Antiviral Therapy 2015; 20:613–621 (doi: 10.3851/IMP2956)
Original article An economic evaluation of conception strategies for heterosexual serodiscordant couples where the male partner is HIV-positive Michelle Letchumanan1,2, Peter C Coyte2, Mona Loutfy1,2,3,4* Women’s College Research Institute, Women’s College Hospital, Toronto, ON, Canada Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, ON, Canada 3 Faculty of Medicine, University of Toronto, Toronto, ON, Canada 4 St Michael’s Hospital, Toronto, ON, Canada 1 2
*Corresponding author e-mail: [email protected]
Background: To conduct an economic evaluation of the three commonly used interventions that reduce sexual HIV transmission when an HIV-negative female aims to conceive with an HIV-positive male on combination antiretroviral therapy (condomless sex restricted to timed ovulation [CS], sperm washing with intrauterine insemination [SW] and condomless sex restricted to timed ovulation with pre-exposure prophylaxis [CS-PrEP]). As SW and CS-PrEP are only privately available for pregnancy planning for this population in Canada, this study was conducted to inform policy decisions concerning potential public health insurance coverage, as well as to inform fertility counselling in settings with adequate combination antiretroviral therapy access globally. Methods: We developed a cohort Markov model with a lifetime horizon and used the perspective of Ontario’s Ministry of Health (MOH). Input parameters were drawn from literature, the MOH’s Schedule of Benefits and a
time trade-off questionnaire designed for this study. Outcome measures included quality-adjusted life-years and incremental cost-effectiveness. Costs and benefits were discounted at annual rates of 3%. Costs were reported in Canadian 2013 dollars and an exchange rate of 1 USD to 1.066 CND was applied where necessary. Sensitivity analysis assessed the uncertainty of model parameters. Results: The base case analysis found that CS-PrEP and SW were each more costly and less effective at conception than CS. The results were robust in the sensitivity analysis and suggest that CS is the dominant conception strategy in this population. Conclusions: Neither CS-PrEP nor SW represent better value for money relative to CS as a conception option for HIV-discordant couples with positive male partners. Based on these findings, CS-PrEP and SW cannot be recommended for public-funding in developed countries.
Introduction The dramatic improvements in quality and length of life for people with HIV due to highly effective combination antiretroviral therapy (cART) have brought about the prioritization of their fertility desires . To achieve natural conception, however, the exchange of bodily fluids between serodiscordant couples, that is, an HIV-positive individual on cART paired with HIVnegative partner, poses the risk of HIV transmission to the uninfected partner as well as the potential cost of care for their HIV management. If horizontal transmission occurs during pregnancy in a heterosexual couple where a man is HIV positive and the female partner is HIV negative, there is the additional risk of vertical ©2015 International Medical Press 1359-6535 (print) 2040-2058 (online)
transmission to the fetus. In developed health-care settings, like Canada, where provincial governments subsidize HIV medication, in addition to the provision of public health services, the lifetime health-care costs of a novel HIV infection can be approximately $250,000 CND. Additional HIV cases would add to constrained health-care budgets and create psychosocial burdens for the infected individuals . Despite the potential risk of horizontal HIV transmission, fertility services in developed countries are limited for HIV-discordant couples. For example, only five out of ten Canadian provinces readily treat HIV-affected couples with specialized fertility care to minimize the likelihood of HIV 613
M Letchumanan et al.
