Conditional cash transfer programs are gaining steam as an intervention to prevent HIV infection. Under these programs, health authorities offer people money to increase incentives for healthy behaviors. For instance, in Malawi, the World Bank recently tested a program for HIV prevention among young women. In this program, if you were a girl randomized to the intervention arm, your parents received $4-10/month and, if you attended school regularly, you received $1-5. In the study, the intervention cut HIV incidence by more than half. Additional studies testing similar interventions are underway.
These interventions are being considered for large-scale rollout. They're also pretty controversial: USAID considered the topic as part of its debate series on emerging issues in HIV response. In particular, opponents object to the conditioning of the cash transfer. Inevitably somebody will use the word "bribe" to describe what happens: some people will alter their behavior in order to obtain the incentive. (If the incentive is unconditional, then the mechanism of action seems a little different: perhaps access to greater resources reduces young women's dependence on men, a likely reason that some of them engage in risky sexual behaviors.) But suppose that the program is cost-effective (or even cost-saving)--what's wrong with using money to get people to do what you want them to?
Well, we can start by looking at a program that looks pretty bad to most people: the attempt by a U.S.-based nonprofit to increase contraceptive use in Kenya by offering HIV-positive women $40 to have an intrauterine device implanted. The founder of the program, Barbara Harris--who gained notoriety in the '90s for offering $300 to drug-using California women to be sterilized--says the goal of the program is to reduce the transmission of HIV to newborns. An observer says this program is "coercive and discriminatory." Can we distinguish it from other conditional cash transfer programs?
Wednesday, June 8, 2011
Monday, June 6, 2011
HPTN 052 results and Partners PrEP (part II)
...is Partners PrEP unethical?
If HIV prevention research requires maximal risk reduction for all participants, then it appears unethical based on the 052 data. Maximal risk reduction would involve treating all of the HIV+ partners in the study, even if doing so were inconsistent with answering the study question.
For those who do not take a hard line on maximal risk reduction in HIV prevention, a study offering submaximal risk reduction might be justified by the value of the study question (and its methodological incompatibility with providing maximal risk reduction). Arguably, HPTN 052 undermines the value of Partners PrEP data: why would we pursue a PrEP strategy--giving ARVs to the HIV- partner for prevention--when giving the ARVs to the HIV+ partner virtually eliminates transmission? Is PrEP a viable strategy in serodiscordant couples?
Some arguments for why it might still be:
1. in some couples, the HIV+ partner will decline therapy or will not take ARVs regularly. In these couples, a method controlled by the HIV- partner has value.
2. the HIV- partner may have risk factors other than sexual contact with the stable romantic partner (e.g. concurrent relationships). PrEP could be a viable strategy for protecting these individuals.
3. outside of stable serodiscordant couples, there are high-risk individuals (e.g. sex workers) who would be candidates for a PrEP strategy. Partners PrEP could provide scientific support for this strategy outside of serodiscordant couples.
Here's one thing we can say about these arguments: for some couples in Partners PrEP, none of the arguments will offer a reason why that couple is better off. Whenever this is true, then potential benefit to the study participant is being traded in favor of scientific knowledge and/or public health benefit. So we are accepting that some couples are probably left worse off by the Partners regimen (especially in the control group) than they would be on the 052 regimen.
Another important (?) factor: Partners PrEP meets or exceeds local standards of care in the communities where it is being conducted. If maximal risk reduction represents one end of the spectrum, the opposite end of the spectrum is the view that local standards of care represent the baseline or minimal standard for research. For those who are sympathetic to this view, the 052 data would not disturb the assumption that Partners PrEP should continue as designed, unless it suddenly appeared that Partners PrEP had no value from a public health standpoint.
So perhaps the question is where the default lies. If the maximal risk reduction is the default, then the ongoing value of Partners PrEP is probably not compelling enough to justify the deviation (assuming that we're willing to allow any deviations). If local standards are the default, then the ongoing value of Partners PrEP is probably sufficient to justify continuation as designed. If we are disposed towards a moderate view--somewhere between the two extremes--it will be tougher to determine whether Partners PrEP should continue as designed.
If HIV prevention research requires maximal risk reduction for all participants, then it appears unethical based on the 052 data. Maximal risk reduction would involve treating all of the HIV+ partners in the study, even if doing so were inconsistent with answering the study question.
For those who do not take a hard line on maximal risk reduction in HIV prevention, a study offering submaximal risk reduction might be justified by the value of the study question (and its methodological incompatibility with providing maximal risk reduction). Arguably, HPTN 052 undermines the value of Partners PrEP data: why would we pursue a PrEP strategy--giving ARVs to the HIV- partner for prevention--when giving the ARVs to the HIV+ partner virtually eliminates transmission? Is PrEP a viable strategy in serodiscordant couples?
Some arguments for why it might still be:
1. in some couples, the HIV+ partner will decline therapy or will not take ARVs regularly. In these couples, a method controlled by the HIV- partner has value.
2. the HIV- partner may have risk factors other than sexual contact with the stable romantic partner (e.g. concurrent relationships). PrEP could be a viable strategy for protecting these individuals.
3. outside of stable serodiscordant couples, there are high-risk individuals (e.g. sex workers) who would be candidates for a PrEP strategy. Partners PrEP could provide scientific support for this strategy outside of serodiscordant couples.
