Table 1.
Key model and calibration parameters in the PopART individual-based simulation model
| Value or range explored | Notes | |
|---|---|---|
| HIV-related parameters | ||
| Start of HIV epidemic, year | 1975 (Zambia), 1980 (South Africa) | .. |
| Average annual hazard of an (uncircumcised) man becoming HIV-positive from an HIV-positive partner who has maximal set-point viral load | 0·05–0·30 | Hollingsworth et al (2008);18 Fraser at al (2007)19 |
| Relative infectivity by HIV stage (relative to CD4 count of ≥500 cells per μL) | 1·00 (CD4 350–500), 1·00 (CD4 200–350), 2·34 (CD4 <200), 5·30 (AEHI) | Bellan et al (2015)20 |
| Duration of AEHI, years | 0·08–0·25 | Bellan et al (2015)20 |
| Relative infectivity of male-to-female transmission (compared with female-to-male) | 1·0–3·0 | Boily et al (2009)21 |
| HIV care-related parameters | ||
| Probability of a women having an HIV test under standard care in 2000–06 | 0·1–0·2 | Estimated in the calibration; probability is for a period of 6 years |
| Annual probability of a women having an HIV test from 2006 onwards under standard care | 0·05–0·40 | Estimated in the calibration |
| Relative probability of a man having an HIV test under standard care (at any time; compared with women) | 0·4–1·1 | Estimated in the calibration |
| Probability of collecting HIV test results from an HIV test under standard care | 0·97–1·00 | Demographic and Health Survey 2013 (Zambia) |
| Probability of collecting a CD4 test result under standard care | 0·75–0·95 | Lower limit is from Mugglin et al (2012);22 higher limit assumed |
| Mean time to starting ART after an HIV-positive test delivered under standard care (conditional on starting ART), years | 0·4–0·7 | Estimated in the calibration |
| Probability of a women staying virally suppressed for life after ART initiation | 0·65–0·90 | Estimated in the calibration |
| Relative probability of a man staying virally suppressed for life (compared with a woman) | 0·6–1·0 | Estimated in the calibration |
| After ART initiation, probability of an individual becoming virally unsuppressed due to suboptimal ART adherence | 0·1 | Vinikoor et al (2014)23 |
| Relative infectivity of an individual on ART (compared with not being on ART) | 0·5 (early ART),* 0·7 (virally unsuppressed), 0·0 (virally suppressed) | Values assumed; no transmission from individuals who are virally suppressed |
| Probability of a man accepting VMMC after an HIV-negative test result† | 0·4 | Assumption cross-checked against population cohort† data |
| Reduction in susceptibility to HIV infection for a circumcised male | 0·6 (VMMC), 0·0 (traditional male circumcision) | Population cohort‡ data |
| Partnership-related parameters | ||
| Risk assortativity | 0·05–0·95 | The propensity for individuals within the same risk group§ to form partnerships with those in the same risk group; estimated in the calibration |
| Relative number of sexual partners (compared with self-report) | 0·625–5·000 | Estimated in the calibration; used to account for misreporting of sexual partners |
AEHI=acute and early HIV infection. ART=antiretroviral therapy. VMMC=voluntary medical male circumcision.
*
The 2-month period after initiating ART when an individual is not fully virally suppressed.
†
Assumed to be a fixed probability of VMMC acceptance across the intervention and counterfactual simulations.
‡
A random sample of ∼2500 individuals aged 18–44 years per trial community (n=21), within which the primary endpoint of the trial was measured.
§
Representing level of sexual activity.