transmission . An even smaller fraction of these services are tailored to cART-treated, HIV-positive males partnered with HIV-negative females and are costly, out-of-pocket endeavors . Without public insurance in place to support safe conception, HIV-negative women are at risk of HIV contraction during attempts to conceive. Specialized reproductive strategies that are recommended for heterosexual serodiscordant couples with HIV-positive male partners on cART report varying safety and effectiveness profiles [4,5]. By European clinical standards, the main method used to conceive in this setting is intrauterine insemination (IUI) with washed sperm since its introduction in 1992 . A meta-analysis of 3,900 cycles of IUI with washed sperm in 1,184 couples found zero HIV seroconversions and a pregnancy rate of 18% per cycle . During IUI with washed sperm, the uninfected female is exposed to isolated sperm pellets separated from seminal fluid and HIV particles whereas during condomless sex, she is exposed to both components of her partner’s ejaculate. While sperm washing is the preferred conception strategy between male-positive, serodiscordant couples in many European countries, condomless sex is still a viable option as listed by the Canadian HIV Pregnancy Planning Guidelines published in 2012 . When used as a conception method, condomless sex is best restricted to the ovulation period and entails 37% chance of successful insemination per menstrual cycle. Another method that is used in Europe , and was recently approved by the United States Federal Drug Association in July 2012, is to administer emtricitabine/ tenofovir disoproxil fumarate (FTC/TDF; Truvada®) as pre-exposure prophylaxis (PrEP) to prevent HIV acquisition in men and women . When administered during pregnancy planning in a small study that was conducted in Switzerland, where 53 HIV-negative women used TDF alone as PrEP during condomless sex with fully suppressed HIV-positive male partners at the time of ovulation for conception .The pregnancy rate was 26% per menstrual cycle with no HIV transmissions to the women . Despite the availability of national recommendations and study findings, there are less well-documented cost considerations of these conception options. Without a comparative analysis of the rates of effectiveness relative to the costs of each conception strategy, policy makers are without evidence to form harm reduction policies that would publically fund any of these conception strategies for a cART-treated HIV-positive male in a relationship with an HIV-negative female partner, to protect against HIV transmission. The notion of subsidizing conception strategies in Ontario, Canada, by the Ministry of Health and Long-Term Care (MOH; the provincial government health ministry responsible 614
for policy and spending on health in Ontario, Canada) in this population received some attention when a provincial Expert Advisory Panel on Assisted Reproductive Therapy was enlisted to increase the accessibility and affordability of fertility treatments, but no final recommendations were made. We are unaware of any economic evaluation of such interventions that could inform this decision in health-care settings with adequate access to cART globally. This study presents a cost–utility analysis to determine the most cost-effective approach for HIV-positive males, on cART, to conceive with an HIV-negative female partner from a healthsystem perspective.
Methods A cohort Markov chain was designed to evaluate conception strategies for a hypothetical population of couples composed of an HIV-positive male, whose plasma viral load was undetectable on cART for at least 6 months, and a 30-year-old HIV-negative female partner. It was assumed that the estimated lifetime risk of HIV transmission for the female partner was negligible based on two assumptions. First, the couples in this model consistently used condoms for sexual engagement outside conception attempts. Second, female partners did not have sex with individuals outside of the relationship in this setting. Also, the couples in the cohort were assumed to be fully fertile. We compared three conception strategies from the perspective of the MOH: condomless sex restricted to timed ovulation (CS); condomless sex restricted to timed ovulation combined with pre-exposure prophylaxis (CS-PrEP) and sperm washing followed by IUI (SW). Another strategy that was considered for this model was natural conception without timed intercourse or fertility assistance. It is the dominant option in some settings (for example, Switzerland) with no signal of a relevant risk of horizontal transmission. The reason we did not include this strategy is less related to the risk of HIV transmission, and more so because it is associated with a decreased chance of conception [9–11]. Costs and benefits were discounted at an annual rate of 3% and were drawn from primary cost data and the literature . Outcome measures included total health system expenditures for each strategy and the incremental cost per qualityadjusted life years (QALYs) gained. This analysis does not consider the costs and benefits downstream from an initial transmission event as this is a Markov cohort model and not a dynamic transmission model.
Comparators CS requires a couple to engage in vaginal intercourse with ejaculation on the day of ovulation, which is clinically defined as the peak of the luteinizing hormone ©2015 International Medical Press
Economic evaluation of conception strategies
during a given menstrual cycle . CS-PrEP combines this method of timed intercourse with the administration of PrEP to the HIV-negative female partner. The regimen entails a single dose of TDF or TDF/FTC (Truvada®) on the morning of the luteinizing hormone peak, followed by a second dose the next morning . Then on the second night, when ovulation is best predicted to be occurring, the couple engages in condomless sex . SW involves a mechanical process that separates the HIV-positive male’s sperm from the seminal fluid, where infectious agents of HIV reside . The washed sperm is inserted into the cervicovaginal canal or uterus, often on the day of the luteinizing hormone peak and/or the next day, which is the most likely day of ovulation. In our model, the undertaking of each conception strategy per couple would consist of three cycle attempts pursued over the course of a year, mirroring published reports of the median number of attempts required to attain conception via CS-PrEP and SW, given that conception did not occur during the two preceding cycles [6,7]. We assumed that the non-occurrence of pregnancy during the first two cycles would not affect the chances of conception during a third attempt, in accordance with methodological assumptions in fertility studies [15,16]. Also, we assumed that any variability in the number of sex acts during set time for insemination to occur during each strategy would not significantly affect the couple’s rates of conception, live births or HIV transmission .
