of acute diarrhea using the international classification of disease 10 (ICD-10) that admitted between July 2009 and June 2010. Outcomes were length of hospital, adverse effects and factors associated with the use of antibiotics. RESULTS There were 409 children included; 18.8% were treated with antibiotics. Most of them were male with the median age of 12 months old. It found that the use of antibiotics were associated with lengthening hospital stay in the bacterial and unspecified diarrhea (P<0.001 for both groups) but not in those with diarrhea caused by virus or parasite. Only one case had adverse effects from the use of antibiotics. White blood cell from the blood analysis, rectal swab culture and time starting antibiotics were found to be associated significantly with the length of stay. CONCLUSION Using antibiotics might prolong hospital stay especially in children with diarrhea caused by bacteria or unspecified diarrhea.

26


Journal of Medical Research and Education

JMRE

Acute diarrhea is a major public health problems worldwide.1,2 Each year diarrheal disease causes more than 3 million deaths worldwide in children under five years of age.2 The World Health Organization (WHO) recommends strategies for treating children with diarrhea including prevention of dehydration through early administration of appropriate fluids available at home such as using oral rehydration salt (ORS) solution, treatment of severe dehydration with an intravenous (IV) electrolyte solution, continued feeding during, and increased feeding after the diarrheal episode, and selective use of antibiotics and refrain from using anti-diarrheal drugs.2 Empirical antibiotic treatment is recommended only for dysentery, suspected cholera and invasive bacterial diarrhea.3,4 Previous surveys have found that 23.8 to 72.6% of Thai physicians overuse antibiotics,5,6 that is similar to data from other countries where antimicrobial therapy is prescribed as high as 16 to 94% of episodes of acute diarrhea.7-12 Inappropriate use of antibiotics can causes antimicrobial drug resistance among enteric pathogens which raises the cost of health care, increases the risk of drug reactions and delays appropriate treatment.13 Moreover, the benefit of using antibiotic in children with acute diarrhea both worldwide and in Thai children is unknown and still controversy whether it can improve the treatment outcome in those patient in term of reduction of mortality rate.14,15 Still, another study shows no significant results in relation to hospital stay comparing the use of antibiotics and IV electrolyte solution for treating acute diarrheal episode.6,18 So that, in this study, we aim to identify the effect of using antibiotics for treating acute diarrhea in children age younger than five years of age in relation to length of hospital stay.

chemotherapy, surgery condition or allergy against antibiotics. Medical record Baseline data including age, gender, presenting symptoms (e.g., fever, diarrhea and vomiting), dehydration condition17 (mild, moderate and severe), laboratory investigation (stool examination, rectal swab culture and complete blood count), stool texture were reviewed and collected from the included medical records. Outcomes including length of hospital stay, adverse effects and prescribed treatment were also determined. Diarrhea from virus

A08 Viral and other specified intestinal infections

Diarrhea from bacteria

A00 Cholera A01 Typhoid and paratyphoid fevers A02 Other salmonella infections A03 Shigellosis A04 Other bacterial intestinal infections A05 Other bacterial food borne intoxications, not elsewhere classified

Diarrhea from parasite

A06 Amoebiasis

Unspecified Diarrhea

A09 Diarrhea and gastroenteritis of presumed infectious origin

BOX 1. International Classification of Disease 10 of diarrhea including in the present study

METHODS

Statistical analysis Before data analysis, all recorded data were cleaned and validated. For descriptive statistics, categorical data were presented in relation to number and percentage. Numeric data were test for their normal distribution using Kolmogorov-Smirnov test and presented in relation to median and Interquartile range if they were not normally distributed. For inferential statistics, chi-square was used for the analysis of categorical data or exact test where appropriate. Mann-Whitney U test was used to compare the different between two groups and Kruskal Wallis test for three or more groups. Multivariable analysis using the linear regression to determine factors associated with length of hospital stay was used at the end of the analysis. All statistics were significant at the level of 0.05.

Study design This is a retrospective cohort study aimed to identify the effect of antibiotics on length of hospital stay Eligibility Medical records of children aged five years old younger with the diagnosis of acute diarrhea using the International Classification of Disease 10 (ICD to 10) and admitted to the Department of Pediatrics, Khon Kaen Hospital between July 2009 and June 2010 were reviewed. See Box 1 for the detail of various types of diarrhea including in the present study. We excluded the records of those with underlying diseases such as thalassemia, congenital anomaly, immune deficiency, cancers, pneumonia or upper respiratory tract infection and those on the treatment of steroid,

27


JMRE

Journal of Medical Research and Education

Table 1. Baseline Characteristics Characteristic Median age (range) – month Male sex–no. (%)

Antibiotics (N=77)

No antibiotics (N=332)

P Value

12 (11-24)

12 (12-28.5)

0.291

49 (63.6)

194 (58.4)

0.393

Symptoms–no. (%)

0.430

Fever

0

1 (0.3)

Vomiting

0

12 (3.6)

Diarrhea

10 (13)

39 (11.7)

Fever with vomiting

5 (6.5)

23 (6.9)

Fever with diarrhea

31 (40.3)

68 (20.5)

4 (5.2)

99 (29.8)

27 (35.1)

90 (27.1)

Vomiting with diarrhea Fever with vomiting with diarrhea Dehydration–no. (%)

0.165

No

13 (16.9)

27 (8.1)

Mild

32 (41.6)

153 (46.1)

Moderate

29 (37.7)

150 (45.2)

3 (3.9)

2 (0.6)

Severe WBC count–no. (%)

0.004

PMN predominate

24 (31.2)

71 (21.4)

Lymphocytosis

25 (32.5)

71 (21.4)

Normal

25 (32.5)

145 (43.7)

3 (3.9)

45 (13.6)

None Stool exam–no. (%)

< 0.001

WBC positive

10 (13)

5 (1.5)

RBC positive

3 (3.9)

4 (1.2)

RBC with WBC positive

8 (10.4)

7 (2.1)

Normal

33 (49.9)

163 (49.1)

None

23 (29.9)

153 (46.1)

Rectal swab culture–no. (%)

0.036

Positive

2 (2.6)

1 (0.3)

Negative

36 (46.8)

94 (28.3)

None

39 (50.6)

237 (71.4)

Stool texture–no. (%)

0.063

Watery

59 (76.6)

292 (88)

Bloody

15 (19.5)

6 (1.8)

Normal

3 (3.9)

34 (10.2)

PMN=Neutrophil; WBC=White blood cell; RBC=Red blood cell.

28


Journal of Medical Research and Education

JMRE

480 Patients preliminary included

71 Were excluded 32 Pneumonia or upper respiratory tract infection 4 With underlying of cancer 8 With underlying of thalassemia

409 Patients

77 Treated with antibiotics

332 Non-treated with antibiotics

77 Were included in the analysis

332 Were included in the analysis

Figure 1. Study Flow Chart

were found associated significantly with the use of antibiotics (Table 1) while ages, genders, presenting symptoms, dehydration conditions (mild, moderate, severe) and stool texture were not associated significantly with the use of antibiotics. Of 409 patients, they were divided into four groups regarding the pathogen; (i) virus, (ii) bacteria, (iii) parasite and (iv) unspecified. About 66% of diarrhea from virus was treated with ORS combined with intravenous (IV) fluid and antiemetic drugs (Table 2). A bit more than 31% of diarrhea from bacteria was

RESULTS A total of 480 medical records of children with diarrhea were included and 71 cases were excluded regarding the exclusion criteria (Figure 1). Seventyseven cases (18.8%) of the 409 were treated with antibiotics compared with 332 cases (81.2%) without the use of antibiotics. Laboratory reports such a white blood cell count from complete blood count, positive stool examination and positive rectal swab culture

29


JMRE

Journal of Medical Research and Education

Table 2. Prescribed Treatment Diarrhea due to Diarrhea due to Diarrhea due to virus bacteria parasite

Treatment

Unspecified Diarrhea

no. (%) ORS only

0

0

0

8 (2.3)

IV fluid only

0

0

0

12 (3.5)

1 (4.2)

1 (2.9)

0

64 (18.7)

Antiemetics

2 (8.3)

0

0

0

Antibiotics

0

1 (2.9)

0

0

Antiemetics

2 (8.3)

0

0

7 (2.0)

Antibiotics

0

2 (5.7)

0

2 (0.6)

Antibiotics plus antiemetics

0

0

0

1 (0.3)

16 (66.7)

9 (25.7)

2 (28.6)

206 (60.1)

ORS plus IV fluid ORS plus drugs

IV fluid plus drugs

ORS plus IV fluid and drugs Antiemetics Antibiotics

0

11 (31.4)

1 (14.3)

17 (5.0)

3 (12.5)

11 (31.4)

4 (51.1)

26 (7.6)

Amoxycillin

0

0

0

4 (9.1)

Augmentin

0

0

1 (20.0)

0

Ceftriaxone

0

17 (68.0)

3 (60.0)

21 (47.7)

Cefotaxime

0

1 (40.0)

0

1 (6.8)

Amplicillin

0

0

0

2 (4.5)

Gentamycin

0

0

0

1 (4.5)

Amplicillin plus gentamycin

0

2 (8.0)

0

2 (4.5)

Co-trimoxazole

0

4 (16.0)

1 (20.0)

1 (11.4)

3 (100.0)

1 (40.0)

0

5 (11.4)

Antibiotics plus antiemetics Type of antibiotics

Norfloxacin

ORS=Oral rehydration salt; IV fluid=Intravenous fluid

Table 3. Length of Hospital Stay, Time to Start Antibiotics and Types of Diarrhea Median Length of Hospital Stay (days) Type of diarrhea

No Antibiotics

P Value

0

1 (1-2.5)

0.778

4.5 (3-4.5)

8 (7-8)

1 (1-1.3)

< 0.001

4 (2-4)

0

7 (6-7)

2.5 (2-2.5)

0.517

4.5 (2-6)

3 (2-3)

5 (3.5-8)

2 (1-3)

< 0.001

Start antibiotics 0 – 24 hr

Start antibiotics 24 – 48 hr

Start antibiotics 48 – 72 hr

Start antibiotics >72 hr

1

2

0

Diarrhea from bacteria

2 (2-3)

2.5 (2-3)

Diarrhea from parasite

2 2.5 (1.3-4)

Diarrhea from virus

Unspecified Diarrhea

30


Journal of Medical Research and Education

JMRE

treated by ORS combined with IV fluid and antibiotics, and others 31% was treated with ORS combined with antibiotics and antiemetic drugs and nearly 51% of diarrhea from parasite was treated by ORS combined with IV fluid, antibiotics and antiemetic drugs. For unspecified diarrhea, about 61% was treated by ORS combined with IV fluid and antiemetic drugs. In relation to the use of antibiotics, three cases of viral diarrhea were prescribed with antibioticsnorfloxacin for all three cases. About 25 cases (68.0%) of bacterial diarrhea and five cases (60%) of diarrhea from parasite were treated with ceftriaxone as antibiotics, as well as 44 cases of unspecified diarrhea (47.7%) Regarding the length of hospital stay, in the cases with diarrhea from bacteria and unspecific diarrhea and treated with antibiotics, they tended to have longer hospital stay (P<0.001). However, diarrhea from virus and diarrhea from parasite were found no association with the length of hospital stay (P=0.385 and 0.281), respectively (Table 3). Furthermore, only one case had adverse effects (skin rash) from the use of cefotaxime. In the present study, there were many factors affected the length of hospital stay aside from the use of antibiotics. Univariable analysis in Table 4 shows that presenting symptoms, positive stool examination, positive rectal swab culture, high white blood cell count and delayed time starting antibiotics were found associated significantly with the length of stay in hospital. From the multivariable analysis, it confirmed that high white blood cell count, positive rectal swab culture and delayed time starting antibiotics were still found associated significantly with longer length of stay in hospital. However, presenting symptoms, stool examination, stool texture, ages, genders, dehydration conditions and types of treatment given were found no associated significantly with the length of hospital stay.

with the length of hospital stay. In the antibiotics using group preferred use of neutrophil predominate, lymphocytosis, red blood cell count, white blood cell positive in stool exam, rectal swab culture positive. Others ages, gender, presenting symptoms, dehydration and stool texture had no significant in decision making for using antibiotics. Most acute diarrhea cases in this study were treated by oral rehydration salt, IV fluid and antiememtic drug that depended on clinical of patient and in this study preferred to use ceftriaxone and norfloxacine than ampicilin, gentamycin, cotrimaxazone, cefotaxime and amoxycillin. Furthermore, from this study there was case had adverse effects from the use of antibiotics. This study purposed to find out a conclusion of the way to treat patient with acute diarrhea that still controversy. For this study, the result was a using of antibiotics did not decreases the length of hospital stay but increase the length of hospital stay and drug adverse also. However, the limitation of database, such as a source of database, the misunderstood of physicians about ICD 10 and the differences about questions of physician during data collection. Furthermore, the use of length of hospital stay to be an outcome might effect on result of the study because there were other factors effected on the length of hospital stay, for example white blood cell count, rectal swab culture, time starting antibiotics and the uncollected data (the needs and doses of IV form antibiotics for treatment, other familial reasons). In additional if this study had collected the length of hospital stay directly from the time starting antibiotics, the result of study would have been more accuracy. In this study, antibiotics using had no effect to decrease the length of hospital stay, but increase the length of hospital, increase the cost of hospital stay16, additional adverse drug effect, the longer time of treatment also related to the previous study that result in the overuse of antibiotics caused many problems such as drugs resistance, cost, drugs adverse and longer time to treat13. Then the use of oral rehydration salt and IV fluid combined with antibiotics were not associated with the length of hospital stay as well as the previous study that the use of antibiotics and oral rehydration salt had no differentiation in the result of treatment6,18. Mostly of acute diarrhea cases were treat by oral rehydration fluid, IV fluid and antiemetic drug and also related to the previous study that treated acute diarrhea in children via water lost prevention by gave replacement fluid suddenly from home or gave oral rehydration salt for treat mild dehydration stage and the treatment of severe dehydration with IV fluid, continued feeding during, and increase feeding after the diarrheal episode, selective use of antibiotics and none use of anti diarrheal drugs.2,19 In conclusion, the use of antibiotics had no effect to decrease the length of hospital stay but increase risk of drug adverses. The appropriated treatment of acute diarrheal was oral rehydration salt, IV fluid and antiemetic drug.

DISCUSSION From the four groups by of diarrhea in children age younger than five years old, length of hospital stay in children with diarrhea from bacteria and unspecified diarrhea and treated with antibiotics had longer length of hospital stay than those without antibiotics. Other factors affected the length of hospital stay included high white cell count, positive rectal swab culture and time delayed starting antibiotics had were found associated significantly with the length of hospital stay. Although, presenting symptoms, stool examination, stool texture, age, gender, dehydration (mild, moderate, severe), white blood cell count and treatment were found no associated significantly with the length of hospital stay. Furthermore, the result from studying found that how physicians choose antibiotics for use was based on laboratory result such complete blood count, rectal swab culture and stool examination that were found associated significantly

31


JMRE

Journal of Medical Research and Education

Table 4. Factor Associated with Length of Hospital Stay Variable Median hospital stay P Value from P Value from logistic (range) univariable analysis regression analysis Age group (month) 0.076 0.211 0-12

2 (1-3)

13-24

2 (1-3)

25-36

1 (1-2)

37-48

2 (1-3)

49-60

2 (1-3)

61-71

2 (1-5.5)

Sex Male

2 (1-3)

Female

2 (1-3)

Symptoms Fever

1

Diarrhea

2 (1-3)

Fever with vomiting

2 (1-3)

Fever with diarrhea

2 (1-3)

Vomiting with diarrhea

1 (1-2)

Fever with vomiting and diarrhea

2 (1-3)

Dehydration No

2 (1-3)

Mild

2 (1-3)

Moderate

2 (1-3)

PMN predominate

2 (1-3)

Lymphocytosis

2 (1-3)

None

0.077

0.816

< 0.001

0.001

<0.001

0.279

<0.001

0.003

2 (1-3)

WBC positive

3 (2-4)

RBC positive

2 (2-3)

RBC with WBC positive

2 (1-3)

Normal

2 (1-3)

None

1 (1-2)

Rectal swab culture Positive

None

0.253

1 (1-1.8)

Stool exam

Negative

<0.001

3 (2-5.5)

WBC count

Normal

0.702

8

Vomiting

Severe

0.077

2 2 (1-4) 1.5 (1-2)

32


Journal of Medical Research and Education

JMRE

Table 4. (Continued.) Variable

Median hospital stay (range)

Univariable model Multivariable model P Value P Value

Stool texture Watery

2 (1-3)

Bloody

3 (2-3)

Normal

1 (1-2)

Time to start antibiotics 0-24 hr

2 (1.3-3.8)

24-48 hr

3 (2-6)

48-72 hr

3 (2.5-6.5)

>72 hr No antibiotic

IV fluid only

<0.001

0.994

0.109

2 (1-3) 2 (1-2.8) 2 (1-2)

ORS plus other drugs*

2 (1-2)

ORS plus IV fluid and ther drugs*

<0.001

2 (1-3)

ORS plus IV fluid

IV fluid plus other drugs*

0.167

7 (4.5-8)

Treatment ORS only

0.082

2 (1.8-2.3) 2 (1-3)

PMN=Neutrophil; WBC=White blood cell; RBC=Red blood cell; ORS=Oral rehydration salt; IV fluid=Intravenous fluid. *Other drug included antiemetics and antibiotics REFERENCES 1. Guerrant RL, Van Gilder T, Steiner TS, Thielman NM, Slutsker L, Tauxe RV, et al. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001;32(3): 331-51. 2. Richards L, Claeson M, Pierce NF. Management of acute diarrhea in children: lessons learned. Pediatric Infect Dis J 1993;12(1):5-9. 3. World Health Organization . The Treatment of Diarrhea . A Manual for Physicians and Other Senior Health Workers. Geneva: Who, 2005; WHO/ CDD/SER/80.2. 4. Levine MM. Antimicrobial therapy for infectious diarrhea. Rev Infect Dis 1986; 8 (suppl 2):S207–16. 5. Osatakul S, Tangadullart C. The appropriate use of empirical antibiotic in children with acute diarrhea in Songklanagarind Hospital. Songkla Med J 1999; 17:25–30. 6. Howteerakul N, Higginbotham N, Dibley MJ . Antimicrobial use in children under five years with diarrhea in a central region province, Thailand. Southeast Asian J Trop Med Public Health 2004; 35:181–7. 7. Beria JU, Damiani MF, dos Santos IS, Lombardi C. Physicians’ prescribing behaviour for diarrhoea in children:an

ethnoepidemiological study in Southern Brazil. Soc Sci Med 1998; 47:341–6. 8. Karras DJ, Ong S, Moran GJ, et al. Antibiotic use for emergency department patients with acute diarrhea: prescibing practices, patient expectations and patient satisfaction. Ann Emerg Med 2003; 42:835–42. 9. Bojalil R, Calva JJ. Antibiotic misuse in diarrhea. A household survey in a Mexican community. J Clin Epidemiol 1994; 47:147–56. 10. Nizami SQ, Khan IA, Bhutta ZA. Drug prescribing practices of general practitioners and paediatricians for childhood diarrhoea in Karachi, Pakistan. Soc Sci Med 1996; 42:1133–9. 11. Singh J , Bora D , Sachdeva V, Sharma RS , Verghese T. Prescribing pattern by doctors for acute diarrhoea in children in Delhi, India. J Diarrhoeal Dis Res 1995; 13:229–31. 12. Gani L, Arif H, Widjaja SK, et al. Physicians’ prescribing practice for treatment of acute diarrhoea in young children in Jakarta. J Diarrhoeal Dis Res 1991; 9:194–9. 13. Osatakul S, Puetpaiboon A. Appropriate use of empirical antibiotics in acute diarrhoea: a cross-sectional survey in southern Thailand. Ann Trop Paediatr 2007;27(2):115-22. 14. Traa BS, Walker CL, Munos M, Black RE. Antibiotics for the treatment of dysentery in children. Int J Epidemiol;39 Suppl 1:i70-4.

33

15. Abba K, Sinfield R, Hart CA, Garner P. Antimicrobial drugs for persistent diarrhoea of unknown or non-specific cause in children under six in low and middle income countries: systematic review of randomized controlled trials. BMC Infect Dis 2009;9:24. 16. Aikins M, Armah G, Akazili J, Hodgson A. Hospital health care cost of diarrheal disease in Nor thern Ghana. J Infect Dis;202 Suppl:S126-30. 17. Duggan C,Santosham M,Glass RI. The management of acute diarrhea in children : oral rehydration, maintenance, and nutritional therapy. MMWR 1992;41(No.RR-16); and World Health Organization. The treatment of diarrhea : a manual fo physicians and other senior health workers. Geneva, Switzerland : World Health Organization, 1995. 18. Sur D, Bhattacharya SK. Acute diarrhoeal diseases--an approach to management. J Indian Med Assoc 2006;104(5):220-3. 19. World Health Organization . Diarrhoea: Why children are still dying and what can be done . The United Nations Children’s Fund (UNICEF)/Wor ld Health Or ganization (WHO), 2009


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Effects of High Dose Vitamin C Supplements on Urinary Oxalate Excretion: A Systematic Review Attakorn Raksasataya 1 Jariya Laiad1 Namthip Imwatthanakul1 Nararat Jantaraboot1 Parntip Wongput1 Thammasorn Piriyasupong2, M.D., Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen, Thailand 2Medical Education Center, Khon Kaen Hospital, Khon Kaen, Thailand

ABSTRACT BACKGROUND There are many studies focusing on the effect of vitamin C supplements on oxalate excretion in calcium-stone forming patients. However, its effect on urinary oxalate excretion is still unknown. METHODS This is a systematic review and meta-analysis of randomized controlled trials investigating the effect of high dose vitamin C supplement on urinary oxalate excretion that published until December 2010 from four electronic databases (PubMed, Scopus, Ovid and ScienceDirect) and reference lists of the trial reports. All included studies were reviewed systematically and extracted. Mean and standard deviation (and 95% confidence intervals) were calculated based on the amount of urinary oxalate excretion between those with vitamin C supplements 2 g/day and no supplement in population of both stone formers and non-stone formers. RESULTS Three trials investigating the effect of high dose vitamin C on urinary oxalate excretion were identified with a total of 120 participants (70 stone formers and 50 non stone formers). Each of them participated in two 6-days experimental periods (ascorbic acid supplement and no supplement). There was statistically significantly increase of urinary oxalate in stone former (standard mean difference 0.78; 95% confidence interval 0.44 to 1.13) but not statistically significant for the non stone former. CONCLUSION Increase of urine oxalate excretion were observed especially in those with stone former. Thus, we do not recommend the use of high dose vitamin C supplement in the stone former and should consider cautiously for non stone formers.

