Evidence based policy-making, a case study: Development of the WHO PMTCT Guidelines
1 like•1,092 views
The document summarizes the WHO guideline development process for recommendations on the use of antiretroviral drugs for treating HIV-positive pregnant women and preventing transmission to infants. It describes the 10-step process including: scoping the topic and convening experts, formulating questions to guide evidence review, systematically reviewing evidence and assessing quality, drafting recommendations considering risks/benefits, conducting peer review, and issuing the 2010 guidelines recommending "triple ARV prophylaxis" or Option B for pregnant and breastfeeding women. The guidelines represented the first time WHO recommended ARVs during breastfeeding to prevent transmission.
Evidence based policy-making, a case study: Development of the WHO PMTCT Guidelines
- 1. WHO guideline development using GRC standards; a practice example from the Kesho Bora study Stanley Luchters MD MPH PhD Centre for International Health – Burnet Institute April 4th 2012
- 2. 1 0 0 p r e g n a n t H I V p o s it iv e w o m e n O n a v e r a g e 3 5 b a b ie s w ill b e in f e c t e d w it h H I V About 7 become infected during the pregnancy About 15 become infected at the time of delivery About 13 become infected through breastfeeding
- 3. Efficacy of MTCT prevention interventions • MTCT rates in developing countries without ARV intervention = 35% • MTCT rates in developing countries with ARV prophylactic regimens = 20% – single-dose NVP or “AZT short course” • MTCT rates in developed countries = 1 - 2% – combination ARV regimens or treatment (ART), elective CS, no breastfeeding more effective regimens needed
- 4. Kesho Bora overall goal • To optimize the use of ARV drugs before, during and after delivery to prevent MTCT of HIV and preserve the health of the mother in settings where the majority of HIV-positive women are breastfeeding.
- 5. Study Rationale • Efficacy of MTCT prevention interventions in developing countries needs to be improved • Alternatives to replacement feeding for children born to HIV-infected mothers need to be identified • Health of HIV-infected mothers needs more attention Triple-ARV prophylaxis during pregnancy and breastfeeding may achieve all 3 purposes
- 6. General study outline Disease status in Intervention mother pregnancy CD4+ cell count < 200 cells/mm3 PROSPECTIVE COHORT: or HIV stage 4 Lifelong antiretroviral therapy CD4+ cell count 200-500 RANDOMIZED TO: cells/mm3 • “triple ARV prophylaxis” up to 6 months post partum during BF OR • “AZT short course” up to 1 weeks post delivery (WHO 2006 recommended regimen) CD4+ cell count >500 cells/mm3 PROSPECTIVE COHORT: “AZT Short-course”
- 7. Participating sites Bobo Dioulasso . • Bobo-Dioulaso, Burkina-Faso Nairobi • Durban, South Africa Mombasa • Mtubatuba, South Africa • Nairobi, Kenya Mtubatuba • Mombasa, Kenya Durban
- 9. Infant HIV-free rates to 12 months of age.RCT, by study stratum 1.00 HIV-free survival of 0.95 Propotion not infected infants from mothers 0.90 with CD4 200-500 Short 0.85 Triple 0.80 Log rank test p = 0.017 0.75 (stratified on centre and intention to BF) 0 1 2 3 4 5 6 7 Age (in months) 8 9 10 11 12 Triple-ARV prophylaxis AZT short course Rate Rate Reduction (95% CI) (95% CI) Birth 2.8 (1.5-4.9) 3.0 (1.7-5.2) 7% 6 weeks 4.8 (3.1-7.4) 6.2 (4.2-9.1) 23% 6 months 8.4 (6.0-11.6) 12.6 (9.7-16.3) 33% 12 months 10.2 (7.6-13.6) 16.0 (12.7-20.0) 36%
- 10. Subgroup analysis: mothers with CD4 350-500 cells/uL Triple-ARV prophylaxis AZT short course Rate Rate Reduc (95% CI) (95% CI) -tion Birth 1.7 (0.6-5.2) 1.7 (0.5-5.1) 0% 6 weeks 2.9 (1.2-6.7) 3.4 (1.5-7.4) 15% 6 months 4.1 (2.0-8.3) 5.8 (3.1-10.5) 29% 12 months 4.7 (2.4-9.3) 7.1 (4.1-12.2) 34% Log rank test p = 0.33 (stratified on centre and intention to BF)
- 11. Results • Statistically significant reduction in HIV-free survival at 12 months among infants in “triple ARV prophylaxis” arm as compared to “AZT short course” • However, no statistically significant effect was seen among women with CD4 350-500 cells/mm3.
