Depot contraception and HIV: an exercise in obfuscationBMJ 2019; 367 doi: https://doi.org/10.1136/bmj.l5768 (Published 07 October 2019) Cite this as: BMJ 2019;367:l5768
All rapid responses
ECHO rings hollow: is DMPA safe or just last among equals?
Rupert Kaul1, Charu Kaushic2, Julie Lajoie3, Adam D. Burgener3,4, Robert Reinhard5, Refilwe Molatlhegi6, Sharon L. Achilles7,8, Keith R. Fowke4,9, Lyle R. McKinnon3,4,9,10
1 Departments of Medicine and Immunology, University of Toronto, Toronto, ON, Canada
2 McMaster University, Hamilton, ON, Canada
3 Department of Medical Microbiology and Infectious Diseases, University of Manitoba, Winnipeg, MB, Canada
4 JC Wilt Infectious Disease Research Centre, National Microbiology Lab, Public Health Agency of Canada, Winnipeg, MB, Canada
5 Public/Global Health Consultant
6 Department of Medical Microbiology, University of KwaZulu-Natal, Durban, South Africa
7 Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Pittsburgh, Pittsburgh, USA
8 Magee-Womens Research Institute, Pittsburgh, PA, USA
9 Department of Medical Microbiology, University of Nairobi, Nairobi, Kenya
10 Centre for the AIDS Programme of Research in South Africa (CAPRISA), Durban, South Africa
Corresponding author: Lyle McKinnon, Department of Medical Microbiology and Infectious Diseases, University of Manitoba, Room 504-745 Bannatyne Ave, Winnipeg, MB Canada R3E 0J9. Email: firstname.lastname@example.org
Recent publication of the ECHO study, designed to test the hypothesis that use of the injectable contraceptive depot medroxyprogesterone acetate (DMPA) was associated with increased HIV risk compared to other contraceptives, has unfortunately raised additional questions that preclude definitive interpretation of the impact of DMPA use on HIV acquisition risk in women. Others have eloquently discussed ethical issues around randomizing women to a drug that is hypothesized to cause harm. Here we focus on another major concern: the contraceptive methods against which DMPA was compared.
Access to safe and effective contraception is an integral component of women’s health. Understanding contraceptive safety in HIV-1 endemic areas is necessary; however, most data evaluating DMPA and HIV risk data are drawn from secondary cohort analyses, and may be confounded by alterations in sexual behaviour after contraceptive initiation. The ECHO study aimed to overcome this, in part, by open-label randomization of women to use of intramuscular DMPA, copper intrauterine device (IUD), or levonorgestrel (LNG) implant for contraception. Since ECHO was unblinded, the possibility of women changing their sexual behavior post-randomization remains a possible source of confounding. ECHO was powered to detect a 50% difference in HIV risk between any of the three method pairs. It follows then that if all three contraceptives were to increase HIV risk equally, then the trial would find no difference. Thus, the conclusion that the evaluated contraceptives did not differ statistically with respect to HIV risk could mean that they are all be safe, or that they all increase risk approximately equally. Given the high HIV incidence observed in all arms of ECHO, is it possible that they indeed all increased risk?
The aHR for HIV acquisition by women using DMPA was 1.50 in an individual participant data meta-analysis, possibly due to DMPA-mediated increases in highly-susceptible CD4+ T cells in the genital mucosa and/or disrupting epithelial barrier properties. There are far fewer data regarding HIV risk in women using the copper IUD. Although the mechanisms of action of the copper IUD are incompletely understood, historically this contraceptive is thought to prevent pregnancy by inducting inflammation. The current generation of copper-IUDs are considered safe and mostly non-controversial since the primary effect appears to be on sperm and prevention of fertilization [7,8]. However, IUD use was also associated with a substantial increase in bacterial vaginosis (BV)  and BV-associated inflammatory genital cytokines, both of which have been linked to HIV acquisition risk[11,12]. Contraceptive implants are relatively new additions to the contraceptive method mix in sub-Saharan Africa and thus there are few available associated data on HIV risk.
