Evidence-based Public Health . . . and some reasons why we need it

1. Evidence-based Public Health . . .and some reasons why we need it Russell S. Kirby, PhD, MS, FACE Professor and Vice Chair Department of Maternal and Child Health School of Public Health University of Alabama at Birmingham

2. Objectives • Describe the evidence-based practice (EBP) paradigm • Identify key characteristics of evidence-based public health (EBPH) • Differentiate between EBP and EBPH • Review several recent controversies and their impact • Speculate on the future uses of evidence

3. Brief Summary for Those Who Are Knitting, Doing Crossword Puzzles, or Discerning the Geometric Pattern in the Carpeting • Evidence-based public health is the leading edge of modern public health practice. • It requires the same level of diligence with understanding principles of study design, sources of bias, internal and external generalizability, and research synthesis as is necessary in evidence-based practice. • Many of the necessary materials are ephemeral, but this is also true of clinical research due to the publication bias. • Several examples serve to show how this can work well, and . . . perhaps, not so well.

4. The Practice of Evidence-based Practice • “integrating individual clinical expertise with the best available external clinical evidence from systematic research” • individual clinical expertise: the proficiency and judgment acquired through experience and practice in clinical settings • external clinical evidence: clinically relevant research, from basic medical science and patient-centered clinical research

5. How Do We Practice EBP? • EBP is a life-long process of self-directed learning, in which caring for patients creates for the clinician a need for clinically important information about diagnosis, therapy, prognosis, and other clinical and health services issues. In this process, we: • Convert information needs into answerable questions (testable hypotheses) • Track down the best evidence with which to answer them • Critically appraise the evidence for validity and usefulness • Apply the results of this appraisal in clinical practice • Evaluate performance

6. Why EBP? • New types of evidence are being generated which, when known and understood, have the potential to create frequent and major changes in the way we care for our patients • Although we need this evidence daily, we usually fail to get it • Because of this, both our up-to-date knowledge and clinical performance deteriorate over time • Trying to remedy this personally through traditional CME/CEU programs generally doesn’t improve clinical performance • A different approach to clinical learning has been shown to keep its practitioners up-to-date. EBP is that different approach.

7. Quality of Evidence I: Evidence obtained from at least one properly randomized controlled trial. II-1: Evidence obtained from well-designed controlled trials without randomization. II-2: Evidence obtained from well-designed cohort or case-control analytic studies, preferably from more than one center or research group. II-3: Evidence obtained from multiple time series with or without the intervention. Dramatic results in uncontrolled experiments (i.e. results of introduction of penicillin treatment in 1940s) could also be regarded as this type of evidence. III: Opinions of well-respected authorities, based on clinical experience; descriptive studies and case reports; or reports of expert committees.

8. What is Evidence-based Public Health?

9. Develop an initial statement of the issue Sequential Framework for Enhancing Evidence-based Public Health (Brownson, et al.) Tools: rates and risks, surveillance data Quantify the issue Evaluate the program or policy Tools: systematic reviews, risk assessment, economic data Implement Search the scientific literature and organize information Re-tool Develop an action plan and implement interventions Develop and prioritize program options Refine the issue

10. Key Differences Between EBP and EBPH Characteristic EBP EBPH Quality of evidence Experimental Studies Observational and quasi- experimental studies Volume of evidence Larger Smaller Time from intervention Shorter Longer to outcome Professional training More formal, with Less formal, certification/licensing no standard certification Decision making Individual Team

11. Comparison of the Types of Scientific Evidence Characteristic Type I Type II Typical data/ Strength of preventable Relative effectiveness of relationship risk-disease relationship public health intervention Common setting Clinic or controlled Socially intact groups or community setting community-wide Quantity of evidence More Less Action “Something should be done” “This should be done”

12. Types of evidence • Type I: ‘something should be done’ • Analytic data on specific health condition and its link to preventable risk factor(s) • Type II: ‘specifically, this should be done’ • Focus on relative effectiveness of specific interventions to address a particular health condition

13. The Realistic Evidence-Based Rating Scale • Class 0: Things I believe • Class 0a:Things I believe despite the available data • Class 1: Randomized controlled clinical trials that agree with what I believe • Class 2: Other prospectively collected data • Class 3: Expert opinion • Class 4: Randomized controlled clinical trials that don’t agree with what I believe • Class 5: What you believe that I don’t

14. Some examples • VBAC and Cesarean section • Folic Acid and prevention of neural tube defects • Back to Sleep • HRTs: the mystery continues

15. Trends in Cesarean Deliveries and VBACs, United States 1990-2002 30.0 25.0 20.0 Percent of Live Births 15.0 Total C- Section Rate 10.0 Primary C-Section Rate 5.0 VBAC Rate 0.0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year

16. Trends • The velocity of the increase in the primary Cesarean section rate and the decline in VBAC rates in the recent past in the US is unprecedented. • In less than five years, more than ten years of increasing VBAC rates has disappeared. • Is this a good thing, or even a matter of concern?

