FIGURES, TABLES, AND EXHIBITS
FIGURE
FIGURE 2.17
FIGURE 2.18
FIGURE 2.19
FIGURE
TABLES, AND EXHIBITS
FIGURE 2.24 Tools to Learn from Variation in Data 72
FIGURE 2.25 Scatter Plots for Data in Table 2.18 74
FIGURE 3.1 Historical Example of a Run Chart 78
FIGURE 3.2 Run Chart Example 78
FIGURE 3.3 Run Chart Leading to Questions 79
FIGURE 3.4 Run Chart with Labels and Median 82
FIGURE 3.5 Run Chart with Goal Line and Tests of Change Annotated 83
FIGURE 3.6 Stat Lab Run Chart with No Evidence of Improvement 84
FIGURE 3.7 Improvement Evident Using a Set of Run Charts Viewed on One Page 85
FIGURE 3.8 Run Charts Used as Small Multiples 86
FIGURE 3.9 Run Chart Displaying Multiple Measures 87
FIGURE 3.10 Run Chart Displaying a Different Measure for Each Axis 87
FIGURE 3.11 Run Chart Displaying Multiple Statistics for the Same Measure 88
FIGURE 3.12 Run Chart with Little Data 88
FIGURE 3.13 Run Chart with Clinic Team Uncertain About Improvement 89
FIGURE 3.14 Four Rules for Identifying Nonrandom Signals of Change 90
FIGURE 3.15 Run Chart Evaluating Number of Runs 92
FIGURE 3.16 Measure with Too Few Runs 93
FIGURE 3.17 Run Chart with Too Many Runs 94
FIGURE 3.18 Run Charts of Clinic Cycle Time 95
FIGURE 3.19 Average Time to Administer Antibiotics 96
FIGURE 3.20 Three Key Uses of Run Charts in Improvement Initiatives 98
FIGURE 3.21 Beginning a Run Chart as Soon as the First Data Are Available 100
FIGURE 3.22 Run Charts for Waiting Time Data 101
FIGURE 3.23 Delay Detecting Signal with Proper Median Technique 102
FIGURE 3.24 Detecting Signal with Proper Median Technique 102
FIGURE 3.25 Detecting Signal of Improvement with Two Medians 103
FIGURE 3.26 Two Cases When Median Ineffective on Run Chart 104
FIGURE 3.27 Run Chart of Incidents Resulting in Too Many Zeros 105
FIGURE 3.28 Run Chart of Cases between an Incident 105
FIGURE 3.29 Starting and Updating Chart of Cases between Undesirable Rare Events 106
FIGURE 3.30 Mature Run Charts Tracking Cases Between Rare Events 107
FIGURE 3.31 Use of Data Line on Run Chart 108
FIGURE 3.32 Data from Unequal Time Intervals Displayed in Usual Run Chart 108
FIGURE 3.33 Data From Unequal Time Intervals Displayed to Reveal Impact of Time 109
FIGURE 3.34 Run Chart from Figure 3.22 With Seventh Week Added 110
FIGURE 3.35 Run Chart with Inappropriate Use of Trend Line 110
FIGURE 3.36 Run Chart of Autocorrelated Data from a Patient Registry 111
FIGURE 3.37 Run Chart with Percentage Doubled in Most Recent Month 112
FIGURE 3.38 Shewhart Control Chart (P Chart) Adjusting Limits Based on Denominator Size 113
FIGURE 3.39 Infant Mortality Data Stratified Using a Run Chart
FIGURE 3.40 Harm Data Stratified Using a Run Chart
FIGURE 3.41 Multi-Vari Chart
FIGURE 3.42 Run Chart and CUSUM Run Chart of Patient Satisfaction Data
FIGURE 4.1 Using Shewhart Charts to Give Direction to an Improvement Effort
FIGURE 4.2 Example of Shewhart Chart with Equal Subgroup Size 131
FIGURE 4.3 Example of Shewhart Chart with Unequal Subgroup Size 131
FIGURE 4.4 Rules for Detecting a Special Cause
FIGURE 4.5 Detecting “Losing the Gains” For an Improved Process
FIGURE 4.6 Depicting Variation Using a Run Chart versus a Shewhart Chart
FIGURE 4.7 Shewhart Charts Common Cause and Special Cause Systems
FIGURE 4.8 Shewhart Chart Revealing Process or System Improvement
FIGURE 4.9 Shewhart Chart Using Rational Subgrouping
FIGURE 4.10 Shewhart Chart Using Stratification
FIGURE 4.11 Shewhart Charts Depicting a Process or System “Holding the Gain”
FIGURE 4.12 Run Charts and Shewhart Charts for Waiting Time Data
FIGURE 4.13 Improper and Proper Extension of Baseline Limits on Shewhart Chart
FIGURE 4.14 Dealing with Special Cause Data in Baseline Limits
FIGURE 4.15 Recalculating Limits After Special Cause Improvement
FIGURE 4.16 Recalculating Limits after Exhausting Efforts to Remove Special Cause
FIGURE 4.17 Stratification of Laboratory Data with a Shewhart Chart
FIGURE 4.18 Disaggregation of ADEs Data
FIGURE 4.19 ADE Rate Rationally Subgrouped in Different Ways
FIGURE 4.20 Shewhart Chart Meeting Goal but Unstable
FIGURE 4.21 Shewhart Chart Stable but Not Meeting Goal
FIGURE 4.22 Special Cause in Desirable Direction