34


Journal of Medical Research and Education

JMRE

Vitamin C as known as ascorbic acid or ascorbate, belongs to the water-soluble class of vitamin.1 The vitamin is easily oxidized to form dehydroascorbic acid (DHAA), and thus oxidation is readily reversible.2 Vitamin C is a generic name for all compounds that exhibits the same biologic activity as ascorbic acid (AA). Consequently, the term includes both AA and DHAA described the process by which vitamin C is eliminated by the kidney via filtration and active tubular re-absorption, and the metabolic conversion of vitamin C to oxalate was demonstrated in the 1960s (Figure 1).2-4 The use of vitamin C was very popular in the 1970s as a means to decrease the incidence and severity of the common cold.5 More recently vitamin C has been shown to be effective for protecting against oxidative tissue damage, particularly with regard to the aging process. 6 The current recommended upper limit for ascorbic acid (AA) is 2,000 mg/day.7 A small percentage (1.5%) of ingested AA is converted in vivo to oxalate, which is excreted without further metabolism quantitatively in the urine over 24 hours. AA supplementation is common in the United States; 12.4% of the U.S. adult population and 12 to 14% of stone formers reported taking 500 mg of ascorbic acid daily.8,-11 Most of urinary stones diagnosed consist predominantly of calcium oxalate12 that originates from a combination of absorbed dietary oxalate and endogenous synthesis from such oxalate precursors as ascorbic acid and glyoxylate.13 There are many studies focusing on the effect of vitamin C supplementation on oxalate excretion including randomized controlled trials, cohort and case-control studies with some reports of significant findings demonstrated the relationship of vitamin and urinary oxalate excretion with no consensus on their findings. Thus, we systematically searched the literatures for randomized controlled trials for the

best evidences of effect of vitamin C on oxalate excretion and performed a meta-analysis.

METHODS Study design This is a systematic review included randomized controlled trials investigating the effect of high dose vitamin C on urinary oxalate excretion. Criteria for eligibility studies PARTICIPANTS Studies investigating the effect of high dose vitamin C on urinary oxalate excretion in both stone formers and non-stone formers, in participants aged 18 years old or over, both male and female, studies in participants with other underlying diseases were included. INTERVENTION Participants of the included studies were prescribed with either vitamin C supplement 2 g/day (intervention group) or non-vitamin C supplement (control group). OUTCOMES 24 to hour urinary oxalate (mg) was measured in both groups of participants. Exclusion criteria We screened the titles and abstracts and identified and excluded all papers not meeting any of the prespecified criteria by consensus. After that we evaluated the remaining studies as full papers. Studies were excluded if they did not prescribed vitamin C orally, used vitamin C derivatives, measured other

Ascorbic acid

Dihydroascorbic acid

2,3-diketogulonic acid

L-threose

Oxalic acid

Figure 1. Metabolic Conversion of Ascorbic Acid to Oxalate

35


JMRE

Journal of Medical Research and Education

Table1. Combination of Keywords Used for Search Strategy Database

Terms used for search

PubMed

"Ascorbic acid"[Mesh] AND "Stone"[Mesh]

Number of studies 114

"Ascorbic acid"[Mesh] AND "Urinary calculi"[Mesh]

80

Scopus

ALL (Ascorbic acid OR Vitamin C) AND  ALL(Stone OR Calculi)

1861

Ovid

(Ascorbic acid OR Vitamin C) AND (Stone OR Calculi)

1532

ScienceDirect

Ascorbic acid" OR “vitamin C” AND “stone” OR “ Urinary calculi” in All Fields

193

Total

3666

outcomes rather than specified or treated with multiple interventions.

28 articles not meeting our inclusion criteria; seven articles used different routes of intervention, an article used another type of intervention, five articles used different outcome measurement, 15 articles had multiple interventions, then we had 17 articles with full text. Only 11 articles were randomized control trial and we further excluded duplicated studies of the four databases. At the end, three trials were included in the analysis

Search strategies We followed the reports of randomized controlled trials through four databases including PubMed, Scopus, Ovid and ScienceDirect (from inception to December 2010), combining key search terms using the Medical Subject Headings (MeSH) strategy where appropriate or similar. For instance, searching with the MeSH term “ascorbic acid” via Mesh databases, the term L-Ascorbic Acid, Sodium Ascorbate, Magnesium Ascorbicum, Magnorbin, Ferrous Ascorbate, Hybrin were also included in the searching. Another combined search term was “Calculi” with the same search term strategies (Table 1). We limited our search to studies in human being and English only. We also searched the reference lists of the identified articles.

Study characteristics The characteristics of the included randomized controlled trials are summarized in the Table 3. The total number of participant in the trials was 120. Each participated in two 6-days experimental periods (ascorbic acid supplement and no supplement). The study by Weiwen Chai et al trial was a randomized crossover design investigating the effect of high dose vitamin C supplement among 48 participants of calcium oxalate stone formers and non-stone formers. They found significantly greater urinary oxalate level during the 24 hours, compared stone formers with non stone formers of both with and without supplement of vitamin C. In stone formers, significant higher urine oxalate was found in those with vitamin C supplement. In the non stone formers, the level of urine oxalate was higher after vitamin C supplement with no significance. The study by Linda K. Massey et al was a randomized crossover controlled design to investigate ascorbate increase human oxaluria and urinary stone risk. All 49 participants were 29 stone formers, 19 non stone formers and one non stone former was dropped from the analysis. In stone formers, there was significantly higher urinary oxalate excretion in vitamin C supplement and not significantly higher in non stone formers. The study by Oliver Traxer et al trial was a double blind, randomized crossover study of effect of ascorbic acid consumption on urinary stone risk factors among 12 stone formers and 12 non stone formers. No difference in urinary oxalate excretion was detected for each phase in stone formers and non stone formers. However, urinary oxalate was statistically significantly higher in the ascorbic acid

Quality assessment and data extraction The included studies were assessed for their quality using Jadad score for randomized controlled trials. After that data of relevant studies were abstracted according to the population of stone formers and non-stone formers who received vitamin C supplement 2 g/day and non-vitamin C supplement. Data synthesis Data were analyzed using the Review Manager 5 software. For continuous data as urinary oxalate level, the analysis was reported as mean and standard deviation compared between ascorbic acid supplement and no supplement in stone formers and non stone formers. Heterogeneity of the combined studies was examined using chi2 and I2. The metaanalysis was performed using random effect model.

RESULTS We searched 3,666 studies for four databases (Figure 2). After title and abstract screening, there were 67 articles left. We excluded 14 more articles of letters,

36


Journal of Medical Research and Education

JMRE

Table 2. Quality Assessment using Jadad Score Weiwen Oliver Linda K. Chai Traxer et Massey et al. al. et al. 1 1 1

Jadad score Was the study described as randomized? The method of randomization was described in the paper, and that method was appropriate.

0

0

0

Was the study described as double blind?

0

1

0

The method of blinding was described, and it was appropriate.

0

0

0

Was there a description of withdrawals and dropouts?

1

1

1

Total

2

3

2

phase compared with the placebo phase. In stone formers, there was significantly higher urinary oxalate excretion in vitamin C supplement and not significantly higher in non stone formers. In the meta-analysis, three clinical trials provided information about ascorbic acid in urinary oxalate total. A total of 70 stone formers and 50 non stone formers. We analyzed data and reported as mean, standard deviation and 95% confidence interval (CI) comparing between ascorbic acid supplement and no supplement in stone formers and non stone formers. In stone formers, we found significantly greater urinary oxalate level during the 24 hour in vitamin C supplement group. In non stone formers, urinary oxalate is higher in vitamin C supplement group, but not significant. In total, 24 hour urinary oxalate in vitamin C supplement group is significantly higher than non supplement group. The overall quality of the study were, however, fairly low (less than two for the Jadad score (Table 2.) We assessed statistical heterogeneity of the combined studies using I2 and chi-square test. In the present reviews, relatively absolute homogeneities were observed in non stone formers (I²=0%; Chi²=1.04; P=0.59), in stone formers (I²=0%; Chi²=0.29; P=0.86) and for total participants (I²=0%; Chi²=4.06; P=0.54). However, high heterogeneity of the combined subgroup was found relatively high(I²=63.3%; Chi²=2.72; P=0.10) (Figure 3).

supersaturation in urine.14 In addition to this, up to one-third of patients with calcium stone former may have increased absorption of dietary oxalate, and in some cases a deficiency of oxalate degradation by the bacterium Oxalobacter formigenes in the gut could be the culprit.15 However, our study showed that no statistically significant increase in urinary oxalate with the high dose supplement of vitamin C in non stone formers. In our review, heterogeneities were found to be very low in the subgroups (stone formers and non stone formers). This suggested the consistency of the effect of vitamin C on urine oxalate across the combined studies as well as the similarity of the included participants, inclusion and exclusion criteria, study design, intervention and collected data were appropriate for combining the results of all studies. In the other hand, after combined the total effect of vitamin C on urine oxalate of both stone former and non stone former, the heterogeneity was found to be relatively high because baseline of urinary oxalate in stone formers is higher than non stone formers. Our study has some limitations; the differences of characteristic such as age. We excluded participants who were younger 18 years old thus we might not generalize the findings to children. Moreover, the synthesis of the outcomes was based on a small amount of participants and this might not represent larger population. In search strategies we selected solely English articles, therefore we missed articles that might be able to be included in our study. However, regarding our search, other language articles in this area were very few. For quality assessment, all studies were scored two points that under cut-offs for high quality (Jadad score 3 or more), and this might affect the trustworthiness of our results. Still, outcome measures for the excretion of urinary oxalate were various; Oliver et al study, they used chromatography technique while the others used Isotope C2 and only Oliver et al study, oxalate load was not described for their participants. The urine collection techniques for prevention of ascorbic acid metabolize to oxalate were different. In Linda et al study used hydrochloric acid (HCl) 100 ml of 3 mmol/ L and collected on fifth day at the adaptation phase

DISCUSSION This meta-analysis of randomized controlled trial, to our knowledge, is one of the first investigating the effect of vitamin C on excretion of urine oxalate. It shows the positive effect of vitamin C supplement on urinary oxalate excretion. In our study, urinary oxalate excretion was statistically significant increase in overall participants especially in stone formers. This was supported by the fact that vitamin C is a precursor of oxalate. Consumption of high dose lead to hyperoxaluria, a major causative factor in the formation of renal calcium oxalate stones by oxalate

37


JMRE

Journal of Medical Research and Education

Table 3. Characteristics of Three Randomized Controlled Trials of Vitamin C Supplement on Urinary Oxalate Total Age Study design No. (years) Sex Health status Calcium stone Weiwen Randomized, forming Chai et al, cross over study 48 >= 18 Mixed participants, non 2004 stone former Oliver Double -blind, Calcium stone Traxer randomized, 24 18-70 Mixed forming patient, et al, 2005 cross over study non stone former Linda K. Calcium stone Massey, Randomized, 49 >=18 Mixed forming patient, cross over study 2005 non stone former Trial

Scopus Articles found (n=1, 861)

load Duration Vitamin Oxalate th C dose (the 6 day) (days) Complete-ness Outcomes 2 gm/d

118 mg

6

100%

2 gm/d

No

6

100%

2 gm/d

136 mg

6

97.9%

Medline Articles found (n=80)

Ovid Articles (n=1,532)

24 hours Urinary oxalate 24 hours Urinary oxalate 24 hours Urinary oxalate

Science Direct Articles found (n=193)

Articles Excluded (n=3,599) On basic of title and abstract (n=3,599)

Records screened (n=67)

Articles Excluded (n=42) Review, editorials, case report, letter (n=14) Route of intervention (n=7) Type of intervention (n=1) Difference outcome measurement (n=5) Multiple intervention (n=15)

Not full text (n=8)

Full articles obtained (n=17)

Randomized controlled trial (n=11)

Duplicate articles included (n=3)

Figure 2. Flow Diagram of Identifying and Included Trials

38


Journal of Medical Research and Education

JMRE

and used HCl 20 ml of 3 mmol/L at metabolic unit phase, In Oliver et al study used undescribed type and concentration of acid for urine collection. In Weiwen et al study, they used 3N of HCl 20 ml. The variation of acid used and collecting time may effect urine oxalate level. For the interpretation, we would like to remind that the included studies were only six days. Nonetheless, vitamin C possibly increase the excretion of urinary oxalate if we study for a longer period and might eventually cause a urinary calculi. We used to believe that, high dose vitamin C was found to decrease the incidence and severity of the common cold but in Douglas et al systematic review concluded that high dose vitamin C has no benefit for preventing common cold.16 Thus, the benefit of high dose of vitamin C for preventing common cold is not envisaged. Furthermore, our data suggest high dose vitamin C supplement on oral route is associated with

increase urinary oxalate especially in stone former. Thus high dose vitamin C supplement should be considered cautiously in this group. Although 80% of urinary stones formed by adults in the U.S. are calcium oxalate stones.17 But cause of urinary oxalate stone is multi-factorial such as low fluid intake, sodium, protein, Oxalate-rich foods and vitamin D.18 Therefore urinary oxalate stone is not dependent on urinary oxalate alone. In conclusion, our systematic review shows statistically significantly increase of urinary oxalate after the use of vitamin C especially in those stone formers. Our study has very low heterogeneities of the findings caused by the method of selecting article, inclusion and exclusion criteria were appropriate for combining the results of all studies. However, interpretation of the results has to be done carefully as discuss above.

Figure 3. Comparison between Three Studies in Non Stone Former and Stone Former REFERENCES 1. Groff, J.L., Gropper S.S., and Hunt S.M. The Water Soluble Vitamins. In: Advanced Nutrition and Human Metabolism. Minneapolis: West Publishing Company, 1995, p. 222-237. 2. Jacob, R.A., Vitamin C. In: Modern Nutrition in Health and Disease. Ninth Edition. Edited by Maurice Shils, James Olson, Moshe Shike, and A. Catharine Ross. Baltimore: Williams & Wilkins, 1999, p. 467-482. 3. Baker EM, Saari JC, Tolbert BM. Ascorbic acid metabolism in man. Am J Clin Nutr 1966; 19:371-8. 4. Kagawa Y. Enzymatic studies on ascorbic acid catabolism in animals. 1. Catabolism of 2,3 -diketo-frgulonic acid. J Biochem (Tokyo) 1962; 51:134-44 5. Pauling, L.: The significance of the evidence about ascorbic acid and the common cold. Proc Natl Acad Sci U S A, 68: 2678, 1971 6. Bode, A. M.: Metabolism of vitamin C in health and disease. Adv Pharmacol, 38: 21, 1997 7. Food and Nutrition Board, Institute of Medicine (2000) Vitamin C. In: Reference

Dietary Intakes for Vitamin C, Vitamin E, Selenium and Carotenoids. National Academies Press, Washington, DC. 8. Urivetzky, M., Kessaris, D. & Smith, A. D. (1992) Ascorbic acid overdosing: a risk factor for calcium oxalate nephrolithiasis. J. Urol. 147: 1215–1218. 9. Radimer K., Bindewald B., Hughes J., Ervin B., Swanson C., Picciano M. F. Dietary supplement use by US adults: data from the National Health and Nutrition Examination Survey, 1999–2000. Am. J. Epidemiol. 2004;160:339-349. 10. Wandzilak T. R., D’Andre S. D., Davis P. A., Williams H. E. Effect of high dose vitamin C on urinary oxalate levels. J. Urol. 1994;151:834-837. 11. Chai W., Liebman M., Kynast-Gales S., Massey L. Oxalate absorption and endogenous oxalate synthesis from ascorbate in calcium oxalate stone formers and non-stone formers. Am. J. Kidney Dis. 2004;44:1060-1069. 12. Asper R, Bigler M, Georgi U, Schmucki O (1987) In vitro-Oxalatbildung im Urin. Fortschr Urol Nephrol 25: 152

39

13. Williams HE,Wandzilak TR: Oxalate synthesis, transport and the hyperoxaluric syndromes. J Urol 141:742-747, 1989 14. Urivertzky M, Kessaris D, Smith AD. Ascorbic acid overdosing: a risk factor for calcium oxalate nephrolithiasis. J Urol 1992; 147: 1215 8 15. Holmes RP, Assimos DG. The impact of dietary oxalate on kidney stone formation. Urol Res. 2004;32(5):311–6 16. Douglas RM, Chalker EB, Treacy B. National Centre for Epidemiology and Population Health, The Australian National University, Canberra, ACT, Australia, 0200. 17. United States Department of Agriculture, Human Nutrition Information Ser vice, Agriculture Handbook Number 8-11 18. Trinchieri A, Mandressi A, Luongo P, et al. The influence of diet on urinary risk factors for stones in healthy subjects and idiopathic renal calcium stone former s. Br J Urol. 1991;67:230-236.


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Effectiveness and Safety of Influenza A (H1NI) Vaccine: A Systematic Review Chalisa Nithuthorn1 Kanokkarn Sawangsrisutigul1 Panthipa Huangprakhon1 Piyathida Klawkla1 Thammasorn Piriyasupong2, M.D., Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen, Thailand 2Medical Education Center, Khon Kaen Hospital, Khon Kaen, Thailand

ABSTRACT BACKGROUND Various randomized controlled trials stated the potential benefit of H1N1 vaccine. However, it efficacy have not been proved by a systematic review. METHODS We conducted a systematic review of the randomized controlled trail investigating the efficacy of all formulation of vaccine against H1N1 infection in healthy population through four databases. Eligible studies were identified, reviewed and extracted. The outcomes were immunogenicity response and its adverse events. RESULTS All formulations of H1N1 vaccine were effective in relation to rising of the immunogenicity compared with placebo. The significantly highest immunogenicity response age were between 12 and 17 years. Our studies demonstrated immunogenicity of the unadjuvant vaccine was raised greater than that of adjuvant vaccine with no significant difference. Moreover its effectiveness was not different between vaccine administration one and two doses. At the dose of 30 µg unadjuvant vaccine, it has shown the highest immunogenicity response in all age group. At the dose of 7.5 µg of unadjuvant vaccine formulation was found to have the least local and systemic reactions whereas the 30 μg of adjuvant vaccine made the worst local reactions and 30 μg of unadjuvant vaccine were the worst for systemic reactions. Comparison between first and second dose administration, it found that first dose had systemic reactions more than second dose significantly. CONCLUSION The vaccine against H1N1 infection of all formulation was effective to increase the immunogenicity of the healthy population at all age group with minimal side effect.

40


Journal of Medical Research and Education

JMRE

A pandemic influenza A 2009 (H1N1) is an emergency and rapid global spreading and can be life threatening that World Health Organization (WHO) prompted to declare a pandemic on June 11, 2009.1 Its worldwide cumulative confirmed cases were 1,483,520 and 25,174 were dead during May 9, 2009  to January 30, 2011.2 Currently used trivalent seasonal-influenza vaccines are unlikely to provide protection against this new virus.3 Hence, the new vaccines for controlling this pandemic at large-scale immunization were developed.1 There are many studies regarding the usage of vaccine against H1N1 influenza viral infection. Nonetheless, the recommendation dosages are still varied as well as their formulation (split-virion vs. whole virion) and adjuvants. For instance, Liang study used whole virion vaccine in 303 and split-virion in 10,917 Chinese people. It found the latter form of vaccine was superior in term of hemagglutinin inhibition titer.4 Later studies to evaluate the efficacy of the vaccine were conducted but they had fewer number of participants.5,6 However, the effectiveness of the vaccine has not been proved by a systematic review. Thus, we systematized the available studies for the best evidence of efficacy of H1N1vaccine and its safety.

within seven days, pregnancy, age group that lesser than three years, history of influenza infection were excluded. INTERVENTION Intervention are all types of vaccine that developed to prevent a pandemic influenza A 2009 (H1N1) whether they are single or multi-dose administration, with adjuvant or unadjuvant, monovalent or polyvalent, any concentration of vaccines, types of vaccines (splitvirus or whole virus) and control is placebo. OUTCOME MEASURES Primary outcome measure was number of people who developed titer more than 1:40 or four-fold rising. At this level, it claims that the vaccine can be protective against the viral infection.4,6 Secondary outcome were adverse events of both local and systemic reactions occurring anytime after vaccination. Search strategies We identify the relevant studies through four databases including PubMed, Scopus, Ovid and Science directs. The search strategies and terms are detailed in (Table 1).

Criteria for inclusion of studies STUDY DESIGN We included only randomized control trials

Data extraction Two authors separately extracted data from the included trials. The extraction included study names, types of design, dates and durations, study population sizes, participants analysis in the different treatment groups, age, sex, inclusion and exclusion criteria, intervention (type of vaccine and dosage), treatment outcome and adverse events both local (e.g., pain, induration, redness, swelling, exanthema and itching) and systemic (e.g., fever, allergy, headache, fatigue, nausea, myalgia and cough).