- 12. Participating Institutions Financial Support UNIVERSITY OF NAIROBI COLLEGE OF HEALTH SCIENCES SCHOOL OF MEDICINE
- 13. The Kesho Bora Team STUDY SITES STUDY COORDINATION Bobo Dioulasso, Burkina-Faso (Centre Muraz) World Health Organization, Switzerland • Nicolas Meda – Principal Investigator • Isabelle de Vincenzi – Study Coordinator • Paulin Fao, Clarisse Gouem – Study Coordinators • Philippe Gaillard – Site Coordinator • Paulin Somda, Hervé Hien, Elysée Ouedraogo – Investigators • Tim Farley – Project Manager • Diane Valea – Laboratory Coordinator • Ndema Habib – Statistician • Roselyne Toure – Data manager • Sihem Landoulsi – Data Management Durban, South Africa (University of KwaZulu-Natal) • Nigel Rollins, Kevi Naidu – Principal Investigators SUPPORTING INSTITUTIONS • Lynne McFetridge – Study Coordinator Université Montpellier 1, EA 4205 and CHU Montpellier, Laboratoire • Johannes Viljoen – Laboratory Coordinator de Bactériologie-Virologie, Montpellier, France Mombasa, Kenya (International Centre for Reproductive Health) • Philippe Van de Perre – Laboratory Coordination • Stanley Luchters, Marcel Reyners – Principal Investigators Centre Muraz, Burkina-Faso • Eunice Irungu – Study Coordinator • Francois Rouet – Site Laboratory Coordination and Quality Assurance • Christine Katingima, Mary Mwaura, Gina Ouattara – Investigators Institut de Recherche pour le Développement, France • Kishor Mandaliya – Laboratory Coordinator • Cecile Cames, Amandine Cournil, Kirsten Simondon – Nutrition • Mary Thiongo – Data manager Coordination Nairobi, Kenya (University of Nairobi) National Institute of Child Health and Human Development, National • Ruth Nduati – Principal Investigator Institutes of Health, USA • Judy Kose – Study Coordinator • Jennifer Read – Sponsor Representative and Co-investigator • Ephantus Njagi – Laboratory Coordinator Agence Nationale de Recherche sur le SIDA, France • Peter Mwaura – Data Manager • Brigitte Bazin, Claire Rekacewicz – Sponsor Representatives and Somkhele, South Africa (Africa Centre) Co-investigators • Marie-Louise Newell – Principal Investigator Centers for Disease Control and Prevention, USA • Steven Mepham, Thembi Blose – Study Coordinator • Mary Glenn Fowler, Denise Jamieson, Allan Taylor, Michael • Londiwe Mthethwa – Data Manager Thigpen – Sponsor Representatives and Co-investigators International Centre for Reproductive Health, Belgium • Patricia Claeys , Marleen Temmerman – Sponsor Representative and Co-investigator
- 14. WHO guideline development using GRC standards
- 15. World Health Organization • WHO has a mandate to: – define evidence-based global health norms and standards and – help countries adapt international recommendations according to their national circumstances
- 16. The problem with WHO guidelines before 2009 • Limited transparency about judgments • Limitations in evidence base – Review of evidence not always systematic – Often expert opinion • Long delay between time taken and when advice needed • Other issues (few country adaptations) Oxman et al, Lancet 2007;369:1883-9
- 17. WHO response • Revised WHO handbook for guidelines • Guideline Review Committee (GRC) • Standards for – Use of evidence – Processes – Reporting
- 18. Revised WHO recommendations (2010) Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants; Recommendations for a public health approach
- 20. Step 1-3 Scoping the document, group composition, conflict of interest • Identify need for policy adaptations: reasons for choosing the topic, problems with existing guidelines, new evidence emerged, etc • WHO convened an expert consultation in November 2008 to review new evidence since the 2006 guidelines • Core writing group (7 members) • 2 WHO • 2 guideline writers • 3 other experts
- 22. Step 4 - WHO prepared four key questions (PICOT) • When to start ART in pregnant women; and what regimen to give to pregnant women eligible for ART • When to start ARV prophylaxis for prevention of MTCT in pregnant women, and what regimen to give HIV-infected pregnant women for ARV prophylaxis? • What ARV prophylaxis regimen to give newborn infants born to HIV-infected mothers in the immediate postpartum period? • What regimen to use for preventing breastfeeding transmission of HIV beyond the immediate postpartum period?
- 24. Step 5 – Evidence retrieval, evaluation and synthesis • The HIV/AIDS Cochrane Collaborative Review Group search strategy was used for each of the four key questions – Protocol developed for each question – Systematic review of peer-reviewed literature, abstracts, conference proceedings and ongoing trials • For each of the selected studies, an assessment of the risk of bias using standard Cochrane format was completed
- 25. Step 5 – Evidence retrieval, evaluation and synthesis (continued) • Formal assessment of the quality of the evidence (GRADE) • GRADE is the Grading of Recommendations Assessment, Development and Evaluation • Review outcomes for each question as defined in the PICO tables as high, moderate, low or very low
- 27. Example of a GRADE table PICOT: Are triple-ARV regimens during pregnancy and continued during breastfeeding (intervention) more effective than regimens using antepartum AZT (comparator) for PMTCT among HIV-1 infected pregnant women (population)?