Model structure and transition probabilities The present model took the form of a five-state Markov chain (Figure 1). We tracked costs and quality of life
over the lifetime of a cohort of 30-year-old HIV-negative women, or until the age 80, which is the average life expectancy of a Canadian woman, that would undergo our conception strategies with an HIV‑positive male partner . Each year, from age 30 to 35 inclusive, the cohort would attempt one of the three possible comparators 3× with the possibility that they: would not acquire HIV and would not give birth to a child; would not acquire HIV and would give birth to an HIV-negative child; would acquire HIV and would not give birth to a child; would acquire HIV and would give birth to an HIV-negative child; or would acquire HIV and would give birth to an HIV-positive child. We assumed that the probability of conception and of live births were the sole variables that determined the probability of giving birth to a healthy child . The model was constructed using TreeAge Pro, 2009 (TreeAge Software Inc., Williamstown, MA, USA).
HIV transmission comparators To estimate the risk of horizontal HIV transmission in this Markov model, it was assumed that the HIV-positive male partners were on cART with regular monitoring by an HIV specialist, and sustained an undetectable plasma viral load for at least six months prior to engaging in each of our conception comparators . Horizontal HIV transmission estimates for the base case analysis were drawn from a meta-analysis of transmission rates between an HIV-positive partner with confirmed virological suppression and an HIV-negative partner, which included the results of the HPTN 052 Study, for CS [19,20]. The above-mentioned Swiss prospective study of HIV-negative female partners on PrEP
Figure 1. Markov state transition chain
A) Mum HIVNo baby
B) Mum HIVBaby HIV-
E) Mum HIV+ Baby HIV+
C) Mum HIV+ No baby
D) Mum HIV+ Baby HIV-
Depicts the outcomes of giving birth to a child and HIV transmission to mother and to child during condomless sex, condomless sex combined with pre-exposure prophylaxis and sperm washing. Antiviral Therapy 20.6
M Letchumanan et al.
during conception attempts with an HIV-positive male partner on cART whose viral load was undetectable was used to estimate horizontal HIV transmission estimates for CS-PrEP. A meta-analysis that summarized the HIV seroconversion rates of eleven studies on discordant couples undergoing SW was used to estimate horizontal HIV transmission estimates for SW (Table 1) . We calculated the probability of vertical HIV transmission as a weighted average of published estimates that report vertical transmission rates in HIV-positive mothers on cART (Table 1) [21–26]. We modelled the probability of giving birth to a healthy child during each comparator as the product of the probability of conception per cycle, drawn from the literature to be associated with CS, CS-PrEP and SW, and a general probability of live birth per cycle (Table 1).
Costs Since a health-system perspective was adopted for this model, we included the direct health system cost of attempting each comparator 3× per year based on initial consultations, fertility health screens, the intervention and follow-up care (Table 1). Patient costs were therefore not included in the analysis (for example, luteinizing hormone-surge urine sticks used to time
ovulation, transportation to fertility clinics). The cost items and prices of these categories were drawn from the Ontario Health Insurance Plan Schedule of Benefits and Fees (2013), the Canadian Medical Laboratory, the Trillium Drug Program, Ontario Public Health and a reproductive clinic based in Toronto, which is one of four servicing HIV-discordant couples in Ontario who offered their billing data for the purpose of this study called, ‘CReATE Fertility Centre’ (Additional file 1). We included the lifetime cost of HIV infection in an adult woman using estimates from a published model that reported on the economic burden of HIV seroconversion in a Canadian woman at age 30 (Table 1) . Modelling the lifetime cost of HIV infection in an infant was derived from an average of relevant studies included in a systematic review of lifetime costs for treating an HIV-positive infant in the absence of AIDSdefining illness . Using the US federal government’s Consumer Price Index and an exchange rate of 1 USD to 1.066 CND, all costs were reported in 2013 Canadian dollars .