PARTICIPANTS The healthy participants aged between 3 and 60 years of both genders with injection of any formulation of H1N1 vaccine or placebo. Studies with participants with history of seasonal influenza vaccine injection

Quality assessment All included studies were assessed using Jadad score for quality assessment. In brief, it assessed whether the study described the methods of randomization, blinding and withdrawal or drop out of the

METHODS Study design We systematized and meta-analyzed clinical trials to identify the efficacy of H1N1 vaccine

Table 1. The Search Strategies and Terms Ised. Combination of keywords used for search strategy Database PubMed Scopus Ovid SciDirects

Terms used for search ("Vaccines"[Mesh] AND "Influenza, Human"[Mesh]) AND "Influenza A Virus, H1N1 Subtype"[Mesh] seroprotection (ALL(pandemic influenza) AND ALL(vaccine) AND ALL(seroprotection) AND ALL(h1n1)) (ovid full text available and yr="2009 -Current") - Search terms used: Vaccine and H1N1 and seroprotection H1N1 vaccine influenza A seroprotection prevention

Total

Number of studies 25 48 335

141 549

41


JMRE

Journal of Medical Research and Education

Screening

Records identified through four database searching (n = 549)

Excluded on title and abstract (n = 563)

Records screened (n = 12)

Full-text articles assessed for eligibility (n = 11) Excluded on content (n = 5)

Included

Research design not met (n = 2) Outcome measures didnâ&#x20AC;&#x2122;t meet the criteria of outcome measure for this review (n=3)

Eligibility

Studies include in qualitative synthesis (n = 3)

Studies include in quantitative synthesis (meta-analysis) and quality assessment (Jadad score â&#x2030;Ľ3) RCTs (n=3)

Figure 1. Flow Diagram of Selection Process of the Included Studies

42


Journal of Medical Research and Education

JMRE

Table.2 Quality Assessment using Jadad Score for Randomized controlled trial Kawsar R. Jadad-Score calculation

Talaat et al. 1

Was the study described as randomized?

Feng-Cia Jiang Wu

Michael E. Greenber

Zhu et al.

et al.

1

1

g et al 1

The method of randomization was described in the paper, and that method was appropriate.

0

1

1

1

Was the study described as double blind?

1

1

1

1

The method of blinding was described, and it was appropriate.

1

1

1

1

Was there a description of withdrawals and dropouts?

1

0

1

1

Total

4

4

5

5

participants and whether the methods were appropriate.7 It assumed a high quality study should have a Jadad score of three or more

seroconversion (four-fold raising) rate and rate of adverse effect either local or systemic reactions. The third study included 1 313 participants, were stratified by age to two age group (young,18 to 64 years and elderly adults,≥65 years). Participants were randomized 1:4:4:4 to receive two dose of placebo or 7.5, 15 or 30 μg of non adjuvant vaccine, respectively. Its outcomes were seroprotection and seroconversion rate. The last study had 240 participants, 120 participants received 15 μg unadjuvant, 120 participants received 30 μg unadjuvant without any control group (placebo). We analyzed data and reported as risk ratio and 95% confidence interval (CI) comparing between as the followings:

Data synthesis We carried out statistical analysis using the Review Manager software (RevMan 2008) by Mantel-Haenszel meta-analysis for combining data where trials were examining the same intervention, and the trials’ populations and methods were judged sufficiently similar. Intervention effects were estimated with risk ratios and 95% confidence intervals for dichotomous data. The heterogeneity was measured using I2 statistic (I2>50% suggests significant heterogeneity), and Q test. We used a random effects model for metaanalysis.

Immunogenicity; vaccine vs. placebo Vaccine found to be superior to placebo at 21 days after first and second dose of immunization in relation to immunogenicity measured by seroprotection and seroconversion of the three combined studies with RR 6.58 (95% CI, 2.98 to 14.57) for the first dose and one study with RR 8.96 (95% CI, 7.47 to 10.75) for the second dose (Figure 2 and Figure 3).

RESULTS We searching through four online databases retrieved 549 citations. After reviewing the titles and abstracts, identified and retrieved 11 references in full text for review information. We excluded seven studies due to their designs; three case-control studies, two cohorts, one cross-sectional survey, and one was not fulfilled in this study as it studied in non-healthy participants. (Figure 1) summarizes the study selection process. Finally, four studies were included, their characteristics were described in Table 3. Their qualities were assessed by Jadad score for randomized controlled trial (RCT) where studies with the cut point of three or more were assumed as high quality study (Table 2). The first study had 12,691 participants; 11,220 were randomly assigned to receive eight vaccine formulation, 1,471 participant received placebo (Table 3). Their outcomes were number of participants who developing titer 1:40 or more and number of participants had adverse effect. The second study had 2,200 participants, 880 participants were randomly assigned to receive 15 μg or 30 μg of unadjuvant vaccine and 110 participants in 18 to 60 years age group received placebo. Its outcomes were seroprotection (defined by HI titer 1:40 or more),

Immunogenicity; first dose vs. second dose This present review found to be no significantly different between the immunogenicity of first and second dose vaccine administration in all ages and all vaccine formulation by three studies demonstrated with RR 2.40 (95% CI, 0.55 to 10.42) as Figure 4. Response between vaccine formulation; adjuvant VS unadjuvant Our present review showed unadjuvant vaccine immunogenicity was greater than adjuvant vaccine but not significantly different by RR 1.09 (95% CI, 0.89 to 1.33) (Figure 5). In the unadjuvant vaccine, the comparison of immunogenicity between 7.5 µg,15 µg and 30 µg, it showed that 30 µg formulation was better than 7.5 µg in term of seroprotection and seroconversion rate RR 1.02 (95%CI, 1.01 to 1.03) while comparing between 7.5µg and 15µg not significant with RR 0.99 (95%CI, 0.98 to 1.01).

43


44

Result

Outcome measurement

Study design Intervention

Participant

Feng-Cai Zhu. 2009

Xiao-Feng Liang. 2009

Number of subjects who has seroprotection and seroconversion

Michael E. Greenberg. 2009

Feng-Cai Zhu. 2009

Randomize control trials Experimental group: adjuvant vaccine with 7.5, 15, 30 µg of hemagglutinin or nonadjuvant vaccine with 15 µg or 30 µg of hemagglutinin as a first dose ,single dose administration or twice dose of schedule Control group: placebo

split-virion formulation, 5 μg and10 μg adjuvant whole-virion formulation Control group: placebo Xiao-Feng Liang. 2009

systemic reaction at day 35 % of subject in HI titer ≥ 1:40,four –fold raising at day 0, 21,35 % of subject in injection site reaction at day 21 % of subject in systemic reaction at day 21, % of subject in injection site reaction at day 35, % of subject in systemic reaction at day 35

Number of people with local % of subjects in seroprotection reactions, Number of people with % of subjects in seroconversion systemic reaction, Number of people with HI titer ≥ 1:40

HI titer ≥ 1:40, seroconversion systemic adverse reaction, local adverse reaction, Observed at baseline, after first dose injection, second dose injection

Michael E. Greenberg. 2009

Randomize control trials Randomize control trials Experimental group: 7·5 μg, 15 μg Experimental group:15 ug and 30ug of hemaglutinin antigen administration and 30 μg non-adjuvant splitreceived two dose of vaccines virion formulation , 7·5 μg, 15 μg and 30 μg adjuvant No control group (placebo)

2200 subjects that healthy, no history of 12691 healthy participants aged 240 healthy, non pregnant participants aged infection and hadn’t received H1N1 vaccine between 3 to ≥ 60 years and 18 to 64 years hadn’t confirmed or between the ages 3 years to 77 years hadn’t received an H1N1 vaccine suspected 2009 H1N1 infection and hadn’t received an experimental influenza vaccine during 6 months.

HI titer ≥ 1:40(seroprotection), four fold HI titer ≥ 1:40,four –fold raising at day 0, 21, HI titer ≥ 1:40 at day 0, 21, 42 rising (seroconversion) 35 Seroprotection rate Observed at baseline before vaccination, Injection site reaction at day 21, Local reaction After vaccination first dose 21 days. systemic reaction at day 21, Systemic reaction, Injection site reaction at day 35,

Kawsar. 2010

1313 healthy non pregnant , hadn’t has a history of an influenza-like illness and hadn’t received seasonal influenza and H1N1 vaccine, Aged range between 18 to more than 75 years Randomize control trials Experimental group: 7.5, 15, 30 µg of H1N1 hemagglutinin administered Control group: placebo

Tables 3. Summaries of Included Studies Kawsar. 2010

JMRE Journal of Medical Research and Education


Journal of Medical Research and Education

JMRE

The comparison between adjuvant vaccines contained 7.5 µg, 15 µg and 30 µg found that there were no significant increase of immunogenicity rate with RR 0.96 (95% CI, 0.93 to 1.00), 1.03 (95% CI, 1.00 to 1.05) and 1.07 (95% CI, 1.00 to 1.14) for the comparison of 7.5 μg vs. 15 μg, 30 μg vs. 15 μg and 7.5 μg vs. 30 μg., respectively (Figure 6).

RR 1.26 (95% CI, 1.10 to 1.44), RR 1.06 (95% CI, 1.04 to 1.07), RR 1.11 (95% CI, 1.00 to 1.23), RR 1.11 (95% CI, 1.08 to 1.15) for the comparison of 3 to 11 years vs. 12 to 17 years,12 to 17 vs. 18 to 60 years, 12 to 17 vs. older than 60 years and 18 to 60 years vs. older than 60 years, respctively (Figure 7).

Immunogenicity response and age group In relation to immunogenicity response among four age groups (3 to 11, 12 to 17,18 to 60 and >60 years), it found that participants aged between 12 and 17 years were the group with the highest response,

Adverse event of all vaccine formulations were found more often than the placebo in all age groups significantly RR 1.69 [95% CI, 1.18 to 2.41]. The adverse events are difference in each age groups such as in 3 to 11 years has RR 2.14 (95% CI, 1.33 to 3.46),

Adverse reaction; vaccine vs. placebo

Figure 2. Comparison between Vaccine and Placebo in All Age Groups and Dose Formulations, 21 Days after First dose.

Figure 3. Comparison the Immunogenicity Response between Vaccine and Placebo in all Age Groups and Dose Formulations, 21 days After Second Dose.

Figure 4. Comparison the Immunogenicity Response between First and Second Dose Vaccine in All Age Groups and All Formulations.

45


JMRE

Journal of Medical Research and Educatio1

in 12 to 17 years has RR 2.16 (95% CI, 1.37 to 3.43), in 18 to 60 years has RR 1.18 (95% CI, 1.03 to 1.35), in ≥60 years has RR 2.66 (95% CI, 1.29 to 5.50) (Figure 8).

vs. 7.5 µg adjuvant, RR 1.78 [95% CI, 1.42 to 2.24] in 30 µg vs. 15 µg adjuvant (Figure 17). SYSTEMATIC REACTION The unadjuvant 7.5 µg vaccine had least systemic reaction RR 0.59 (95% CI, 0.48 to 0.72) in 7.5 µg vs. 15 µg unadjuvant, RR 0.47 (95% CI, 0.38 to 0.57) in 7.5 µg vs. 30 µg unadjuvant, RR 0.74 (95% CI, 0.57 to 0.96) in 7.5 µg unadjuvant vs. 7.5 µg adjuvant, RR 0.73 (95% CI, 0.56 to 0.95) in 7.5 µg unadjuvant vs. 15 µg adjuvant, RR 0.67 [95% CI, 0.49 to 0.93] in 7.5 µg unadjuvant vs. 30 µg adjuvant (Figure 10). While 30 µg unadjuvant vaccine has much systemic adverse effect more than others RR 1.26 (95% CI, 1.16 to 1.37). Our present review also showed first dose administrations making more systemic adverse reactions than second dose administration notably showed that RR 1.59 (95%CI, 1.25 to 2.04). (Figure 11)

LOCAL REACTION The estimating of local adverse reaction in 7.5 µg, 15 µg and 30 µg with or without adjuvant seemed to be 7.5 µg unadjuvant vaccine had least local adverse effects RR 0.02 (95% CI, 0.00 to 0.13) in 7.5 µg vs. 15 µg unadjuvant, RR 0.01 (95% CI, 0.00 to 0.09) in 7.5 µg vs. 30 µg unadjuvant, RR 0.01 (95% CI, 0.00 to 0.04) in 7.5 µg unadjuvant vs. 7.5 µg adjuvant, RR 0.01 (95% CI, 0.00 to 0.05) in 7.5 µg unadjuvant vs. 15 µg adjuvant, RR 0.00 (95% CI, 0.00 to 0.03) in 7.5 µg unadjuvant vs. 30 µg adjuvant (Figure 9). While 30 µg adjuvant vaccine has much local adverse effect more than others RR 1.66 [95% CI, 1.32 to 2.07] in 30 µg

Between unadjuvant and adjuvant vaccine dose vaccine in all age groups.

Between 7.5µg unadjuvant and 30µg adjuvant vaccine in all age groups.

Between 7.5µg unadjuvant and 15µg unadjuvant vaccine in all age groups.

Figure 5. Comparison of Immunogenicity Response in All Type of Vaccine and Age Group

46


Journal of Medical Research and Education

JMRE

Between 7.5µg adjuvant and 15µg adjuvant vaccine in all age groups.

Between 15µg adjuvant and 30µg adjuvant vaccine in all age groups.

Between 30µg adjuvant and 7.5µg adjuvant vaccine in all age groups.

Figure 6. Comparison the immunogenicity Response in Adjuvant Subgroup

dose of 30 µg unadjuvant, it had shown the highest immunogenicity response in all age group. The 7.5 µg of unadjuvant vaccine formulation was found to have the least local and systemic reactions whereas the 30 μg of adjuvant vaccine made the worst local reactions and 30 μg of unadjuvant vaccine were the worst for systemic reactions. Comparison between first and second dose administration, It found that first dose had systemic reactions more than second dose significantly.

DISCUSSION Principal findings Our present review collected and analyzed from four randomized control trials aimed to determine effectiveness and safety in healthy participants receiving H1N1 vaccine. It found to be H1N1 vaccine all formulation significantly raising immunogenicity response when comparing with placebo in all ages, thus, vaccine administration was superior to placebo in term of seroprotection and seroconversion rate notably. The significantly highest immunogenicity response ages were between 12 and 17 years. Our studies demonstrated immunogenicity of the unadjuvant vaccine was raised greater than that of adjuvant vaccine with no significant difference. Moreover its effectiveness was not different between vaccine administration one and two doses. At the

Strengths and limitations To our knowledge, this is the first systematic review regarding the effectiveness of vaccine against H1N1 infection. We reviewed and summarized the evidences from four databases. However, unpublished studies from conference or small studies might not be able to be retrieved and this can cause publication bias. From the included studies in the current review, for

47


JMRE

Journal of Medical Research and Education

Between 3-11 years and 12-17 years in all formulations.

Between 12-17 years and 18-60 years in all formulations.

Between 12-17 years and older than 60 years in all formulations.

Between 18-60 years and older than 60 years in all formulations.

Figure 7. Comparison the immunogenicity response Subgroup by Age

48


Journal of Medical Research and Education

JMRE

instance, in Kawsar et al. study we were unable to access to individual data for the more complete data of the par ticipants regarding second dose administration of the vaccine, as we used only the secondary data, thus, we had to exclude the study and were unable to meta-analyzed the findings with other studies for more accurate estimation of the effect of the second dose administration of the vaccine although the attempt to contact the author had been made.5 In our review, high variety of age groups were included as well as formulation and doses of the vaccine, all of these can cause high heterogeneity of

which seemed to be greater than 50% measured by I2. Thus, subgroup interpretation regarding the age, dose and formulation is more appropriate. As our review used only surrogate outcome to measure the efficacy of the vaccine, thus, long term follow up for the real event in relation to incident of H1N1 infection is required. Comparison with other studies There are three studies mentioned to effectiveness of H1N1 vaccine measured using another method-real time reverse transcription polymerase chain reaction

Figure 8. All Adverse Event Comparison between All Vaccine Formulations and Placebo in Each Age Group

49


JMRE

Journal of Medical Research and Education Between 7.5 µg unadjuvant vaccine formulation and others

Between 30 µg and 7.5 µg adjuvant, 30 µg and 15 µg adjuvant, respectively

Figure 9. Comparison of Local Reactions

50


Journal of Medical Research and Education

JMRE

Between 7.5 µg unadjuvant vaccine fotmulation and others

Between 30 µg unadjuvant and 15 µg unadjuvant

Figure 10. Comparison of Systemic Reactions

51


JMRE

Journal of Medical Research and Education

Figure 11. Comparison of Systemic Reactions between First and Second dosage Administration

(rRT-PCR). All three studies confirmed that the vaccine were effective to prevent the H1N1 infection in healthy population.8-10 However, this measure is still a surrogate outcome. In our review, 12 to17 year of age was the best age range with highest immunogenicity response to the vaccine. This similar finding was found in one Chinese study using the PCR technique to measure the outcome. At the age of 12 to17 years old, it found that number of negative PCR is the highest in this group.10 In this Chinese study, the rate of complication was also similar to our findings (12-17%) for both local and systemic reaction. A study from Canada also stated that the failure rate was high in older adult.11 However, in one Korean study, the rate response rate of the vaccine measured by PCR was found to be similar in all age group.9 Still, this study suffered from the recall bias as they used phone survey as a method to determine whether the participants received intervention and the participants were both healthy and non healthy. In our review, immunogenicity response in adjuvant and unadjuvant vaccine was found to be no significant difference. However, a study from the UK revealed that adjuvant vaccine can produce higher HI titer more than that of unadjuvant vaccine.12 However, this study was done in only 176 healthy participants. Its results, thus, may not be precise compared to our findings. In the current review, second dose of the vaccine should be given in the people age 3-11 years old. The findings was confirmed by a phase II study from the USA which they suggested to have a booster dose for the healthy young people age nine years old

or younger.13 In the present review, at the dose of 30 ug, the vaccine can trigger the highest immune response and side effects. This suggests that the vaccine is a dose related for both efficacy and adverse effects. However, these findings were contrary with the previous study where 7.5 ug of unadjuvant vaccine was the best dose for inducing HI titer. 14 Nonetheless, the participants in this study were age 18 years old or older with fewer sample compared to our findings where the outcome of approximately 9, 000 were meta-analyzed regardless age group. Conclusion and implication From our review, we recommend to administration the 30 Îźg either adjuvant or unadjuvant vaccine as it generates higher immunogenicity compared to other formulation. However, this has to be trade off with its higher adverse event both local and systemic. The vaccine seems to be effective for those healthy at all age group. As this review used only the surrogate outcome of immunogenicity response at the 21 and 42 days, longer follow up randomized controlled trial with large sample to identify the real event of H1N1 infection should be conducted to establish the efficacy of the vaccine. As first dose administration is adequate in term of immunogenicity, we recommend only single dose administration was appropriate. Moreover, single dose of vaccine had the least local and systemic adverse effects. In conclusion, the vaccine against H1N1 infection of all formulation was effective to increase the immunogenicity of the healthy at all age group with minimal side effect.

52


Journal of Medical Research and Education

JMRE

REFERENCES 1. New influenza A (H1N1) virus: global epidemiological situation, June 2009. Wkly Epidemiol Rec 2009;84(25):249-57. 2. pandemic. wsotHNiA. 3. Swine influenza A (H1N1)infection in two children--Southern California, March-April 2009. MMWR Morb Mor tal Wkly Rep 2009;58(15):400-2. 4. Liang XF, Wang HQ, Wang JZ, Fang HH, Wu J, Zhu FC, et al. Safety and immunogenicity of 2009 pandemic influenza A H1N1 vaccines in China: a multicentre, double-blind, randomised, placebo-controlled trial. Lancet;375(9708): 56-66. 5. Kawsar R. Talaat MEG, Michael H. Lai. A Single Dose of Unadjuvanted Novel 2009 H1N1Vaccine Is Immunogenic and Well Toleratedin Young and Elderly Adults. A Single Dose of Unadjuvanted Novel 2009 H1N1 Vaccine Is Immunogenic and Well Tolerated in Young and Elderly Adults 2010;202(9):1327â&#x20AC;&#x201C; 1337.

6. Zhu FC, Wang H, Fang HH, Yang JG, Lin XJ, Liang XF, et al. A novel influenza A (H1N1) vaccine in various age groups. N Engl J Med 2009;361(25):2414-23. 7. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17(1):1-12. 8. Hardelid P, Fleming D, McMenamin J, Andrews N, Robertson C, Sebastianpillai P, et al. Effectiveness of pandemic and seasonal influenza vaccine in preventing pandemic influenza A(H1N1)2009 infection in England and Scotland 2009-2010. Euro Surveill;16(2). 9. Song JY, Cheong HJ, Heo JY, Noh JY, Choi WS, Park DW, et al. Effectiveness of the pandemic influenza A/H1N1 2009 monovalent vaccine in Korea. Vaccine;29(7):1395-8. 10. Wu J, Xu F, Lu L, Lu M, Miao L, Gao T, et al. Safety and effectiveness of a 2009 H1N1 vaccine in Beijing. N Engl J Med;363(25): 2416-23.

53

11. Skowronski DM, Janjua NZ, De Serres G, Hottes TS, Dickinson JA, Crowcroft N, et al. Effectiveness of AS03 adjuvanted pandemic H1N1 vaccine: case-control evaluation based on sentinel surveillance system in Canada, autumn 2009. Bmj;342:c7297. 12. Clark TW, Pareek M, Hoschler K, Dillon H, Nicholson KG, Groth N, et al. Trial of 2009 influenza A (H1N1) monovalent MF59adjuvanted vaccine. N Engl J Med 2009;361(25):2424-35. 13. Plennevaux E, Sheldon E, Blatter M, Reeves-Hoche MK, Denis M. Immune response after a single vaccination against 2009 influenza A H1N1 in USA: a preliminary report of two randomised controlled phase 2 trials. Lancet; 375(9708):41-8. 14. Talaat KR, Greenberg ME, Lai MH, Hartel GF, Wichems CH, Rockman S, et al. A single dose of unadjuvanted novel 2009 H1N1 vaccine is immunogenic and well tolerated in young and elderly adults. J Infect Dis;202(9): 1327-37.