Here's one thing we can say about these arguments: for some couples in Partners PrEP, none of the arguments will offer a reason why that couple is better off. Whenever this is true, then potential benefit to the study participant is being traded in favor of scientific knowledge and/or public health benefit. So we are accepting that some couples are probably left worse off by the Partners regimen (especially in the control group) than they would be on the 052 regimen.
Another important (?) factor: Partners PrEP meets or exceeds local standards of care in the communities where it is being conducted. If maximal risk reduction represents one end of the spectrum, the opposite end of the spectrum is the view that local standards of care represent the baseline or minimal standard for research. For those who are sympathetic to this view, the 052 data would not disturb the assumption that Partners PrEP should continue as designed, unless it suddenly appeared that Partners PrEP had no value from a public health standpoint.
So perhaps the question is where the default lies. If the maximal risk reduction is the default, then the ongoing value of Partners PrEP is probably not compelling enough to justify the deviation (assuming that we're willing to allow any deviations). If local standards are the default, then the ongoing value of Partners PrEP is probably sufficient to justify continuation as designed. If we are disposed towards a moderate view--somewhere between the two extremes--it will be tougher to determine whether Partners PrEP should continue as designed.
Implications of HPTN 052 for standards of prevention in serodiscordant couples (part I)
HPTN 052 recently found that, in stable relationships consisting of an HIV positive partner and an HIV negative partner ("serodiscordant couples"), early antiretroviral treatment of the HIV+ partner dramatically reduces transmission of the virus. These study data support the idea that "test and treat"--identifying and treating HIV+ individuals as early as possible, post-infection--will reduce HIV transmission rates on a community level. Of course, an important question is whether the will and resources exist even to implement this strategy in the U.S., let alone in lower-resource settings where HIV incidence is higher.
While the public health implications of this research raise numerous ethical issues (which may be considered later on this blog), here's a research ethics question: among serodiscordant couples, does early treatment of the HIV+ partner now represent the correct standard of care? This question could affect other studies among serodiscordant couples, especially the Gates Foundation's Partners PrEP study. In Partners PrEP, SD couples are randomized such that in two thirds of the couples, the HIV- partner receives oral PrEP--that is, taking antiretroviral pills for prophylaxis. In the control group, the HIV- partner receives a placebo. In all of these couples, the HIV+ partner was enrolled at a CD4 count where he/she was not eligible for ART under national guidelines (although the couple stays on-study if the HIV+ partner becomes eligible for ART and begins taking it).
Therefore, many Partners PrEP participants, whose CD4 counts are in the range where 052 found a 96% reduction in transmission from treating the HIV+ partner, are not receiving ARVs. In the experimental groups of the study, the HIV- partner is receiving an intervention which has shown effectiveness in preventing incidence among men who have sex with men (the iPrEx study) but not among heterosexual women (FHI stopped the FEM-PrEP study in April due to futility). The placebo group will be receiving no intervention.
This issue represents a twist on the "standards of prevention" problem that is vexing HIV prevention researchers: once a new modality is known to be effective for HIV prevention, is it ethically acceptable to conduct a study which withholds that modality from any study participants? The Declaration of Helsinki seems to prohibit this practice, and recent HIV prevention trials have provided a comprehensive "background package" of all known effective interventions (condoms, risk reduction counseling, etc.). So, suppose WHO issues a strong recommendation based on the 052 data, recommending that serodiscordant couples should start early treatment of the HIV+ partner whenever possible. Is Partners PrEP unethical?
While the public health implications of this research raise numerous ethical issues (which may be considered later on this blog), here's a research ethics question: among serodiscordant couples, does early treatment of the HIV+ partner now represent the correct standard of care? This question could affect other studies among serodiscordant couples, especially the Gates Foundation's Partners PrEP study. In Partners PrEP, SD couples are randomized such that in two thirds of the couples, the HIV- partner receives oral PrEP--that is, taking antiretroviral pills for prophylaxis. In the control group, the HIV- partner receives a placebo. In all of these couples, the HIV+ partner was enrolled at a CD4 count where he/she was not eligible for ART under national guidelines (although the couple stays on-study if the HIV+ partner becomes eligible for ART and begins taking it).
Therefore, many Partners PrEP participants, whose CD4 counts are in the range where 052 found a 96% reduction in transmission from treating the HIV+ partner, are not receiving ARVs. In the experimental groups of the study, the HIV- partner is receiving an intervention which has shown effectiveness in preventing incidence among men who have sex with men (the iPrEx study) but not among heterosexual women (FHI stopped the FEM-PrEP study in April due to futility). The placebo group will be receiving no intervention.
This issue represents a twist on the "standards of prevention" problem that is vexing HIV prevention researchers: once a new modality is known to be effective for HIV prevention, is it ethically acceptable to conduct a study which withholds that modality from any study participants? The Declaration of Helsinki seems to prohibit this practice, and recent HIV prevention trials have provided a comprehensive "background package" of all known effective interventions (condoms, risk reduction counseling, etc.). So, suppose WHO issues a strong recommendation based on the 052 data, recommending that serodiscordant couples should start early treatment of the HIV+ partner whenever possible. Is Partners PrEP unethical?
Subscribe to:
Posts (Atom)