FIGURE 4.23 Shewhart Chart with Special Cause in Undesirable Direction
FIGURE 4.24 Shewhart Chart for LOS
FIGURE 4.25 Percentage of Patients with an Unplanned Readmission
FIGURE 5.1 Shewhart Chart Selection Guide 161
FIGURE 5.2 I Chart for Volume of Infectious Waste 167
FIGURE 5.3 I Chart Extended and Updated with New Limits 167
FIGURE 5.4 Rational Ordering for an I Chart for Intake Process 168
FIGURE 5.5 I Chart for Budget Variances 170
FIGURE 5.6 Xbar S Chart for Radiology Test Turnaround Time 172
FIGURE 5.7 Xbar S Chart for LOS 173
FIGURE 5.8 Xbar S Chart for LOS by Provider 174
FIGURES, TABLES, AND EXHIBITS
FIGURE 5.9 Xbar and S Chart Subgrouped by Provider and Quarter 175
FIGURE 5.10 Xbar S Chart Showing Improvement in Deviation from Start Times 176
FIGURE 5.11 P Chart for Percentage of Patients Harmed 182
FIGURE 5.12 Extended P Chart for Percentage of Patients Harmed 183
FIGURE 5.13 P Chart Showing Second Phase After Improvement 184
FIGURE 5.14 P Chart for Percentage of Unplanned Readmissions 185
FIGURE 5.15 P Chart for Percentage of MRSA for Hospital System 186
FIGURE 5.16 Funnel Plot of P Chart for Percentage of MRSA for Hospital System 187
FIGURE 5.17 P Chart with Funnel Limits for Systemwide Medication Compliance 188
FIGURE 5.18 C Chart for Employee Needlesticks 191
FIGURE 5.19 C Chart for Issues by Surgeon 192
FIGURE 5.20 U Chart for Flash Sterilization 193
FIGURE 5.21 U Charts Showing the Effect of Choosing the Standard Area of Opportunity 195
FIGURE 5.22 U Chart for Complaints by Clinic with Funnel Limits 196
FIGURE 5.23 Comparison of G Chart to U Chart 199
FIGURE 5.24 G Chart for ADEs 201
FIGURE 5.25 T Chart for Number of Days between ADEs 202
FIGURE 5.26 Different Formats for Displaying a T Chart 204
FIGURE 5.27 T Chart for Retained Foreign Objects 205
FIGURE 5.28 Process Capability: Typical Situations and Actions 207
FIGURE 5.29 Capability From an I Chart 208
FIGURE 5.30 Capability Analysis from an Xbar S Chart 209
FIGURE 6.1 Tools to Learn from Variation in Data 224
FIGURE 6.2 Histogram, Dot Plot, and Stem-and-Leaf Plot for Age at Fall 225
FIGURE 6.3 Frequency Plot (Dot Plot) of Patient Satisfaction Data 226
FIGURE 6.4 Age of Children with Head Injury 228
FIGURE 6.5 Shewhart Chart of Average Minutes to Initiate Antibiotics for Sepsis Patients 229
FIGURE 6.6 Histogram of Minutes to Antibiotic Start for Patients with Sepsis 230
FIGURE 6.7 Stable Shewhart Chart of Patient Fall Rate 230
FIGURE 6.8 Histogram of Age of People Who Fell 231
FIGURE 6.9 Distribution of Data without and with Skew 231
FIGURE 6.10 Frequency Plot of Clinic Patient Wait Time 232
FIGURE 6.11 Stratified Histograms of Patient Falls by Time of Day 233
FIGURE 6.12 Histogram of Antibiotic Start Time Stratified by Location 234
FIGURE 6.13 Shewhart Chart of Average Patient Satisfaction 235
FIGURE 6.14 Histograms Stratified by Common Cause and Special Cause Timeframes 236
FIGURE 6.15 Example of a Pareto Chart 237
FIGURE 6.16 Pareto Chart with Cumulative Percentage Line 239
FIGURE 6.17 Stable Shewhart Chart of SMC Readmission 240
FIGURE 6.18 Pareto Chart of Cited Reasons for SMC Adult Readmission 240
FIGURE 6.19 Factors Noted with Late Antibiotic Administration on Nursing Units 241
FIGURE 6.20 Shewhart Chart of Hospital Mortality Percentage 242
FIGURE 6.21 Pareto of Opportunities to Improve 242
FIGURE 6.22 Unweighted Pareto Chart of Nosocomial Infections 243
FIGURE 6.23 Weighted Pareto Chart of Nosocomial Infections 244
FIGURE 6.24 Stratified Pareto Charts of Health Status Stratified by Race 245
FIGURE 6.25 Stratified Pareto Charts of Factors Associated with Pediatric Head Injuries 246
FIGURE 6.26 Stratified Pareto Charts of Factors Associated with Patient Falls 247
FIGURE 6.27 Shewhart Chart of Adverse Drug Event Rate 248
FIGURE 6.28 Stratified Pareto Charts of Medications Associated with ADEs 248
FIGURE 6.29 Stratified Pareto Charts Contrasting Common Cause to Special Cause Timeframe 249
FIGURE 6.30 Scatterplot of Time with Provider Related to Patient Satisfaction 250
FIGURE 6.31 Scatterplot of Wait Related to Patient Satisfaction 252
FIGURE 6.32 Scatterplot with Trend Line and Statistics Added 253
FIGURE 6.33 Interpreting Patterns on the Scatterplot 254
