Table 2. Factors associated with missing antenatal testing for HIV infection and with first identification of such infection 4–26 weeks postpartum, Malawi, October 2014–May 2016.
| Characteristic | Unweighted denominator | Missed antenatal HIV testing |
Newly identified HIV infectionsa |
|||
|---|---|---|---|---|---|---|
| Unweighted, no. (%) | Weighted aORb (95% CI) | Unweighted, no. (%) | Weighted aORb (95% CI) | |||
| Location (n = 33 744) | NIIMc | |||||
| Rural facility in Central or North regions | 8 991 | 73 (0.8) | 54 (0.6) | 0.8 (0.4–1.6) | ||
| Urban facility in Central or North regions | 13 748 | 80 (0.6) | 96 (0.7) | 0.5 (0.2–1.1) | ||
| Rural facility in South region | 6 106 | 278 (4.6) | 52 (0.9) | 1.0 (0.4–2.2) | ||
| Urban facility in South region | 4 899 | 37 (0.8) | 42 (0.9) | Reference | ||
| Mother’s age in years (n = 33 669) | ||||||
| ≤ 19 | 6 427 | 119 (1.9) | 4.9 (2.3–10.6) | 24 (0.4) | 0.6 (0.3–1.3) | |
| 20–24 | 11 504 | 232 (2.0) | 2.0 (1.4–2.8) | 93 (0.8) | 1.1 (0.6–2.2) | |
| 25–29 | 7 423 | 105 (1.4) | 1.3 (0.8–1.5) | 66 (0.9) | 1.1 (0.6–2.1) | |
| ≥ 30 | 8 315 | 11 (0.1) | Reference | 61 (0.7) | Reference | |
| Parity (n = 33 680) | ||||||
| 1 | 10 943 | 139 (1.3) | Reference | 42 (0.4) | 1.0 (0.5–2.0) | |
| 2–3 | 14 179 | 186 (1.3) | 2.1 (1.2–3.7) | 148 (1.0) | 2.0 (1.2–3.6) | |
| ≥ 4 | 8 558 | 143 (1.7) | 2.5 (1.3–4.6) | 54 (0.6) | Reference | |
| ANC attendance (n = 33 744) | NIIMd | NIIMd | ||||
| Yes | 33 681 | 425 (1.3) | 244 (0.7) | |||
| No | 63 | 43 (68.3) | 0 (0.0) | |||
| Location of ANC (n = 33 674) | ||||||
| At study site | 26 225 | 349 (1.3) | Reference | 163 (0.6) | Reference | |
| At different site | 7 449 | 77 (1.0) | 0.7 (0.3–1.9) | 81 (1.1) | 2.2 (1.5–3.1) | |
ANC: antenatal clinic; aOR: adjusted odds ratio; CI: confidence interval; HIV: human immunodeficiency virus; NIIM: not included in model.
a Mother missed HIV diagnosis, she claimed to be HIV-negative or not to know her HIV status, but was found infected in the study.
b Weighted according to sampling interval and the probabilities of districts, clusters and subjects being selected and adjusted for all other variables in the model.
c Excluded from model because of skewed data caused by three rural health facilities in South region having no test kits available when women would have been seeking antenatal care.
d Too few events to be included in the model.