In ECHO, women randomized to the LNG implant did show a trend towards lower HIV acquisition compared to DMPA-IM (adjusted HR 1·29, 0·98–1·71, p=0·060), although the observed HIV incidence in this arm was also very high (3.31 per 100 person-years). We raise concern that two of the selected contraceptives studied in ECHO have documented mucosal effects that may be associated with increased HIV susceptibility [14,15], and the third contraceptive has scant data. We suggest that a major design flaw of ECHO is the lack of inclusion of a comparator contraceptive that has robust cohort data suggesting safety with respect to HIV risk. Combined oral contraceptives (COC) might have been a safe comparator since they demonstrate no risk for increased HIV risk, even though compliance monitoring of COC use would have been more challenging than injectables or IUD. Another option may have been norethindrone enanthate (Net-En) as an alternative injectable with some available cohort comparative data with DMPA demonstrating significantly lower risk of HIV acquisition.
Perhaps the most glaring design flaw is the lack of power to rule out a harmful effect on which to definitively make safety recommendations. Non-superiority is not the same as safety, and there is ample reason in ECHO to worry about type II error. The ~30% difference between DMPA and LNG is not that dissimilar from the effect size of 40% in the previous meta-analyses of cohort data[4,17], which is based on a much larger sample size compared to ECHO.
Although there were no significant differences in proportions of HIV infections observed in each arm of ECHO, a higher number of infections were observed in the DMPA-IM group. This was despite the fact that women in the DMPA-IM arm self-reported significantly lower HIV-risk behaviors when compared to women in either the copper IUD or LNG implant arms, including lower frequency of having multiple sex partners and engaging in unprotected sex. The trend to reduced HIV incidence in the LNG implant arm highlights important concerns regarding the design and analysis of ECHO, and in the subsequent public health implications of the results.
The ECHO results have been widely interpreted as demonstrating safety of all three contraceptive methods. This has already led to warnings around DMPA use being eliminated from messaging to women at risk of HIV acquisition. However, we argue that there are still grounds for concern that all contraceptive methods increased risk in this study, and so this may prevent much-needed future research into which contraceptives might be safer and the mechanisms by which current comparators might alter HIV risk. This in turn may lead to complacency among researchers working in this field. Safe contraception is critical for development in Africa, and one of the clearest messages from ECHO is that women in this region are interested and willing to use new contraceptive methods with which they have little familiarity. Perhaps one of our best defenses in the face of uncertainty around impact on HIV susceptibility is to increase the method mix, improve the quality of counseling, and empower women to make choices that work for them.
Overall, we are concerned that hasty interpretation of the ECHO outcome, in the absence of its acknowledged weaknesses, will “close the door” to finding alternatives to DMPA for safe and effective family planning in this region. The ECHO results also reinforce the need to effectively deliver HIV prevention, including newer biomedical strategies such as pre-exposure prophylaxis, while evaluating contraceptive efficacy.
1 Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium. HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet 2019;394:303–13. doi:10.1016/S0140-6736(19)31288-7
2 Sathyamala C. In the name of science: Ethical violations in the ECHO randomised trial. Glob Public Health 2019;182:1–16. doi:10.1080/17441692.2019.1634118
3 Hofmeyr GJ, Morrison CS, Baeten JM, et al. Rationale and design of a multi-center, open-label, randomised clinical trial comparing HIV incidence and contraceptive benefits in women using three commonly-used contraceptive methods (the ECHO study). Gates Open Res 2017;1:17. doi:10.12688/gatesopenres.12775.2
4 Morrison CS, Chen P-L, Kwok C, et al. Hormonal contraception and the risk of HIV acquisition: an individual participant data meta-analysis. PLoS Med 2015;12:e1001778. doi:10.1371/journal.pmed.1001778
5 Byrne EH, Anahtar MN, Cohen KE, et al. Association between injectable progestin-only contraceptives and HIV acquisition and HIV target cell frequency in the female genital tract in South African women: a prospective cohort study. Lancet Infect Dis 2016;16:441–8. doi:10.1016/S1473-3099(15)00429-6
6 Zalenskaya IA, Chandra N, Yousefieh N, et al. Use of contraceptive depot medroxyprogesterone acetate is associated with impaired cervicovaginal mucosal integrity. J Clin Invest 2018;128:4622–38. doi:10.1172/JCI120583
7 International Planned Parenthood Federation IPPF. International Medical Advisory Panel IMAP. Statement on intrauterine devices (IUDs). IPPF Med Bull 1987;21:3–5.