FIGURE 6.34 Shewhart Chart of Patient Willingness to Recommend the Clinic 255
FIGURE 6.35 Scatterplots Related to Willingness to Recommend the Clinic 255
FIGURE 6.36 Scatterplot Relating Arrival Time and Time to Start Antibiotic 256
FIGURE 6.37 Scatterplot Relating Days between Case Worker Visits and QL Scores 257
FIGURE 6.38 Stratified Scatterplots Case Load and Sick Leave Use 258
FIGURE 6.39 Stratified Scatterplots Relating Case Worker Visits and QL Scores 259
FIGURE 6.40 Stratified Scatterplots Relating Wait Time to Satisfaction 260
FIGURE 6.41 Radar Chart of Satisfaction with Health Care 261
FIGURE 6.42 Patient Satisfaction with Urgent Care 262
FIGURE 6.43 Radar Chart of Satisfaction with Urgent Care by Element 262
FIGURE 6.44 Patient Satisfaction with Urgent Care Showing Special Cause 263
FIGURE 6.45 Radar Charts of Satisfaction with Health Care Stratified by Common and Special Cause Timeframes 264
FIGURE 6.46 Radar Charts of Patient Satisfaction Stratified by Race 264
FIGURE 7.1 Showing Data Points: (a) With Dots, (b) No Dots 269
FIGURE 7.2 Vertical Scale: (a) Just Right, (b) Too Wide, (c) To Narrow 270
FIGURE 7.3 (a) Inappropriate Vertical Scale, (b) Appropriate Scale 271
FIGURE 7.4 Including 0% and 100% on Vertical Scale 272
FIGURE 7.5 Overuse of Gridlines and Illegible Data Display 273
FIGURE 7.6 Example of Shewhart Chart with Appropriate Annotations 273
FIGURE 7.7 Extending Limits “Backward” on a Shewhart Chart 278
FIGURE 7.8 Importance of Freezing Limits on Shewhart Charts 280
FIGURE 7.9 I Charts with Limits Calculated Both with and without Screening 282
FIGURE 7.10 I Chart Compared to C Chart for Stable Count Measure 285
FIGURE 7.11 I Chart Compared to C Chart for Unstable Count Measure 286
FIGURE 7.12 I Chart Compared to P Chart for Unstable Classification Data 287
FIGURE 7.13 Published Shewhart Chart Using a Descriptive Strategy for Phasing 288
FIGURE 7.14 Deductive and Inductive Statistical Approaches 292
FIGURE 7.15 Comparison of Shewhart Chart and Statistical Inference 294
FIGURE 7.16 Comparison of Shewhart Chart with Special Cause and Statistical Inference 295
FIGURE 8.1 Expanded Chart Selection Guide to Include Alternative Charts 300
FIGURE 8.2 Example of an NP Chart 302
FIGURE 8.3 Example of an Xbar R Chart 304
FIGURE 8.4 Example of a Median Chart 306
FIGURE 8.5 P Chart with Limits that Appear “Very Tight” 307
FIGURE 8.6 Same Data as Figure 8.5 On a P’ Chart 308
FIGURE 8.7 U Chart and U’ Chart for Medication Errors 310
FIGURE 8.8 Improper Use of P Prime Chart for Self-Management Goals 311
FIGURE 8.9 Percentage of State Populations Fully Vaccinated for COVID-19 313
FIGURE 8.10 Comparison of Negative Binomial Chart to C Chart 315
FIGURE 8.11 C Chart and Negative Binomial Chart for Infections 316
FIGURE 8.12 Weighting Schemes for Alternative Charts 317
FIGURE 8.13 MA Charts Compared to I Chart 319
FIGURE 8.14 CUSUM Run Chart of Patient Satisfaction Data from Figure 3.42 321
FIGURE 8.15 CUSUM Charts for Patient Satisfaction 323
FIGURE 8.16 CUSUM Chart and Run Chart of HbA1c Values 326
FIGURE 8.17 Comparison of a C Chart and CUSUM Chart for the Same Data 327
FIGURE 8.18 EWMA Chart for Patient Satisfaction Data 329
FIGURE 8.19 EWMA Chart for HbA1c Values for Diabetic Patient 330
FIGURE 8.20 EWMA Chart with Two Phases for HbA1c Values for Diabetic Patient 330
FIGURE 8.21 Regular and Standardized Xbar Chart 332
FIGURE 8.22 Regular and Standardized P Chart 333
FIGURE 8.23 Regular and Standardized U Chart 333
FIGURE 8.24 General Form of Multivariate Control Chart (T2 Chart for Five Measures) 334
FIGURE 8.25 T2 Chart for First Two Years’ Financial Data 337
FIGURE 8.26 T2 Chart for Three Financial Measures—Baseline Limits Extended to Phase 2 337
FIGURE 8.27 I Charts for Three Financial Measures (Limits Based on 2018–2019 Data) 338
FIGURE 9.1 Shewhart Chart with Slanted Centerline for Obesity Data 342
FIGURE 9.2 I Chart with Regression Centerline for Opioid Deaths 344
FIGURE 9.3 Shewhart Chart with Nonlinear Regression Centerline 344
FIGURE 9.4 Shewhart Chart for Wait Times for the Next Appointment 345
FIGURE 9.5 Wait Times for Appointment Subgrouped by Month of Year 346
FIGURE 9.6 Individual Chart for Adjusted Wait Times 347
