Table 3.
Service availability and uptake, and socioeconomic and behavioural characteristics of people living inside the low HIV prevalence cluster between Bonda and Rusape compared to those living in intermediate HIV prevalence areas, Manicaland, east Zimbabwe.
| Inside cluster |
Inside cluster vs. outsidea |
Overallb |
||||||
|---|---|---|---|---|---|---|---|---|
| Characteristicc | Units | Value | N | aOR | (95% CI) | p-Value | Value | N |
| HIV prevalence | (%) | 11.3 | 1381 | 0.62 | (0.51–0.75) | <0.001 | 15.9 | 7976 |
| Service availability (median distance to service providers) | ||||||||
| Reported distance VMMC services (males 15–29) (≤13 vs. >13 km) | km | 10 | 143 | 1.16 | (0.81–1.68) | 0.420 | 13 | 849 |
| Reported distance HTC services (≤3 vs. >3 km) | km | 3 | 1306 | 0.83 | (0.74–0.94) | 0.003 | 3 | 7334 |
| Reported distance ART services (≤4 vs. >4 km) | km | 5 | 576 | 0.87 | (0.72–1.04) | 0.129 | 4 | 3434 |
| Euclidian distance health facility (≤2.7 vs. >2.7 km) | km | 2.8 | 1397 | 0.63 | (0.56–0.71) | <0.001 | 2.74 | 8095 |
| Euclidian distance hospital (≤16.7 vs.>16.7 km) | km | 13.1 | 1397 | 1.47 | (1.31–1.66) | <0.001 | 16.7 | 8095 |
| Service uptake | ||||||||
| VMMC (circumcised vs. not circumcised) (males 15–29) | % | 2.82 | 319 | 0.68 | (0.33–1.39) | 0.290 | 3.76 | 1755 |
| HTC uptake (HIV test in past three years vs. no HIV test in past three years) | % | 64.7 | 1393 | 1.03 | (0.91–1.17) | 0.664 | 63.5 | 7833 |
| ART uptake (ever on ART vs. never on ART)d | % | 53.7 | 121 | 1.04 | (0.69–1.56) | 0.869 | 50.1 | 1024 |
| Sociodemographic | ||||||||
| Sex ratio (% female) (female vs. male) | % | 63.9 | 1397 | 1.19 | (1.05–1.35) | 0.005 | 60.7 | 8092 |
| Median age (≤30 years vs. age >30 years) | Years | 29 | 1397 | 1.11 | (0.98–1.24) | 0.094 | 30 | 8095 |
| Currently married (vs. not currently married) | % | 55.7 | 1394 | 0.85 | (0.75–0.97) | 0.012 | 58.9 | 8069 |
| In-migration in past three years (vs. no in-migration in past three years) | % | 11.0 | 1391 | 0.85 | (0.70–1.02) | 0.083 | 12.6 | 7986 |
| Socioeconomic | ||||||||
| Urban or peri-urban (vs. rural) | % | 2.6 | 1397 | 0.06 | (0.04–0.08) | <0.001 | 31.1 | 8095 |
| Primary or no education (vs. secondary or higher) | % | 23.8 | 1380 | 0.89 | (0.76–1.03) | 0.119 | 24.5 | 7870 |
| Poorest wealth tercile (vs. intermediate or least poor) | % | 57.1 | 1397 | 1.81 | (1.61–2.03) | <0.001 | 43.8 | 8095 |
| Sexual behaviour | ||||||||
| Age at first sex 18 years or older (vs. ≤17 years)e | % | 25.4 | 1038 | 1.16 | (1.00–1.36) | 0.056 | 28.2 | 6085 |
| More than 1 sexual partner in past 12 months (vs. ≤1 partner) | % | 5.72 | 979 | 0.68 | (0.50–0.92) | 0.013 | 8.7 | 6112 |
| Condom use at last sex (vs. no use) | % | 19.4 | 1056 | 0.93 | (0.78–1.10) | 0.386 | 21.4 | 6195 |
aParticipants inside the low-prevalence cluster were compared to the reference group of participants outside the high- and low-prevalence clusters in the logistic regressions, adjusting for age and sex.
bThe overall data refers to the study area as a whole, including all clusters.
cFor the logistic regressions, the characteristic was the outcome. Continuous variables were divided along the median of the variable across the study area. The reference category is the category listed last. For example, the first service availability or should be read as “Those inside the cluster are 1.16 times as likely to report distances to VMMC services of 13 km or less.”
dUptake of ART in all individuals with a positive HIV test result in the survey.
eAmongst participants aged 18 and over.
aOR: odds ratio adjusted for sex and age; CI: confidence interval.