Illinois Breastfeeding Blueprint - Data Appendix

Page 1

Illinois Breastfeeding Blueprint Quantitative Data Appendix: Data Sources, Methodology, and Additional Analyses Table of Contents

Section 1. Illinois Breastfeeding Data Sources…………………………………………………………………………………………1

Section 2. Notes about Race/Ethnicity………………………………………………….……………………………………………….. 2

Section 3. Creation of Low Income Index………………………………………………………………………………………………..3

Section 4. Hispanic Sub-Group Analyses………………………………………………………………………………………………… 4

Section 5. WIC Client Breastfeeding Continuation Analysis……………………………………………………………………. 8

Section 6. Effect of Hospital Practices on Breastfeeding Continuation and Exclusivity………………………….10


Section 1: Illinois Breastfeeding Data Sources ** indicates data source used in the Illinois Breastfeeding Blueprint report

Birth Certificates (starting with 2010 births) As of 2010, Illinois revised the birth certificate to include a question about breastfeeding. The question is worded: ““Is the infant being breastfed at discharge?” with the option of responding yes or no. Once 2010 birth certificate data are complete, the Illinois Department of Public Health will be able to examine the responses and report differences in breastfeeding rates throughout the state, including by delivery hospital. This will provide the state with important information about geographic and demographic differences in early breastfeeding behaviors, as well as documenting the association of early breastfeeding with later infant and child health. Hospital level data on the birth certificate breastfeeding question is publically available on the Illinois Hospital Report Card website (http://www.healthcarereportcard.illinois.gov).

Cornerstone Cornerstone is the administrative database for many Illinois Department of Human Services programs, including the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) and Family Case Management (FCM). For infants enrolled in WIC, breastfeeding information is recorded and updated over time, so this system can provide information about breastfeeding initiation, duration, and length of exclusive breastfeeding.

Maternity Practices in Infant Nutrition and Care (mPINC) ** mPINC is a survey of hospital representatives about policies and practices related to breastfeeding conducted biannually by the Centers for Disease Control and Prevention (CDC). Questions are asked about specific practices related to: labor and delivery care, feeding of breastfed infants, breastfeeding assistance, contact between the mother and infant, facility discharge care, staff training, and structural/organizational aspects of care. The results of this survey are publicly available by state, so Illinois’ scores may be compared to those of other states and the nation overall. The first version of the survey was administered in 2007 and all hospitals in Illinois were invited to participate. Illinois, however, had a low response rate compared to other states, with only 59% of hospitals completing the survey compared to a national average of 80%. The mPINC survey was repeated in 2009, and data will be publicly available in summer 2011. Please check the CDC mPINC website for updated data (http://www.cdc.gov/breastfeeding/data/mpinc/index.htm).

National Immunization Survey (NIS) NIS is an annual survey conducted by the Centers for Disease Control and Prevention. It surveys a representative sample of families of children 19-35 months old. While the major focus of the survey is immunizations, NIS asks questions about breastfeeding initiation, duration, and exclusivity. Estimates of breastfeeding for all women giving birth in each State are available yearly and may be compared to those in other states and the nation overall.

Pregnancy Risk Assessment Monitoring System (PRAMS) ** PRAMS is an annual survey of Illinois women who recently delivered a live birth that collects information about maternal attitudes and experiences before, during, and shortly after pregnancy. Participants are typically surveyed by mail or phone 3 to 6 months after delivery. Women are asked about their breastfeeding behaviors (such as initiation, duration, and length of exclusivity), reasons why they chose not to breastfeed or to stop breastfeeding, and whether or not they experienced specific hospital practices during their delivery hospitalization. There is a rich amount of information available in this survey, which is linked to the birth certificate for infants of participating mothers. 1


Section 2: Notes about Race/Ethnicity In the Illinois Breastfeeding Blueprint, we examine the data by race/ethnicity after dividing women into four groups: non-Hispanic white, non-Hispanic black, Hispanic, and Asian. These classifications were made based on the woman’s race and ethnicity as noted on her infant’s birth certificate. The birth certificate asks two separate questions about race and Hispanic ethnicity; these responses were combined to create the racial/ethnic groups.