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Weight Gain during Pregnancy and Risk of Cesarian Section Sarawut Sriburin1 Teeraporn Suttayamook1 Thanut Boonyaleepun1 Yanin Sripanuskul1 Thammasorn Piriyasupong 2, M.D, Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen 2Medical Education Center, Khon Kaen Hospital, Khon Kaen

ABSTRACT BACKGROUND The rate of caesarean section had increased globally. There were many studies about the relationship between rates of caesarean delivery and weight gain during pregnancy. However, most of studies were conducted in Europe and North America. Consequently, it might not be explained that relationship in Asian population. METHODS Retrospective cohort study was conducted to explain this association. The samples were 1,430 women carrying singleton pregnancy who delivered a live-born infant at Khon Kaen Hospital Between June, 1st and September, 4th 2010. Women who had multiple pregnancies, previous history of caesarean section, congenital fetal anomaly and intrauterine fetal dead were all excluded. RESULTS The principle outcome occurred in 1,039 patients. We found that mothers with weight gain during pregnancy excess the Institute of Medicine (IOM) recommendation had significantly higher rate of caesarean section than those whom gained weight within the IOM recommendation (relative risk (RR), 1.26; 95% confidence interval (CI), 1.07 to 1.48). When we examined their relationship adjusted by maternal age and body mass index (BMI). We found that the risk of caesarean delivery was significantly increased in the overweight mothers, age 20-35, (RR, 1.99; 95% CI, 1.11 to 3.58) CONCLUSION Excessive weight gain during pregnancy had influences on caesarean section rate, but it was not the only predictor of caesarean section. Maternal age and BMI were also the important confounders in this relationship. Thus, excess weight gain during pregnancy alone was weakly associated with higher rate of caesarean delivery.

54


Journal of Medical Research and Education

JMRE

The global rate of caesarian section has been steadily rising, especially in high income countries.1 The CDC’s National Center for Health Statistics revealed that 4,317,119 babies born in 2007 (almost all birth re c o rd s re p o r t e d by t h e u n i t e d s t a t s o f America), 1,372,844 (31.8 percents) were delivered via cesarean section.2 The caesarian section (Csection) incurs high cost and morbidity in health care system in most countries, and can lead to unpredictable complications e.g., post operative adhesion, intestinal obstruction and more than 1 liter of intraoperative hemorrhage. 3,5 Although the complications are found occasionally, it is suggested that the caesarian section should not be operated in non-medically necessary cases due to the increase in many of the risks and cost.6 It has been long known that caesarian section is associated with cephalopelvic disproportion (15.5%), previous classical caesarean section (14.7%), fetal distress (9.6%), eclampsia (7.2%), unsuccessful trial of forceps or vacuum (5.5%), placenta previa totalis (5.1%), some types of abnormal presentation (4.7%), fetal macrosomia (4.3%) active genital herpes, multiple pregnancy (4.2%).7,10 In the recent systematic review and previous studies, maternal body mass index (BMI) and the weight gain during pregnancy was found to be predictor for higher risk of caesarian section in both elective and emergency conditions, especially in obese and overweight women. However, their findings were inconsistent and relied on only fair to moderate strength of evidences from solely European countries and North America Continent.11,14 Also there is no strong evidence that support their association.14 Thus, in this study, we aimed to examine the association between gestational weight gain and cesarean delivery rate in the field of Asian population due to the difference between previous studies’ samples and Asian e.g., With the same body mass index (BMI) the Asian give the higher percentage of fat than the European which lead to higher risk of type 2 diabetes, hypertension, dyslipidemia and lower cut-off point of BMI, which make it cannot be applied to Asian population.15

deliveries without medical indications and dead fetus in utero were excluded.

METHODS

RESULTS

Data collection Variables regarding the mothers including age, parity, gestational age estimated either by ultrasonographic method or last menstrual period (LMP), height, pregravid and pre-delivery weight, smoking habit and alcohol consumption, as well as variables regarding the infants including sex, birth weight and APGAR score at 10 minutes after birth will be collected from medical records. Statistical analysis The sample size calculation was based on the assumption that the overall caesarean section rate would be at least 20% difference between the group with normal weight gain during pregnancy and those with weight gain excess the Institute of Medicine (IOM) recommendation. From the preliminary analysis of the collected sample, the rate of cesarean delivery was about 65% in those with weight gain excess the IOM recommendation while only 22% in those with normal weight gain. Thus, with 80% of power and 5% alpha error, the required sample would be at least 15 in both groups. However, the included up to 1,430 pregnant women for the best approximate of the results. Statistical Analysis Summary data will be expressed as median and the interquartile range (IQR). Comparisons the baseline characteristics of the mothers in four study groups were performed by Kruskal-Wallis test. We used Chisquare test or Fisher’s exact test to compare the two groups with respect to the primary outcome-the proportion of women who had caesarean delivery. The relative risk (RR) of caesarean delivery and 95% confidence intervals (CI) were calculated for the weight gain within the IOM recommendation group as compared with the weight gain excess the IOM recommendation group. To answer questions regarding proper gestational weight gain in Thai women, we used receiver operating characteristic (ROC) curve.

Study Design This is a retrospective cohort study.

There were 1,430 women included in the study, 203 were excluded as the mother had history of previous caesarean section, had multiple pregnancy, death fetus in utero (DFIU) or fetal anencephaly. Finally, there were 1,227 left for the analysis (Figure 1). Their baseline characteristics are shown in Table 1. They had a median age of 25 years old (IQR, 20.0 to 29.0). Most of them were nulliparous (61.1%). Their median gestation ages and BMI were about 39 weeks (IQR, 37.8 to 40.0) and 20.5 (IQR, 18.7 to 23.1) respectively.

Participants and study site This study was conducted in Khon Kaen Hospital. Participants were women carrying singleton pregnancies who delivered a live-born infant between June, 1stand September, 4th 2010. All pregnant women who had history of previous caesarian section, elective cases that were willing to have caesarian

55


JMRE

Journal of Medical Research and Education

1,430 Pregnant women underwent delivery. All from Khon Kaen Hospital. From June, 1st 2010 â&#x20AC;&#x201C; September, 4th 2010

203 Were excluded 177 Had history of previous caesarean section. 21 Had multiple pregnancy 3 Had dead fetus in utero 2 Fetal anencephaly

1227 Pregnant women were eligible

382 Underwent Caesarean section

845 Underwent Non caesarean

67 Had Incomplete record

121 Had Incomplete record

315 Were primary analyzed

724 Were primary analyzed

Figure1. Study Flow Chart The relationship between weight gain during pregnancy and risk of caesarean delivery was assessed. The study was analyzed in women carrying singleton pregnancy who underwent delivery at Khon Kaen Hospital. These excluded the women who had dead fetus in utero, a history of previous caesarean section, severe congenital fetal anomaly and incomplete data of pregravid weight.

56


Journal of Medical Research and Education

JMRE

Table1. Baseline Characteristics of the Patients. Underweight Normal Overweight Obese (BMI < 18.5) (BMI ≥ 18.5 - (BMI ≥ 23 - < (BMI ≥ 27) Variable (N=237) < 23.0) 27) (N=85) (N=532) (N=186) Age-years Median

23.0

25.0

27.0

27.0

19.0-26.0

20.0-29.0

23.0-31.0

22.0-30.5

Nulliparous

169 (71.3)

335 (63.0)

85 (45.7)

46 (54.1)

Multiparus

68 (28.7)

197 (37.0)

101 (54.3)

39 (45.9)

Less than 37 weeks.

48 (20.3)

87 (16.4)

27 (14.5)

5 (5.9)

37 to 42weeks.

184 (77.6)

441 (82.9)

155 (83.3)

80 (94.1)

5 (2.1)

4 (2.2)

4 (2.2)

0

Interquartile range Parity-no. (%)

Median Interquartile range

0.019

38.8

39.0

39.0

39.1

37.3-40.0

38.0-39.8

37.9-40.0

38.1-40.1

158

156.5

157.0

158.0

155.0-163.0

153.0-161.0

153.0-161.0

153.0-163.0

Height-centimeters Median Interquartile range Gestational weight gain regarding to IOM* recommendation-no. (%) Normal weight gain Excessive Weight gain

0.060

0.000 192 (81.0)

378 (71.2)

69 (37.3)

25 (29.4)

45 (19.0)

153 (28.8)

116 (62.7)

60 (70.6)

Smoking habit-no. (%) Never

0.418 207 (100.0)

524 (98.5)

185 (99.5)

83 (97.6)

Stopped before pregnancy

0

3 (0.6)

0

1 (1.2)

Still Smoking

0

5 (0.9)

1 (0.5)

1 (1.2)

233 (98.3)

522 (98.1)

185 (99.5)

85 (100.0)

Stopped before pregnancy

2 (0.8)

5 (0.9)

0

0

Still drinking

2 (0.8)

5 (0.9)

1 (0.5)

0

Male

126 (53.2)

261 (49.1)

108 (58.1)

40 (47.1)

Female

111 (46.8)

271 (50.9)

78 (41.9)

45 (52.9)

Alcohol drinking-no. (%) Never

0.739

Infant sex-no. (%)

0.144

Infant birth weight (g)-no. (%)

0.027

Very low birth weight (<1,500 g)

3 (1.3)

7 (1.3)

2 (1.1)

0

Low birth weight (1500-2,500 g)

23 (9.7)

63 (11.8)

17 (9.1)

4 (4.7)

Normal birth weight (2500-4000)

76 (89.4)

209 (88.2)

455 (85.5)

159 (85.5)

High birth weight (>4,000 g)

2 (0.8)

7 (1.3)

8 (4.3)

5 (5.9)

Median

2,950.0

3,030.0

3,190.0

3,260.0

Interquartile range

2,730.0-3,205.0 2,730.0-3,310.0 2,795.0-3,430.0 2,915.0-3,945.0

APGAR score-no. (%) 0

0.000

0.000

Gestational age-no. (%)

More than 42 weeks.

P Value

0.638 1 (0.4)

0

1-5

0

6-9

5 (21)

10

231 (97.5)

* Institute of Medicine14

57

1 (0.5)

0

3 (0.6)

0

1 (1.2)

7 (1.3)

1 (0.5)

1 (1.2)

532 (98.1)

184 (98.9)

83 (97.6)


JMRE

Journal of Medical Research and Education

Table 2. Primary Outcomes Outcome Caesarean section-no. (%)

Weight gain Weight gain within excess than IOM IOM recommendation recommendation (N=374) (N=664)

Relative risk (95% CI)

132 (35.3)

182 (24.4)

1.26 (1.07-1.48)

Very low birth weight (<1,500 g)

3 (0.8)

9 (1.4)

0.69 (0.26-1.85)

Low birth weight (1501-2,500 g)

28 (7.5)

79 (11.9)

0.70 (0.51-0.98)

332 (88.8)

565 (85.1)

0.48 (0.43-0.53)

11 (2.9)

11 (1.7)

1.40 (0.91-2.14)

0

1 (0.3)

1 (0.2)

1.39 (0.35-5.57)

1-5

1 (0.3)

3 (0.5)

0.69 (0.13-3.79)

6-9

3 (0.8)

11 (1.7)

0.59 (0.22-1.62)

10

369 (98.7)

649 (97.7)

1.45 (0.68-3.11)

50 (13.4)

117 (17.6)

0.80 (0.63-1.03)

5 (1.3)

8 (1.2)

1.07 (0.53-2.14)

Infant birth weight

Normal birth weight (2501-4000 g) High birth weight (>4,000 g) APGAR score

Preterm labor Post term

Table 3. Secondary Outcomes Mode of delivery Outcome

Relative risk (95% CI)

Caesarean delivery

Vaginal delivery

Underweight (BMI<18.5)

3 (37.5)

14 (18.9)

2.29 (0.61-8.66)

Normal (BMI≥18.5-22.9)

5 (20.8)

36 (31.1)

0.64 (0.25-1.59)

Overweight (BMI≥23-26.9)

9 (81.8)

17 (81.0)

1.04 (0.30-3.62)

Obese (BMI≥27)

6 (85.7)

5 (71.4)

1.64 (0.30-8.86)

Underweight (BMI<18.5)

11 (23.9)

17 (15.9)

1.40 (0.82-2.41)

Normal (BMI≥18.5-23.9)

28 (24.8)

70 (28.9)

0.86 (0.60-1.24)

Overweight (BMI ≥26.9)

33 (75.0)

47 (52.8)

1.99 (1.11-3.58)

Obese (BMI≥27)

25 (86.2)

21 (63.6)

2.17 (0.90-5.29)

Normal (BMI≥18.5-23.9)

6 (40.0)

8 (38.1)

1.05 (0.48-2.30)

Overweight (BMI ≥26.9)

5 (45.5)

5 (55.6)

0.83 (0.38-1.86)

Obese (BMI ≥ 27)

1 (20.0)

2 (50.0)

0.50 (0.09-2.73)

Weight gain excess than Institute of Medicine recommendation-no% Teenage (<20)

Normal (20-<35)

Elderly (>35)

58


Journal of Medical Research and Education

JMRE

!

A

B 1.0 1.0

0.8

Sensitivity

0.8

Sensitivity

0.6

0.6

0.4

0.4

0.2 0.2

0.0 0.0

0.2

0.4

0.6

0.8

0.0

1.0

0.0

1 - Specificity

C

0.2

0.4

0.6

0.8

1.0

!

1 - Specificity

D

! 1.0

1.0

0.8

Sensitivity

Sensitivity

0.8

0.6

0.6

0.4

0.4

0.2

0.2

0.0

0.0 0.0

0.2

0.4

0.6

0.8

0.0

1.0

0.2

0.4

0.6

1 - Specificity

1 - Specificity

0.8

1.0

!

Figure 2. Receiver Operating Characteristic (ROC) curves for Maternal Weight Gain During Pregnancy in Relation to caesarean delivery rate. Underweight (BMI<18.5) (Panel A). Normal weight (BMI = 18.5-22.9) (Panel B). Overweight (BMI 23-26.9) (Panel C). Obese (BMI≥27) (Panel D).

Almost all of them were non-smoker and did not drink alcohol. The infant sex tended to be male (51.4%). The median of infant birth weight was 3,030 grams (IQR, 2,730 to 3,330). Furthermore, nearly all of the APGAR score of the infants at 10 minutes after birth were 10 (98.1%). We categorized the mothers by their BMI into four groups. The mothers with high BMI tended to gain weight excess the IOM recommendation and likely to be older. The mothers with low BMI tended to have greater chance of preterm labor. Anyway, there were no statistical differences in maternal height and the infant’s APGAR score The optimal cut-off value for the weight gain during pregnancy in predicting delivery route was estimated using ROC curve, illustrated in (Figure 2). However, at any cut off point value, there seems to be no large difference between true positive (sensitivity) and false positive (1-specificity). Moreover, the overall areas under the ROC curve were approximately 0.6. Thus, using the weight gain during pregnancy alone is not likely to be able to predict the rout of delivery.

Neonatal and delivery outcomes are shown in (Table 2). Most of the newborn were normal weight and had APGAR score of ten in both groups. The infants of mother who had weight gain within IOM recommendation tended to have normal birth weight more than those of mothers with weight gain excess than IOM recommendation. Moreover, there were no differences between those with weight gain within and excess the IOM recommendation in relation to preterm and postterm of deliveries. Regarding the primary outcomes, the mother with weight gain during pregnancy excess the IOM recommendation had significantly higher rate caesarean section compared with those whom gained weight within IOM recommendation (RR, 1.26; 95% CI, 1.07 to 1.48) (Table 2). However, when we examined the relationship between weight gain and rate of caesarean delivery regarding BMI and maternal age, we found that only the over weight mothers aged between 20 and 35, the risk of caesarean delivery was significantly increased (RR, 1.99; 95% CI, 1.11 to 3.58) (Table 3).

59


JMRE

Journal of Medical Research and Education singleton pregnancies who delivered a live-born infant and categorized them into four groups of BMI and three groups of maternal age in order to focus on the effect of maternal weight gain on caesarean section rate without confounding by maternal age and BMI. However, our study is a retrospective cohort study, missing data cannot be recollected in some cases, consequently cannot be analyzed may lead to biased study results. Nonetheless, the attempts to verify and complete all missing data were made. Still, approximately 14% of the primary outcome (weight gain during pregnancy) was missing. From our study, the important problem is many missing data could not be recorded due to retrospective cohort study design. Subjective recorded data such as smoking habit and alcohol consumption might not be exactly accurate. For the confirmation of the relationship between weight gain and caesarean delivery as well as other maternal and neonatal outcomes, the prospectively cohort study with larger samples should be conducted. We thought that cross sectional study is a better option but it may take a long time. Although our study shows only weak relationship between weight gain during pregnancy and risk of caesarean section, it is undeniable that weight control still be important, especially in overweight woman with age from 20 to 35. In summary, excessive weight gain during pregnancy had influences on caesarean section rate, but it was not the only predictor of caesarean section. Maternal age and BMI were also the important confounders in this relationship. Thus, excess weight gain during pregnancy alone was weakly associated with higher rate of caesarean delivery.

DISCUSSION The rate of caesarean delivery was significantly associated with varies of maternal conditions including weight gain during pregnancy.1,7-15 According to many previous studies, despite they provided the association between maternal weight gain during pregnancy and caesarean section rate, there was no studies that are likely to be able to generalize to Thai population. Even more there is no standard guideline for Asian population. The IOM recommendation cannot be applied to use with the Asians due to differences of BMI.15 In this study we focused on maternal weight excess or within the IOM recommendation and used of Asianâ&#x20AC;&#x2122;s BMI instead of the World Health Organization (WHO) BMI. We found that the maternal weight gain had influence on caesarean section rate, but when we adjusted all the confounders (BMI and maternal age), surprisingly the outcomes suggested that there was weakly statistically significant in differences between the mother who had weight gain excess the IOM recommendation and those who gained within the recommendation. The results of this study suggested that pregnant women those who had weight gain during pregnancy excess the IOM recommendation were likely to have more risk of caesarean section, especially in overweight women. As in many studies, there were varies of factors that had influence on caesarean section rate.1,7-14 but there was no study focused exclusively on weight gain during pregnancy. In our analyses, we selected a group of women with REFERENCES 1. Bondok WM, El-Shehry SH, Fadllallah SM. Trend in cesarean section rate. Saudi Med J; 32(1):41-5. 2. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Mathews TJ, Kirmeyer S, et al. Births: final data for 2007. Natl Vital Stat Rep;58(24):1-85. 3. Andolf E, Thorsell M, Kallen K. Cesarean delivery and risk for postoperative adhesions and intestinal obstruction: a nested casecontrol study of the Swedish Medical Birth Registry. Am J Obstet Gynecol;203(4):406 e1-6. 4. Bergholt T, Stenderup JK, Vedsted-Jakobsen A, Helm P, Lenstrup C. Intraoperative surgical complication during cesarean section: an observational study of the incidence and risk factor s. Acta Obstet Gynecol Scand 2003;82(3):251-6. 5. Henderson J, McCandlish R, Kumiega L, Petrou S. Systematic review of economic aspects of alternative modes of delivery. Bjog 2001;108(2):149-57. 6. Souza JP, Gulmezoglu A, Lumbiganon P, Laopaiboon M, Carroli G, Fawole B, et al. Caesarean section without medical indications

is associated with an increased risk of adverse shor t-term maternal outcomes: the 2004-2008 WHO Global Survey on Maternal and Perinatal Health. BMC Med;8:71. 7. Chanthasenanont A, Pongrojpaw D, Nanthakomon T, Somprasit C, Kamudhamas A, Suwannarurk K. Indications for cesarean section at Thammasat University Hospital. J Med Assoc Thai 2007;90(9):1733-7. 8. Geidam AD, Audu BM, Kawuwa BM, Obed JY. Rising trend and indications of caesarean section at the university of Maiduguri teaching hospital, Nigeria. Ann Afr Med 2009;8(2): 127-32. 9. Khunpradit S, Patumanond J, Tawichasri C. Risk indicators for cesarean section due to cephalopelvic dispropor tion in Lamphun hospital. J Med Assoc Thai 2005;88 Suppl 2:S63-8. 10. Surapanthapisit P, Thitadilok W. Risk factor s of caesarean section due to cephalopelvic disproportion. J Med Assoc Thai 2006;89 Suppl 4:S105-11. 11. Chu SY, Kim SY, Schmid CH, Dietz PM, Callaghan WM, Lau J, et al. Maternal obesity

60

and risk of cesarean delivery: a meta-analysis. Obes Rev 2007;8(5):385-94. 12. Kominiarek MA, Vanveldhuisen P, Hibbard J, Landy H, Haberman S, Learman L, et al. The mater nal body mass index: a strong association with delivery route. Am J Obstet Gynecol;203(3):264 e1-7. 13. Poobalan AS, Aucott LS, Gurung T, Smith WC , Bhattachar ya S. Obesity as an independent risk factor for elective and emergency caesarean delivery in nulliparous women--systematic review and meta-analysis of cohort studies. Obes Rev 2009;10(1): 28-35. 14. Viswanathan M, Siega-Riz AM, Moos MK, Deierlein A, Mumford S, Knaack J, et al. Outcomes of maternal weight gain. Evid Rep Technol Assess (Full Rep) 2008(168):1-223. 15. Zheng W, McLerran DF, Rolland B, Zhang X, Inoue M, Matsuo K, et al. Association between body-mass index and risk of death in more than 1 million Asians. N Engl J Med; 364(8):719-29.


Journal of Medical Research and Education

JMRE

Photo by Viranuj Sueblinvong, M.D.

61


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Validity of Symphysis-Fundal Height Measurement for Estimating Infant Birth Weight Nichapat Thanthanet Phanuphong Kaeochaiya Walaipan Paopad Thammasorn Piriyasupong 2, M.D, Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen 2Medical Education Center, Khon Kaen Hospital, Khon Kaen

ABSTRACT BACKGROUND Symphysis-fundal height (SFH) is one of the methods to estimate infant birth weight usually perform. However, its validity is still varied especially in the Asian population. METHODS Diagnostic study was conducted in pregnant women who admitted in the labor room in Department of Obstetrics and Gynecology, Khon Kaen Hospital for delivery between September and November 2010. The multivariable analysis was used to construct the model estimating infant birth weight (IBW). RESULTS There were 535 pregnant women were included with the average age of 25 years old and median gestational age of 39 weeks. The mean IBW was 3,172 grams. The factors predicting the IBW were gestational age, pregravid and predelivery BMI, weight gain during pregnancy and number of parity (p-value < 0.01) aside from the SFH while fetal head engagement before admission was not found to be associated with the infant birth weight.Validities for predicting fetal weight using the various models were 65.9%, 66.1% and 69.0%. CONCLUSION Using SFH yield high validity for estimating IBW alongside with gestational age and body mass index.