8 Ortiz ME, Croxatto HB, Bardin CW. Mechanisms of action of intrauterine devices. Obstet Gynecol Surv 1996;51:S42–51. doi:10.1097/00006254-199612000-00014
9 Achilles SL, Austin MN, Meyn LA, et al. Impact of contraceptive initiation on vaginal microbiota. Am J Obstet Gynecol 2018;218:622.e1–622.e10. doi:10.1016/j.ajog.2018.02.017
10 Sharma P, Shahabi K, Spitzer R, et al. Cervico-vaginal inflammatory cytokine alterations after intrauterine contraceptive device insertion: A pilot study. PLoS ONE 2018;13:e0207266. doi:10.1371/journal.pone.0207266
11 Low N, Chersich MF, Schmidlin K, et al. Intravaginal practices, bacterial vaginosis, and HIV infection in women: individual participant data meta-analysis. PLoS Med 2011;8:e1000416. doi:10.1371/journal.pmed.1000416
12 Masson L, Passmore J-AS, Liebenberg LJ, et al. Genital inflammation and the risk of HIV acquisition in women. Clin Infect Dis 2015;61:260–9. doi:10.1093/cid/civ298
13 Brown A, Harries J, Cooper D, et al. Perspectives on contraceptive implant use in women living with HIV in Cape Town, South Africa: a qualitative study among primary healthcare providers and stakeholders. BMC Public Health 2019;19:1003–9. doi:10.1186/s12889-019-7312-1
14 Shanmugasundaram U, Hilton JF, Critchfield JW, et al. Effects of the levonorgestrel-releasing intrauterine device on the immune microenvironment of the human cervix and endometrium. Am J Reprod Immunol 2016;76:137–48. doi:10.1111/aji.12535
15 Horcajadas JA, Sharkey AM, Catalano RD, et al. Effect of an intrauterine device on the gene expression profile of the endometrium. J Clin Endocrinol Metab 2006;91:3199–207. doi:10.1210/jc.2006-0430
16 Noguchi LM, Richardson BA, Baeten JM, et al. Risk of HIV-1 acquisition among women who use diff erent types of injectable progestin contraception in South Africa: a prospective cohort study. Lancet HIV 2015;2:e279–87. doi:10.1016/S2352-3018(15)00058-2
17 Polis CB, Curtis KM, Hannaford PC, et al. An updated systematic review of epidemiological evidence on hormonal contraceptive methods and HIV acquisition in women. AIDS 2016;30:2665–83. doi:10.1097/QAD.0000000000001228
Competing interests: No competing interests
Before posting my reaction to the BMJ-editorial by C. Sathyamala “Depot contraception and HIV: an exercise in obfuscation”, I have waited, expecting that the WHO would respond to Sathyamala’s contentions. And react in the only way considered scientifically responsible: by giving valid and convincing counter-arguments why the view in this editorial is invalid. But to my astonishment, no such response has come forward in the weeks that have followed. Now my waiting-time is up. Here is my belated ‘quick’ response. Was there not at least one (‘N=1’) learned person among the many scientists available to this huge and richly funded institution, being able to defend WHO’s honor?
