TABLE 2.
The heterogeneity and pooled correct identification ratios by subgroup analyses
| Subanalysis | No. of isolates (no. of studies) | Within-group heterogeneity |
Correct identification ratio (95% CI)a | P comparison with overall ratio (χ2 value) | |
|---|---|---|---|---|---|
| P (Q value) | I2 (95% CI [%]) | ||||
| Category of strains | |||||
| Yeasts | 8,842 (27) | <0.0001 (291.38) | 91.1 (88.2–93.2) | 0.959 (0.943–0.973) | 0.129 (2.309) |
| Molds | 1,135 (8) | <0.0001 (45.71) | 84.7 (71.6–91.7) | 0.934 (0.888–0.969) | 0.002 (9.961) |
| Systems | |||||
| Andromas | 2,130 (4) | <0.0001 (33.68) | 91.1 (80.3–96.0) | 0.972 (0.936–0.994) | <0.001 (13.202) |
| Biotyper | 7,289 (24) | <0.0001 (303.10) | 92.4 (89.9–94.3) | 0.954 (0.933–0.971) | 0.725 (0.124) |
| Saramis | 1,926 (6) | <0.0001 (63.71) | 92.2 (85.7–95.7) | 0.938 (0.881–0.978) | 0.001 (10.477) |
| Vitek MS | 1,818 (4) | <0.0001 (31.79) | 90.6 (78.9–95.8) | 0.933 (0.887–0.968) | <0.001 (16.923) |
| Source of strains | |||||
| Clinical isolates only | 9,549 (29) | <0.0001 (332.42) | 91.6 (89.0–93.5) | 0.950 (0.932–0.965) | 0.097 (2.755) |
| Clinical isolates plus reference strains | 428 (4) | 0.0304 (8.92) | 66.4 (1.5–88.5) | 0.988 (0.960–1.000) | 0.001 (10.962) |
| System database | |||||
| Commercial database only | 8,527 (25) | <0.0001 (282.37) | 91.5 (88.7–93.6) | 0.955 (0.937–0.970) | 0.988 (0) |
| Commercial database plus self-established database | 1,450 (8) | <0.0001 (63.44) | 89.0 (80.6–93.7) | 0.955 (0.916–0.983) | 0.962 (0.002) |
| Modified outcomes | |||||
| All systems | 3,366 (11) | 0.0003 (32.41) | 69.1 (42.4–83.5) | 0.987 (0.978–0.994) | <0.001 (73.927) |
| Biotyper | 2,501 (8) | 0.1269 (11.28) | 37.9 (0–72.6) | 0.992b (0.987–0.995) | <0.001 (74.183) |
a
Random-effects pooled ratios, except where noted otherwise.
b
Fixed-effect pooled ratio.