Table.
Changes in two measures of high risk sexual behaviour among gay men completing anonymous questionnaires in London. Odds ratios calculated using logistic regression and adjusted for age and place of recruitment
| Year of questionnaire
|
Significance
|
|||||
|---|---|---|---|---|---|---|
| 1996 | 1997 | 1998 | χ2 (df) | P | ||
| % response | 75 (2482/3318) | 80 (2121/2639) | 85 (2068/2427) | 94.7 (1) | <0.001* | |
| Median (range) age (years) | 30 (16-71) | 31 (16-73) | 32 (15-78) | 27.9 | <0.001† | |
| % who had unprotected anal intercourse with >1 partner in past year: | 32 (729/2263) | 36 (694/1943) | 38 (730/1916) | 14.7 (2) | <0.001‡ | |
| Adjusted odds ratio (95% CI) | 1 | 1.17 (1.03 to 1.33) | 1.23 (1.12 to 1.45) | |||
| % who had sexual partner with unknown or discordant HIV status in past year: | 18 (375/2147) | 19 (348/1851) | 21 (378/1796) | 8.43 (2) | <0.02‡ | |
| Adjusted odds ratio (95% CI) | 1 | 1.10 (0.94 to 1.3) | 1.26 (1.08 to 1.49) | |||
| % who had an HIV test in past year: | 29 (683/2347) | 31 (614/2006) | 33 (654/1994) | 3.5 (2) | 0.17‡ | |
| Adjusted odds ratio (95% CI) | 1 | 1.03 (0.9 to 1.19) | 1.14 (0.99 to 1.31) | |||
*
Using χ2 test for trend (1 df).
†
Using Kruskal-Wallis test.
‡
Using χ2 test for variation in adjusted odds between years from logistic regression (2 df).