62


Journal of Medical Research and Education

JMRE

Accurate estimation of infant birth weight (IBW) is important to determine proper management for those pregnant women at the initiation of labor.1,2 There are two common methods of estimation of IBW; ultrasonography and clinical examination that includes abdominal palpation and measurement of fundal height.1 Measure of symphysis-fundal height (SFH) is usually performed due its inexpensive cost and ultrasonography is used for more accurate result especially in those with abnormal size of fetus.3 SFH has been verified to be beneficial in approximation of IBW.1,2,4-18 However, its sensitivity and specificity can be various due to characteristics of nature of the population, ethnicities, health service systems and other factors.6,7 Moreover, studies revealed the validity of using SFH in Asian population are relatively scant. Most of them recommended that SFH can be used to approximate the fetal weight in term and non-complicated pregnancy.1,5,8,9 There were two previous studies conducted in Thailand and Iran using the SFH to estimate the IBW but their conclusions were relied on only fair to moderate strength of evidences.1,9 Thus, in present study, we aim to validate the SFH measurement and construct a valid equation for the precise approximation of fetal weight in the subgroup Thai population.

approximate of the validity of the SFH in estimating of IBW as in general, the sample size required for the diagnostic study should be at least one - two hundred for the test in place where the outcome is highly prevalent.19

METHODS

Statistical analysis All data were double entered and cleaned before preceding the analysis. All numeric data were test for their distribution using Kolmogorov-Smirnov test. Non-normal distributed data were expressed as median and interquartile range (IQR). We used multivariable analysis to assess the association among many variables and IBW and constructing the proper equation. Two-sided P values of less than 0.05 were considered to indicate statistical significance. Correlations between the two non- normally distributed numeric variables, were performed using the Spearman correlation coefficient. Finally, the validity of the formula in predicting IBW was presented in term of percentage of true positive estimation of the IBW.

Measurement tool SFH was measured in centimeters using non-elastic tape while subject lie in supine position on a firm surface with an empty bladder immediately on admission for intrapartum period. Study outcome The primary study outcome is actual IBW in gram which measured promptly after birth of the infants. Data collection Variables regarding the mothers including age, GA, parity, height, pre-gravid and pre-delivery weight, smoking habit and alcohol consumption, underlying disease, estimated fetal weight by ultrasonographic method and fetal head engagement examined by Leopoldâ&#x20AC;&#x2122;s maneuvers as well as variables regarding the infants including sex, mode of delivery, height, actual birth weight and Apgar score at 5 minutes were reviewed and collected from medical records onto a designed excel spreadsheet.

Study design This is a diagnostic study to ascertain the validity of SFH in approximation of IBW. Study site This study was conducted in Department of Obstetrics and Gynecology, Khon Kaen Hospital. Participants All pregnant women in all ages who admitted in the labor room for delivery between September and November 2010 were included. We excluded those with uncertain gestational age or gestational age (GA) below 28 weeks, without vertex presentation and intact membrane. Pregnant women who had history of untreated abdominal or gynecologic neoplasm and obstetrical complications in antepartum period including antepartum haemorrhage, heart disease, hypertensive diseases in pregnancy, nephrotic syndrome, cirrhosis and others affecting SFH measurement were also excluded as well as those with the diagnosis of molar pregnancy, dead fetus in utero, any fetal malformations, multiple pregnancy, o l i go hy d r a m n i o s a n d p o ly hy d r a m n i o s by ultrasonographic method. In the present study, the samples were included up to 500 cases for the best

RESULTS There were 1,315 pregnant women included in the present study, 780 were excluded as they were without intact membrane, and vertex presentation, hy p e r t e n s i ve d i s e a s e , m u l t i p l e p re g n a n c y, oligohydramnios or polyhydramnios, untreated gynecologic neoplasms and others. Moreover, 341 had incomplete records and were excluded as well (Figure

63


JMRE

Journal of Medical Research and Education

1,315 Thai pregnant women underwent delivery at Khon Kaen Hospital between September and November 2010

780 Were excluded 341 Were excluded due to incomplete record 255 Were excluded due to without intact membrane 72 Were excluded due to without vertex presentation 48 Were excluded due to hypertensive disease 17 Were excluded due to multiple pregnancy 16 Were excluded due to oligohydramnios or polyhydramnios 9 Were excluded due to birth before admission (BBA) 8 Were excluded due to untreated gynecologic neoplasm 14 Were excluded due to other Conditions*

535 Pregnant women were eligible

Underweight

Normal range

Overweight

133 Were primary analyzed

268 Were primary analyzed

134 Were primary analyzed

Figure 1. Patients Flow Chart All 535 pregnant women were formed into three groups of their predelivery BMI; underweight (BMI < 24.4%), normal range (BMI 24.4 â&#x20AC;&#x201C; 29.1%) and overweight (BMI >29.1%). * Other conditions including fetal malformations, dead fetus in utero, antepartum haemorrhage, abortion and maternal systematic lupus erythematosus.

64


Journal of Medical Research and Education

JMRE

1). Finally, there were 535 left for the analysis and thier baseline characteristics are shown in (Table 1). Their median maternal age was 25 years (IQR, 21 30) with the median gestational age of 39 weeks (IQR, 38 - 40). Most of them were primigravida (82.8%) with the average height of 157 centimeters (IQR, 153 - 160) with 51 kilograms of pre-gravid weight (IQR, 45 - 58) and 66 kilograms of pre-delivery weight (IQR, 60 - 73). The median pre-gravid and predelivery BMI are 20.58 (IQR, 18.83 – 23.34) and 26.64 Kg/m² (IQR, 18.83 – 23.34) respectively. It was unlikely to have the low birth weight infant in the group of mother with BMI in the fourth quartile regarding their weights before delivery. In contrary, it seemed that the large for gestation neonates were not from the mothers from the first quartile. Almost all of them were non- smoker and did not drink alcohol. About 32.9 % and 2.2 % of participants had thalassemia and diabetes mellitus respectively. The other underlying diseases included viral hepatitis, thyroid disease and HIV infection were found in less than 2% of the participants. Of these, 441 (77%) already had fetal head engagement at the time of admission. The median SFH is 32 centimeters (IQR, 30 - 33)They were a bit more male infant (50.3%) than female. The median of infant birth height was 53 centimeters (IQR, 52 - 54) and 3,171.5 grams (± 412.2) of mean IBW correspondingly. Mode of delivery, normal labor was 65.8% and 29.0% of cesarean section. Moreover, nearly all of the APGAR score of the infants at 5 minutes after birth were 10 (99.8%). In Table 2, most of the data of the IBW were from mother with gestational age between 37 and 41 weeks. SFH of the mother were ranged between 27 and 37centimeters. Only seven cases had SFH more beyond 37 centimeters and we considered as the outlier data. In Table 2, we stratified the IBW regarding SFH and gestational age as they were the two confounders found to be associated with IBW (Table 3). We also found that IBW from the mothers in the same body mass index (BMI) and SFH tended to increase as the gestational age increase. Still, the IBW from the mothers in the same BMI and gestational age, were also inclined to increase as the SFH increase and this confirmed the relationship between IBW and SFH as well as gestational age and BMI. The factors associated with the IBW from the multivariable analysis are shown in Table 3. It found that that GA, BMI, pre-gravid and pre-delivery BMI, weight gain, and parity were statistical significance for predicting IBW (P value < 0.01) aside from the SFH while fetal head engagement before admission was not found to be associated with the IBW. Various equations predicting the IBW is shown in Box 1. In the present study, our equation is made from data to estimate fetal weight of mother with gestational age between 37 and 41 weeks with SFH between 27 and 37 centimeters. Validity for predict fetal weight ± 300

gm is 65.9%, 66.1% and 69.0% for the equation 1, 2 and 3 respectively. This suggested that the three main determinants for prediction of IBW were GA, predelivery BMI and SFH and for the simplicity to use of the equation, the validity in term of true positive prediction of IBW given the ± 300 grams was calculated based on Equation 1 where only three variables were taken into consideration (Table 4). Moreover, we trimmed down the outlier data mentioned earlier for the more precision of the estimated results. Hence, Table 4 includes only validity of IBW from the mothers with gestational age between 37 and 41 with SFH between 27 and 37 centimeters. Comparing to previous studies from Thailand and Iran, the estimated IBW based on our equation dealt the highest validity in the various ranges (estimated IBW from the equation ± 100, ± 200 and ± 300 grams) amongst the three models; at the estimated birth weight ± 300 grams, the validity was found to be as high as 65.9% comparing to 5.2% and 61.9% from the previous two studies. However, at the gestational age of 42 weeks, it seemed that equation from the Iranian study had higher validity than our equation. Nonetheless, as mention earlier, we consider the data from this point were outliers.

DISCUSSION Our study is the first to our knowledge that estimates the validity in term of accurate prediction of fetal birth weight using the FHM which is commonly used in the primary care settings. However, accuracy can be various because there are differences in the population group due to characteristics of nature of the population. 6,7 Moreover, the estimation of IBW is also depended on several factors not solely the SFH. In the present study, we specifically focused on the Thai subgroup of Asian population and found that GA and BMI were strongly related to the IBW and yield more accurate estimation of IBW when GA and BMI were taken into account. The study had several strengths. Firstly, we conducted the study in nearly all groups of age, gestational age and weight of Thai pregnant women. Secondly, the equation in predicting IBW had the highest validity amongst the three models proposed from the previous studies as we put GA and BMI into the equation. For more accurate estimation of IBW, other variables included pregravid BMI, weight gain during pregnancy, number of parity should be considered. However, this has to be trade-off between the uncomplicated implementation and validity of the estimation. The main limitation of this study is that the SFH was measured by various health personnel. Therefore, it

65


20.3-29.0

2 (20) 25.5 20.5-34.5

35 to older years

Median

38

Median

66 0 0

Still Smoking

10 (100)

51.3-59.1

0

0

10 (100)

63.5-69.3

66.0

48.0-58.0

41.0-49.5 57.0

50.5

153.5-161.3

45.5

154.8-160.0

157.0

4 (40)

2 (20) 156.0

6 (60)

8 (80)

Stopped before pregnancy

Never

Smoking habit-no. (%)

IQR

Median

Predelivery weight

IQR

Median

Pregravid weight

Weight-kilogram

IQR

Median

Height-centimeter

Multiparous

Nulliparous

Parity-no. (%)

38.0-40.3

39

0

0

42 weeks to more than 37.0-39.0

10 (100)

9 (90)

37 to 41 weeks.

IQR

0

1 (10)

0

Less than 37 weeks.

Gestational age-no. (%)

IQR

20.3

6 (60)

8 (80)

2 (20)

20-34 years

2 (20) 82 (66.7)

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

123 (100)

52.3-59.9

57.0

42.0-48.0

45.0

153.0-162.0

157.0

56 (45.5)

67 (54.5)

38.0-40.0

39

2 (1.6)

119 (96.8)

2 (1.6)

19.0-27.0

24.0

4 (3.3)

37 (30.1)

< 24.4 (n=123)

0 0

1 (0.4)

122 (100) 2 (0.8)

250 (98.8)

79.0 73.9-85.0

65.4

56.0-71.6

62.0

62.0-70.0

47.0-55.0

51.0

157.0 153.0-160.0

157.0

72 (59.0)

50 (41.0)

38.0-40.0

39

2 (1.6)

117 (95.9)

3 (2.5)

23.0-31.0

28.0

16 (13.1)

91 (75.6)

15 (12.3)

> 29.1 (n=122)

153.0-160.0

134 (53.0)

119 (47.0)

38.0-40.0

39

7 (2.8) 240 (94.8) 6 (2.4)

21.0-30.0

25.0

20 (7.9)

192 (75.9)

41 (16.2)

24.4.-29.1 (n=253)

Maternal body mass index (Kg/m²)*

0 0

Infant birth weight 2,500- 4,000 grams

0

Maternal body mass index (Kg/ m²)* < 24.4 24.4-29.1 > 29.1 (n=10) (n=10) (n=0)

< 2,500 grams

Young than 20 years

Age -no. (%)

Variable

Table 1. Baseline Characteristics of the Participants

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

< 24.4 (n=0)

0

0

5 (100)

62.0-70.9

67.6

44.5-52.0

51.0

153.0-166.5

161.0

3 (60)

2 (40)

39.0-41.0

41

0

5 (100)

0

21.5-29.5

26.0

0

4 (80)

1 (20)

24.4.-29.1 (n=5)

0

0

12 (100)

74.7-89.1

83.5

58.5-78.5

62.5

155.0-164.5

158.0

8 (66.7)

4 (33.3)

39-40.8

39.5

1 (8.3)

11 (91.7)

0

22.5-32.0

28.5

1 (8.3)

10 (83.4)

1 (8.3)

> 29.1 (n=12)

Maternal body mass index (Kg/m²)*

> 4,000 grams

1 (0.2)

2 (0.4)

532 (99.4)

60.0-73.0

66.0

45.0-58.0

51.0

153.0-160.0

157.0

92 (17.2)

443 (82.8)

38.0-40.0

39.0

11 (2.1)

511 (95.5)

13 (2.4)

21.0-30.0

25.0

43 (8)

393 (73.5)

99 (18.5)

Total (n=535)

JMRE Journal of Medical Research and Education


1 (10)

Stopped before pregnancy

67 8 (80) 5 (50)

Male infant― no. (%)

1 (10) 0

Vacuum extraction

Forceps extraction

Infant birth weight 2,500- 4,000 grams

> 4,000 grams

10

7 - 10

Median

10

10 (100)

0

0

48.0-51.0

49.0

0

1 (1)

3 (3)

6 (6)

5 (50)

7 (70)

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

5 (1.98)

253 (100) 10

10

0

0 123 (100)

0

0

53.0 52.0-54.0

53.0

2 (0.8)

51.0-54.0

1 (0.8)

14 (3.3)

69 (27.3)

19 (15.4) 4 (3.3)

168 (66.4)

120 (47.4)

99 (80.5)

58 (47.2)

202 (79.8)

3 (2.4) 92 (74.8)

12 (4.7)

5 (2.0)

5 (2.0)

5 (2.0)

2 (0.8)

77 (30.4)

0

7 (2.8)

246 (97.2)

5 (4.1)

2 (1.6)

2 (1.6)

1 (0.8)

1 (0.8)

45 (36.6)

0

1 (0.8)

122 (99.2)

10

120 (99.2)

1 (0.8)

0

52.0-55.0

53.0

1 (0.8)

4 (3.3)

51 (41.8)

66 (54.1)

68 (55.7)

89 (73.0)

2 (1.6)

1 (0.8)

2 (1.6)

2 (1.6)

2 (1.6)

5 (4.1)

39 (32)

0

1 (0.8)

121 (99.2)

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

10

5 (100)

0

0

54.5-57.5

57

0

0

3 (60.0)

2 (40.0)

3 (60)

4 (80)

0

1 (8.3)

0

0

0

0

2 (40)

10

12 (100)

0

0

54.3-57.0

56

0

0

8 (66.7)

4 (33.3)

10 (83.3)

10 (83.3)

1 (8.3)

1 (5.9)

0

0

1 (8.3)

1 (8.3)

3 (25)

0

0

1 (20) 0

12 (100)

4 (80)

10

533 (99.8)

1 (0.2)

0

52.0-54.0

53.0

4 (0.7)

24 (4.5)

155 (29.0)

352 (65.8)

269 (50.3)

411 (77)

12 (2.2)

19 (3.6)

9 (1.7)

9 (1.7)

10 (1.9)

12 (2.2)

176 (32.9)

0

11 (2.1)

534 (97.9)

Total (n=535)

IQR 10-10 10-10 0 10-10 10-10 10-10 0 10-10 10-10 10-10 * Maternal body mass index (BMI) is defined as the individual's predelivery body weight divided by the square of her height. Group of BMI < 24.4 Kg/m² is pregnant women in the first quartile and Group of BMI > 29.1 Kg/m² is pregnant women in the forth quartile. † Other underlying diseases including epilepsy, asthma, migraine, limb amputation, allergic rhinitis, anemia and heart disease.

0 10 (100)

4-7

0

48.8-51.0

0-3

APGAR score at 5 minutes

IQR

Median

50.0

2 (20)

Cesarean section

Infant birth height

7 (70)

Normal labor

Mode of delivery― no. (%)

1 (10)

Fetal head engagement―no. (%)

0

EFW by ultrasonography-no. (%)

0

0

Thyroid disease

Others†

0

1 (10)

Viral hepatitis

HIV infection

0

2 (20)

1 (10)

6 (60)

Diabetes mellitus

4 (40)

0

0

10 (100)

Thalassemia

Underlying disease-no. (%)

0

9 (90)

Still drinking

< 2,500 grams Maternal body mass index (Kg/m²)* Maternal body mass index (Kg/m²)* Maternal body mass index (Kg/m²)* 24.4-29.1 > 29.1 < 24.4 24.4.-29.1 > 29.1 < 24.4 24.4.-29.1 > 29.1 < 24.4 (n=10) (n=10) (n=0) (n=123) (n=253) (n=122) (n=0) (n=5) (n=12)

Never

Alcohol drinking-no. (%)

Variable

Table 1. (Continued.)

Journal of Medical Research and Education JMRE


GA 36 weeks

68

-

-

-

-

-

-

-

41

42

43

44

45

46

47

-

-

-

-

-

-

-

-

-

-

-

-

-

-

2500

-

3290

2940

-

-

2540

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

2590

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

2660

-

-

1790

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

2960

-

-

2830

-

-

-

-

-

-

-

-

-

-

-

-

-

-

3350

-

-

-

-

-

2640

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

3250

3110

-

2820

2300

2740

2450

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

3290

3265

-

3195

-

-

3190

-

-

-

-

-

3310

-

2820

2570

3270

2745

3200

2880

2725

2930

UW NW OW

GA 37 weeks

-

-

-

-

-

-

-

-

-

-

-

-

-

3080

3330

3030

2810

2795

2890

2840

2390

-

-

-

-

-

-

-

-

-

-

-

-

3230

3210

3170

3210

2790

2875

2920

2835

2620

-

-

-

-

-

-

-

3650

-

4070

3475

3380

3650

3470

3480

3200

3710

3050

3250

3290

-

UW NW OW

GA 38 weeks

-

-

-

-

-

-

-

-

-

-

-

-

2820

3570

3270

3160

2958

2770

2900

2665

2280

-

-

-

-

-

-

-

-

-

-

-

3830

3580

3230

3155

3095

3260

3080

3020

2760

2220

4840

-

-

-

-

-

-

4190

-

4020

3740

3400

3400

3410

3150

3230

-

-

-

-

-

-

-

-

-

-

-

2790

-

-

3060

3070

3440

3050

3510 3000

2930

3000

2930

-

-

-

-

-

-

-

-

-

-

3190

-

3360

3470

3425

3020

3140

3245

-

2980

3100

-

-

-

-

-

-

4520

-

-

-

3820

3720

3460

3400

3205

3290

-

2835

3020

-

-

UW NW OW

GA 40 weeks

2800

-

-

UW NW OW

GA 39 weeks

Median IBW (grams)

-

-

-

-

-

-

-

-

-

-

-

-

3550

3405

3320

2890

2995

2930

-

2720

-

-

-

-

-

-

-

-

-

-

-

-

3850

3510

3560

3480

3450

3365

3350

-

2430

2430

-

-

-

-

-

4180

-

-

-

-

-

-

3605

3675

3475

3400

-

2880

-

-

-

UW NW OW

GA 41 weeks

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

3150

3170

3370

-

-

-

-

-

-

3325

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

3545

-

3530

-

-

-

-

-

UW NW OW

GA 42 weeks

SFH symphysis fundal height, GA gestational age,- was used where data are not available. * Underweight (UW) means maternal BMI < 24.4 Kg/m², Normal weight (NW) means maternal BMI 24.4 to 29.1 Kg/m² and Overweight (OW) means maternal BMI > 29.1 Kg/m²

-

40

-

35

-

-

34

39

-

33

-

-

32

38

-

31

-

3470

30

37

-

29

-

-

28

36

-

UW* NW* OW* UW NW OW

GA 35 weeks

27

SFH (cm.)

Table 2. Median IBW Stratified by SFHand Gestational Age.

JMRE Journal of Medical Research and Education


Journal of Medical Research and Education

JMRE

may cause measurement bias affecting validity of our study. However, those personnel were trained and qualified to do this simple task. Moreover, the model we used was exclusively based on pregnant women with gestational age 37-41 weeks and SFH 27-37 centimeters. Thus, generalization of the findings to other group of population might be limited. Nonetheless, the sample size was relatively small which inevitably limited the estimation of IBW. However, the number of sample was enough to ensure the validity of the proposed model.19 Generally, the characteristics of the mothers and infants in the present study were comparable to those in previous studies from various ethnic groups. It found that the maternal ages were about 25 years old and relatively similar to the previous studies.1,18 In term of parity of the mother, however, most of the

mothers were nulliparous (82.8%) while 50-55% in the previous studies.1,2,18 In the present study, women with underlying diseases such as diabetes, anemia, thalassemia were included whereas earlier studies excluded .7,9 Regardless different ethnicity, parity of the mother and underlying diseases, BMI and GA of the women in the present study seemed to be similar to the Brazilian study and French study respectively. 2,18 Moreover, the SFH in our study were in the same range of that from the Iranian study.1 In relation to the IBW, its mean was 3,172 grams in the present study and was also comparable to the other previous studies that had means of IBW ranged from 2,909- 3,212 grams.1,7,9 At this point, it might be able to conclude that ethnicity and parity were likely to have fairly small effect on arthopometric index of women as well as the IBW across the studies and

Table 3. Factors Associated with Infant Birth Weight from Multivariable Analysis. Factors

Coefficient

95% Confidence Interval

P value

Body mass index

3.7

6.3-20.6

< 0.001

Gestational age

5.4

38.1-81.5

< 0.001

Symphysis fundal height

13.3

71.8-96.7

< 0.001

IBW (g) = 13.47 (pre-delivery BMI) + 59.81 (GA) + 84.26 (SFH) – 2,211…………………………………………………………………….(Equation 1) IBW (g) = 13.62 ( pre-gravid BMI ) + 9.02 (WG ) + 59.78(GA) + 82.77 (SFH) – 2215.19 …………………………………………………...(Equation 2) IBW (g) = 135.06 ( pre-gravid BMI) – 122.49( pre-delivery BMI ) + 58.70 (WG ) + 62.41 (GA) + 50.95 (parity) +79.20 (SFH) – 2221.08....(Equation 3)

Box 1. Equation for Predicting Infant Birth Weight. This box presents the four equations predicting from the multivariable analysis, IBW= infant birth weight, BMI= body This box presents the four equations predicting from the multivariable analysis, IBW= infant birth weight, BMI= body mass index, GA= gestational age, WG= weight gain during pregnancy, parity= number of parity and SFH= symphysis-fundal height Table 4. The Validity of Infantile Birth Weight Given ± 300 grams. SFH (cm.)