Now I ask myself: HOW and WHY? HOW is it possible that such a fundamentally damaging editorial about the flaws of an important and far-reaching, hugely funded and clamorously propagated research-project of the WHO has not been countered? WHY is it that nobody from the research-team (ECHO) or from the WHO has come forward to explain why it is a responsible policy for WHO to make such far-reaching, highly-implicating announcements about the reliability and validity of this research-project, its findings and its implications?
Is it arrogance? Is it shame for involvement in a research-project that has resulted in a highly biased outcome-report that will have devastating effects on millions of women from the non-Western world, who are putting their trust in WHO? Is it fear of losing future funding by Bill Gates cum suis? Is it the rigid discipline of a ‘total institution’ that on questionable policy-grounds has simply decided that everybody concerned has to keep quiet? Or is it simply the absence of valid scientific arguments to put forward to Sathyamala’s critique? I’d say, what a pitiable performance by the WHO. What a way to squander one’s reputation: presenting knowingly this questionable analysis from second-rate research with such confidence.
Competing interests: No competing interests
Heidi E. Jones, PhD MPH,1 Erica Gollub, DrPH MPH,2 Janneke van de Wijgert, MD PhD MPH,3
1. Dept. of Epidemiology & Biostatistics, CUNY School of Public Health (New York, NY)
2. Health Science Program, College of Health Professions, Pace University (Pleasantville, NY)
3. Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht (Utrecht, Netherlands) & Institute of Infection and Global Health, University of Liverpool (Liverpool, UK)
C. Sathyamala argues that results from the ECHO trial  do not support the World Health Organization’s (WHO) downgrading of medical eligibility criteria for women at high risk of HIV for the use of depot medroxyprogesterone acetate (DMPA) from category 2 (benefits generally outweigh risks) to category 1 (no restrictions). We agree with her for the myriad reasons she and others responding to this commentary list, most notably: 1. The lack of a true comparison group in the ECHO trial as DMPA was compared to levonorgestrel (LNG) implant and copper intrauterine device (c-IUD)—two contraceptive methods for which there is little data on their impact on HIV risk; 2. Lack of masking which may have resulted in differential sex behaviors between arms, i.e. the benefits of a randomized control trial in preventing unmeasured confounding are largely derived from masking, which this open label study was not able to do; 3. Underpowering the study to detect a minimum effect size of 1.5, when three meta analyses had previously found a hazard ratio of 1.2-1.4 comparing DMPA to no hormonal contraceptive method use;[3-5] and 4. “Dichotomania” in interpretation of results, leading to a misleading interpretation of the HR of 1.29 (96% CI 0.98, 1.71) as not statistically significant and therefore not an important finding worthy of sharing with women to inform their decision making, when the finding suggests increased HIV risk for DMPA compared to LNG implant.
We would like to add that the voices of drug regulatory bodies have been noticeably absent in this debate. In the United States (US), a group of concerned scientists submitted a Citizens Petition to the US Food and Drug Administration (FDA) in May 2015 asking that the label to DMPA be updated to include in the “Warnings and Precaution” section that, “Depo-Provera may increase the risk of HIV infections (AIDS). Patients at risk of HIV infection (AIDS) should be counseled to only use Depo-Provera CI in combination with an HIV-prevention method.” The FDA acknowledged receipt of the petition in November 2015, stating, “FDA has been unable to reach a decision on your petition because it raises complex issues requiring extensive review and analysis by Agency officials.” This petition remains open today, more than four years later, despite increasing evidence for biologically plausible mechanisms. Based on the results comparing DMPA to the LNG implant in the ECHO trial, the most likely interpretation is that DMPA use increases HIV risk compared to the LNG implant. All of the data to date, combined with the results from the ECHO trial, should be considered sufficient information to warrant a warning on the DMPA label alerting women to the possibility of an increase in HIV risk when using this product.
