Table 1.
HIV testing assumptions.
| Parameter | Value | Range* | Source |
|---|---|---|---|
| RR of testing in previously-diagnosed ART-naïve | 0.50 | 0.025–0.0975† | — |
| RR of testing in ART patients (relative to diagnosed ART-naïve) | 0.36 | 0.07–0.74† | 54 |
| Probability that newly-diagnosed woman discloses her HIV status to a short-term partner | 0.50 | — | 55– 57 |
| OR for effect of marriage on probability of disclosure | 2.50 | — | 57– 59 |
| OR for effect of male sex on probability of disclosure | 1.25 | — | 58, 59 |
| Probability of referral for testing if disclosure occurs | 0.42 | — | 60** |
| OR for effect of assisted partner notification on partner referral for HIV testing | 3.00 | 1.79–4.52 | 61, 62 |
| Fraction of adult population tested for each round of home-based testing | 0.70 | 0.55–0.83 | 12 |
| Ratio of male to female uptake of testing through home-based testing | 0.67 | — | 12 |
| Annual rate of testing in women attending FP clinics if HIV testing is integrated into FP | 0.45 | 0.28–0.67 | 63, 64 |
| Annual rate of HIV testing through mobile testing in communities with mobile testing services | 0.055 | 0.023–0.101 | 65, 66 |
| Relative rate of testing through mobile clinics in the presence of community mobilization | 2.0 | — | 67 |
| Increase in annual rate of testing in MSM if MSM-focused HIV testing is introduced | 0.40 | 0.11–0.88 | 39, 40 |
| Increase in annual rate of testing in FSWs if FSW-focused HIV testing is introduced | 1.35 | 0.14–3.90 | 68– 70 |
| Probability of test acceptance by sexually-experienced adolescents if offered testing in school | 0.67 (M) 0.76 (F) | — | 71 |
| Relative rate of school testing in virgins | 0.50 | 0.13–0.87 | 72 |
| Fraction of employed population reachable through annual workplace testing programmes | 0.30 | 0.03–0.70 | 41 |
| Probability of testing through workplace testing if workplace testing is offered | 0.22 (M) 0.28 (F) | — | 73 |
| Fraction of husbands of married pregnant women who get tested if given invitation letters | 0.33 | 0.15–0.53 | 74– 78 |
| OR for effect of marriage on probability of pregnant women’s partner getting tested | 4.00 | — | 76, 79, 80 |
| OR for effect of self-testing offer on uptake of testing | 8.17 | 6.33–10.24 | 16 |
| Probability of starting ART soon after diagnosis | |||
| Diagnosed in ANC setting | 0.93 | — | 81 |
| Diagnosed in OI/TB clinic | 0.78 | — | 81 |
| Diagnosed in other clinic-based settings | 0.40 | — | 82 |
| RR of starting ART if diagnosed through community-based testing (relative to other clinic-based testing) | 0.68 | 0.36–0.93 | 18, 49– 51 |
| Cost assumptions | |||
| Average health worker time (minutes) per facility-based test | |||
| HIV-positive test | 27.98 | — | S2.6.4.1 |
| HIV-negative test | 21.63 | — | S2.6.4.1 |
| Cost per rapid test (first test) (USD) | 0.52 | — | S2.6.4.1 |
| Cost per rapid test (second test) (USD) | 0.54 | — | S2.6.4.1 |
| Cost per self-testing kit (USD) | 2.40 | — | S2.6.4.6 |
| Cost per person year of ART (USD) | 280.07 | — | 83 |
*Ranges are specified only for the parameters that vary in the uncertainty analysis; the specified ranges correspond to the 2.5 and 97.5 percentiles of the distributions from which the parameter values are sampled. †The parameter is fixed for the baseline testing modalities, but a different parameter value is randomly sampled for each of the new testing modalities. **The parameter is estimated by dividing the reported fraction of partners who get tested (30%60) by the assumed probability of disclosure (0.71 in the case of married women). FP = family planning; FSW = female sex worker; MSM = men who have sex with men; OI = opportunistic infection; OR = odds ratio; RR = relative rate.