Validity (%) GA 37 weeks

GA 38 weeks

GA 39 weeks

GA 40 weeks

GA 41 weeks

UW

NW

OW

UW

NW

OW

UW

NW

OW

UW

NW

OW

UW

NW

OW

27

100

0

-

100

66.7

-

0

50

-

100

100

-

-

0

-

28

100

50

-

100

75

0

66.7

100

-

0

100

-

100

0

-

29

0

100

-

100

60

33.3

77.8

71.4

0

100

-

0

-

-

-

30

100

33.3

50

75

50

66.7

77.8

100

0

83.3

33.3

50

60

66.7

50

31

-

50

-

75

66.7

0

85.7

60

42.9

50

90.9

-

100

75

-

32

100

75

-

100

46.2

66.7

60

60

33.3

40

61.5

80

20

66.7

85.7

33

100

0

100

50

77.8

66.7

100

54.5

57.1

57.1

57.1

75

100

57.1

100

34

-

0

-

100

75

100

100

87.5

57.1

-

75

60

100

100

66.7

35

-

-

50

-

100

71.4

0

100

50

-

66.7

100

100

75

0

36

-

100

100

-

-

50

-

100

40

0

-

50

-

100

-

37

-

-

-

-

-

100

-

-

100

-

0

100

-

-

-

- was used where data are not available.

69


JMRE

Journal of Medical Research and Education

Table 5. The Validity of the Three Formulas in Predicting Infantile Birth Weight with Symphysis-Fundal Height between 27 and 37 Centimeters. Deviation

Formulas in predicting IBW Regression model of birth weight on SFH, GA and BMI 27.4 %

(SFH - 11.5) x 115*

(SFH x 87) + 515†

0.8 %

22.4 %

35

16.7%

16.7%

16.7%

36

28.6%

14.3%

.0%

37

17.6%

5.9%

17.6%

38

27.3%

.0%

21.8%

39

30.3%

.6%

21.7%

40

27.5%

.8%

26.7%

41

24.6%

.0%

20.0%

± 100 grams Gestational age-weeks

42

36.4%

.0%

27.3%

47.2 %

2.6 %

43.3 %

35

33.3%

16.7%

33.3%

36

57.1%

28.6%

14.3%

37

41.2%

8.8%

38.2%

38

48.2%

.9%

39.1%

39

48.6%

2.9%

45.7%

40

45.0%

2.5%

45.0%

41

49.2%

1.5%

43.1%

65.9%

5.2%

61.9 %

35

33.3%

16.7%

33.3%

36

57.1%

28.6%

42.9%

37

58.8%

17.6%

47.1%

38

67.3%

5.5%

58.9%

39

65.1%

4.0%

63.4%

40

66.7%

4.2%

65.0%

41

67.7%

3.1%

64.6%

42 45.5% *The formula from previous study in Thailand. † The formula from previous study in Iran.

9.1%

54.5%

± 200 grams

± 300 grams Gestational age-weeks

might be irrelevant to the validity of the model estimating the IBW. In a study from Iran abdominal girth (abdominal circumference at the umbilicus) was also used for the estimation of IBW .1 However, the findings did not yield a better estimation in the small or normal weight infant and should be used only those with weight 4,000 grams or higher. As mention earlier, our model yield higher validity than that from this previous study.

In summary, high validity was found using the SFH predicting the IBW together with GA and BMI in the Thai pregnant women from the various model purposed and give the highest accuracy rate than those from previous studies. For more precise estimating model, study with larger sample should be conducted and for more insight in those with underlying disease such as mothers with diabetes, a further study of the application of the model in this specific subgroup should be examined

70


Journal of Medical Research and Education

JMRE

REFERENCES 1. F.Mor tazavi and A.Akaberi (2010). Estimation of birth weight by measurement of fundal height and abdominal girth in parturients at term. EMHJ 5, 553-558. 2. G.KAYEM, G.GRANGÉ, G.BRÉART and F.GOFFINET (2009) Comparison of fundal height measurement and sonographically measured fetal abdominal circumference in the prediction of high and low birth weight at term. Ultrasound Obstet Gynecol 2009; 34: 566–571. 3. Kate Morse, Amanda Williams, Jason Gardosi (2009) fetal growth screening by fundal height measurement. Best Practice & Research Clinical Obstetrics and Gynecology 23, 809-818. 4. Neilson JP.Symphysis-fundal height measurement in pregnancy. Cochrane Database Syst Rev. 2000; 2: CD000944. 5. Indraccolo U,  Chiocci L,  Rosenberg P,  Nappi L,  Greco P. Usefulness of symphysisfundal height in predicting fetal weight in healthy term pregnant women. Clin Exp Obstet Gynecol. 2008; 35(3):205-7. 6. Euans DW, Connor PD, Hahn RG, Rodney WM,  Arheart KL. A comparison of manual and ultrasound measurements of fundal height. J Fam Pract. 1995 Mar; 40(3):233-6.

7. Kenneth Challis, Nafissa Bique Osman, Lennarth Nyström, Gunnar Nordahl and Staffan Bergström (2002) Symphysis-fundal height growth chart of an obstetric cohort of 817 Mozambican women with ultrasounddated singleton pregnancies. Tropical Medicine and International Health 7, 678-684. 8. Jinda Pooprasat. Prediction of birth weight form mesurement of SFH and anthropometric measurement at term gestation at ramathibodi hospital.2003; p. 81 9. [Chanchai Maleepan. Easy to estimate fetal weight, Abstract published in conferences of the ministry of public health 2544 p. 67] Thai 10. Woo JS, Ngan HY, Au KK, Fung KP, Wong VC.Estimation of fetal weight in utero from symphysis-fundal height and abdominal girth measurements.Aust N Z J Obstet Gynaecol. 1985: 268-271. 11. Geirsson RT,  Agustsson P.Total intrauterine volume and symphysis fundus height. Is there a relation? Acta Obstet Gynecol Scand. 1987; 345-8. 12. Wikström I,  Bergström R,  Bakketeig L,  Jacobsen G,  Lindmark G.Prediction of high birthweight from maternal characteristics, symphysis fundal height and ultrasound biometry. Gynecol Obstet Invest. 1993; 27-33.

71

13. Walraven GE, Mkanje RJ,  van Roosmalen J,  van Dongen PW,  van Asten HA,  Dolmans WM.Single pre-deliver y symphysis-fundal height measurement as a predictor of birthweight and multiple pregnancy. Br J Obstet Gynaecol. 1995 ; 525-9. 14. Mohanty C, Das BK, Mishra OP.Parturient fundal height as a predictor of low birth weight. J Trop Pediatr. 1998 ; 222-4. 15. Bothner BK,  Gulmezoglu AM,  Hofmeyr GJ.Symphysis fundus height measurements during labour: a prospective, descriptive study. Afr J Reprod Health. 2000 Apr;4(1):48-55. 16. Kraiem J,  Chiha N,  Bouden S,  Ounaissa F, Falfoul A.[Clinical fetal weight estimation and prediction of macrosomia]. Tunis Med.  2004 Mar;82(3):271-5. 17. Buchmann E,  Tlale K. A simple clinical formula for predicting fetal weight in labour at term--derivation and validation. S Afr Med J. 2009 Jun;99(6):457-60. 18. P.H.C. Rondo, N.L. Maia Filho, K.K. Valverde. Symphysis–fundal height and size at birth. International Journal of Gynecology and Obstetrics 81 (2003); 53–54 19. Richard K. Riegelman.Studying a study & Testing a test.2005; 141-142


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

The Association Among Capillary Blood Glucose Levels, Sepsis, and Death in Patients with Melioidosis and Diabetes Mellitus: A Retrospective Cohort Jaturong Thongyaem1 Nathamon Homchampa1 Pattaraporn Insawang1 Sakunchai Ranok1 Thammasorn Piriyasupong 2 , M.D, Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen 2Medical Education Center, Khon Kaen Hospital, Khon Kaen

ABSTRACT BACKGROUND The present study aims to determine the association between capillary blood glucose levels in relation to sepsis and mortality in patients with melioidosis and diabetes mellitus. METHODS This is a retrospective cohort study including medical records of patients from Khon Kaen Hospital, Thailand that aged between 21 and 65 years and had melioidosis with diabetes mellitus during January 2009 through October 2011. Serial determinations of capillary blood glucose were collected starting at admission (random capillary blood glucose) and continuing at a six-hour interval (before each meal, and before sleep) for the first three days of admission. The mean capillary blood glucose levels were calculated and categorized into three groups as follows, less than 200 mg%, between 201 - 250 mg% and more than 251 mg%, afterwards the proportion of patients with sepsis, death and other complications were analyzed regarding these three groups. RESULTS There were 244 patients recruited. The levels of mean capillary blood glucose showed no statistical difference in relation to sepsis (adjusted odds ratio (AOR), 1.00; 95% confidence interval (CI) 1.00 to 1.01) or death (AOR, 1.00; 95% CI 1.00 to 1.01) in patients with melioidosis and diabetes mellitus. Similar associations were observed, and we found that patients who had upper gastrointestinal bleeding as a further complication had a significant increase in death rates (AOR, 12.55; 95% CI 2.16 to 72.95). Male sex was found to be the only significant risk factor associated with the higher proportion of sepsis (AOR, 0.98; 95% CI 0.94 to 1.02). We found that patients who suffered from other complications also had a significant increase in death rates (AOR, 0.51; 95% CI 0.14 to 1.80), the explanation for this event might be due to various severe complications such as cardiac arrest and respiratory failure. CONCLUSION Capillary blood glucose levels were not associated with the proportion of sepsis and mortality in patients with melioidosis and diabetes mellitus.

72


Journal of Medical Research and Education

JMRE

Melioidosis is an infectious disease caused by a saprophytic bacterium, Burkholderia pseudomallei. The organism is widely distributed in soil and water in tropical regions, such as Southeast Asia and Northern Australia.1,2 The disease is most frequently reported from the Northeast of Thailand as the most common cause of community acquired bacteremia.3 Infection in adulthood usually occurs in people with one or more predisposing factors and the strongest associated risk factor is diabetes mellitus.3,4 Not only do these factors predispose patients to melioidosis, but they also increase the severity of the disease.5 In the septicemic form, melioidosis is a serious and life threatening condition which requires early detection and specific treatment to avoid case-fatality. 6 Furthermore, septicemic melioidosis has been reported to be significantly associated with preexisting diabetes mellitus or renal failure.7 If glycemic control were to have an effect on the rate of sepsis and mortality, knowing the suitable range may improve the outcome of the disease, leading to increased survival rates. At present, no known research has been done to identify the exact association between blood glucose levels, the rate of sepsis, and mortality. Further studies should be attempted to find the optimal level of blood glucose that may significantly lower the rate of sepsis and mortality in melioidosis patients with diabetes mellitus. We therefore conducted a retrospective cohort study to determine the association.

melioidosis diagnosis such as negative culture tests or negative melioidosis titers, patients with history of readmission and were already included in the study, no capillary blood glucose records for the first 72 hours since admission, and patients who had missing computerized medical records from the hospital database were excluded from the study (Figure 1). On the assumption, to achieve a type one error of 5% for a two sided test, and a power of 80%, a sample of 102 patients per group would be required to detect the difference between three groups of mean capillary blood glucose. To assemble all required eligible records, we performed the study by searching the patients from the computerized Khon Kaen Hospital database using the code A241-A244 and E100-E149. These codes represented melioidosis with diabetes mellitus by ICD-10 classification. Later we identified patients that were diagnosed with melioidosis and diabetes mellitus, by the evidence of either positive culture for Burkholderia pseudomallei or positive melioidosis titer with a level of 1:160 or more. For hemoculture, we used BacT/ALERT® 3D system as a diagnostic machine for hemoculture tests. MacConkey agars, chocolate agars or blood agars were used with other specimens for culture test. Indirect hemagglutination tests were used to determine melioidosis titers. The solutions were produced by the Center for Immunodiagnostic Production (CIP) Immunological Laboratory Department of Pathology Faculty of Medicine Ramathibodi Hospital Mahidol University Thailand.

METHODS

Outcome measures The primary outcome was sepsis defined by the presentation of at least two of the following; fever (oral temperature > 38°c) or hypothermia (< 36°c), tachypnea (> 24 breaths/min), tachycardia (> 90 beats/ min), leukocytosis (> 12,000/µL), leukopenia (< 4,000/ µL), or > 10% band forms,2 from computerized medical records. The secondary outcomes were death and other complications including acute respiratory failure, acute renal failure, disseminated intravascular coagulopathy, cardiac arrest, septic shock and ventilator acquired pneumonia.

Study design Our study was a retrospective cohort conducted on computerized medical records of patients diagnosed with melioidosis and diabetes mellitus to verify the association between capillary blood glucose levels and outcomes of melioidosis treatment in relation to sepsis. Patients Medical records of patients were from Khon Kaen Hospital, Thailand. Eligible patients were aged between 21 and 65 years and had melioidosis with diabetes mellitus during January 2009 through October 2011. Patients were diagnosed with diabetes mellitus by demonstrating any one of the following: fasting blood glucose ≥ 126 mg/dl (7.0 mmol/l) or HbA1C ≥ 6.5% or two-hour plasma glucose ≥ 200 mg/dl (11.1 mmol/ l) during on OGTT or classic symptoms of hyperglycemia or hyperglycemic crisis, a random plasma glucose ≥ 200 mg/dl (11.1 mmol/dl) 8 or previously diagnosed with diabetes mellitus from computerized medical records. The patients were excluded by the following criteria: no evidence of

Data collection Data collection was also performed using a computerized database of the selected hospital. Demographic data, suspected risk factors related to melioidosis, sepsis, and death, such as occupation, type of diabetes, method of melioidosis diagnosis, patient’s diagnosis, previous diabetic control, previously prescribed antibiotics, treatment on admission, comorbidities and complications were recorded retrospectively. Diabetic control was classified into seven types consisting of the use glibenclamide, metformin, glipizide, mixtard insulin, regular insulin,

73


JMRE

Journal of Medical Research and Education comorbidities and complications were also obtained from each patient’s medical record. Laboratory results were confirmed and verified from the hospital’s computerized database (Laboratory Information(s) Version 1.00.00). Serial determinations of capillary blood glucose were collected starting at admission (random capillary blood glucose) and continuing at a six-hour interval (before each meal, and before sleep) for the first three days of admission. On Call® Advanced Blood Glucose Test Strips were used as capillary blood glucose detectors in all patients. The mean capillary blood glucose level was calculated in each patient and was categorized into three groups as follows, less than 200 mg%, between 201 - 250 mg% and more than 251 mg%, afterwards the proportion of patients with sepsis, death and other complication were analyzed regarding these three groups (Table 1).

329 Patients diagnosed with melioidosis and diabetes mellitus during January 2009 - October 2011 and aged between 21 - 65 years were assessed for eligibility 42 Were exclude due to no evidence of melioidosis diagnosis 25 Were exclude due to readmission 17 Were exclude due to no capillary blood sugar records on admission 1 Were excluded due to no medical records from the hospital’s database

Statistical analysis All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) software, version 15.0. We analyzed any outcomes that were related to mean capillary blood glucose in each category using either Pearson’s Chi-square or Fisher’s exact test where appropriate if there were categorical variables and Kruskal Wallis test if there were scale variables. All factors were analyzed for risk association using odds ratio and 95% Confidence Interval. However, to adjust for confounders, logistic regression analysis (Table 2) was used to calculate adjusted odds ratio. To determine time and affects, Kaplan-Meier estimator (Figure 2) was used.

244 Were  included  in  the  study  

Serial determinations of capillary sugar were performed in patients starting at admission and continuing at 6- hour interval during the first 3 days of their stay

RESULTS In the present study, 329 patients were preliminarily included (Figure 1). Eighty-five cases were excluded due to no evidence of melioidosis infection, readmission, and no capillary blood glucose in the first 72 hours after admission. At the end, 244 patients were included in the analysis and arranged in to three groups according to mean capillary glucose levels. Most patients were male (67.8%) with a median age of 52. Most of them aged between 50 and 59 (Table 1). The majority of them were farmers (41.8%). A bit more than half were diagnosed using positive titer higher than 1:160. Approximately 55% had a positive culture for Burkholderia pseudomallei. Almost 98% of the study group was diagnosed with type 2 diabetes; only six patients had type 1 diabetes. Fifty-nine patients (28.1%) had a first-admission diagnosis of diabetes mellitus. As for diabetic control, the majority of the patients were using glibenclamide, metformin, and mixtard insulin; less than 5% using diet control. In relation to antibiotics, the three most commonly prescribed were ceftazidime, ceftriaxone and doxycycline (Table 1). However, very few cases were

244 Were divided into 3 categories by the range of mean capillary blood glucose levels

86 Had blood glucose levels ≤ 200 mg%

64 Had blood glucose levels 201-250 mg%

94 Had blood glucose levels 251 mg%

Figure 1. Screening and Patients Flow Chart

and no previous treatment in cases of newly diagnosed diabetes patients. Occupation was classified into five groups, farmers, employees, businessmen, government officers, and the unemployed. The diagnosis was recorded by the patient’s principal diagnosis in their medical record; the patient’s

74


Journal of Medical Research and Education

JMRE

Table 1. Characteristics of the Melioidosis Patient with Diabetic Mellitus Mean capillary glucose-mg % Characteristic

≤ 200

201-250

≥ 251

Male sex – no. (%)

61 (70.9)

47 (73.4)

59 (62.8)

21 – 30

1 (1.2)

2 (3.1)

3 (3.2)

31 – 39

3 (3.5)

4 (6.3)

16 (17.0)

40 – 49

23 (26.7)

25 (39.1)

27 (28.7)

50 – 59

38 (44.2)

22 (34.4)

35 (37.2)

60 – 65

21 (24.4)

11 (17.2)

13 (13.8)

Age – yr (%)

Median

53.5

50

50

48.0 - 59.3

45.0 - 58.0

41.8 - 57.3

Farmer

30 (34.9)

32 (50.0)

41 (43.6)

Employee

30 (34.9)

19 (29.7)

32 (34.0)

Unemployed

23 (26.7)

10 (15.6)

19 (20.2)

2 (2.3)

2 (3.1)

1 (1.1)

1(1.2)

1 (1.6)

1 (1.1)

44 (51.2)

35 (54.7)

55 (58.5)

Interquartile range Occupation – no. (%)*

Government officer Business Laboratory test for melioidosis – no. (%)† Culture positive

58 (67.4)

37 (57.8)

49 (52.1)

Type 2 diabetes – no. (%)‡

Melioidosis titer positive

84 (97.7)

62 (96.9)

92 (97.9)

New case of diabetes – no. (%)

18 (26.9)

15 (29.4)

22 (27.2)

Glibenclamide

23 (34.3)

18 (35.3)

25 (30.9)

Metformin

19 (28.4)

15 (29.4)

22 (27.2)

Mixtard insulin

9 (13.4)

8 (15.7)

18 (22.2)

Glipizide

7 (10.4)

2 (3.9)

7 (8.6)

Regular insulin

3 (4.5)

5 (9.8)

1 (1.2)

Diet control

2 (3.0)

0

1 (1.2)

Ceftazidime

32 (41.0)

19 (31.7)

22 (26.5)

Ceftriaxone

31 (39.7)

24 (40.0)

28 (33.7)

Doxycyclin

11 (14.1)

6 (10.0)

11 (13.3)

Cotrimoxazole

8 (10.3)

0

2 (2.4)

Metronidazole

8 (10.3)

5 (8.3)

5 (6.0)

Cloxacillin

5 (6.4)

4 (6.7)

3 (3.6)

Penicillin G sodium

2 (2.6)

0

2 (2.4)

Roxythromycin

3 (3.8)

2 (3.3)

1 (1.2)

Sulperazone

1 (1.3)

2 (3.3)

0

Others

1 (1.2)

1 (1.6)

2 (2.1)

Type of Diabetic control – no. (%)‡

Previous treatment – no. (%)

75


JMRE

Journal of Medical Research and Education

Table 1. (Continued.) Mean capillary glucose - (mg %) Characteristic

≤ 200

201-250

Anemia

38 (44.2)

27 (42.2)

45 (47.9)

Hypertension

27 (31.4)

11 (17.2)

20 (21.3)

Chronic kidney disease

11 (12.8)

8 (12.5)

14 (14.9)

Upper gastrointestinal bleeding

11 (12.8)

4 (6.3)

6 (6.4)

Urinary tract infection

7 (8.1)

1 (1.6)

6 (6.4)

Thrombocytopenia

3 (3.5)

1 (1.6)

12 (12.8)

≥ 251

Comorbidity – no. (%)¶

Renal insufficiency

5 (5.8)

3 (4.7)

6 (6.4)

29 (33.7)

29 (45.3)

44 (46.8)

Melioidosis with septicemia

10 (11.6)

16 (25.0)

29 (30.9)

Melioidosis with pneumonia

16 (18.6)

7 (10.9)

14 (14.9)

Melioidosis with septic shock

4 (4.7)

5 (7.8)

9 (9.6)

Melioidosis with liver abscess

2 (2.3)

2 (3.1)

3 (3.2)

Disseminated melioidosis

5 (5.8)

9 (14.1)

5 (5.3)

Acute fulminant melioidosis

4 (4.7)

2 (3.1)

3 (3.2)

Others Diagnosis – no. (%)

Others 20 (23.3) 12 (18.8) * Occupation was determined by the investigators on the basis of the hospital records. † The laboratory tests such as melioidosis titers or cultures determined on laboratory records. ‡ Diabetic control and types of diabetes mellitus were obtained from medical records during admission. Previous treatment was determined by investigator on transfer medical notes. ¶ Diagnosis and comorbidities were made from ICD-10 on hospital records.

given ampicillin, meropenam, amoxicillin clavulanate, and gentamycin. There were high proportions of anemia, hypertension, chronic kidney disease and upper gastrointestinal bleeding in the study group. Other uncommon comorbidities included dyslipidemia, thalassemia, hepatitis, cholecystitis, cholestatic jaundice, splenic abscesses, lung abscesses, pulmonary tuberculosis, supraventricular tachycardia, atrial fibrillation, myocardial infarction, ischemic heart disease, and systemic sclerosis. On admission, most of the patients were diagnosed with melioidosis with septicemia, melioidosis with pneumonia and disseminated melioidosis. Other diagnoses found in this study were melioidotic septic arthritis, melioidosis with pyomyositis, melioidosis with splenic abscess, melioidosis with lung abscess, melioidosis with renal abscess, melioidosis with diarrhea, melioidosis with hepatic encephalopathy, melioidosis with parapneumonic effusion and melioidosis with pleural effusion.