FIGURE 9.7 Wait Time Chart with Centerline and Limits Adjusted for Monthly Effects 347
FIGURE 9.8 P Chart for Emergency Asthma Visits 348
FIGURE 9.9 P Chart to Study Monthly Effects 348
FIGURE 9.10 Asthma P Chart—Adjusted Data and Adjusted Centerline and Limits 349
FIGURE 9.11 Initial I Chart for Average Monthly Wait Times in an ED 351
FIGURE 9.12 Scatterplot of Wait Time vs. Volume 352
FIGURE 9.13 I Chart for Average Monthly Wait Times in an ED with Adjusted Data 353
FIGURE 9.14 Monthly Electric Bill Before and After Solar Installation 353
FIGURE 9.15 Scatterplots Exploring the Relationship between the Electric Bill and Temperature 354
FIGURE 9.16 Adjusted Monthly Electric Bill Before and After Solar Installation 354
FIGURE 9.17 Ineffective I Chart for Time to Complete Weekly Report 357
FIGURE 9.18 Frequency Plots for Data and Transformations (Time for Weekly Task) 357
FIGURE 9.19 Alternative Displays of I Charts for Transformed Data 358
FIGURE 9.20 Xbar S Chart for ED Times from Arrival to Treatment 359
FIGURE 9.21 Frequency Plots for Original and Log10 Transformed Time to Treatment 360
FIGURE 9.22 Xbar S Chart Based on Log10 Transformed ED Times 360
FIGURE 9.23 Chart for Average HbA1c Values from Registry 362
FIGURE 9.24 Scatter Plot to Evaluate Autocorrelation of Registry Data 363
FIGURE 9.25 I Charts for Visit Cycle Time in Specialty Clinic 365
FIGURE 9.26 Scatterplot Implying Spurious Autocorrelation 365
FIGURE 9.27 Including Case-Mix Adjustments on Shewhart Chart 368
FIGURE 9.28 Example of Comparison Chart for Perioperative Mortality 369
FIGURE 9.29 Xbar S Chart for LOS 371
FIGURE 9.30 95% Confidence Intervals for Average LOS by Month 372
FIGURE 10.1 Shewhart Chart Revealing Improvement Not Sustained 379
FIGURE 10.2 The Drill Down Pathway 380
FIGURE 10.3 Comparison of Aggregated and Disaggregated Mortality Data 382
FIGURE 10.4 Mortality Rate Using a Different Sequencing Strategy 383
FIGURE 10.5 Rational Subgrouping Strategy for Mortality Data 384
FIGURE 10.6 Shewhart Chart at the Aggregate Level 387
FIGURES, TABLES, AND EXHIBITS xx
FIGURE 10.7 Shewhart Chart Displaying All Eight Hospitals on the Same Chart 389
FIGURE 10.8 Separate Shewhart Chart for Each Hospital Special Cause to the System 391
FIGURE 10.9 Separate Shewhart Chart for Each Hospital Common Cause to the System 392
FIGURE 10.10 ADE Rate Subgrouped by Day of the Week 394
FIGURE 10.11 Aggregate Shewhart Chart Rationally Subgrouping Common Cause Data by Shift 395
FIGURE 10.12 ADE Rate Subgrouped by Shift for Common Cause Hospitals 396
FIGURE 10.13 Pareto Chart of ADE Occurrence by Medication Name 398
FIGURE 10.14 Pareto Chart of ADEs Associated with Various Factors 399
FIGURE 10.15 Shewhart Chart Used to Determine Impact of Changes Implemented 400
FIGURE 11.1 Xbar S Charts for Peak Flow Readings from Patient with Asthma 408
FIGURE 11.2 Run Chart for PSA Test Results for a Colleague 410
FIGURE 11.3 PSA Test Results for One of the Authors 410
FIGURE 11.4 Run Chart of Ultrasound Measures on Whiteboard in Patient Room 412
FIGURE 11.5 Run Charts for Patient Bone Density Test 413
FIGURE 11.6 I Charts for Patient BMD Tests at Two Locations 414
FIGURE 11.7 Run Chart of Temperatures for Hospitalized Patient with Fever 415
FIGURE 11.8 I Chart for Temperature Readings for Patient with Fever 416
FIGURE 11.9 Xbar S Chart for Temperature Readings for Patient with Fever 416
FIGURE 11.10 CUSUM Chart for Patient Temperatures 417
FIGURE 11.11 Run Chart of Patient Monitoring Data (Half-Hour Averages) 418
FIGURE 11.12 I Charts for Patient Heart Function Variables Monitored in the ICU 419
FIGURE 11.13 Run Chart/I Chart for an Individual’s Weighings—Two Horizontal Scales 421
FIGURE 11.14 I Chart for Monitoring HbA1c for Patient with Diabetes 422
FIGURE 11.15 I Chart for Patient Pain Assessments during Hospital Stay 423
FIGURE 12.1 Excerpts from HCAHPS Survey 427
FIGURE 12.2 Excerpt from NHS Survey 428
FIGURE 12.3 Shewhart Charts for One Question from Patient Satisfaction Survey 431
FIGURE 12.4 Patient Satisfaction Data Summarized with Multiple Negative Replies 434
FIGURE 12.5 Patient Satisfaction Percentile Ranking 435
FIGURE 12.6 Pareto Chart of Types of Patient Complaints 436