- 28. 1 eligible study? • Triple has been used for over a decade in developed countries. • Studies in developing countries available, but largely observational • Sometimes regimen not specified…
- 30. Step 6 - Risk-benefit profile • Risk-benefit tables – Draft recommendations (incl. strength) – Evidence (incl. quality) – Benefits and risks (AEs; drug resistance) – Values and acceptability (complexity; adherence) – Cost – Feasibility (integration)
- 31. Step 7 - Guideline review meeting • 2-day expert meeting (50 participants) • Final recommendations formulated, taking into consideration quality of evidence, balance between benefits and harms, balance between values and preferences, cost, feasibility, and related factors. • Final agreement on recommendations was achieved by active consensus of the participants in the guidelines meeting.
- 33. Step 8 – Additional feasibility assessment • Various meetings and presentations addressed the feasibility of potential new recommendations. – An informal two-day meeting was held in September 2009 with implementing agencies and other key stakeholders – rapid assessment of relevant country-specific policies and barriers (structured questionnaire to 12 selected WHO country offices) – Review of health systems issues of PMTCT programmes
- 34. Step 9 – Formulating additional research needs • Predominantly done by Guideline review meeting and the core writing group
- 35. Step 10 - Peer review • The summary recommendations and Grade Evidence Profile Tables were sent for peer review to six independent peer reviewers and the WHO regional offices.
- 36. Revised WHO recommendations (2010) Antiretroviral drugs for treating “ triple ARV prophylaxis" pregnant women and preventing (Option B) HIV infection in infants; Mother Recommendations for a public Maternal triple ARV prophylaxis starting from as early health approach as 14 weeks gestation and continued until delivery, or if breastfeeding, continued until one week after all infant exposure to breast milk has ended. Recommended regimens include: AZT+3TC+LPV/r or AZT+3TC+ABC or AZT+3TC+EFV or TDF+3TC(or FTC)+EFV Infant Daily NVP or twice daily AZT from birth until 4 to 6 weeks of age (irrespective of mode of infant feeding). "For the first time, WHO recommends that HIV-positive mothers or their infants take ARVs while breastfeeding to prevent HIV transmission"
- 37. Eligibility criteria for initiating antiretroviral treatment (ART) or prophylaxis in HIV-infected pregnant women based on CD4 cell count and WHO clinical stage.
- 38. Eligibility criteria for initiating antiretroviral treatment (ART) or prophylaxis in HIV-infected pregnant women based on CD4 cell count and WHO clinical stage.
- 39. The problem with WHO guidelines before 2009 • Limited transparency about judgments • Limitations in evidence base – Review of evidence not always systematic – Often expert opinion • Long delay between time taken and when advice needed • Other issues (few country adaptations) Oxman et al, Lancet 2007;369:1883-9
- 40. Conclusions • Transparency about judgments? • Evidence based recommendations? – Review of evidence systematic – yes – Often expert opinion – still yes • Time between need and availability reduced?
- 41. Conclusions (2) • Important to have standardized process for developing evidence-based guidelines • More formal consideration of observational and other relevant data would be helpful • Allow forward thinking recommendations • Countries are implementing the 2010 PMTCT guidelines (PNG is currently doing country adaptation)
- 42. Thank you
Editor's Notes
- #3: Many people believe that all babies of HIV positive women become infected with HIV – because they think that the mother and baby share the same blood when the baby is in the womb. In fact the mother ’s and the baby’s blood do not mix – although they come close together they are separated by a few layers of cells. This pictogram shows the average risk of MTCT – ie 35% - this comes from a meta-analysis by de Cock et al, JAMA 2000.
- #10: Major difference seen between 6 wks and 6 mths (6 infections in triple compared with 14 infections in short) Several infections after 6 mths in both arms (some mothers having not been able to stop breastfeeding by 6 months, and infants had continued exposure after mother had stopped ARVs)
- #11: Major difference seen between 6 wks and 6 mths (6 infections in triple compared with 14 infections in short) Several infections after 6 mths in both arms (some mothers having not been able to stop breastfeeding by 6 months, and infants had continued exposure after mother had stopped ARVs)
- #23: PICOT=Population, Intervention, Comparator, Outcome, Time Are regimens using triple-drug combinations in the antepartum, intrapartum and postpartum periods (intervention) for HIV-1 infected pregnant women (population) more effective than regimens using antepartum AZT without 3TC (comparator)?