17. Trends in Induction of Labor, United States, 1980-2002 25.0 Induction NHDS Medical Induction NHDS 20.0 Surgical Induction NHDS Induction-Birth Certificates 15.0 Percent of Live Births/Deliveries 10.0 5.0 0.0 1980 1985 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year

18. Trends • Rates of induction have increased dramatically across the nation. • There are differences based on data source, but no one can dispute the direction of the trend. • Let’s look at some specifics for Alabama:

19. Trends in Induction, C-Section, and VBAC, Alabama 1998-2002 35 30 25 C-Section Rate 20 VBAC Rate Primary C-S Rate Percent 15 Repeat C-S Rate Induction % 10 5 0 1998 1999 2000 2001 2002 Year

20. Is this a public health concern? • Con: public health does not focus on clinical management of patients. That is in the responsibility of the health care system, peer review, quality compliance, and provider organizations. • Pro: Cesarean section is among the most common surgical procedures. It is more expensive per total hospital stay than vaginal delivery, and leads to more complications and re-hospitalizations.

21. Is this a public health concern?(continued) • The Public Health Service has established goals for the year 2010 promoting continued reduction in overall Cesarean section rates and increases in VBAC rates for the United States. • Objective 16-9a: Reduce C-S among low-risk nulliparous women • Objective 16-9b: Reduce C-S among women with prior Cesarean birth

22. Where do Alabama and Wisconsin fit in? • Historically, Wisconsin has had one of the lowest C-section rates in the US. Alabama, on the other hand, generally has one of the highest. • In 1960, the national rate was 4%, and from the 1970s on the C-section rate has tended to be 25-33% lower than the national rate. • Wisconsin has also been a leader in the use of vaginal birth after Cesarean section, while Alabama has been comparatively slow to adopt.

23. Total Cesarean Section Rate and VBAC Rateby Race of Mother, 2001United States Compared to Wisconsin and Alabama Wisconsin US Alabama Rate Rate State Rank Rate State Rank Total C-Section Rate 19.1 45th highest 27.6 4th highest 24.4 White Non-Hispanic 24.5 19.7 28.5 Black Non-Hispanic 25.9 16.9 26.8 Hispanic 23.6 18.4 21.5 VBAC Rate 16.4 23.0 43rd lowest 11.8 6th lowest White Non-Hispanic 16.8 22.3 11.0 Black Non-Hispanic 16.7 28.8 13.5 Hispanic 14.7 22.9 12.3

24. Risk Factors Associated with Cesarean Delivery • Many patient, health care system, and physician characteristics are associated with higher or lower rates of Cesarean section. • A partial list includes maternal age (increased risk), parity (decreased risk), obesity and short stature (increased risk), estimated fetal weight > 4000g (increased risk), breech presentation (increased risk), delivery in teaching hospital (decreased risk), private insurance (increased risk), fear of malpractice suits (greatly increased risk).

25. Method of Delivery by Body Mass Index (BMI)Sinai Samaritan CNM Patients, 1994-1998 BMI Cesarean Vaginal Total No. % No. % No. % < 20 9 3.2 271 97.1 279 15 20 - 24.9 31 3.9 759 96.1 790 42 25 - 25.9 28 6.5 407 93.8 434 23 30 + 28 7.4 348 92.6 376 20 Total 96 5.1 1785 94.9 1881 Chi-Square (3 df) = 10.19, p<0.018

26. Adjusted Odds of Cesarean Delivery, SSMC CNM Patients, 1994-1998 Characteristic Odds Ratio 95 % C.I. p-value Obesity (BMI 30 +) 3.26 (1.60, 6.67) 0.0012 Weight Gain > Recommended 2.09 (1.06, 4.11) 0.0326 Short Stature (< 155 cm) 2.52 (1.12, 5.64) 0.0252 No Previous Live Births 4.30 (1.78, 10.37) 0.0012 Age 35 + 4.93 (1.08, 22.61) 0.0399 60.42 (29.86, 122.24) 0.0001 Failure to Progress 458.34 (133.74, 999) 0.0001 Breech Presentation 82.56 (19.00, 358.67) 0.0001 Placental Abruption 0.0001 5.71 (2.58, 12.64) Fetal Distress 0.0412 8.68 (1.09, 69.20) Severe Pre-eclampsia Adjusted for race of mother (black), marital status, primigravidity and very low birth weight.

27. Clinical Documentation of Previous Cesarean Section • Most clinicians practice in settings that do not have comprehensive, unified clinical informatics applications. • In a patient who’s previous delivery was with another provider, how likely is it that the patient’s history will document the type of incision, the position of the uterine scar, whether single- or double-suturing was used, etc?