TABLE 2.
Summary measures of accuracy of commercial and in-house tests for all studiesa
| Reference standard | Test property | % sensitivity (95% CI, I2) | % specificity (95% CI, I2) | PLR (95% CI) | NLR (95% CI) | DOR (95% CI) | AUC (95% CI) |
|---|---|---|---|---|---|---|---|
| All studies (63 studies) | Culture (71 data sets with 1,492 TBM cases) | 82 (75–87, 82.4) | 99 (98–99, 85.0) | 58.6 (35.3–97.3) | 0.19 (0.14–0.25) | 314 (169–584) | 98 (96–99) |
| CRS (24 data sets with 652 TBM cases) | 68 (41–87, 83.6) | 98 (95–99, 76.2) | 36.5 (15.6–85.3) | 0.32 (0.15–0.70) | 113 (39–331) | 98 (96–99) | |
| Culture (71 data sets) | In-house tests (43 data sets with 950 TBM cases) | 87 (80–92, 82.0) | 99 (97–99, 88.5) | 64.6 (28.4–147.0) | 0.13 (0.08–0.20) | 372 (165–839) | 98 (97–99) |
| Commercial tests (28 data sets with 543 TBM cases) | 67 (58–75, 64.8) | 99 (98–99, 48.3) | 46.1 (28.3–75.0) | 0.33 (0.25–0.43) | 139 (71–274) | 98 (96–99) | |
| CRS (24 data sets) | In-house tests (21 data sets with 626 TBM cases) | 68 (38–88, 83.5) | 98 (95–100, 78.0) | 44.4 (16.0–123.2) | 0.32 (0.14–0.75) | 138 (41–468) | 98 (96–99) |
| Commercial tests (3 data sets with 26 TBM cases) | 53 (33–73, 84.7) | 90 (82–95, 52.2) | 70.0 (40.0–124.2) | 0.57 (0.24–0.31) | 21 (4.2–104) | 94 (90–97) |
a
CRS, combined reference standard; PLR, positive likelihood ratio; NLR, negative likelihood ratio; DOR, diagnostic odds ratio; AUC; area under the curve; I2, I-square statistic.