1 / 55

Unmet need and safety of hormonal contraceptive use in HIV+ women in Rakai, Uganda

Unmet need and safety of hormonal contraceptive use in HIV+ women in Rakai, Uganda. Chelsea Polis and Ron Gray PFRH Seminar October 14, 2009. Rakai Health Sciences Program. The Rakai Community Cohort (RCCS) 1994-2009. Surveillance of >12,000 adults aged 15-49

corentine
Télécharger la présentation

Unmet need and safety of hormonal contraceptive use in HIV+ women in Rakai, Uganda

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Unmet need and safety of hormonal contraceptive use in HIV+ women in Rakai, Uganda Chelsea Polis and Ron Gray PFRH Seminar October 14, 2009

  2. Rakai Health Sciences Program

  3. The Rakai Community Cohort (RCCS) 1994-2009 • Surveillance of >12,000 adults aged 15-49 • in 50 rural communities with annual: • Census • Interviews (sociodemographic, behaviors and health characteristics including family planning) • Sample collection for HIV, STI and other infections • Open cohort Rural Trading Center Towns

  4. Data Collection GPS Mapping Identification number Photo ID Interviews in private Blood collection

  5. Contraceptive use in HIV+ women

  6. For HIV+ women who wish to prevent pregnancy, contraception helps… • Avoid unintended pregnancies, pregnancy complications, and adverse birth outcomes • Maintain access to a wider selection of antiretroviral drugs • Avoid pregnancy/breastfeeding, which may accelerate HIV disease progression

  7. Contraception has been called “The best kept secret in PMTCT” • Prevention of perinatal and lactational HIV • As cost-effective* as ART prophylaxis • Avoids concerns about viral mutation and resistance • Reduction in number of AIDS orphans * Reynolds 2006 Sexually Transmitted Diseases

  8. WHO comprehensive strategy to reduce HIV in infants Primary HIV prevention in women of childbearing age Preventing unintended pregnancies to women living with HIV Prevention of MTCT (ART, safer delivery, feeding counseling) Treatment, care and support for HIV+ women and their families

  9. Estimated impact of combined strategies on percent reduction of HIV+ infants Slide courtesy of Michael Sweat

  10. Why focus on hormonal contraception?

  11. Can HIV+ women safely use hormonal contraception?

  12. A randomized controlled trial caused concern, but had limitations… • High crossover • High dropout rate • Not based on a priori hypothesis • No control for baseline differences in disease stage HR: 1.6 (1.04-2.3) HC IUD Stringer 2007, AJOG

  13. What other evidence exists on the effect of HC on HIV progression? ? HC accelerates HIV progression 3 observational studies (+ RCT) HC has no effect on HIV progression 6 observational studies HC is protective against HIV progression 1 observational study Effect of HC depends on time 1 observational study

  14. WHO meeting, March 2007 “There is enormous need to have some information to work with right now to answer the question: what is the best contraceptive choice for HIV+ women?“

  15. Rakai data overcomes many limitations of previous studies • Population-based • Largest sample of incident HIV seroconverters ever analyzed (n=625) • Over 10 years of follow-up time • Yearly information on contraceptive use • Well-established research infrastructure • Information available immediately to address a pressing issue

  16. Objectives

  17. Overview of Dissertation Aims Aim 1: Describe trends in and predictors of HC use among HIV+ women. Aim 2: Assess the effect of HC use on HIV progression in terms of clinical outcomes, specifically time from infection to death. Aim 3: Assess the effect of HC use on HIV progression in terms of immunologic outcomes, specifically time from infection to AIDS or death.

  18. Methods

  19. Aim 1: Methods • Calculated prevalence of HC use from 1994-2006 among all HIV+ women at each round • Pooled data using GEE and controlled for within-individual correlation. Used logistic regression to obtain ORs for predictors of HC use at 3 time intervals.

  20. Aims 2 and 3: Methods • 625 incident seroconverters contributed 1294 interviews between 1994 and 2006 • Kaplan-Meier and Cox regression compared women by HC use on survival • Time-varying exposure and lagged exposure (HC use at previous round)

  21. Aims 2 and 3: Methods (cont.) • Hormonal users compared against: • (1) All women not using HC • (2) Contraceptors using non-hormonal methods • (3) Non-contraceptors • Censored at availability of ART in 2004 • Considered OCs and DMPA separately

  22. ResultsAim 1Aim 2Aim3

  23. Trends in HC use in Rakai, 1994-2006

  24. Prevalent pregnancies in HIV+ women in Rakai, 1994-2006

  25. Proportion of unintended pregnancies to HIV+ women is on the rise… Pregnancy intention info unavailable Rounds 1-6

  26. Predictors of HC use in HIV+ women in Rakai over time (1994-2006) * Information on this variable not collected at all rounds, but always associated with HC use at rounds when it was collected.