Quality of life A review of the literature did not locate a quality of life instrument that was tailored to HIV-negative females
Table 1. Input parameters: clinical probabilities, costs and utilities Model parameters
Base case value
Clinical Conception per cycle CS (sd) Conception per cycle CS-PrEP Conception per cycle SW Live birth per cycle Male to female HIV transmission CS Male to female HIV transmission CS-PrEP Male to female HIV transmission SW Mother to child HIV transmissionb (sd) Utilities A. HIV- mum, no baby (95% CI) B. HIV- mum, HIV- baby C. HIV+ mum, no baby (95% CI) D. HIV+ mum, HIV- baby (95% CI) E. HIV+ mum, HIV+ baby (95% CI) Costs ($CND 2013)d CS CS-PrEP SW Lifetime cost of HIV infection in female adultd Lifetime cost of HIV infection in infantd
0.360 (0.337)  0.229 (0.152 [0.091]a)  0.165 (0.184 [0.021]a)  0.150  0.000 (0.014 [0.0368]a)  0.000 (0.014 [0.0368]a)  0.000  0.010 (0.74) [21–26] 0.594 (0.714, 0.473) 1.00c 0.578 (0.456, 0.700) 0.850 (0.698, 1) 0.838 (0.689, 0.986) 313 Additional file 1 417 Additional file 1 3,848 Additional file 1 365,111.00  17,315 
Under the subheading, ‘Utilities’, each outcome (A to E) was assigned a utility value to represent a measure of health preferences on a scale of 0 to 1, where 1 is equal to perfect health (defined in this model as having a healthy baby without HIV in transmission). These values were obtained through a time trade-off questionnaire and do not reflect complimentary probabilities, but utility measures. aMean/sd used in probability sensitivity analysis. Values are base case value (mean value in probabilistic sensitivity analysis [sd value in probabilistic sensitivity analysis]). bWeighted mean and sd of published studies using REVMAN. cTime trade-off survey defined state B as perfect health. dCosts discounted by 3%. Adjusted for inflation using US 2013 Consumer Price Index and converted using an exchange rate of 1 USD to 1.066 CND dollars. CS, condomless sex; CS-PrEP, condomless sex combined with pre-exposure prophylaxis; SW, sperm washing; $CND, Canadian dollars. 616
©2015 International Medical Press
Economic evaluation of conception strategies
who aimed to conceive with HIV-positive males and encompassed the health preferences of their unborn child. Hence, a five-item time trade-off questionnaire was designed to collect utilities of each health state in this study and was approved by the Research Ethics Boards of the University of Toronto and Women’s College Hospital. The target population was HIV-negative women who wished to conceive with an HIV-positive male partner. Participants were sought out from Toronto Maple Leaf Medical Clinic, St Michael’s Hospital, ISIS Fertility Care and University Hospital Network. After encountering major recruitment challenges, we searched for participants through the Canadian Hemophiliac Society and revised our inclusion criteria to include HIV-negative females whose social circles included one or more HIV-positive individual. Our survey data was used to assign a quality-of-life weight, or utility for giving birth to a healthy child as well as possible HIV transmission to the mother and/or child, to each state based on survey data. We multiplied the utility weight of the health state by its duration to calculate QALYs, our measure for effectiveness (Table 1).
Cost-effectiveness analysis The incremental cost-effectiveness ratio (ICER) was calculated by dividing the incremental cost by the incremental effectiveness (incremental effectiveness is measured in QALYs). We determined that a strategy would be cost-effective if this ratio did not exceed the maximum estimate a policy-maker would commit to paying in order to acquire one unit of benefit . A willingness-to-pay threshold of $50,000/QALY was used . The dominance of a strategy was established, relative to another, if it offered fewer (or equal) expected benefits for greater (or equal) expected costs .
Sensitivity analysis Three sensitivity analyses were performed. Firstly, we investigated the effect of parameter uncertainty on model outcomes by varying the probabilities of conception per cycle for each comparator, the rates of horizontal and vertical HIV transmission, and utilities across plausible ranges using a Monte Carlo probabilistic sensitivity analysis (Table 1). Beta distributions were specified for each of these probabilities given that these parameters were bound to values between 0 and 1. Since the literature consistently reports zero HIV seroconversions during SW administered in the ‘cART’ era, SW was fixed at zero while CS and CS-PrEP varied across a 95% CI of 0.04 to 0.31 per 100 personyears reported in a meta-analysis of transmission rates between an HIV-negative partner and an HIV-positive partner whose plasma viral load was unconfirmed, although on cART . For vertical transmission, the risk ranged from 0.009 to 0.012 [21–26]. Antiviral Therapy 20.6
A second probabilistic sensitivity analysis compared a fourth strategy of natural conception with unrestricted condomless sex (UCS) to CS. The health system costs and HIV transmission risks (horizontal and vertical) associated with UCS were assumed to be equal to CS. However, it was expected that UCS would produce a 0.15 chance of conception per menstrual cycle based on untimed intercourse averaging once per week . Given that the literature reports this probability as a point estimate, we assumed a range of ±25% to calculate a standard deviation of 0.1593 for our probabilistic sensitivity analysis. Thirdly, a deterministic sensitivity analysis assessed alternate annual discount rates of 0, 5 and 7% based on the range of values observed in standard economic evaluations .