16 (17.0)

significantly among the three categories of mean capillary blood glucose (P=0.024) (Table 2). Patients with a mean capillary blood glucose higher than 250 mg% had the highest proportion of sepsis (46.8%). Secondary outcome At first, there was no significant difference in mortality among the three groups. Complications such as acute respiratory failure, acute renal failure, disseminated intravascular coagulopathy, cardiac arrest, and ventilator acquired pneumonia were also found to have no significant differences amongst the three groups (Table 2). However, septic shock was the only secondary outcome that showed significant difference, in which patients with capillary blood glucose levels higher than 251 mg% had the highest proportion of sepsis (27.7%, P= 0.048). In spite of this, after confounding factors were adjusted in the logistic regression analysis, the levels of mean capillary blood glucose showed no statistical difference in developing sepsis in patients with melioidosis and diabetes mellitus (adjusted odds ratio (AOR) 1.0, 95% confidence interval (CI) 1.00 to 1.01; P=0.063) (Table 3). Female sex was found to be the only significant protective factor associated with the lower proportion of sepsis (AOR 0.37; 95% CI 0.16 to

Primary outcome Out of the 244 patients included in the study, 96 patients (39.3%) had sepsis; among them 51 (20.9%) developed septic shock, and unfortunately 43 patients (17.6%) died. The proportions of sepsis differed

76


Journal of Medical Research and Education

JMRE

Table 2. Treatment Outcomes Mean capillary glucose Categories - mg %

Outcome

≤ 200

201 – 250

≥ 251

P Value

no. (%) Primary outcome Sepsis‡

24 (27.9)

28 (43.8)

44 (46.8)

0.024*

14 (16.3)

12 (18.8)

17 (18.1)

0.915†

Acute respiratory failure

18 (20.9)

16 (25.0)

21 (22.3)

0.839†

Acute renal failure

14 (16.3)

9 (14.1)

16 (17.0)

0.879†

4 (4.7)

6 (9.4)

4 (4.3)

0.416†

Secondary outcomes Death§ Other complications¶

Disseminated intravascular coagulopathy Cardiac arrest Septic shock Ventilator acquired pneumonia * † ‡

§ ¶

5 (5.8)

3 (4.7)

5 (5.3)

1.000†

11 (12.8)

14 (21.9)

26 (27.7)

0.048*

3 (3.5)

3 (4.7)

7 (7.4)

0.517†

Others 44 (51.2) 27 (42.2) 48 (51.1) The P value was calculated with the use of chi-square test. The P value was calculated with the use of Fisher exact test. Sepsis was defined by the presence of at least 2 of the following : Fever (oral temperature > 38°c) or hypothermia (< 36°c), tachypnea (> 24 breaths/min), tachycardia (> 90 beats/min), leukocytosis (> 12,000/µL), leukopenia (< 4,000/µL), or > 10% bands. The etiology of SIRS may be infectious or noninfectious. Death was recorded from ICD-10 hospital records that had both deaths in hospital and against advice. Other complications were obtained from hospital records.

0.86, P= 0.020). The presence of upper gastrointestinal bleeding (AOR, 12.55; 95% CI 2.16 to 72.95), having other comorbidities (AOR 5.48, 95% CI 1.81 to 16.60; P=0.005) including dyslipidemia, thalassemia, hepatitis, cholecystitis, cholestatic jaundice, splenic abscesses, lung abscesses, pulmonary tuberculosis, supraventricular tachycardia, atrial fibrillation, myocardial infarction, ischemic heart disease, and systemic sclerosis and, other complications (AOR 15.06; 95% CI 3.17 to 71.55) such as acute respiratory failure, acute renal failure, disseminated intravascular coagulopathy, cardiac arrest, septic shock, and ventilator acquired pneumonia showed significant differences in mortality rate. In the analysis using Kaplan-Meier estimator to assess factors and their risk with death, either mean capillary blood glucose or sex showed no significant associations with time since admission till death (P=0.676, P=1.537 respectively) (Figure. 2).

0.471†

found that capillary blood glucose levels did not cause a significant difference in the proportion of sepsis and mortality among the three groups. Male sex was found to be the only significant risk factor associated with the higher proportion of sepsis. As for mortality rates, factors that gave rise to increased death rates were the presence of upper gastrointestinal bleeding, having other comorbidities, (dyslipidemia, thalassemia, hepatitis, cholecystitis, cholestatic jaundice, splenic abscesses, lung abscesses, pulmonary tuberculosis, supraventricular tachycardia, atrial fibrillation, myocardial infarction, ischemic heart disease, and s y s t e m i c s c l e ro s i s ) a n d d eve l o p i n g o t h e r complications such as acute respiratory failure, acute renal failure, disseminated intravascular coagulopathy, cardiac arrest, septic shock, and ventilator acquired pneumonia. Having multiple comorbidities and complications increases the severity of the disease, leading to increased mortality rates. Comparison with other studies Since this is the first study attempted to identify the association between capillary blood glucose levels and the rate of sepsis and mortality in patients with melioidosis and diabetes, there had been no studies that focused on the same primary outcome. Even so,

DISCUSSION Initially, our results showed that sepsis and septic shock differed significantly among the three groups. However, after adjusting the confounding factors, we

77


JMRE

Journal of Medical Research and Education

Table 3. Factors Associated with Sepsis and Death Sepsis

Death

Adjusted odds ratio* (95% CI)

P Value

Adjusted odds ratio (95% CI)

P Value

Reference

1.000

Reference

1.000

0.37 (0.16 - 0.86)

0.020

0.51 (0.14 - 1.80)

0.294

0.98 (0.94 - 1.02)

0.284

1.01 (0.95 - 1.08)

0.670

Reference

0.694

Reference

0.346

Farmer

0.61 (0.24 - 1.53)

0.291

0.78 (0.21 - 2.95)

0.710

Employee

0.60 (0.22 - 1.66)

0.329

0.56 (0.13 - 2.41)

0.433

Government officer

1.82 (0.13 - 25.37)

0.656

5.04 (0.26 - 96.83)

0.284

Business

1.74 (0.12 - 25.07)

0.684

8.65 (0.37 - 204.60)

0.181

Variable Sex Male Female Age Occupation Unemployed

Type of diabetic mellitus Type 1 diabetes

Reference

1.000

Reference

1.000

Type 2 diabetes

1.09 (0.15 - 8.02)

0.933

0.41 (0.02 - 7.32)

0.540

1.01 (0.34 - 2.98)

0.984

0.23 (0.04 - 1.20)

0.081

Glibenclamide

0.64 (0.23 - 1.76)

0.382

0.22 (0.04 - 1.15)

0.072

Metformin

1.18 (0.47 - 2.95)

0.724

0.96 (0.21 - 4.39)

0.959

Mixtard insulin

0.96 (0.29 - 3.13)

0.944

0.42 (0.07 - 2.50)

0.343

Glipizide

0.44 (0.09 - 2.19)

0.315

0.37 (0.03 - 5.27)

0.466

Regular insulin

0.95 (0.17 - 5.29)

0.948

3.95 (0.44 - 35.61)

0.221

5.99 (0.34 - 105.38)

0.221

3.50 (0.13 - 96.40)

0.460

Anemia

1.19 (0.59 - 2.37)

0.629

1.53 (0.52 - 4.53)

0.444

Hypertension

0.40 (0.14 - 1.18)

0.096

0.21 (0.04 - 1.20)

0.079

Chronic kidney disease

1.48 (0.42 - 5.17)

0.542

1.68 (0.33 - 8.49)

0.533

UGIB

2.57 (0.66 - 10.00)

0.173

12.55 (2.16 - 72.95)

0.005

UTI

2.69 (0.60 - 12.10)

0.197

0.48 (0.05 - 4.77)

0.532

Thrombocytopenia

5.39 (0.97 - 29.87)

0.054

0.69 (0.10 - 4.63)

0.706

Renal insufficiency

1.17 (0.27 - 5.01)

0.838

0.15 (0.01 - 2.46)

0.181

Others

1.83 (0.93- 3.60)

0.082

5.48 (1.81 - 16.60)

0.003

1.00 (1.00 - 1.01)

0.063

1.00 (1.00 - 1.01)

0.818

-

-

15.06 (3.17 - 71.55)

0.001

New case of diabetes Type of Diabetic control

Diet control Comorbidity

Mean capillary glucose level Complications

* Odds ratio were calculated using logistic regression which compared the proportion of sepsis and other confounders such as sex, age, occupation, type of diabetes mellitus, new case of diabetes, type of diabetic control, comorbidity, complications, and mean capillary glucose.

78


Journal of Medical Research and Education

JMRE

other data such as epidemiology and patients’ characteristics supported previous studies reported elsewhere. Our study shows that males have a higher risk of acquiring melioidosis than females do, which supports two other studies including one conducted in Kedah, Malaysia9 and another in Ubon Ratchathani, Thailand.10 Both studies suggested that male sex were independent risk factors for melioidosis.10 A possible explanation might be that males tended to work or have more activities related to higher contact of soil and water which increased the susceptibility to melioidosis infection. The actual reason why males tended to develop higher proportions of sepsis is still unknown, further studies may be needed. In this study, 97.5% of the patients had type 2 diabetes which was similar to the study in Darwin Australia,9 but in contrast with that of Wood et al. who observed that patients with type 1 diabetes had a higher susceptibility to melioidosis.11 The median age found in this study appeared to be similar with recent studies in Australia,12 Thailand9, Singapore9 and Pahang, Malaysia9 with a median age of 49, 45, 55, and 51 years respectively. Our study also supported previous studies which stated that farming is still the leading occupation of patients suffering from melioidosis. 4, 9, 10, 12 In addition to this, ceftazidime was the most commonly prescribed drug used in patients with melioidosis from our cohort study; this was also consistent with the recommended therapy in previous studies.10 Other choices of antibiotics including cotrimoxazole, doxycycline, amoxicillin-clavulanate, ceftriaxone, ampicillin, genta mycin, and meropenam were found in the present study as well. 10, 13, 14

A

B

Figure 2. Kaplan-Meier Estimates Panel A Kaplan – Meier estimates the survival function between time till death and capillary blood glucose levels. The Kaplan-Meier estimates for patient in three groups were similar to those shown in the Figure 2. P value was calculated by the Log-Rank test (P=0.676). The number of patients who died was 14 in groups with a mean blood capillary ≤ 200 mg% (16.3%) and 12 in groups with a mean blood capillary 201-250 mg% (18.8%) and 17 in groups with a mean blood capillary ≥ 251 mg% (18.1%). Data were censored for 72 in group that mean blood capillary ≤ 200mg% (83.7%) and 52 in group that mean blood capillary 201-250 mg% (81.3%) and 77 in group that mean blood capillary ≥ 251 mg% (81.9%). Panel B Kaplan – Meier Estimates the survival function between time till death and sex. (136 in male group and 65 in female group) The Kaplan-Meier estimates for both sexes were similar to those shown in the figure 3. (P=1.537). The number of patient who died was 31 in male group (22.8%) and 12 in female group (18.5%). Data were censored for 136 in male group (81.4%) and 65 in female group (84.4%).

Strength and limitations A major strength of this study is that it is the first study to our knowledge to try to determine the association between blood glucose levels and the proportion of sepsis and mortality in patients with melioidosis and diabetes mellitus. Furthermore, it was conducted in the Northeast of Thailand, an endemic area with the highest incidence of melioidosis (21.31 per 100,000 people per year).10 There were several limitations in our study. Despite the fact that this study was conducted in an endemic area with the highest incidence of melioidosis, the number of population included in the studies were not as high as expected which was due to various reasons including lost medical records and no evidence for melioidosis diagnosis. Because Khon Kaen Hospital serves as a tertiary care hospital, patients admitted are those who have been partially treated and referred from community hospitals. Thus, patients included in the study tended to have a higher severity of the disease which might have an effect on our results. Patients with positive melioidosis titers or culture that had incomplete medical records in relation to melioidosis might be underdiagnosed, leading to a biased result. Moreover, many of the patients had multiple

comorbidities and complications which significantly lead to an increased mortality rate. Unfortunately, we were unable to identify the precise factors in the other comorbidities or complication groups that significantly increased death rates. The majority of the patients didn’t have a complete record of their

79


JMRE

Journal of Medical Research and Education

capillar y blood glucose results which were attributable either to missing medical records or to the length of hospital stay of less than three days.

d i a b e t e s m e l l i t u s . T h e p re s e n c e o f u p p e r gastrointestinal bleeding, having other comorbidities, and other complications were significant independent risk factors for death. Taken together, we would like to suggest that male patients, patients with upper gastrointestinal bleeding, and patients with other comorbidities should be closely monitored. Anyhow, there are still little studies on the association among blood glucose levels, the rate of sepsis and death, further studies with larger sample sizes may be needed.

Conclusion and implications In conclusion, this retrospective-cohort study suggested that the capillary blood glucose levels were not associated with the rate of sepsis and mortality rates in patients with melioidosis and diabetes mellitus. Male sex was a significant independent risk factor for sepsis in patients with melioidosis and

REFERENT 1. Wuthiekanun V, Smith MD, Dance DA, White NJ. Isolation of Pseudomonas pseudomallei from soil in north-eastern Thailand. Transactions of the Royal Society of Tropical Medicine and Hygiene. Jan-Feb 1995;89(1):41-43. 2. Harrisonâ&#x20AC;&#x2122;s principle of internal medicine edition 17. Vol Chapter 145. 3. Suputtamongkol Y, Hall AJ, Dance DA, et al. The epidemiology of melioidosis in Ubon Ratchatani, northeast Thailand. International journal of epidemiology. Oct 1994;23(5): 1082-1090. 4. Currie BJ, Jacups SP, Cheng AC, et al. Melioidosis epidemiology and risk factors from a prospective whole-population study in northern Australia. Trop Med Int Health. Nov 2004;9(11):1167-1174. 5. Schwarzmaier A, Riezinger-Gepper t F, Schober G, Karnik R, Valentin A. Fulminant septic melioidosis after a vacation in Thailand. Wiener klinische Wochenschrift. Oct 27 2000;112(20):892-895.

6. Chaowagul W, White NJ, Dance DA, et al. Melioidosis: a major cause of communityacquired septicemia in northeastern Thailand. The Journal of infectious diseases. May 1989;159(5):890-899. 7. Jesudason MV, Anbarasu A, John TJ. Septicaemic melioidosis in a tertiary care hospital in south India. The Indian journal of medical research. Mar 2003;117:119-121. 8. Standards of Medical Care in Diabetes-2011. Diabetes care 34. . 9. Hassan MR, Pani SP, Peng NP, et al. Incidence, risk factors and clinical epidemiology of melioidosis: a complex socio-ecological emerging infectious disease in the Alor Setar region of Kedah, Malaysia. BMC infectious diseases. 2010;10:302. 10. Limmathurotsakul D, Wongratanacheewin S, Teerawattanasook N, et al. Increasing incidence of human melioidosis in Northeast Thailand. The American journal of tropical

80

medicine and hygiene. Jun 2010;82(6): 1113-1117. 11. Woods DE, Jones AL, Hill PJ. Interaction of insulin with Pseudomonas pseudomallei. Infection and immunity. Oct 1993;61(10): 4045-4050. 12. Suputtamongkol Y, Chaowagul W, Chetchotisakd P, et al. Risk factors for melioidosis and bacteremic melioidosis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. Aug 1999;29(2):408-413. 13. Cheng AC, Currie BJ. Melioidosis: e p i d e m i o l o g y, p a t h o p h y s i o l o g y, a n d management. Clin Microbiol Rev. Apr 2005;18(2):383-416. 14. Inglis TJ, Rolim DB, Rodriguez JL. Clinical guideline for diagnosis and management of melioidosis. Rev Inst Med Trop Sao Paulo. JanFeb 2006;48(1):1-4.


Journal of Medical Research and Education

JMRE

Photo by Viranuj Sueblinvong, M.D.

81


JMRE

Journal of Medical Research and Education

ORIGINAL ARTICLE ESTABLISHED IN 2012

January-June 2012

VOL. 1 NO. 1

Maternal Urinary Tract Infection is Independently Associated with Preterm Delivery Kesinee promkote1 Narat Virotjanawat1 Pholawat Ouemphancharoem1 Thanatcha Apiratmontri1 Thammasorn Piriyasupong 2 , M.D, Ph.D. 1Fifth

year medical student, Department of Social Medicine, Khon Kaen Hospital, Khon Kaen 2Medical Education Center, Khon Kaen Hospital, Khon Kaen

ABSTRACT BACKGROUND Association between urinary tract infection (UTI) and preterm delivery were frequently mentioned. However, evidences to establish the association were scarce and even contradicted with each other. METHODS This is a hospital-based retrospective cohort study conducted in Khon Kaen Hospital reviewing the records of pregnant women delivered between January 2009 and March 2012. Those with UTI were included in the study and matched for the control with the ratio of 1:1. Primary outcome was preterm delivery. Association of risk factor for preterm delivery was interpreted using odds ratio. RESULTS Of the 17,589 pregnant women were screened, 181 pregnant women had history of UTI and were matched with 184 controls, 365 in total. History of UTI was significantly associated with multiple pregnancy (3.3% and 0%; P=0.014), active UTI at delivery (65.2% and 0.5%; P<0.001), Pregnancy induced hypertension (15.5% and 3.8%; p<0.001), thalassemia (33.7% and 15.2%), preterm delivery (28.7% and 5.4%; P<0.001), intrapartum fetal distress (9.2% and 2.7%; P=0.009), lower Apgar score at 1st minute (6.9% and 1.6%; P=0.013), low birth weight infants (14.5% and 5.4%; P=0.004) and admission to neonatal intensive care unit (13.3% and 3.8%; P=0.001). In the logistic regression analysis, independent factor for preterm delivery were history of UTI (AOR 8.66; 95% confidence interval(CI) 3.17-23.63), multiple pregnancy (AOR 13.27; 95% CI1.24-141.82), previous preterm delivery (AOR 16.53; 95% CI 2.37-115.24) and number attending adequate ANC (AOR 0.45 ; 95% CI 0.32-0.64) CONCLUSION Urinary tract infection is associated with adverse perinatal outcomes and specifically is an independent risk factor for preterm delivery.

82


Journal of Medical Research and Education

JMRE

Delivery before term preterm birth could be qualified according to birth weight with large variations depending on the studied populations or to gestational ages strata (such as at less than 37 week of gestation), if this definition is used, it represents 5 to 10% of the deliveries in the US.1 Preterm birth could also be categorised by its clinical presentation: medically induced, preterm premature rupture of membranes (PPROM) and spontaneous preterm labour leading to preterm delivery and is an important cause of perinatal mortality and morbidity.2 Various factors are mentioned to be associated with preterm delivery including socioeconomic like lower social class or less education or low income or little contact with neighbours,3 pre-eclampsia or eclampsia, intrauterine growth restriction, cholesterol concentration with the U shape association, smoking, previuos induced abortion, progressive periodontal disease of the mother, high pre-pregnancy body mass index (BMI), short cervical length, short interval between pregnancy, gestational diabetes (GDM)9,12 mother with history of coronary artery disease. Maternal with complicated thalassemia, hyperthyroidism, bleeding at first trimester of pregnant, placenta abruption,induction of labor.3-16 Many types of infection during pregnancy are also stated to be related to preterm delivery such as intrauterine infection, genital tract infection, and urinary tract infection (UTI).17-22 For the UTI very few studies are conducted to examine its relationship with preterm delivery. A Norwegian study in 304 pregnant women with UTI, it found no association between UTI and spontaneous preterm delivery.23 However, the data were collected using questionnaire that might prone to subjectivity and bias. In a Taiwanese study in 135 women suggested that acute pyelonephritis did not increase the risk of preterm delivery.21 The contradicted findings were observed in another study from Israel that also found significant relationship between acute pyelonephritis and the increased risk of preterm with odds ratio (OR) of 2.6.22 However, these two studies focused on only those with acute pyelonephritis. In those with subclinical UTI, one previous study from the US in nearly 750 pregnant women, positive urine culture was not associated with preterm delivery.25 Nonetheless these findings were contradicted with a study from Bangladesh in 200 pregnant women that suggested the high risk of preterm in the those with asymptomatic bacteriuria, UTI.26 From the current literatures, there are scarce evidence to establish the association between UTI and preterm delivery due t h e s m a l l s a m p l e s i z e . F u r t h e r m o re , h i g h heterogeneity of findings are also observed. Thus, in the present study, we aim to examine the association

between UTI and preterm delivery as well as to identify factors that contribute to preterm delivery.