FIGURE 12.7 Small Multiples of Patient Satisfaction Data 437
FIGURE 12.8 Pareto Chart of Clinic Patient Feedback 439
FIGURE 12.9 Clinic Patient Feedback Shewhart Charts for Three Areas of Focus 441
FIGURE 12.10 Scatterplots for Three Areas of Focus 441
FIGURE 12.11 Shewhart Chart of Willingness to Recommend the Clinic 442
FIGURE 12.12 Xbar S Chart of Average Self-Reported Patient Pain Assessment 443
FIGURE 12.13 P Chart Summarizing Patient Feedback Regarding Pain 444
FIGURE 12.14 Employee Feedback Upon Exit Interview 445
FIGURE 12.15 Importance and Satisfaction Matrix 446
FIGURE 12.16 Using an Interim of Surrogate Measure to Avoid Lag Time 448
FIGURE 12.17 Data Not Used When Treating Continuous Data as Classification 449
FIGURE 13.1 Tabular VOM Using Green, Yellow, and Red Formatting 456
FIGURE 13.2 Shewhart Chart of Safety Error Rate 457
FIGURE 13.3 Percentage of Perfect Care Displayed on a Shewhart Chart 458
FIGURE 13.4 Shewhart Chart of Percentage of Areas Meeting Appointment Goal 459
FIGURE 13.5 Infection Rate Data Color-Coded Monthly 460
FIGURE 13.6 Average Physician Satisfaction 461
FIGURE 13.7 Appropriate Display of VOM 462
FIGURE 13.8 Graph with Appropriate Space for Future Data 465
FIGURE 13.9 Graph with Excessive Number of Data Points 466
FIGURE 13.10 Graph Updated to Provide More Readable Number of Data Points 467
FIGURE 13.11 Graph with Historical Data Summarized 467
FIGURE 14.1 Example of Typical Epidemiological Curve with Four Epochs 478
FIGURE 14.2 Example of Hybrid Shewhart Chart for Epidemic Data 478
FIGURE 14.3 Initial C Charts for COVID-19 Deaths in Three Countries 480
FIGURE 14.4 C Chart of COVID-19 Deaths for Maine (First Half of 2021) 481
FIGURE 14.5 Initial Attempt at Charts for Epoch 2 482
FIGURE 14.6 Charts for Epoch 2 Based on Log-Regression I Charts 483
FIGURE 14.7 Hybrid Shewhart Chart for COVID-19 Daily Deaths 484
FIGURE 14.8 Chart for COVID-19 Daily Deaths Showing End of Epoch 2 485
FIGURE 14.9 Chart of US COVID-19 Daily Deaths Showing Epoch 3 Chart 486
FIGURE 14.10 Italy Daily COVID-19 Deaths Showing Epoch 4 Chart 486
FIGURE 14.11 Bar Chart Showing Variation in Reporting COVID-19 Deaths by Day of the Week 488
FIGURE 14.12 Comparison of Raw Data and Adjusted Data on the Hybrid Shewhart Charts 489
FIGURE 14.13 COVID-19 Daily Reported Deaths and Cases for the United Kingdom 490
FIGURE 14.14 Family of Measures for COVID-19 from Ireland (March 2020 to July 2021) 491
FIGURE 14.15 Family of Measures for COVID-19 from Ireland (Recent Ninety Days) 492
FIGURE 15A.1 Baseline Data for Clinic Access Project 496
FIGURE 15A.2 24-Week Data for Clinic Access Project 499
FIGURE 15A.3 Urology Services Regional Demand Versus Capacity 501
FIGURE 15A.4 One Year Data for Clinic Access Project 502
FIGURE 15B.1 Significant Revisions Project: P Chart for Significant Revisions in Reading Films 505
FIGURE 15B.2 Turnaround Time Project: Xbar S Chart for Turnaround Times for Routine X-Rays 505
FIGURE 15B.3 Start Time Project: Xbar S Chart for Procedure Start Times (Actual-Scheduled) 506
FIGURE 15B.4 Scatterplot of Revisions and Turnaround Time and Revisions 507
FIGURE 15B.5 Significant Revisions Project: T Chart for Days between Significant Revisions 508
FIGURE 15B.6 Significant Revisions Project: G Chart for Films between Significant Revisions 509
FIGURE 15B.7 Significant Revisions Project: Updated G Chart for Revisions 510
FIGURE 15B.8 Turnaround Time Project: Xbar S Charts for Turnaround Times for Routine X-Rays 510
FIGURE 15B.9 Turnaround Time Project: Frequency Plot for Turnaround Times after Change 511
FIGURE 15B.10 Start Time Project: Updated Xbar S Chart for CT Scan Start Times 512
FIGURE 15C.1 Shewhart Chart of Post CABG Infection Rate Prior to Improvement Project 516
FIGURE 15C.2 Shewhart Chart of CABG Infection Data after Testing New Glucose Protocol 517
FIGURE 15C.3 Shewhart Chart of CABG Infection Data after Protocol with Trial Limits 518
FIGURE 15C.4 Shewhart Chart of CABG Infection Rate after New Protocol Stratified by Hospital 519
FIGURE 15C.5 Stratified Histograms: Common Versus Special Cause Time Frames 521
FIGURE 15C.6 Shewhart Chart of CABG Infection Data after Protocol Subgrouped by Physician 523