Results Base case analysis We evaluated the cost-effectiveness of CS, CS-PrEP and SW as strategies to conceive between HIV-negative women and HIV-positive men. Due to major recruitment challenges, utility measures were obtained from eight HIV-negative women (four who wanted to conceive with their HIV-positive partner and four whose social circles included one or more HIV-positive individual). Despite the small sample size, these utility data were still useful since the health preferences of the outcomes in this model are under-investigated in the literature. Compared to CS in our base case analysis, CS-PrEP was associated with a loss of 0.15 QALYs and an added expense of $438 (Table 2). This means that CS-PrEP cost an extra $438 for our base case and yielded a loss of QALYs of 0.15; with QALYs in our analysis representing QALYs of the HIV-negative female partner taking into consideration: if she gave birth to a live child; if she acquired HIV and if the child acquired HIV. Relative to the commonly used willingness-to-pay threshold of $50,000/QALY gained, CS-PrEP does not represent a better value for money than CS given that a greater payout is required and yields negative benefits. In an analysis that compared SW to CS, an even larger addition to costs was observed for a greater loss in benefits. The cost associated with SW was $14,910 more than CS and was less effective by a difference of 0.38 QALYs (Table 2). Overall, CS dominated CS-PrEP and SW.
Sensitivity analysis CS’s cost-effectiveness proved robust in our probabilistic sensitivity analysis as the differences in costs and benefits relative to its comparators were associated with little uncertainty. As shown in Figure 2, the variability surrounding the average ICER between CS-PrEP and CS is consistently more costly and less effective. Similarly, a smaller range spanned the net dollars and QALYs incurred when 617
M Letchumanan et al.
Table 2. Discounted lifetime costs and benefits: base case and sensitivity analysis Outcomes Incremental changes Costs, $CND QALYs Costs, $CND QALYs Base case analysis CS CS-PrEP SW Sensitivity analysis CS, mean (95% CI) CS-PrEP, mean (95% CI) SW, mean (95% CI) UCS, mean (95% CI)
Incremental cost effectiveness ratio, cost ($)/QALY gained
1,319 15.56 – – Undominated 1,757 15.41 438 -0.15 Dominated 16,229 15.24 14,910 -0.32 Dominated 1,577 (1,509, 1,643) 15.57 (3.23, 27.90) – – Undominated 2,012 (1,999, 2,024) 15.06 (2.73, 27.40) 435 (423, 447) -0.50 (-12.84, 11.83) Dominated 16,485 (16,482, 16,489) 15.24 (15.24, 15.24) 14,909 (14,904, 14,913) -0.32 (-0.32, -0.33) Dominated 1,577 (-67, 1,643) 15.28 (-0.00, 15.28) 50 (-89, 119) -0.27 (-0.01, -0.27) Dominated
For the sensitivity analysis, the factors that were varied included the risk of HIV transmission from male to female, from mother to child during pregnancy and utilities. Dominated strategy is more costly and less effective than comparator. Undominated strategy is less costly and more effective than comparator. CS, condomless sex; CS-PrEP, condomless sex combined with pre-exposure prophylaxis; QALYs, quality-adjusted life-years; SW, sperm washing; UCS, unrestricted condomless sex.
Figure 2. Scatter plots of 1,000 simulated pairs of incremental costs and effects in the cost-effectiveness plane: SW and CS, CS-PrEP and CS analyses
CS-PrEP versus CS 15,200
15,000 14,900 14,800
Cost difference, $
Cost difference, $
SW versus CS
15,000 14,900 14,800 14,700
14,700 14,600 -0.6
QALY difference Individual results
Positive costs indicate that the comparators to condomless sex (CS) were more expensive and resulted in negative quality-adjusted life-years (QALYs) indicating that the comparators to CS were associated with fewer benefits. CS-PrEP, condomless sex combined with pre-exposure prophylaxis; SW, sperm washing.
comparing SW to CS. In summary, there was minimal variation from our base case ICERs in our sensitivity analysis. CS was consistently more effective and less costly than CSPrEP and SW. Relative to our base case analysis, the average cost of each option increased approximately by $250 and there were 0.35 fewer QALYs generated by CS-PrEP, 0.1 QALYs lost by CS, while the QALYs associated with SW remained unchanged. As depicted in Table 2, CS remained cost-effective in a comparison to a strategy of natural conception with unrestricted condomless sex. On average, unrestricted 618
condomless sex incurred an additional cost of $30 (95% CI -89, 119) and -0.27 (95% CI -0.01, -0.27) fewer QALYs than CS. Varying the annual discount rates in a one-way sensitivity analysis did not yield significant variation from base case ICERs.