1. Evidence for Contraceptive O, Consortium HIVOT. HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet. 2019;394(10195):303-13.
2. Sathyamala C. Depot contraception and HIV: an exercise in obfuscation. BMJ. 2019;367:l5768.
3. Morrison CS, Chen PL, Kwok C, Baeten JM, Brown J, Crook AM, et al. Hormonal contraception and the risk of HIV acquisition: an individual participant data meta-analysis. PLoS Med. 2015;12(1):e1001778.
4. Ralph LJ, McCoy SI, Shiu K, Padian NS. Hormonal contraceptive use and women's risk of HIV acquisition: a meta-analysis of observational studies. Lancet Infect Dis. 2015;15(2):181-9.
5. Polis CB, Curtis KM, Hannaford PC, Phillips SJ, Chipato T, Kiarie JN, et al. An updated systematic review of epidemiological evidence on hormonal contraceptive methods and HIV acquisition in women. AIDS. 2016;30(17):2665-83.
6. Jones HE. Require that the following warning be added to the prescribing infromation section labeled “”Sexually Transmitted Diseases” in the “Warnings and Precautions” section of Depo-ProveraL “Depo-Provera CI may increase the risk of HIV infections (AIDS). Patients at risk of HIV infection (AIDS) should be counseled to only use Depo-Provera CI in combination with an HIV-prevention method 2015 [Available from: https://www.regulations.gov/docket?D=FDA-2015-P-1900.
7. Hapgood JP, Kaushic C, Hel Z. Hormonal Contraception and HIV-1 Acquisition: Biological Mechanisms. Endocr Rev. 2018;39(1):36-78.
Competing interests: No competing interests
Majority of the general public are neither scientists nor statisticians. They have not been taught how to interpret p-values or confidence intervals and they have not been trained on how to critically analyse data. As scientists that is our job. Educating others about the world of science at a level that is understandable to all is crucial. If we want our research to positively impact other's lives and the environment, we need to communicate it in an honest and reliable manner.
The danger is misinterpretation and bad reporting.
The ECHO trial was well performed with rigorous analyses and statistics, but there were a number of limitations which is inherent to any trial. The ECHO trial authors concluded that they "'did not find a substantial difference in HIV risk among the methods evaluated". This is true if by "substantial" they mean a difference of "at least 50%". However, the results were misreported by a number of parties with claims that there was "no link" and "no difference between HIV-1 infection and type of contraceptive used."
This misinterpretation of the data may have severe health consequences for women who are unaware of the risks associated with the use of certain contraceptives. The reproductive health of women is vital. Women deserve the right to be properly informed with regard to the risks and benefits of different contraceptives before making their choices. As scientists it is our duty to be the voice of the voiceless and in a country where HIV-1 prevalence is so high, we need to make a stand for change.
Competing interests: No competing interests
We read with equal interest the first publication from the ECHO Trial that reported there were no significant differences in HIV acquisition among women using intramuscular depot medroxyprogesterone acetate (DMPA), levonorgestrel (LNG) implant, or copper intrauterine device (IUD) and the more recent editorial penned by Dr. Sathyamala passionately arguing for re-interpretation of these results [1, 2]. The most cogent opinion advanced in this editorial is simply that while the ECHO Trial publication will profoundly impact public health and family planning policies in countries where disproportionate numbers of young women are living with HIV, results and conclusions from the study generated important new questions and no definitive answers.
Indicative of current uncertainties, the ECHO Trial identified disparate sexual behavior in women randomized to DMPA vs. the long-acting reversible contraceptives. While study participants reported comparable behavior at enrollment, self-report of sexual behavior during follow-up identified significantly higher frequencies of multiple partners, new sex partners, and unprotected sex in women randomized to LNG implant and copper IUD groups than women randomized to DMPA (p-values < 0.001). Though authors concluded these dissimilar behaviors did not influence data interpretation, it seems possible that these dramatic between-group differences in high risk activity impacted rates of HIV transmission. Further clouding data interpretation from the ECHO Trial is the observation that about 13% of women enrolled had serum MPA levels denoting DMPA use in the prior 3 months. Although this recent use of DMPA excluded women from ECHO Trial enrollment and participation, it was unclear if data from these particular women was censored from data analyses.