FIGURE 15C.7 CABG Infection Rates after Intervention with Physician E 524
FIGURE 15C.8 Shewhart Chart of CABG Infection Rate Post Protocol— Sustained Improvement 525
FIGURE 15D.1 The Drill Down Pathway 527
FIGURE 15D.2 P Chart of Aggregate Percentage of C-Section Deliveries 529
FIGURE 15D.3 Percentage of C-Section Stratified and Sequenced by Physician 530
FIGURE 15D.4 Percentage of C-Sections by Physician with Special Cause Removed 531
FIGURE 15D.5 Shewhart Chart for Each Physician of Percentage of C-Section 532
FIGURE 15D.6 Pareto Chart of Documented Factors Associated with C-Sections 533
FIGURE 15D.7 Monthly Versus Quarterly Aggregate Percentage of C-Section 535
FIGURE 15E.1 Xbar S Chart for Length of Stay Outcome Measure 538
FIGURE 15E.2 Xbar S Chart for Cost Outcome Measure 539
FIGURE 15E.3 Balancing Measures: Complications and Readmissions 540
FIGURE 15E.4 Scatterplot of Length of Stay and Total Cost 541
FIGURE 15E.5 Funnel Plots (U Charts) for Surgical Complications and Readmissions 542
FIGURE 15E.6 Funnel Plot for Length of Stay Sub-grouped by Surgeon 542
FIGURE 15E.7 Length of Stay Subgrouped by Gender and Month 543
FIGURE 15E.8 Length of Stay Subgrouped by Race and Quarter 544
FIGURE 15E.9 Length of Stay Subgrouped by Day of the Week and Quarter 544
FIGURE 15E.10 Xbar S Chart for LOS with Limits Extended for Factorial Test 547
FIGURE 15E.11 Analysis of Factorial Study 548
FIGURE 15E.12 Xbar S Chart for LOS with Limits Extended from February, 2020 549
FIGURE 15E.13 Xbar S Chart for Total Cost with Limits Extended from February, 2020 549
FIGURE 15E.14 U Charts for Complications and Readmissions with Extended Limits 550
FIGURE 15F.1 P Chart of Monthly CHF Patient Admissions 553
FIGURE 15F.2 Funnel Plot of Hospitals Admitting CHF Patients 2018–2019 553
FIGURE 15F.3 P Chart of Hospital Admissions for CHF with Special Cause 555
FIGURE 15F.4 P Chart of Hospital Admissions for CHF with Updated Limits 556
FIGURE 15F.5 P Chart of Hospital Admissions for CHF One Year Post Improvement 557
FIGURE 15G.1 The Drill Down Pathway 559
FIGURE 15G.2 Aggregate APL Rate per Surgery 562
FIGURE 15G.3 Aggregate APL Rate per Surgery with Special Cause Data Excluded 563
FIGURE 15G.4 APL Rate Disaggregated by Site 563
FIGURE 15G.5 Separate Shewhart Charts of APL Rate for Each Site 564
FIGURE 15G.6 Rational Subgrouping Scheduled Versus Emergency Surgery APL Rate 565
FIGURE 15G.7 Rational Subgrouping Laparoscopic Versus Open Surgery APL Rate 566
FIGURE 15H.1 P Charts of Outcome Measures 569
FIGURE 15H.2 Telemedicine Data Charted Weekly Rather than Monthly 571
FIGURE 15H.3 P’ Charts of Outcome Data 571
FIGURE 15H.4 Percentage of No Shows with Special Cause Data Ghosted 572
FIGURE 15H.5 Percentage of Failed Calls by Service 573
FIGURE 15H.6 Percentage of Failed Calls from Internal Medicine by Provider 574
FIGURES, TABLES, AND EXHIBITS
FIGURE 15H.7 Percentage of Failed by Calls Zip Code 576
FIGURE 15H.8 Scatterplot of SES Score and Percentage of Failed Calls by Zip Code 577
FIGURE 15H.9 Percentage of Failed Calls by Age Group 579
FIGURE 15H.10 Daily Percentage of Failed Calls Used in PDSA Test Cycle 580
FIGURE 15H.11 Daily Percentage of No Shows Used in PDSA Test Cycle 581
FIGURE 15H.12 Telemedicine Family of Measures 582
FIGURE 15I.1 Run Chart of Pneumonia Charges over Time (All Physicians in Practice) 583
FIGURE 15I.2 I Chart of Pneumonia Charges over Time (All Physicians in Practice) 584
FIGURE 15I.3 I Chart of Pneumonia Charges Reordered by Physician Experience 585
FIGURE 15I.4 Xbar S Chart for Pneumonia Charges Ordered by Date of Diagnosis 586
FIGURE 15I.5 Xbar S Chart for Pneumonia Charges—Subgrouped by Physician 587
FIGURE 15I.6 Scatterplot for Pneumonia Charges and Length of Stay in Days 588
FIGURE 15I.7 Run Chart for Charges per Day 589
FIGURE 15I.8 Xbar and S Charts for Charges per Hospital Day 590
FIGURE 15I.9 Xbar S Charts (Funnel Plot) for Charges per Day by Physician 591
FIGURE 15I.10 Scatterplots for Comorbidities versus Days and Charges 592
TABLES
Table 1.1 Overview of Methods for Improvement 13
Table 1.2 Overview of Tools for Improvement 14
Table 1.3 Initial Team Plan for PDSA Cycles 23
Table 1.4 Some Additional Plans for PDSA Cycles 25