Discussion In our study, we assessed the cost-effectiveness of three common strategies to conceive between an HIV-positive ©2015 International Medical Press
Economic evaluation of conception strategies
man and an HIV-negative female: CS, CS-PrEP and SW. We found that conception achieved through timed intercourse ($101 per QALY gained) was less costly and more effective than each alternative. Therefore, CS dominated the use of PrEP during condomless sex and IUI with washed sperm costing. Our findings are important as current HIV conception guidelines worldwide are mainly recommending SW or CS-PrEP as the preferred methods of conception for serodiscordant couples with an HIV-positive man and HIV-negative female partner with limited available data on the topic [13,33,34]. In 2010, the American Society for Reproductive Medicine endorsed that there is reasonable data to support the recommendation that, if the male partner is HIV-positive, sperm washing should be performed to decrease HIV-infected white blood cells and free virus in the fluid used during IUI . In addition, it was suggested that the use of PrEP may reduce further the susceptibility of the uninfected female partner . In Canada, the Canadian HIV Pregnancy Planning Guidelines indicate that all available insemination options be reviewed with the serodiscordant couple and the best option that is recommended is that the HIV-positive man be referred to a fertility specialist for sperm washing with IUI . In Europe, the standard of care is to recommend sperm washing or PrEP as safer alternatives in the place of condomless sex . Our findings can help to adapt these guidelines or can guide the physician’s way of counselling. Our findings consider not only the extremely low risk of horizontal transmission, but also the high cost of procedures such as SW, and the utility that a woman in such a situation places on having a baby, becoming HIV-positive or her baby becoming HIV-positive. With all these variables being considered, we found that condomless sex which is limited to the time of ovulation with the HIV-positive male partner being on effective cART with a suppressed viral load for greater than 6 months was the most viable conception option for these couples. In recent years, there has been increasing support in HIV literature for the use of condomless sex as a viable option to conceive between heterosexual HIV-discordant couples in monogamous relationships [20,33]. In 2008, the Swiss statement proposed this notion with three conditions that would nullify the risk of HIV transmission . They included maintaining supervised adherence to cART by an HIV specialist, undetectable viraemia for at least 6 months, and an absence of sexual coinfections. The controversy that followed the Swiss recipe for becoming ‘sexually non-infectious’ was largely resolved when data from the HPTN Study 052 emerged . In this multicentre, randomized control trial of 1,763 participants, a 96% reduction in linked HIV-1 transmissions was achieved in those whose cART regimen started earlier than participants Antiviral Therapy 20.6
in whom the onset was delayed until subsequent clinical developments. The only case of transmission was estimated to be at day 8 after the start of antiretroviral therapy in the HIV-positive partner resulting in a positive HIV test result in the partner at the 3-month time point . Most recently, a systematic review of the risk of HIV transmission between heterosexual serodiscordant couples that met the Swiss criteria determined a summary rate of zero transmission events . Given these data, it is reasonable to suggest that condomless sex restricted to timed ovulation is a trusted medical option to conceive for discordant couples with an HIVpositive male partner in whom the Swiss conditions have been met. Our analysis has several limitations. We must acknowledge that the decision to have a child is of individual importance. The desire to conceive and preferences on how to approach conception can widely vary between couples. Issues of advanced maternal age and infertility, though outside the scope of this model, are more common in male-positive, HIV-discordant couples. Also, there are potential psychological implications that are associated with CS that may discourage some women from undertaking this option. Therefore, specialized counselling to guide the conception process is necessary for each couple. Consequently, the generalizability of our results is confined to the scope of our study. We did not consider serodiscordant, heterosexual couples composed of an HIV-positive female partner and HIV-negative male partner, heterosexual HIV-concordant couples or same-sex couples affected by HIV. The clinical parameters that we used did not reflect degrading probabilities of conception and live birth that commonly occur as women approach their mid-30s. We did not model alternative predictors of giving birth to a child. Second, this model does not reflect the safety of each conception strategy amongst couples whose HIV-positive male partners fail to comply with the HIV management conditions necessary for negligible risk of sexual transmission. Specifically, men are not on cART with regular monitoring by an HIV specialist, or do not sustain an undetectable plasma viral load for at least six months prior to engaging in each of our comparators may exacerbate the risk of transmission to the female partner. Another limitation was the inclusion of a full reproductive screen prior to performing SW in couples that were defined as not having fertility issues, which contributed to an overall cost of $3,500. However, fertility clinics require reproductive screens in male-positive, HIV-discordant couples seeking to undergo SW based on the rationale that it is not worth the effort and risk of proceeding if there is a fertility issue. This could be a limitation of fertility clinics’ policies. Another source of bias was the number of women that informed our 619
M Letchumanan et al.