Any woman reporting Hispanic ethnicity on the birth certificate was considered to be “Hispanic”, regardless of her race. We recognize that Hispanics in the United States are a heterogeneous group of people, representing many different cultures and varying levels of acculturation. Section 4 of this appendix has more detailed data about Hispanic women who gave birth in Illinois. Pacific Islander women were included in the “Asian” category. While Asian and Pacific Islander women in the United States also represent many diverse cultures, we could not examine sub-groups of Asian women. First, information on country of origin, immigration date, or levels of acculturation were not available in the PRAMS dataset. Secondly, even if additional information were available, the sample size would not have been sufficient for sub-group analysis because Asians comprise only a small proportion (4.5%) of the Illinois population. Women who were reported to be of American Indian, Alaska Native, or “other” race/ethnicity were excluded from the race/ethnicity analyses because only a very small sample of women (0.2%) fell into this category. For all our race/ethnicity group labels (“non-Hispanic white”, “non-Hispanic black”, “Hispanic”, and “Asian”), we opted to match the race and ethnicity language on the birth certificate. For some of these groups, other labels, such as “Caucasian”, “African-American”, and “Latina” could have been used. We opted, however, to remain consistent with the wording of the original birth certificate questions to avoid confusion or changing the meaning of the constructs. For example, a recent immigrant from Africa might describe herself as might describe herself as “black” and “non-Hispanic”, but not as “African-American”. We recognize that there may be errors in birth certificate information on race/ethnicity because the mother may not have answered the race and ethnicity questions herself. Race and ethnicity could have selected for the woman by the person completing the birth certificate. If given the chance, the woman might have selected a different race or ethnicity for herself, or selected multiple races. This could result in misclassification of a woman’s race/ethnicity, though we expect that such misclassification would be rare. Additionally, these measures of race and ethnicity were the only ones available to us in the PRAMS dataset.

2


Section 3: Creation of Low Income Index To compare women of different social situations and explore the relationship between social class and breastfeeding, we classified women as either low income or non-low income. This classification was made on the basis of responses to several questions in the PRAMS survey. Women were classified as low income if they reported at least one of the following: 1) Medicaid as the health insurance payer for prenatal care and/or delivery, 2) Supplemental Nutrition Program for Women, Infants, and Children (WIC) participation during pregnancy and/or after delivery, 3) Temporary Assistance for Needy Families (TANF) receipt during pregnancy, 4) Food Stamps receipt during pregnancy, or 5) an annual household income less than $25,000 per year. If women reported that they did not participate in any of the listed programs and that their household income was $25,000 or higher, they were classified as non-low income.

As shown in Figure A-1, low income status varied across racial/ethnic groups in the population of women surveyed. Thirty-seven percent (37%) of non-Hispanic white women who gave birth and 36% of Asian women who gave birth were considered low income by our classification scheme. In contrast, 91% of non-Hispanic blacks who gave birth and 88% of Hispanics who gave birth were classified as low income by our index. This reveals the strong correlation between race/ethnicity and income in our state, particularly among women of reproductive age, and should be taken into consideration when comparing racial groups overall.

3


Section 4: Hispanic Sub-Group Analyses Hispanics in the United States are a heterogeneous group of people, representing many different cultures. Many studies have demonstrated the association of country of origin, immigration status, and level of acculturation with health behaviors, utilization of health services, and the health status of Hispanics in the U.S. To try to capture the differences among Hispanics, we wanted to find a way of examining Hispanic sub-groups for our analysis. We did not have country of origin, immigration status, or primary language available from the PRAMS survey. We did, however, know which language the mother chose to take the survey in. PRAMS may be taken in either English or Spanish and women are given the option of which survey language they prefer. While survey language may serve as an indication of Hispanic women’s comfort with English, it is clearly not a comprehensive measure for acculturation. Because it was the only option available to us, we used survey language as a way of dividing Hispanic women who gave birth into two groups. Overall, Hispanic women made up about 25% of women who give birth in Illinois each year. Of Hispanic women who took the PRAMS survey, about 35% took the survey in English and 64% took the survey in Spanish.