Table 2.1 Data for Improvement, Accountability, Research 29
Table 2.2 Useful Characteristics When Developing Measurement for Improvement 31
Table 2.3 Traditional Data Typologies 39
Table 2.4 Traditional Data Categories with Science of Improvement Categories 39
Table 2.5 Forms of Data 41
Table 2.6 Outcome, Process, and Balancing Measures 42
Table 2.7 Dimensions of System Performance 42
Table 2.8 FOM Including Outcome, Process, and Balancing Measures 45
Table 2.9 IOM’s Six Dimensions of Care 46
Table 2.10 Operational Definition Percentage of Residents Experiencing One or More Falls with Major Injury (Long Stay)1 (NQF: 0674) (CMS ID: N013.01) 48
Table 2.11 Asthma Severity Operational Definition 48
Table 2.12 Examples of Enumerative and Analytic Studies 53
Table 2.13 Judgment Sampling Data Collection Strategy 57
Table 2.14 Deciding the Scale of a Test 60
Table 2.15 Number of Falls 65
Table 2.16 Data for Rate of Falls 65
Table 2.17 Examples of Useful Ratios 67
Table 2.18 Wait Time and Satisfaction Data from Four Clinics 73
Table 3.1 Run Chart Data 79
Table 3.2 Percentage of On-Time Appointments in Clinic by Month 82
Table 3.3 Method 1: Run Chart Data Reordered and Median Determined 83
Table 3.4 Runs Rule Guidance—Table Checking for Too Many or Too Few Runs on a Run Chart 92
Table 3.5 Data for Percentage of Unplanned Returns to OR 112
Table 3.6 Harm Rate Data for Multi-Vari Chart 116
Table 3.7 Patient Satisfaction CUSUM Using Process Average as Target 118
Table 4.1 Balancing the Mistakes Made in Attempts to Improve 134
Table 5.1 Applications of Shewhart Charts for Continuous Data in Health Care 163
Table 5.2 Five Types of Attribute Shewhart Charts 177
Table 5.3 Minimum Subgroup Size for an Effective P Chart 178
Table 5.4 Examples of Common P Chart Applications in Health Care 181
Table 5.5 Example of Area of Opportunity for Count Data 188
Table 5.6 Applications of a C Chart and U Chart 189
Table 5.7 Data on Number of ADEs by Month 194
Table 5.8 Data on Infections From ICU 200
Table 5.9 Methods Used to Obtain Limits Based Solely on Common Cause 212
Table 6.1 Paired Samples for a Scatterplot 251
Table 7.1 Characteristics to Consider When Selecting SPC Software 283
Table 8.1 Symbols Used with NP Charts 301
Table 8.2 Symbols and Factors Used with Xbar and R Charts 303
Table 8.3 Symbols Used with Median Charts 305
Table 8.4 Symbols Used with P’ or U’ Charts 309
Table 8.5 Symbols Used with Negative Binomial Charts 314
Table 8.6 Example of Calculated MAs of Three and Five 318
Table 8.7 Factors Used with MA Chart 318
Table 8.8 Calculation of CUSUM Statistic and Limits 325
Table 8.9 Creating Standardized Statistics for Shewhart Charts with Variable Limits 332
Table 8.10 Monthly Hospital System Financial Data ($ Million Units) for Four Years 336
Table 9.1 Average Deviations from Centerline (Month Average – CL) for Each Month 346
Table 10.1 Clinical Quality Measures Comparative Summar y 376
Table 10.2 Excerpt from Long-Term Care Report Card 377
Table 10.3 Excerpt from Medical Center Balanced Scorecard 378
Table 10.4 Aggregate Monthly ADE Data 386
Table 10.5 Initial Drill Down Log for Aggregate ADE Data 386
Table 10.6 Completed Drill Down Log for Aggregate ADE Data 387
Table 10.7 Initial Drill Down Log for Disaggregation by Unit on One Chart 388
Table 10.8 ADE Data Disaggregated for Eight Hospitals and Subgrouped by Quarter 388
Table 10.9 Completed Drill Down Log for Disaggregation by Unit on One Chart 389
Table 10.10 Initial Drill Down Log Studying Special Cause Units 390
Table 10.11 Completed Drill Down Log Studying Special Cause Units 391
Table 10.12 Initial Drill down Log with Each Unit on Separate Chart 392
Table 10.13 Completed Drill Down Log with Each Unit on Separate Chart 393
Table 10.14 Initial Drill Down Log Rationally Subgrouping Aggregate Data by Day of Week 394
Table 10.15 Completed Drill Down Log Rationally Subgrouping Aggregate Data by Day of Week 394
Table 10.16 Initial Drill Down Log Rationally Subgrouping Aggregate Data by Shift 395
Table 10.17 Completed Drill Down Log Rationally Subgrouping Aggregate Data by Shift 395
Table 10.18 Initial Drill Down Log by Unit Rationally Subgrouping Shift 396
