Table 6. Results of meta-analyses and test for heterogeneity of provider delay.
| Variable | k | OR (95% CI) a | Heterogeneity between studies | Test for overall effect (P) | ||
|---|---|---|---|---|---|---|
| Q statistic | P-value | I2 | ||||
| Male sex | 24 | 0.96 (0.93, 1.00) | 30.26 | 0.14 | 24% | 0.03 |
| Consult with a public hospital | 12 | 0.43 (0.20, 0.91) | 286.39 | <0.00001 | 96% | 0.03 |
| Long travel time/distance to the first healthcare provider | 7 | 1.68 (1.12, 2.51) | 36.13 | <0.00001 | 83% | 0.01 |
| Age (older) | 24 | 1.17 (0.91, 1.51) | 962.61 | <0.00001 | 98% | 0.21 |
| Education (illiterate) | 10 | 0.88 (0.74, 1.05) | 7.32 | 0.60 | 0% | 0.15 |
| Rural residence | 8 | 1.20 (0.76, 1.88) | 62.37 | <0.00001 | 89% | 0.44 |
| Income (low) | 10 | 1.13 (0.80, 1.58) | 27.44 | 0.0010 | 67% | 0.50 |
| Unemployed | 4 | 1.68 (0.72, 3.94) | 38.51 | <0.00001 | 92% | 0.23 |
| Haemoptysis | 8 | 0.78 (0.42, 1.44) | 53.68 | <0.00001 | 87% | 0.42 |
| Positive smear | 8 | 0.61 (0.13, 2.96) | 2313.32 | <0.00001 | 100% | 0.54 |
| Chest radiographs | 5 | 0.72 (0.31, 1.70) | 182.38 | <0.00001 | 98% | 0.46 |
a. P-value for homogeneous studies (Q statistic, P≥0.10, or P<0.10, but I2≤50%)—fixed-effect model; otherwise, P-value for significantly heterogeneous studies (Q statistic, P<0.10, but I2>50%)—random effects model.