Table 2.
Sex | Age-group | Incidence rate (per 100 py) | IRR | |
2004 to 2015 | 2016 to 2019 | |||
Women | 15 to 19 | 5.32 (4.87 to 5.79) | 3.00 (2.33 to 3.77) | 0.56 (0.43 to 0.73) |
20 to 24 | 6.73 (6.27 to 7.20) | 5.12 (4.25 to 6.37) | 0.76 (0.62 to 0.96) | |
25 to 29 | 5.96 (5.37 to 6.65) | 4.53 (3.52 to 5.88) | 0.76 (0.56 to 1.02) | |
30 to 34 | 4.17 (3.50 to 4.99) | 4.01 (2.78 to 5.66) | 0.96 (0.64 to 1.42) | |
35 to 39 | 1.94 (1.50 to 2.47) | 1.74 (0.95 to 3.09) | 0.89 (0.46 to 1.71) | |
40 to 49 | 1.55 (1.32 to 1.83) | 1.00 (0.59 to 1.73) | 0.64 (0.36 to 1.15) | |
Men | 15 to 19 | 0.92 (0.72 to 1.16) | 0.32 (0.14 to 0.71) | 0.36 (0.15 to 0.82) |
20 to 24 | 2.74 (2.40 to 3.13) | 0.87 (0.51 to 1.53) | 0.32 (0.18 to 0.59) | |
25 to 29 | 3.91 (3.32 to 4.55) | 2.13 (1.24 to 3.64) | 0.54 (0.31 to 0.94) | |
30 to 34 | 3.26 (2.59 to 4.12) | 2.78 (1.68 to 4.57) | 0.85 (0.49 to 1.46) | |
35 to 39 | 2.62 (1.87 to 3.60) | 1.71 (0.65 to 4.24) | 0.66 (0.22 to 1.71) | |
40 to 54 | 1.43 (1.12 to 1.80) | 0.86 (0.00-Inf) | 0.68 (0.00-Inf) | |
Men (circumcised)* | 15 to 19 | 0.52 (0.35 to 0.76) | 0.22 (0.11 to 0.48) | 0.43 (0.21 to 0.95) |
20 to 24 | 1.58 (1.12 to 2.24) | 0.60 (0.33 to 1.06) | 0.38 (0.22 to 0.67) | |
25 to 29 | 2.24 (1.52 to 3.24) | 1.38 (0.76 to 2.41) | 0.62 (0.36 to 1.09) | |
30 to 34 | 1.89 (1.26 to 2.84) | 1.74 (0.99 to 3.08) | 0.92 (0.54 to 1.56) | |
35 to 39 | 1.50 (0.94 to 2.37) | 1.11 (0.41 to 2.94) | 0.75 (0.27 to 2.06) | |
40 to 54 | 0.82 (0.53 to 1.20) | 0.18 (0.00-Inf) | 0.21 (0.00-Inf) | |
Men (uncircumcised)* | 15 to 19 | 0.93 (0.74 to 1.18) | 0.40 (0.19 to 0.84) | |
20 to 24 | 2.82 (2.48 to 3.21) | 1.08 (0.63 to 1.84) | ||
25 to 29 | 4.02 (3.38 to 4.69) | 2.46 (1.49 to 4.20) | ||
30 to 34 | 3.37 (2.64 to 4.32) | 3.06 (1.99 to 5.04) | ||
35 to 39 | 2.68 (1.96 to 3.66) | 1.98 (0.78 to 5.20) | ||
40 to 54 | 1.44 (1.13 to 1.84) | 0.30 (0.00-Inf) |
IRR = incidence rate ratio comparing age-specific incidence in the UTT era (2016 to 2019) relative to age-specific incidence in the pre-UTT era (2004 to 2015). Bold IRRs indicate significance at alpha < 0.05.
Incidence rates for circumcised and uncircumcised men are estimated by fitting a generalized linear model with an indicator for individual-level circumcision status (circumcision model). Fitted IRR’s from the circumcision model are adjusted for circumcision status without an interaction term and so are equivalent for both uncircumcised and circumcised men.