health utilities. These scores were based on eight participants, four of which were actively in a relationship with an HIV-positive male and four of which were not. With this limited sample base, our time trade-off instrument could not be feasibly tested for accuracy and reliability. Thus, severe bias may accompany the utility scores used in this study. Also, we could not test the uncertainty surrounding the probability of live birth in our probability sensitivity analysis due to limited data in the literature. Lastly, as this analysis was not built using a dynamic transmission model, we did not consider downstream costs and benefits from an initial transmission event. In summary, neither CS-PrEP nor SW represent a better value for money relative to timed condomless sex for fertile HIV-negative women who wish to conceive with an HIV-positive male partner whose plasma viral load is undetectable on cART. Based on these findings, we cannot recommend CS-PrEP or SW for public-funding in developed health-care settings globally. However, further evaluation of alternate couple settings that capture the impact of changing reproductive capacity over time and additional correlates of HIV transmission risk will be beneficial in confirming these findings.
Vitorino RL, Grinsztejn BG, de Andrade CA, et al. Systematic review of the effectiveness and safety of assisted reproduction techniques in couples serodiscordant for human immunodeficiency virus where the man is positive. Fertil Steril 2011; 95:1684–1690.
Vernazza PL, Graf I, Sonnenberg-Schwan U, Geit M, Meurer A. Preexposure prophylaxis and timed intercourse for HIV-discordant couples willing to conceive a child. AIDS 2011; 25:2005–2008.
Aaron E, Cohan D. Pre-exposure prophylaxis for the prevention of HIV transmission to women. AIDS 2013; 27:F1–F5.
Stanford JB, White GL, Hatasaka H. Timing intercourse to achieve pregnancy: current evidence. Obstet Gynecol 2002; 100:1333–1341.
15. Butler WJ, Kalasinski LA. Statistical analysis of epidemiologic data of pregnancy outcomes. Environ Health Perspect 1989; 79:223–227.
We thank CReATe Fertility Clinic for providing their provincial billing information for our cost-analysis and the Canadian Hemophiliac Society for their assistance in recruiting participants for our time trade-off survey.
10. Dunson DB, Baird D, Wilcox A, Weinberg C. Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovulation. Hum Reprod 1999; 14:1835–1839. 11. Wilcox AJ, Weinberg CR, Baird DD. Timing of sexual intercourse in relation to ovulation. Effects on the probability of conception, survival of the pregnancy, and sex of the baby. N Engl J Med 1995; 333:1517–1521. 12. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russel LB. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996; 276:1253-1258. 13. Sherman BM, Korenman SG. Hormonal characteristics of the human menstrual cycle throughout reproductive life. J Clin Invest 1975; 55:699–706. 14. Bujan L, Hollander L, Coudert M, et al. Safety and efficacy of sperm washing in HIV-1-serodiscordant couples where the male is infected: results from the European CREAThE network. AIDS 2007; 21:1909–1914.
16. Olsen J, Skov T. Design options and methodological fallacies in the studies of reproductive failures. Environ Health Perspect 1993; 101 Suppl 2:145–152. 17. Government of Canada. Statistics Canada. (Accessed 25 October 2013.) Available from http://www.statcan.gc.ca/ tables-tableaux/sum-som/l01/cst01/health26-eng.htm
18. Boklage CE. Survival probability of human conceptions from fertilization to term. Int J Fertil 1990; 35:75–94.
The authors declare no competing interests.
19. Loutfy MR, Wu W, Letchumanan M, et al. Systematic review of HIV transmission between heterosexual serodiscordant couples where the HIV-positive partner is fully suppressed on antiretroviral therapy. PLoS ONE 2013; 8:e55747.