Table 10.19 Completed Drill Down Log by Unit Rationally Subgrouping by Shift 397
Table 10.20 Initial Drill Down Log Studying Medications Related to ADEs 398
Table 10.21 Completed Drill Down Log Studying Medications Related to ADEs 398
Table 10.22 Initial Drill Down Log Studying Common Factors Related to ADEs 399
Table 10.23 Completed Drill Down Log Studying Common Factors Related to ADEs 399
Table 12.1 Summary Statistics, Issues, and Tools Used with Patient Satisfaction Data 429
Table 12.2 Shewhart Charts for One Question from Patient Satisfaction Survey 433
Table 12.3 Mankoski Pain Scale 443
Table 13.1 A Summary of Some of the Categories Used to Develop a VOM 454
Table 13.2 Concepts for Measures of a System from Different Perspectives 455
Table 13.3 WSM 2.0: Measures to Assess Health System Performance on the Triple Aim 472
Table 15.1 Summary of Use of Tools and Methods to Learn from Variation in the Case Studies 494
Table 15C.1 Post CABG Infection Rate Data Prior to Improvement Project 514
Table 15C.2 CABG Infection Data After Glucose Protocol Testing 516
Table 15C.3 CABG Infection Data after New Protocol Stratified by Hospital 519
TABLES, AND EXHIBITS xxvii
Table 15C.4 CABG Infection Data after Protocol Subgrouped by Physician 522
Table 15C.5 CABG Infection Rates after New Protocol by Physician with Additional Data 524
Table 15D.1 C-Section Data 528
Table 15E.1 Format of Database for Surgery Team 538
Table 15E.2 Study Design for 3-Factor PDSA Test 546
Table 15F.1 Baseline Data for Hospital Admissions for Current CHF Patients from the Health Plan 552
Table 15G.1 Clinical Quality Measures Comparative Summary 558
Table 15G.2 APL Data Monthly 560
Table 15I.1 Data for Comorbidities versus Days and Charges 592
EXHIBITS
EXHIBIT 1.1 Documentation for Initial Self-Management PDSA Cycle 24
EXHIBIT 2.1 Operational Definition Aspirin at Arrival 49
EXHIBIT 3.1 Constructing a Run Chart 81
EXHIBIT 3.2 On-Time Appointments: Run Chart Example 82
WHO IS THIS BOOK FOR?
This book is designed for those who want to use data to help improve health care. Specifically, this book focuses on deepening skills related to using data for improvement. Our goal is to help those working in health care to make improvements more readily and have greater confidence that their changes truly are improvements. Using data for improvement is a challenge and source of frustration to many. The book is designed to meet this challenge and alleviate frustration.
This book is a good companion to The Improvement Guide: A Practical Approach to Enhancing Organizational Performance, 2nd Edition, Langley and others (JosseyBass, 2009), which provides a complete guide to improvement. Our Chapter 1 summarizes the key content from The Improvement Guide and specific references to The Improvement Guide are made throughout this book. If any of these questions sound familiar, then this book is for you:
● How many measures should I be using with improvement projects?
● What kind of measures do I need? Why should I have outcome, process, and balancing measures for an improvement project?
● What methods do I use to analyze and display my data? How do I choose the correct chart?
● How can I better interpret data from my individual patients?
● How do I know that changes I’ve made are improvements? Do I need to use research methods for improvement projects?
● Why don’t I just look at aggregated data before and after my change? Why use a run or Shewhart chart?
● How do I choose the correct Shewhart chart? How do I interpret it? Where do the limits come from? How do I make limits?
● What are 3-sigma limits? Are they different from confidence intervals?
● When do I create and then revise limits on Shewhart control charts?
● I work with rare events (such as infections, falls, or pressure ulcers). What graphs do I use?
● My data are impacted seasonally. How do I display them appropriately?
● I work with huge databases. How do I use Shewhart charts wisely when working with such large amounts of data?
● How do we learn from patient satisfaction data?
● How do we better understand data from an epidemic?
● How do I best display key organizational measures for the board and other senior leaders?
