Skip to main content

Table 2 Meta-analysis of the association between prevalence of UAI with MSM recruited online vs. offline

From: Prevalence of unprotected anal intercourse in men who have sex with men recruited online versus offline: a meta-analysis

Subgroups Sample size No. of studies Test of association# Test of heterogeneity Egger’s test for publication bias
Online group Offline group OR (95% CI) Z Pvalue Q Pvalue I 2(%) t Pvalue
Overall effects 17,580 10,853 14 1.35 (1.13, 1.62) 3.28 0.001 93.06 <0.001 86.0 0.16 0.880
European subsample 10,718 5,141 6 1.38 (1.17, 1.63) 3.84 <0.001 14.88 0.011 66.4 0.82 0.457
American subsample 4,181 4,866 5 1.21 (0.81, 1.82) 0.93 0.354 49.60 <0.001 91.9 −0.57 0.606
Asian subsample 2,681 846 3 1.63 (0.74, 3.62) 1.20 0.229 28.20 <0.001 92.9 1.37 0.402
Sample size > 500 16,984 10,205 11 1.32 (1.09, 1.61) 2.79 0.005 85.48 <0.001 88.3 -0.09 0.930
Sample size ≤ 500 596 648 3 1.51(0.89, 2.56) 1.52 0.128 7.40 0.025 73.0 2.62 0.232
UAI in the last 6 or more months 9,564 6,557 6 1.28 (0.94, 1.76) 1.56 0.12 62.11 <0.001 92.0 -0.79 0.850
UAI in the last 3 or less months 8,016 4,296 8 1.40 (1.13, 1.74) 3.03 0.002 30.44 <0.001 77.0 1.11 0.645
  1. OR, odds ratios, CI, confidence interval, UAI, unprotected anal intercourse. #The overall pooled estimate of risk OR obtained using DerSimonian and Laid method in the random effect model.