The charts on the following pages show the differences in breastfeeding initiation, duration, and exclusivity for Hispanic women based on their survey language. These results demonstrate that not all Hispanic women are equal in terms of breastfeeding.

4


In 2008, approximately 88% of the Hispanic women who took the Spanish survey started breastfeeding their infants, compared to 84% of Hispanic women who took the English survey. This difference was not statistically significant and both groups met the Healthy People 2020 objective of 81.9% for breastfeeding initiation. Over time, the breastfeeding initiation rates for both groups of Hispanics increased.

Among low income Hispanic women, 78% of those who took the English survey and 88% of those who took the Spanish survey started breastfeeding their infants. This difference was statistically significant. Among higher income Hispanic women, there was no difference in breastfeeding initiation by survey language. Both groups had high initiation rates, meeting the Healthy People 2020 objective. Additionally, among Hispanic women who took the Spanish survey, income did not affect breastfeeding initiation. For those who took the English survey, lower income was associated with lower breastfeeding rates.

5


Among Hispanic women who started breastfeeding, there were differences in breastfeeding duration by survey language. The difference becomes pronounced around 4 weeks after delivery, when a greater percent of those who took the English survey stopped breastfeeding compared to those who took the Spanish survey. At 12 weeks after delivery, only 52% of Hispanic women who took the English survey were still breastfeeding, compared to 59% of Hispanic women who took the Spanish survey.

Among low income Hispanic women who started breastfeeding, 41% of those who took the English survey, compared to 60% of those who took the Spanish survey were still breastfeeding their infants 12 weeks after delivery. Among higher income Hispanic women who started breastfeeding, 66% of those who took the English survey, compared to 52% of those who took the Spanish survey were still breastfeeding their infants 12 weeks after delivery.

6


Survey language did not make a difference in breastfeeding exclusivity among Hispanic women who started breastfeeding. At 12 weeks after delivery, among Hispanic women who started breastfeeding, 23% of those who took the Spanish survey were exclusively breastfeeding their infants compared to 21% of those who took the English survey.

Survey language and income made very little difference in the proportion of Hispanic women who exclusively breastfed their infants for at least 12 weeks. Among low income Hispanic women who started breastfeeding, 22% of those who took the English survey and 23% of those who took the Spanish survey were exclusively breastfeeding their infants 12 weeks after delivery. Among higher income Hispanic women who started breastfeeding, 24% of those who took the English survey and 28% of those who took the Spanish survey were exclusively breastfeeding their infants 12 weeks after delivery.

7


Section 5: WIC Client Breastfeeding Continuation Analysis Data on breastfeeding is available from Cornerstone for WIC Participants in Illinois. Using these data, a method called “life tables� was used to describe the pattern of breastfeeding duration among WIC participants who initiated breastfeeding. The Hazard function shows the likelihood of women stopping breastfeeding at particular points in time. If women were more likely to stop breastfeeding at a particular point in time, the hazard rate at that point would be elevated.

8


Results Among WIC participants who initiated breastfeeding: • Less than 50% overall continued for 12 weeks • Less than 20% overall continued for at least one year • There were differences in breastfeeding continuation by race/ethnicity • Black infants were least likely to continue breastfeeding and Hispanic infants were most likely to continue breastfeeding • More than 50% of black and white infants had already stopped breastfeeding by 8 weeks • Only about 10% of black and white infants breastfed for at least one year, compared to 20% of Hispanic infants • Within each racial-ethnic group, the hazard rate is highest in the first six weeks. This demonstrates that the fastest drop-off in breastfeeding occurs during this time period. • The differences in hazard rates between racial/ethnic groups decrease the longer breastfeeding is continued

