Table 2.
Stratified meta-analyses of the proportion of substance abuse among MSM in China.
| Category | Subgroup | Number of studies | Proportion estimate (%) and 95% CI | Heterogeneity | |
|---|---|---|---|---|---|
| I 2 /% | P -value | ||||
| Study period | 2010–2017 | 24 | 19.0 (13.0–29.0) | 99 | < 0.01 |
| 2018–2022 | 23 | 27.0 (19.0–38.0) | 99 | < 0.01 | |
| Study location | Only one city | 34 | 23.0 (17.0–31.0) | 99 | < 0.01 |
| More than two cities or multiple provinces | 13 | 22.0 (12.0–39.0) | 99 | < 0.01 | |
| 11 cities with strong comprehensive strengths | 26 | 20.0 (14.0–28.0) | 99 | < 0.01 | |
| Sample size | < 1,000 | 29 | 28.0 (21.0–36.0) | 98 | < 0.01 |
| 1,000–2,000 | 11 | 20.0 (11.0–35.0) | 99 | < 0.01 | |
| >2,000 | 7 | 12.0 (4.0–34.0) | 99 | < 0.01 | |
| Recall window of substance abuse (months) | 3 months | 12 | 20.0 (11.0–38.0) | 99 | < 0.01 |
| 6 months | 28 | 25.0 (19.0–34.0) | 99 | < 0.01 | |
| ≥12 months | 7 | 17.0 (7.0–43.0) | 99 | < 0.01 | |