Additional file Additional file 1: Cost analysis of conception strategies can be found at http://www.intmedpress.com/uploads/ documents/3360_Letchumanan_Add_file1.pdf
2. 3. 4. 5.
Newmeyer T, Tecimer SN, Jaworsky D, et al. Case series of fertility treatment in HIV-discordant couples (male positive, female negative): the Ontario experience. PLoS ONE 2011; 6:e24853. Kingston-Riechers J. The economic cost of HIV in Canada. Ottawa, Canada: Canadian AIDS Society 2011; pp. 1–17. Yudin MH, Shapiro HM, Loutfy MR. Access to infertility services in Canada for HIV-positive individuals and couples: a cross-sectional study. Reprod Health 2010; 7:7. Semprini AE, Levi-Setti P, Bozzo M, et al. Insemination of HIV-negative women with processed semen of HIV-positive partners. Lancet 1992; 340:1317–1319. Loutfy MR, Margolese S, Money DM, Gysler M, Hamilton S, Yudin MH. Canadian HIV pregnancy planning guidelines. J Obstet Gynaecol Can 2012; 34:575–590.
20. Vernazza P, Hirschel B, Bernasconi E, Flepp M. [The Swiss Statement]. BMS 2009; 89:165–169. French. 21. Forbes JC, Alimenti AM, Singer J, et al. A national review of vertical HIV transmission. AIDS 2012; 26:757–763. 22. Townsend CL, Cortina-Borja M, Peckham CS, de Ruiter A, Lyall H, Tookey PA. Low rates of mother-tochild transmission of HIV following effective pregnancy interventions in the United Kingdom and Ireland, 20002006. AIDS 2008; 22:973–981. 23. Blanche S, Newell ML, Mayaux MJ, et al. Morbidity and mortality in European children vertically infected by HIV‑1. The French Pediatric HIV Infection Study Group and European Collaborative Study. J Acquir Immune Defic Syndr Hum Retrovirol 1997; 14:442–450. 24. Dorenbaum A, Cunningham CK, Gelber RD, et al. Twodose intrapartum/newborn nevirapine and standard antiretroviral therapy to reduce perinatal HIV transmission: a randomized trial. JAMA 2002; 288:189–198. 25. Cooper ER, Charurat M, Mofenson L, et al. Combination antiretroviral strategies for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr 2002; 29:484–494. ©2015 International Medical Press
Economic evaluation of conception strategies
26. Warszawski J, Tubiana R, Le Chenadec J, et al. Motherto-child HIV transmission despite antiretroviral therapy in the ANRS French Perinatal Cohort. AIDS 2008; 22:289–299. 27. Coyte P, Thavorne K, Dhanji H. The economic burden of immigrants with HIV/AIDS: when to say no? Journal for Global Business Advancement 2010; 3:60–78. 28. Johri M, Ako-Arrey D. The cost-effectiveness of preventing mother-to-child transmission of HIV in low- and middleincome countries: systematic review. Cost Effectiveness and Resource Allocation 2011; 9:3. 29. Bureau of Labour Statistics. Washington: U.S. Bureau of Labor Statistics Division of Consumer Prices and Price Indexes; 2013. (Accessed 27 November 2013.) Available from http://www.bls.gov/cpi/ 30. Briggs A, Fenn P. Confidence intervals or surfaces? Uncertainty on the cost-effectiveness plane. Health Econ 1998; 7:723–740.
31. Winkelmayer WC, Weinstein MC, Mittleman MA, Glynn RJ, Pliskin JS. Health economic evaluations: the special case of end-stage renal disease treatment. Med Decis Making 2002; 22:417–430. 32. Cantor SB. Cost-effectiveness analysis, extended dominance, and ethics: a quantitative assessment. Medical Decision Making 1994; 14:259–265. 33. Practice Committee of American Society for Reproductive Medicine. Recommendations for reducing the risk of viral transmission during fertility treatment with the use of autologous gametes: a committee opinion. Fertil Steril 2013; 99:340–346. 34. Loutfy MR, Margolese S, Money DM, Gysler M, Hamilton S, Yudin M. Canadian HIV Pregnancy Planning Guidelines. Int J Gynaecol Obstet 2012; 119:89–99. 35. Cohen MS, McCauley M, Gamble TR. HIV treatment as prevention and HPTN 052. Curr Opin HIV AIDS 2012; 7:99–105.
Accepted 22 March 2015; published online 7 April 2015
Antiviral Therapy 20.6