Conclusions • •

• •

The majority of Illinois infants born to mothers who participate in WIC are not benefiting from the recommended duration of breastfeeding. Even among WIC infants who start breastfeeding, disparities in duration exist. Black infants are least likely to continue breastfeeding, though white infants are not far behind them. Hispanic infants are most likely to continue breastfeeding. The first 6 weeks after delivery is when all groups of women are most likely to stop breastfeeding. The early postpartum period sets the trajectory for the continuation curve. Helping women continue breastfeeding through this early period will improve overall breastfeeding duration.

9


Section 6: Effect of Hospital Practices on Breastfeeding Continuation and Exclusivity Women who initiated breastfeeding were included in a study of how maternity care practices impact breastfeeding duration and exclusivity at 6 weeks postpartum. Among these women who started breastfeeding, women who reported experiencing certain maternity hospital practices were compared to women did not experience those practices.

Table A-1. Adjusted Odds Ratios (a-OR) of Continuing Breastfeeding and Exclusive Breastfeeding for at Least 6 Weeks among Women Who Initiated Breastfeeding

Delivery Hospital Practice

BF ≼6 Wks a-OR* (95% CI)

Exclusive BF ≼6 Wks a-OR* (95% CI)

Breastfed in the hospital

2.32 (1.89 - 2.84)

1.79 (1.44 - 2.22)

Breastfed in the first hour after delivery

1.79 (1.57 - 2.05)

1.82 (1.61 - 2.06)

Baby fed only breast milk in the hospital

3.42 (2.93 - 3.99)

Not assessed**

Hospital staff gave a BF support phone number

1.31 (1.10 - 1.57)

1.22 (1.03 - 1.45)

Hospital gave information about breastfeeding

0.90 (0.68 - 1.19)

0.90 (0.70 - 1.16)

Baby stayed in mom's hospital room

1.41 (1.20 - 1.65)

1.45 (1.25 - 1.69)

Hospital helped with breastfeeding

0.85 (0.72 - 1.02)

0.87 (0.75 - 1.01)

Hospital told mother to breastfeed on demand

1.51 (1.29 - 1.77)

1.39 (1.19 - 1.62)

Hospital gave a formula gift pack

0.74 (0.64 - 0.99)

0.62 (0.53 - 0.73)

Baby given pacifier in hospital

0.62 (0.54 - 0.71)

0.60 (0.54 - 0.68)

*Final regression models adjusted for: maternal race/ethnicity, age, education, marital status, parity, and pregnancy intention. No race/ethnicity*hospital practice interaction terms were statistically significant. **This relationship could not be assessed because infants fed anything other than breast milk in hospital were all not exclusively breastfed at six weeks postpartum

10


About Odds Ratios: Odds Ratios (OR) show the strength of the relationship between maternity care practices and breastfeeding continuation and exclusivity. ORs higher than 1 indicate that a hospital practice is associated with improved breastfeeding continuation and exclusivity; ORs lower than 1 indicate that a hospital practice is associated with reduced breastfeeding continuation and exclusivity. An OR shown in bold indicates that the finding is statistically significant.

Eight delivery hospital practices were significantly associated with improved breastfeeding duration for WIC participants: • Breastfeeding in hospital • Breastfeeding within first hour after delivery • Feeding infant only breast milk in hospital • Giving a breastfeeding support phone number • Rooming-in • Encouraging breastfeeding ‘on-demand’ • NO formula gift pack • NO pacifier use Seven delivery hospital practices were significantly associated with improved exclusive breastfeeding duration for WIC participants: • Breastfeeding in hospital • Breastfeeding within first hour after delivery • Giving a breastfeeding support phone number • Rooming-in • Encouraging breastfeeding ‘on-demand’ • NO formula gift pack • NO pacifier use Illinois hospitals can support breastfeeding by encouraging these practices that are shown to be effective at improving breastfeeding continuation and exclusivity.

11


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.