Table 2.
Study results
| Citation | Study Design | Comparison (C) | Intervention (I) | Sero-conversions/Person years; Incidence rate per 100 Person Years | Unadjusted and Adjusted Results (I versus C) (95% confidence intervals) | Interpretation |
|---|---|---|---|---|---|---|
| A. Sites receiving HCT versus sites not receiving HCT | ||||||
|
| ||||||
| Corbett et al., 2007 | cluster randomised controlled trial | Pre-test counselling, risk assessment, and a voucher for results and post-test counselling at a free-standing clinic; uptake 5.2% | Pre-test counselling, risk assessment, testing, results, post-test counselling, and risk reduction planning in workplace; uptake 70.7% | C: 25/2462; 0.95 (mean/cluster) I: 36/2560; 1.37 (mean/cluster) |
IRR: 1.44 (0.79, 2.80), p=0.4 aIRR: 1.49 (0.77, 2.71), p=0.5 |
no effect of HCT on HIV acquisition |
|
| ||||||
| B. Individuals receiving HCT versus individuals not receiving HCT | ||||||
|
| ||||||
| Machekano et al., 1998 | pre/post cohort | Pre-test counselling, tested for HIV but did not receive HIV test results or post-test counselling. | Pre-test counselling, tested for HIV, and received results and post-test counselling. | C:16/332; 4.82 I: 20/657; 3.04 |
IRR: 0.63 (0.31, 1.30), p=0.2 | trend towards HCT leading to lower HIV acquisition |
|
| ||||||
| Matovu et al., 2005 | exposed/unexposed cohort | Participants who provided blood but did not receive HIV test results and post-test counselling at the first household survey. | Participants who provided blood and received HIV test results and post-test counselling at the first household survey. | C: 35/2441; 1.4(overall) I: 42/2631; 1.6 (overall) C: 11/1001; 1.1 (males) I: 18/1166 1.5 (males) C: 24/1439; 1.7 (females) I: 24/1464; 1.6 (females) |
IRR:1.11 (0.71, 1.75), p=0.6 (overall)** IRR: 1.40 (0.67, 3.08), p=0.4 (males)** IRR: 0.98 (0.55, 1.74), p>0.9 (females)** |
no effect of HCT on HIV acquisition |
|
| ||||||
| Matovu et al., 2007 | exposed/unexposed cohort | Participants who did not accept home-based HCT results. (Entire population) | Participants who received home-based HCT results once or more than once (Entire population) | C: 66/4038; 1.6 (never) I: 76/4658; 1.6 (once); I: 48/3488; 1.4 (repeat) |
IRR: 1.00 (0.72, 1.39), p>0.9 (once v. never)** aIRR: 1.00 (0.72, 1.39) (once v. never) IRR: 0.84 (0.58, 1.22), p=0.4 (once v. never)** aIRR: 0.85 (0.58, 1.23) (repeat v. never) |
no effect of HCT on HIV acquisition |
| Participants who did not accept home-based HCT results. (Those with ≥2 partners) | Participants who received home-based HCT results once or more than once (Those with >2 partners) | C: 16/560; 2.9 (never) I: 9/631; 1.4 (once) I: 7.641; 1.1 (repeat) |
IRR: 0.50, (0.21, 1.12), p=0.1 (once v. never)** aIRR: 0.58 (0.25, 1.37) (once v. never) IRR: 0.38, (0.15, 0.91), 0.03 (repeat v. never)** aIRR: 0.49 (0.21, 1.17) (repeat v. never) |
trend towards HCT leading to lower HIV acquisition | ||
| Participants who did not accept home-based HCT results. (Those with only 1 partner) | Participants who received home-based HCT results once or more than once (Those with only 1 partner) | C: 50/3478; 1.4 (never) I: 67/4027; 1.7 (once) I: 41/2846; 1.4 (repeat) |
IRR: 1.16, (0.80, 1.67), p=0.4 (once v. never)** aIRR: 1.15, (0.79, 1.67) (once v. never) IRR: 1.00, (0.66, 1.51), p>0.9 (repeat v. never)** aIRR: 1.00 (0.66, 1.51) (repeat v. never) |
no effect of HCT on HIV acquisition | ||
|
| ||||||
| Sherr et al., 2007 | exposed/unexposed cohort | Persons who had never been tested or counseled | Participants who received pre-test counselling, testing, and post-test counselling | C: 147/8401; 1.75 (overall) I: 18/801; 2.25 (overall) C: 61/2950; 2.07 (males) I: 10/462; 2.16 (males) C: 86/5451; 1.58 (females) I: 8/339; 2.36 (females) |
IRR: 1.28 (0.77, 2.05), p=0.3 (overall)** aIRR: 1.30 (0.79, 2.14) (overall) IRR: 1.05 (0.51, 1.98), p=0.9 (males)** aIRR: 1.08 (0.62, 1.82) (males) IRR: 1.50 (0.68, 2.95), p=0.3 (females)** aIRR: 1.55 (0.63, 3.84) (females) |
no effect of HCT on HIV acquisition |
|
| ||||||
| Rosenberg et al., 2013 | exposed/unexposed and pre/post cohort | Participants who had not been tested for HIV and learned their results | Participants who been tested for HIV and learned their results | C: 131/4702; 2.79 I: 117/3834; 3.05 |
HR: 1.02 (95% CI: 0.79, 1.31), p=0.5 aHR: 0.65 (95% CI: 0.50, 0.86), p<0.01 ipwHR: 0.59 (95% CI: 0.45, 0.78), p<0.01 |
HCT leads to lower HIV acquisition, but only after adjustment |
|
| ||||||
| C. Individuals receiving HCT alone versus individuals receiving HCT as a couple | ||||||
|
| ||||||
| Allen et al., 1992 | pre/post cohort | HIV-uninfected women before undergoing couple HCT | HIV-uninfected women after undergoing couple HCT | HCT: 12/293 4.1* cHCT: 5/278; 1.8* |
IRR: 0.44 (0.14, 1.22), p=0.1** | trend towards cHCT being more protective than no HCT |
| exposed/unexposed cohort | HIV-uninfected women who underwent individual HCT | HIV-uninfected women who underwent couple HCT | HCT: 24/706; 3.4* cHCT: 5/278; 1.8 |
IRR: 0.53 (0.18, 1.32), p=0.2** | trend towards cHCT being more protective than individual HCT | |
|
| ||||||
| Okiria et al., 2013 | exposed/unexposed cohort | Individual pre- and post-test counselling | Couple pre- and post-test counselling | HCT: 82/NA; 0.81 cHCT: 24NA; 0.25 HCT: 27/NA; 0.85 (women) cHCT 3/NA; 0.14 (women)*** HCT: 11/NA; 0.76 (men) cHCT: 7/NA; 0.38 (men) |
IRR: 0.31 (0.19, 0.48), p<0.01** IRR: 0.4 (0.22–0.75), p=<0.01 (women) aIRR: 0.4 (0.22–0.75) (women) IRR: 0.5 (0.24–1.05), p=0.07 (men) |
cHCT was more protective than individual HCT. |
*
based on hand calculations by the review authors.
**
confidence intervals calculated by review authors using Open Epi
***
Review authors believe original article may not have correctly reported this strata as the hand-calculated incidence rate ratio does not equal the reported rate ratio and the person years seem insufficient.
HCT=HIV counselling and testing; cHCT=couple HIV counselling and testing; C=comparison; I=intervention; IRR=incidence rate ratio; HR=hazard ratio; CI=confidence interval; PY=person year; a=adjusted; ipw=inverse probability weighted; NA=not available