Bloodborne HIV: Don't Get Stuck!

The WHO and UNAIDS promote their Modes of Transmission model[1] to estimate numbers of HIV infections that adults get from various risks. The model has a simple mistake in its design – causing anyone who uses it to overlook crucial data on HIV in married couples and leading thereby to grossly inflated estimates of numbers of HIV infections acquired from spouses.[2,3]

Several experts recently used WHO’s and UNAIDS’s Modes of Transmission model to identify important risks in Malawi’s HIV epidemic. Their published results[4] provide another illustration of ridiculous, stigmatizing, and anti-family estimates produced by the model. Here’s the gist of what they conclude: Infections from spouses account for 81% of new HIV infections in Malawi (76,688 out of an estimated total of 94,455 infections; see Table).

Simple logic says this is absurd, even without looking at any evidence or data: The number of people getting HIV from their spouses cannot exceed the number of spouses bringing HIV into their families (from any source, such as unsafe health care or non-spousal sex partner). This is logically necessary year-by-year in an epidemic, such as Malawi’s, that has been more or less stable over time. Furthermore, a large percentage of people who are married and HIV-positive die without ever infecting their spouses. So the number of new infections coming from spouses will be much less than half of all new infections.

But that’s not what the model says. Where’s the problem?

The model starts out OK: Using data for 2007, the model finds 2,095,000 married men and 2,497,000 married women (roughly 20% of women were in polygamous marriages).[4] Next, using data from Malawi’s 2004 national Demographic and Health Survey,[5] the model recognizes that more than 10% of married men and women were HIV-positive. So far so good.

But then the model falls off the rails. It assumes that almost all HIV-positive married adults were a risk to infect their spouses. This ignores the well-documented fact that many people who are married and HIV-positive have partners who are also HIV-positive – no one is going to infect anyone in such couples. According to the same 2004 national survey, only 4% of married men in Malawi were at risk to get HIV from their wives (that is, 4% were HIV-negative with an HIV-positive wife), and only 5.7% of married women were at risk to get HIV from their husbands.

Table 1: Estimated number of HIV infections acquired from spouses

Thus, only 226,000 married adults (4% of husbands and 5.7% of wives) were at risk to get HIV from their spouses (see next to last row in the Table). How many of these 226,000 will get HIV from their spouses in a year? During the 1990s, 5 studies in Africa followed discordant couples (only one spouse HIV-positive) to watch HIV transmission from one to the other – distressingly, these studies did not routinely warn participants that they or their partner was infected. With few couples taking care to avoid transmission, 6.6% of HIV-positive wives infected husbands in a year, and 9.5% of HIV-positive husbands infected wives in a year (these rates are from a recent review [2]). With these rates of transmission, 226,000 HIV-positive married men and women in Malawi infected an estimated 19,000 spouses in 2007 – only 20% of the estimated 94,454 new HIV infections in Malawi in 2007.

The Modes of Transmission model’s gross and logically absurd overestimate of numbers of HIV infections from spouses is not harmless. Consider these damaging consequences:

1. Diverting attention from HIV risk in unsafe health care: If sex in marriage accounts for only 20% of new HIV infections instead of 81% as estimated by the Modes of Transmission model, then most infections need to be explained by other risks. What are those other risks? The Modes of Transmission model estimates that all non-spousal sex – casual, commercial, and male-male sex – accounts for a combined total of only 18% of infections. If all sexual risks account for only 38% of infections – 20% from spouses and 18% from other sex partners – what non-sexual risks account for the remaining 62% of infections? The Modes of Transmission model avoids this question by grossly overestimating numbers of HIV infections from spouses.

2. Stigmatizing HIV-positive adults: The estimate produced by the Modes of Transmission model – that sex accounts for more than 99% of HIV infections among adults – stigmatizes all HIV-positive adults with the charge they got it from sex. Publishing such estimates contributes to what could be considered a form of sexual abuse – spouses, relatives, and others accusing people of sexual behavior for which there is no evidence. The estimate coincides with racist stereotypes of sexual behavior, which protect it from critical review.

3. Undermining families: Stigmatizing all HIV-positive adults with the charge they got if from sex breeds suspicion among married adults when one or both learn they and/or their partner are HIV-positive. Lack of trust between spouses weakens families and harms children.

The motivation for such misinformation may be traced to a conflict of interest common among health care professionals – who do not want people to know that unsafe health care contributes to Africa’s HIV epidemics. Rather than admitting the obvious (and doing something about it), health care professionals have been blaming victims, insinuating that almost all African adults with HIV got it from sex. The Modes of Transmission model is part of that stigmatizing and racist smear.

References

1. UNAIDS. Modes of Transmission spreadsheet. Geneva: UNAIDS, 2012. Available at: http://www.unaids.org/en/dataanalysis/datatools/incidencebymodesoftransmission/ (accessed 24 April 2014).

2. Gisselquist D. UNAIDS’ Modes of Transmission model misinforms HIV prevention efforts in Africa’s generalized epidemics. Social Science Research Network, 24 August 2013. Available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2315554 (accessed 24 April 2014).

3. Gisselquist D. Misinformation from UNAIDS’ flawed Modes of Transmission model. dontgetstuck, 14 September 2013. Available at: https://dontgetstuck.wordpress.com/2013/09/14/misinformation-from-unaids-flawed-modes-of-transmission-model/ (accessed 24 April 2014).

4. Maleta K, Bowie C. Selecting HIV infection prevention interventions in the mature HIV epidemic in Malawi using the mode of transmission model. BMC Health Services Research 2010; 10: 243. Available at: http://www.biomedcentral.com/content/pdf/1472-6963-10-243.pdf (accessed 22 April 2014). At the end of this article, see the link to Additional file 1: Data sources used to populate the Mode of Transmission model – Malawi 2007.

5. ORC Macro. Malawi Demographic and Health Survey 2004. Calverton: ORC Macro, 2005. Available at: http://dhsprogram.com/pubs/pdf/FR175/FR-175-MW04.pdf (accessed 26 April 2014).