Table 4. Outlier outcomes in the subgroups.
| Test of association | Test of heterogeneity | ||||||
|---|---|---|---|---|---|---|---|
| n | SMD | 95% CI | Pa | Pb | I2 (%) | AM | |
| T1 DM | |||||||
| CN v MA | 2 | 0.90 | 0.75, 1.04 | 10−5 | 0.40 | 0 | F |
| NO v MI | 2 | 0.37 | 0.11, 0.63 | 0.006 | 0.47 | 0 | F |
| NO v MA | 3 | 0.76 | 0.63, 0.88 | 10−5 | 0.51 | 0 | F |
| MI v MA | 2 | 0.61 | 0.45, 0.77 | 10−5 | 0.46 | 0 | F |
| T2 DM | |||||||
| CN v MA | 2 | 0.75 | 0.34, 1.16 | 0.0004 | 0.85 | 0 | F |
| NO v MA | 4 | 1.15 | 0.86, 1.43 | 10−5 | 0.54 | 0 | F |
| MI v MA | 3 | 0.56 | 0.23, 0.90 | 0.001 | 0.85 | 0 | F |
T1DM: Type 1 diabetes mellitus; T2DM: Type 2 diabetes mellitus; CN: control; NO: normoalbuminuria; MI: microalbuminuria; MA: macroalbuminuria; v: versus; n: number of studies; SMD: standardized mean difference; CI: confidence interval; Pa: P-value for association; Pb: P-value for heterogeneity; AM: analysis model; F: fixed-effects; CN versus MI in T1DM did not change with outlier treatment.