METHODS Study design and overview A hospital-based retrospective cohort study comparing all pregnancies of women with and without urinary tract infection (UTI). The study was conducted of those who delivered between January 2008 and March 2012. at the Khon Kaen Hospital, Thailand. Patientsâ&#x20AC;&#x2122; records Records of women with history of the UTI of both lower tract, i.e. cystitis, and upper tract , i.e. pyelonephritis during pregnancy were reviewed. The clinical diagnosis of urinary tract infection was based on physical examination, urinalysis and urine culturing. Patient who delivered less than 20 weeks were classified as abortion and excluded from the study. Unverified and incomplete records were also excluded. The controls were chosen and matched from the records of those without history of UTI who delivered in the same with age (Âą5 years) with the ratio of 1:1. In the sample size calculation, with the alpha error of 5%, beta error 20%, expected outcome regarding rate of preterm delivery in the group of UTI and without UTI of 30% and 5% respectively, the required sample would be 36 for each group, 72 in total. However we included more than 180 for each group for the best approximation. Data collection Data regarding pregnancy outcomes were available from the perinatal database of the hospital and were recorded.Those included characteristics of the mother e.g., maternal age, gravidity, multiple pregnancy, history of abortion, history of previous preterm, placenta plevia, premature rupture of membrane (PROM), active UTI at delivery, number of attending antenatal care (ANC), hormonal contraception used, GDM, pregnancy induced hypertension (PIH), thalassemia, hyperthyroidism, hepatitis B infection, human immunodeficiency virus (HIV) infection, and body mass index. Outcomes Primary outcome of the present study was preterm delivery. Secondary outcomes were categorized into (i) intrapartum outcome e.g., intrapartum fetal distress, and (ii) postpartum neonatal outcomes including Apgar score less than 7 at 1st, 5th and 10th

83


JMRE

Journal of Medical Research and Education

17,589 Pregnant woman admitted for delivery

Pregnant woman with urinary tract infection (N=192)

Pregnancy without urinary tract infection (N=17,397)

11 Were excluded 3 duplicated data 1 Unverifed 7 Incomplete records Matched with same date delivery and age (-/+ 5 years)

181 included in analysis

184 included in analysis

Figure1. Retrospective Cohort of Pregnant women with Urinary Tract infection Analyzed for Preterm Delivery

minute, low birth weight infants, stillbirth, admission to neonatal Intensive care unit.

were primigravida. Six patients had twins and about 17% had history of abortion. Only 7 patients used to had preterm delivery. Four were diagnosed with placenta previa during the pregnancy. Thirty had premature rupture of membrane.More than half of them were active UTI. Their median ANC and adequate ANC was 9 (IQR 6 to 11) and 4 (IQR, 3 to 4) respectively. History of hormonal contraception used was found in 42.6%. Eleven women were diagnosed as GDM and 35 women had PIH. A bit less than half were thalassemic patients. Six were had hyperthyroidism. Four were hepatitis B infected and 2 were HIV infected. Their median BMI before pregnancy was 20.8 (IQR, 18.8 to 23.4) and 18% were obesity. Comparing those with history of UTI during pregnancy and without UTI, it found that the UTI group tended to have multiple pregnancy (P=0.014) (Table 1). Moreover they were likely to have active UTI at delivery (P<0.001), and pregnancy induced hypertension (P<0.001). In addition to this, they tended to be thalassemia (P<0.001). However, there were no significant differences between the two groups in relation to age, gravidity, history of abortion, history of previous preterm, history of placenta previa, history of PROM, number of attending ANC, hormonal contraception used, their underlying

Statistical analysis Statistical analysis was performed with PASW statistics18. Statistical significance was calculated using the chi square test for differences in qualitative variable and Mann-whitney U test for differences in continuous variables. Crude odds ratios (CORs) and t h e i r 9 5 % c o n fi d e n c e i n t e r v a l ( C I ) we re computed.Multivariable logistic regression models were constructed to identify adjusted odds ratio (AOR) in order to find independent risk factors associated with preterm delivery while controlling for the confounders. P<0.05 was considered statistically significant.

RESULTS In the present study, the records of 17,589 pregnant women were screened. Of this, 181 pregnant women had history of UTI and were matched with 184 controls(3 patient in UTI group, duplicated data were exclude), 365 in total Their median age was 24 years old (interquartile range (IQR), 20 to 29). Half of them

84


Journal of Medical Research and Education

JMRE

Table 1. Characteristics of Cohort Members Characteristic Maternal age -- yr (7)(8) Median Interquartile range Gravidity --no. (%)(7)(8) 1 2 3 or more Multiple pregnancy --no. (%) History of abortion --no.(%) History of previous preterm --no. (%) History of placenta previa --no. (%) History of premature rupture of membrane --no. (%) Active urinary tract infection at delivery --no. (%) Number attending antenatal care Total Median Interquartile range Adequate Median Interquartile range Hormonal contraception used --no./total no. (%) Underlying disease --no. (%) Gestational diabetes mellitus Pregnancy induced hypertension Thalassemia Hyperthyroidism Hepatitis B infection --no./total (%) Human immunodeficiency virus infection --no. (%) Body-mass index before pregnancy Median Interquartile range Maternal obesity --no./total (%)

Urinary tract infection (N=181)

Non-urinary tract infection (N=184)

24 20-29

24 19.2-29.0

95 (52.5) 54 (29.8) 32 (17.7) 6 (3.3) 34 (18.8) 3 (1.7) 4 (2.2) 19 (10.5) 118 (65.2)

90 (48.9) 61 (33.2) 33 (17.9) 0 27 (14.7) 4 (2.2) 0 11 (6.0) 1 (0.5)

9 7-11

8 6-10

4 3-4 69/173 (39.9)

4 3-4 77/170 (45.3)

6 (3.3) 28 (15.5) 61 (33.7) 5 (2.8) 1/177 (0.6) 2 (1.1)

5 (2.7) 7 (3.8) 28 (15.2) 1 (0.5) 3/183 (1.6) 0

21.1 19.1-24.0 21/102 (20.6)

20.2 18.7-22.7 15/98 (15.3)

P Value 0.84

0.759

0.014 0.293 1.000 0.059 0.116 <0.001 0.074

0.797

diseases such as GDM, hyperthyroidism, hepatitis B infection and HIV infection and BMI. In relation to the outcome of the present study, patient with UTI had significant higher rates of preterm delivery (P<0.001), intrapartum fetal distress (P=0.009) lower Apgar score at 1st minute (P=0.013), low birth weight infants (P=0.004) and admission to neonatal intensive care unit (P=0.001) (Table 2). However, there were no significant differences between the two groups in relation to lower Apgar score at 5th and 10th minutes, birth weight and rate of stillbirth. In the logistic regression analysis with preterm delivery as the outcome variable, following condition were found to be an independent risk factor for preterm delivery; history of UTI (AOR 8.66; 95% confidence interval (CI) 3.18-23.63), multiple

0.311 0.738 <0.001 <0.001 0.119 0.623 0.245 0.068

0.331

pregnancy (AOR 13.28; 95% CI 1.24-141.82), previous preterm delivery (AOR 16.54; 95% CI 2.37-115.25) and number attending adequate ANC (AOR 0.46 ; 95% CI 0.33-0.65) (Table 3). Nevertheless, maternal age, gravidity, history of abortion, history of PROM, active UTI at delivery, hormonal contraception used, PIH and thalassemia were found not associated significantly with preterm delivery.

DISCUSSION The main findings of the present study is that in women having a history of UTI during pregnancy was an independent risk factor for preterm delivery. It also found to be associated with intrapartum fetal

85


JMRE

Journal of Medical Research and Education

Table 2. Neonatal and Maternal Outcomes of the Cohort Urinary tract Outcome infection (N=181) Preterm labor --no. of patient (%)

Non-urinary tract infection P Value (N=184)

52 (28.7)

10 (5.4)

<0.001

16/173 (9.2)

5/184 (2.7)

0.009

<7 at 1st minute

12/173 (6.9)

3/184 (1.6)

0.013

<7 at 5th minute

4/173 (2.3)

0/184 (0)

0.054

<7 at 10th minute

2/173 (1.2)

0/184 (0)

0.234

2950

3050

Intrapartum fetal distress --no./total no. (%) Apgar score --no./total no. (%)

Birth weight (gram)

0.060

Median Interquartile range Low birth weight (<2500 g) --no./total no. (%) Stillbirth --no./total no. (%) Admission to neonatal intensive care unit --no./total no. (%)

2675-3365

2830-3280

25/173 (14.5)

10/184 (5.4)

0.004

2/175 (1.1)

0/184 (0)

0.237

23/173 (13.3)

7/184 (3.8)

0.001

Table 3. Risk Factor for Preterm Delivery Results of a Binary Logistic Regression with Preterm Delivery as the Outcome Variable 95% Crude Confidence Adjusted 95% Confidence odds ratio odds ratio interval interval History of urinary tract infection 7.01 3.43-14.32 8.66 3.17-23.63 Maternal age

N/A

N/A

1.01

Gravidity

N/A

N/A

0.76

0.48-1.20

Multiple pregnancy

26.49

3.03-231.01

13.27

1.24-141.82

History of abortion

0.95

0.45-1.99

1.00

0.36-2.76

History of previous preterm delivery

6.89

1.50-31.63

16.53

2.47-115.24

History of premature rupture of the membrane

2.72

1.20-6.14

2.32

0.87-6.17

Active urinary tract infection at delivery

3.98

2.26-7.03

0.87

0.39-1.95

Number of attending adequate antenatal care

N/A

N/A

0.45

0.32-0.64

Hormonal contraception used

0.87

0.49-1.55

1.00

0.50-1.99

Pregnancy induced hypertension

2.50

1.15-5.43

0.88

0.32-2.40

Thalassemia

1.33

0.72-2.45

1.02

0.48-2.13

distress, low Apgar score at 1st minute, low birth weight of the infant, and admission to intensive neonatal care. Other factors that found to be significantly related to preterm delivery included multiple pregnancy, history of previous preterm delivery, number of attending adequate ANC. The clinical course of UTI in pregnancy was describe well as 20-30 years ago. However previous studies investigating the UTI during pregnancy outcomes have been studied in western countries, so we decided to study pregnancy outcomes among women with UTI and non UTI in Asia,Thailand. Our study based on people in Khon Kaen hospital and have much more adequate sample size. We found that after adjusting for potential confounders exposure to UTI or non UTI during pregnancy were significantly increase the risk of preterm birth (<37 wks) compare with mother whom were not diagnosis to UTI. Our study was parallel from prior study. 22,26 Furthermore this study have shown multiple pregnancy27, history of previous preterm28 are significantly associated with preterm delivery same to number of adequate antenatal care but has not in other journal. AOR of number of adequate antenatal care is too low so we need more sample size to confirm the association with preterm

0.94-1.07

delivery. We also found delivery rate was much (17%) than among others(5-10%,USA). It is probably this hospital is a tertiary care unit, so there are a lot of complicated case which continue to preterm delivery refer to here. Although there is some study suggest that UTI is not a risk factor for preterm delivery 21,23,25 . These based on relative inadequate sample size and inadequate control of confounders which could undermine the strength of finding. For example recent studies have indicated that the most common complications during pregnancy are anemia, hypertensive disorders, and UTIs; in particular, anemia was common in women with adverse pregnancy outcomes.29 However,unlike our study, where these comorbidities were taken into consideration. By the way, we do not adjust antibiotic UTI, alcohol and smoking30, maternal education, and other which in relation to preterm delivery3,31 because we have unverified data. Finally, our study is hospital-based study which each patient has different and more complication than population, so it should be careful in inferential the result to the population. To our knowledge, this is the first study examining the association of UTI and preterm delivery with 86


Journal of Medical Research and Education

JMRE

adequate sample to established the association. However, as this is a retrospective cohort study, missing data were not avoidable. Still, we attempted to verify all retrieved data as much as possible. Very small percentage of the data were left missing. Even the sample was adequate, however, the estimation of the relationship of the UTI and preterm delivery might be more precise with larger sample in the prospective cohort study.

In this study, UTI was found to be the independent risk factor for preterm delivery. We suggested that those with history of UTI during pregnancy should be closely monitoring for the chance of preterm labor and delivery. However, prescribed antibiotics might pose another risk factor for preterm delivery as well. For the further study, a larger prospective cohort study should be conducted for the more precise results.

REFERENCES 1. Moutquin J, Moutquin J,. Classification and heterogeneity of preterm birth, Classification and heterogeneity of preterm birth. BJOG: An Inter national Jour nal of Obstetr ics & Gynaecology, BJOG: An International Journal of Obstetrics & Gynaecology. 2003 April 1;110, 110(s20,):30, 30–3, 33. 2. Goldenberg RL, Culhane JF, Iams JD, Romero R, . Epidemiology and causes of preterm birth. The Lancet. 2008 march;371(9606):75–84. 3. Ananth C, Savitz D, Luther E, Bowes, Jr. W, author. Preeclampsia and Preterm Bir th Subtypes in Nova Scotia, 1986 to 1992. American Journal of Perinatology. 2008 March 4;14(01):17–23.10. Peacock JL, Bland JM, Anderson HR, . Preterm delivery: effects of socioeconomic factors, psychological stress, smoking, alcohol, and caffeine. BMJ. 1995 aug 26;311(7004):531–5. 4. Zeitlin J, Ancel PY, Saurel-Cubizolles MJ, Papiernik E, Zeitlin J, Ancel PY. The relationship between intrauterine growth restriction and preterm delivery: an empirical approach using data from a European case-control study, The relationship between intrauterine growth restriction and preterm delivery: an empirical approach using data from a European casecontrol study. BJOG: An International Journal of Obstetr ics & Gynaecology, BJOG: An Inter national Jour nal of Obstetr ics & Gynaecology. 2000 May 1;107, 107(6, 6):750, 750–8, 758. 5. Grote NK, Bridge JA, Gavin AR, Melville JL, Iyengar S, Katon WJ, . A Meta-analysis of Depression During Pregnancy and the Risk of Preterm Bir th, Low Bir th Weight, and Intrauterine Growth Restriction. Arch Gen Psychiatry. 2010;67(10):1012–24. 6. Maršál K, Persson P, Maršál K, Persson P, . Ultrasonic measurement of fetal blood velocity wave form as a secondary diagnostic test in screening for intrauterine growth retardation, Ultrasonic measurement of fetal blood velocity wave form as a secondary diagnostic test in screening for intrauterine growth retardation. Journal of Clinical Ultrasound, Journal of Clinical Ultrasound. 1988 may 1;16, 16(4, 4):239, 239– 44, 244. 7. Catov JM, Ness RB, Wellons MF, Jacobs DR, Roberts JM, Gunderson EP, . Prepregnancy Lipids Related to Preterm Birth Risk: The Coronary Artery Risk Development in Young Adults Study. JCEM. 2010 Jan 8;95(8):3711–8. 8. Moreau C, Kaminski M, Ancel PY, Bouyer J, Escande B, Thiriez G. Previous induced abortions and the risk of very preterm delivery: results of the EPIPAGE study, Previous induced abortions and the risk of very preterm delivery: results of the EPIPAGE study. BJOG: An Inter national Jour nal of Obstetr ics & Gynaecology, BJOG: An International Journal of Obstetrics & Gynaecology. 2005 april 1;112, 112(4, 4):430, 430–7, 437. 9. Ancel P, Lelong N, Papiernik E, SaurelCubizolles M, Kaminski M, . History of Induced

Abortion as a Risk Factor for Preterm Birth in European Countries: Results of the EUROPOP Survey. Hum. Reprod. 2004 Jan 3;19(3):734–40. 10. Lao TT, Ho LF, . Induced Abortion Is Not a Cause of Subsequent Preterm Delivery in Teenage Pregnancies. Hum. Reprod. 1998 Jan 3;13(3):758–61. 11. Offenbacher S, Boggess KA, Murtha AP, Jared HL, Lieff S, McKaig RG. Progressive Periodontal Disease and Risk of Very 13. 8Johnson TS, Rottier KJ, Luellwitz A, Kirby RS, . Maternal Prepregnancy Body Mass Index and Delivery of a Preterm Infant in Missouri 1998-2000. Public Health Nursing. 2009 Jan; 26(1):3–13.Preterm Delivery. Obstetrics & Gynecology. 2006 Jan;107(1):29–36. 12. Heath VCF, Southall TR, Souka AP, Elisseou A, Nicolaides KH, Heath VCF. Cervical length at 23 weeks of gestation: prediction of spontaneous preterm delivery, Cervical length at 23 weeks of gestation: prediction of spontaneous preterm delivery 13. Smith GCS, . Interpregnancy interval and risk of preterm birth and neonatal death: retrospective cohort study. BMJ. 2003 Aug 9;327(7410):313–0.. Ultrasound in Obstetrics & Gynecology, Ultrasound in Obstetrics & Gynecology. 1998 Nov 1;12, 12(5, 5):312, 312– 7, 317. 14. Lykke JA, Paidas MJ, Damm P, Triche EW, Kuczynski E, Langhoff-Roos J. Preterm delivery and risk of subsequent cardiovascular morbidity and type-II diabetes in the mother, Preterm delivery and risk of subsequent cardiovascular morbidity and type-II diabetes in the mother. BJOG: An International Journal of Obstetrics & Gynaecology, BJOG: An International Journal of Obstetrics & Gynaecology. 2010 Feb 1;117, 117(3, 3):274, 274–81, 281. 15. F A, L R, Sc R, Ft K, . Maternal placental vasculopathy and infection: two distinct subgroups among patients with preterm labor and preterm ruptured membranes. American journal of obstetrics and gynecology. 1993 Feb; 168(2):585. 16. Luewan S, Chakkabut P, Tongsong T, . Outcomes of pregnancy complicated with hyperthyroidism: a cohort study. Arch. Gynecol. Obstet. 2011 Feb;283(2):243–7. 17. Dh W, Ma K, Sl H, Da E, . The association of occult amniotic fluid infection with gestational age and neonatal outcome among women in preterm labor. Obstetrics and gynecology. 1992 march;79(3):351. 18. Ww A, Rl G, Jc H, . Preterm labor: emerging role of genital tract infections. Infectious agents and disease. 1995 Dec;4(4): 196. 19. Agrawal V, Hirsch E, . Intrauterine infection and preterm labor. Seminars in Fetal and Neonatal Medicine. 2012;17(1):12–9. 20. Taylor D, Kenyon S, Tarnow-Mordi W, Taylor D, Kenyon S, Tarnow-Mordi W.. Infection and preterm labour, Infection and preterm labour. BJOG: An International Journal of

87

Obstetr ics & Gynaecology, BJOG: An Inter national Jour nal of Obstetr ics & Gynaecology. 1997 Dec 1;104, 104(12, 12): 1338, 1338–40, 1340. 21. Chen Y-K, Chen S-F, Li H-C, Lin H-C. No increased risk of adverse pregnancy outcomes in women with urinary tract infections: a nationwide population-based study. Acta Obstet Gynecol Scand. 2010 Junly;89(7):882–8. Taiwan pyelonephitis 22. Breugelmans M, Vancutsem E, Naessens A, Laubach M, Foulon W, Breugelmans M. Association of abnormal vaginal flora and Ureaplasma species as risk factors for preterm birth: a cohort study, Association of abnormal vaginal flora and Ureaplasma species as risk factors for preterm birth: a cohort study. Acta Obstetricia et Gynecologica Scandinavica, Acta Obstetricia et 23. Morken N-H, Gunnes N, Magnus P, Jacobsson B. Risk of spontaneous preterm delivery in a low-risk population: the impact of maternal febrile episodes, urinar y tract infection, pneumonia and ear-nose-throat infections. Eur. J. Obstet. Gynecol. Reprod. Biol. 2011 Dec;159(2):310–4. 24. Farkash E, Weintraub AY, Sergienko R, Wiznitzer A, Zlotnik A, Sheiner E. Acute antepartum pyelonephritis in pregnancy: a critical analysis of risk factors and outcomes. Eur. J. Obstet. Gynecol. Reprod. Biol. 2012 May; 162(1):24 25. Naresh A, Simhan HN. Association of polymicrobial growth from urine culture with adverse pregnancy outcomes. Am J Perinatol. 2011 Aug;28(7):537–42. 26. Ullah MA, Barman A, Siddique MA, Haque AKME. Prevalence of asymptomatic bacteriuria and its consequences in pregnancy in a rural community of Bangladesh. Bangladesh Med Res Counc Bull. 2007 Aug;33(2):60–4. 27. Ziadeh SM. The Outcome of Triplet versus Twin Pregnancies. Gynecologic and Obstetric Investigation. 2000;50(2):96–9. 28. Berkowttz GS. An Epidemiologic Study of Preterm Delivery. Am. J. Epidemiol. 1981 มค 1;113(1):81–92. 29. Bruce FC, Berg CJ, Hornbrook MC, Whitlock EP,Callaghan WM, Bachman DJ, et al. Maternal morbidity rates in a managed care population. Obstet Gynecol.2008;111:1089–95. 30. Wisborg K, Henriksen TB, Hedegaard M, Jergen N. Smoking during pregnancy and preterm birth. BJOG: An International Journal of Obstetrics & Gynaecology. 1996 สค 1;103(8):800–5. 31. Anderson BL, Simhan HN, Simons K, Wiesenfeld HC. Additional antibiotic use and preter m bir th among bacter iur ic and nonbacteriuric pregnant women. Int J Gynaecol Obstet. 2008 สค;102(2):141–5.


Coming Up Next


JMRE

Journal of Medical Research and Education Volume I

Number 2

2012


JMRE

Jmre V.1 No.1  
Read more
Read more
Similar to
Popular now
Just for you