In addition, the inappropriateness of randomizing a group of women in the ECHO trial to no contraceptive was a study design limitation that made identifying use of DMPA, LNG implant, or copper IUD as important HIV risk factors dependent on differential outcomes. In other words, ability of the ECHO Trial to precisely define risks associated with these contraceptive methods became more restricted when use of DMPA, LNG implant, and Cu-IUD by study participants similarly affected HIV susceptibility . This result is particularly relevant as effects of using the LNG implant and copper IUD on HIV transmission are much less explored than DMPA and essentially unknown. It also made it especially important to acknowledge the possibility that all three treatment groups comparably increased HIV susceptibility. Likewise, as regions such as sub-Saharan Africa, where adolescent girls and young women represent 10% of the population but 25% of new infections, any factor increasing HIV susceptibility less than the 30% increase the ECHO Trial was statistically powered to detect may still have devastating consequences. We therefore suggest that while it eventually may be identified as an accurate statement, it is an overly broad interpretation of the data to conclude at this point in time that ECHO trial results definitively support increased access to all three contraceptive methods examined .
1. Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium. HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet 2019; 394:303-313.
2. Sathyamala C. Depot contraception and HIV: an exercise in obfuscation. BMJ, 2019.
RDVM, NEQC, and TLC receive support from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (under Award Number R01HD094634). Content of this correspondence is solely the authors’ responsibility and does not necessarily represent official National Institutes of Health views. Authors have no competing financial interests to declare.
Competing interests: No competing interests
In her editorial, “Depot contraception and HIV: an exercise in obfuscation,” Dr. C. Sathyamala raises issues that are material to any sound scientific assessment of the findings from the ECHO trial. The three respondents raise additional valid and insightful comments with Dr. Hapgood providing evidentiary arguments.
There is a history to the premature guidance issued by the WHO committee that determines the MEC criteria. In 2007, citing a study conducted in Uganda and Zimbabwe that concluded that “no association was found between hormonal contraception and HIV acquisition overall”, (1) the WHO removed hormonal contraception from the risk category. Three years later, the lead authors of the study, who had documented the need to further examine the modifying effects of age, extended their analysis using marginal structural modeling. They then concluded: “We found that the association between hormonal contraceptive use and HIV risk was modified by age. Young women using DMPA and COCs (combined oral contraceptives) were at increased HIV risk compared to young women not using hormonal contraception.” (2) This underscores the importance of not instituting guidance for contraception based on preliminary findings alone, unsupported with more extended and rigorous analysis from a larger circle of HIV and population health researchers. The newly issued guidance endangers the lives of DMPA users, particularly teenage girls and young women. That is unacceptable.
1. Morrison CS, Richardson BA, Mmiro F, Chipato T, Centano DD, Luoto J, Mugerwa R, Padian N, Rugpao S, Brown JM, Cornelisse P, and Salata RA for the Hormonal Contraception and the Risk of HIV Acquisition (HC_HIV) Study Group AIDS (2007): 1(1): 85-95.
2. Morrison CS, Chen PL, Kwok C, Richardson BA, Chipato T, Mugerwa R, Byamugisha J, Padian N, Celentano DD, and Salata RA Hormonal contraception and HIV acquisition: reanalysis using marginal structural modeling. (2010) AIDS, 24(11): 1778-1781.