  27. ResultsAim 1Aim 2Aim3

  28. HC use borderline sig assoc. with reduced mortality in time-varying K-M analysis Median survival: 7.06 years (IQR 4.86-9.55) HC No HC P-value for logrank: 0.067

  29. Multivariate analyses…HC not associated with increased hazard of death overall • Time-varying exposure • HR: 0.72 (0.39-1.32), p=0.293 • Lagged exposure • HR: 0.86 (0.46-1.61), p=0.645 Analyses controlled for education, number of sex partners, breastfeeding , and time period

  30. Multivariate analysis…HC equivalent to non-contracepting or using a non-hormonal method • vs. non-contraceptors • HR: 0.68 (0.36-1.27), p=0.222 • vs. non-hormonal contraceptors • HR: 0.72 (0.33-1.59), p=0.419 Analyses controlled for education, number of sex partners, breastfeeding , and time period

  31. ResultsAim 1Aim 2Aim 3

  32. HC use associated with reduced HIV disease progression in time-varying K-M analysis p-value for logrank test: 0.03 Median survival: 4.45 years (IQR 2.37-6.33) HC No HC

  33. Multivariate analyses…HC still assoc. with reduced hazards of progression(particularly after median survival) • Time-varying exposure • HR: 0.70 (0.50-0.97), p=0.03 • >4.4 yrs: HR: 0.46 (0.27-0.81), p=0.01 • Lagged exposure • HR: 0.76 (0.50-1.14), p=0.19 Analyses controlled for SES and number of sex partners in past year

  34. Multivariate analyses…HC still assoc with reduced hazards of progression • vs. non-contraceptors • HR: 0.68 (0.49-0.96), p=0.03 • vs. non-hormonal contraceptors • HR: 0.73 (0.47-1.13), p=0.16 Analyses controlled for SES and number of sex partners in past year

  35. Subgroup analysis among 128 women with baseline CD4 counts • Small, non-significant differences in baseline CD4 counts between never users and ever users • Controlling for baseline CD4 count still suggested a significant association of HC with reduced risk of progression, despite a much smaller sample size • Adjusted HR 0.22 (0.06-0.82), p=0.02

  36. Conclusions and Implications

  37. Conclusions AIM 1 • HC use is increasing among HIV+ women in Rakai, but prevalent pregnancies did not decline, and unintended pregnancies are rising. • There has been little change over time in terms of predictors of HC use among HIV+ women in Rakai

  38. Conclusions AIMS 2 and 3 • Our findings do not support the concern that HC accelerates HIV progression • HC users and non-users had similar hazards of mortality • Potential association between HC use and lower hazards of progression to AIDS or death • Caveat: difficult to control for the possibility that healthier women selectively use HC

  39. Implications • HIV+ women need safe, effective FP methods they can use consistently • HC appears safe for HIV+ women • Stronger integration of FP and HIV is needed

  40. Thank you! • Ron Gray • Maria Wawer • Heena Brahmbhatt • Defense committee members • Rakai Health Sciences Program staff • Rakai cohort participants

  41. Supplemental slides

  42. Censoring at ART availability • HC may decrease the efficacy of ART, which would differentially accelerate disease progression after ART initiation in HC users • Time to death • HR: 0.64 (0.27-1.53), p=0.315 • Time to AIDS or death • HR: 0.36 (0.19-0.69), p=0.00

  43. Trends in contraceptive use in Rakai

  44. Theorized mechanisms for why HC might slow HIV progression • Avoidance of pregnancy and/or breastfeeding • Seems possible given that differences seemed larger between HC vs. non-contraceptors than between HC vs. non-hormonal contraceptors • Recent study suggested that pregnancy and bfding at SC are associated with higher plasma viral setpoint

  45. Does pregnancy influence HIV progression? • Early studies suggested yes, but poorly conducted • Systematic review found weak association, potential for bias too great to draw conclusions • Most studies (developed world) find no effect or a protective effect (but…fertility, health care?) • Pregnancy’s potential protective effect may be lost in developing world, burden of maternal mortality/morbidity • 5 studies in developing countries, 3 found no effect, 2 found acceleration • Also need to consider effects of breastfeeding on HIV progression in developing countries

  46. Does breastfeeding influence HIV progression? • Nduati 2001 suggested detrimental impact of breastfeedingon maternal health • Several subsequent studies & a meta-analysis suggested no impact (Taha, Coutsoudis, Kuhn, Sedgh) • Recent study (Otieno 2007) suggested BMI and CD4 deteriorated faster in bfding women, but no difference in VL or mortality • Probably no significant effect

  47. Loss to follow up • How to define? • Date of death could be acquired outside of cohort, so even if missing a few years, not necessarily “lost” • Analysis used multiple records, so even if LTFU, they still contributed information during certain periods • LTFU = did not die and did not provide any information during rounds 10 or 11 • 137/625 (21.9%)

  48. Loss to follow up, cont’d • LTFU women similar to others at baseline in terms of: • Current age, age at seroconversion, CD4 count, education, SES, parity, non-marital relationships, number of sex partners, current pregnancy, sex work, condom use, contraceptive use • LTFU were: • More likely to be breastfeeding, younger, and previously married at baseline

More Related