Competing interests: No competing interests
The editorial by Dr Sathyamala on the ECHO trial results raises multiple valid concerns. The issue regarding use of a control warrants further discussion. Understanding the relevance of the absence of a control in the ECHO trial is a key point in interpretation of the results and their implications for public health. Is it only possible to incorrectly claim that the ECHO trial showed that DMPA is not linked to increased HIV acquisition if misinterpretation or obfuscation are involved.
The authors concluded that: “We did not find a substantial difference in HIV risk among the methods evaluated, and all methods were safe and highly effective” (1). The ECHO trial was not a randomly controlled trial, due to the absence of a placebo or a no contraception or an infrequent condom use control (1). The authors do not claim that the LNG implant or Cu-IUD were used as controls. However, this may have been incorrectly inferred by many and some of the statement made in the Lancet paper may have been interpreted as such (1). It is unknown whether LNG implant or Cu-IUD increase HIV risk relative to no contraception or infrequent condom use due to very limited and uncertain data (2-4). However, published data do suggest several plausible biological mechanisms whereby they may increase HIV acquisition (5-7). For Cu-IUD, concerns regarding its inflammatory effects are reinforced by its current classification in the WHO MEC risk 2/3/4 categories for initiation of use for women at increased risk for STIs (3). Increased inflammation is associated with increased HIV acquisition (5). Multiple mechanisms are likely to have an effect on HIV acquisition (5) and each of the three methods used in the ECHO trial could theoretically contribute to any one or more of these mechanisms. The control groups for the previous observational studies showing that DMPA is associated with a 40-50% increased risk of HIV acquisition either involved no contraception in some studies, or included some women using non-hormonal contraception, which included non-hormonal intrauterine devices making up only at most about 1% of the participants in the control group in other studies (2,8,9).
In a bizarre resort to speculation, the latest WHO guidelines suggest that there is indirect evidence that DMPA is not linked to increased HIV-1 risk by noting that the high incidence of HIV infection experienced by each contraceptive group during the ECHO trial was similar to the background incidence assumed when designing the trial (4). This argument is fundamentally flawed for multiple reasons.
Background incidence in the ECHO trial would have been impossible to predict with this level of accuracy. Baseline data would not have been available for the exact same participants as in the ECHO trial. Furthermore, HIV incidence is known to decrease during a trial and would have likely decreased to unpredictable levels in the ECHO trial given the extensive efforts made to provide HIV prevention counselling and interventions resulting in an unusually high level of adherence. The ECHO study was designed to determine whether or not a difference between HIV risk of more than 50% occurs between any two of the three methods of contraception. It was not powered to address differences less than 50% and was not designed nor could it address the question of HIV risk associated with any one method relative to no contraception or infrequent condom use. It should be noted that even if Cu-IUD or LNG implant are associated with, for example, a 10% increased HIV risk, and if DMPA is associated with a 40-50% increased risk, the relative risk between DMPA and the other method would be 27-36%, i.e. far below the 50% threshold.
No conclusions can be drawn from the ECHO trial on whether or not DMPA is associated with increased risk of HIV acquisition relative to no contraception or infrequent condom use, and the ECHO results do not refute the previous observational data showing a 40-50% increased HIV risk associated with DMPA.
1. Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium. (2019) HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet 394, 303-313
2. Polis, C. B., Curtis, K. M., Hannaford, P. C., Phillips, S. J., Chipato, T., Kiarie, J. N., Westreich, D. J., and Steyn, P. S. (2016) An updated systematic review of epidemiological evidence on hormonal contraceptive methods and HIV acquisition in women. Aids 30, 2665-2683
3. WHO. (2015) Medical eligibility criteria for contraceptive use, 5th edition. World Health Organisation, Geneva, Switzerland
4. WHO. (2019) Contraceptive eligibility for women at high risk of HIV. Guidance statement: recommendations on contraceptive methods used by women at high risk of HIV. World Health Organisation, Switzerland
5. Hapgood, J. P., Kaushic, C., and Hel, Z. (2018) Hormonal Contraception and HIV-1 Acquisition: Biological Mechanisms. Endocrine reviews 39, 36-78
6. Quispe Calla, N. E., Vicetti Miguel, R. D., Boyaka, P. N., Hall-Stoodley, L., Kaur, B., Trout, W., Pavelko, S. D., and Cherpes, T. L. (2016) Medroxyprogesterone acetate and levonorgestrel increase genital mucosal permeability and enhance susceptibility to genital herpes simplex virus type 2 infection. Mucosal Immunol 9, 1571-1583
7. Achilles, S. L., Austin, M. N., Meyn, L. A., Mhlanga, F., Chirenje, Z. M., and Hillier, S. L. (2018) Impact of contraceptive initiation on vaginal microbiota. Am J Obstet Gynecol 218, 622 e621-622 e610
8. Morrison, C. S., Chen, P. L., Kwok, C., Baeten, J. M., Brown, J., Crook, A. M., Van Damme, L., Delany-Moretlwe, S., Francis, S. C., Friedland, B. A., Hayes, R. J., Heffron, R., Kapiga, S., Karim, Q. A., Karpoff, S., Kaul, R., McClelland, R. S., McCormack, S., McGrath, N., Myer, L., Rees, H., van der Straten, A., Watson-Jones, D., van de Wijgert, J. H., Stalter, R., and Low, N. (2015) Hormonal contraception and the risk of HIV acquisition: an individual participant data meta-analysis. PLoS medicine 12, e1001778
9. Crook, A. M., Ford, D., Gafos, M., Hayes, R., Kamali, A., Kapiga, S., Nunn, A., Chisembele, M., Ramjee, G., Rees, H., and McCormack, S. (2014) Injectable and oral contraceptives and risk of HIV acquisition in women: an analysis of data from the MDP301 trial. Human reproduction 29, 1810-1817
Competing interests: No competing interests
This is an important article which outlines many of the concerns I have had around the interpretation of the results of the ECHO trial. Dr Sathyamala has a wealth of medical and epidemiological expertise that make her highly qualified to speak on this issue. To-date, discussion of ECHO has not engaged critically with the design of the trial, in particular, the weaknesses that result from choosing to statistically power the study to detect only a 50% percent difference in the relative risks between the different contraceptive methods. As Dr Sathyamala outlines, the authors of the ECHO results paper focus their attention on p-values in a manner that is over-simplified, and the discussion of their paper is over-interpreted and does not acknowledge the nuance in the results. I welcome this paper by Dr Sathyamala and hope that it will stimulate greater critical engagement with the ECHO trial.
Competing interests: No competing interests
In my view this is an apposite and well-crafted piece by a respected medical doctor of many years outstanding practice, and one of the world's most persistent and careful analysts of the impact of depot medroxyprogesterone (known as "DEPO" across Southern Africa), specifically, on the health of women in India. Even if you do not agree with her analysis I think you can agree with the overarching point made by Dr Sathyamala here -- that the way the ECHO trial results are being used and reported is flawed. In the view of many of us working in historical/ anthropological / sociological and philosophical fields, who support science, and work to build confidence in science into our teaching and work, while maintaining a critical attitude to the many perverting forms of power around the scientific endeavour, and who are following ECHO very closely, the outflow from this trial could lead to suspicion about the way in which scientific knowledge gets used, especially in the fields of clinical research vitally important to women on the African continent. By over stating results and indulging in hyperbolic result-launches -- in the style of business marketing events -- and by exaggerating the strength of trial findings, or aiding and abetting news organisations doing the same, the suspicions of many social scientists, civic organisations and the interested public, are once again raised concerning the power of money and status and unstated objectives surrounding many clinical trials. Even as we know that complex and important trials, like this one, involve so many dedicated and ethical scientists and researchers, we also know they cannot always control the use of their work after its completion. I urge scientists and clinical researchers to take a very close look at the trial and the claims being made from it.
Competing interests: No competing interests