Table 3. Diagnostic performance of RPA-LF in reference laboratory and field scenarios compared with composite gold standard (n = 118).
| Scenario | TP | FP | FN | TN | Sensitivity | p | Specificity | p | PPV | p | NPV | p | LR+ | p | LR- | p |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | |||||||||||
| Reference laboratory | 72 | 5 | 11 | 30 | 87 | 0.04* | 86 | 0.7* | 94 | 0.9£ | 73 | 0.06£ | 6.1 | 0.90 | 0.2 | 0.7¶ |
| (79–94) | (74–97) | (88–99) | (60–87) | (2.7–13.7) | (0.1–0.3) | |||||||||||
| Field | 62 | 4 | 21 | 31 | 75 | 89 | 94 | 60 | 6.5 | 0.3 | ||||||
| (65–84) | (78–99) | (88–100) | (46–73) | (2.6–16.6) | (0.2–0.42) |
PPV: Positive predictive value. NPV: Negative predictive value. LR: Likelihood Ratio.
* McNemar Test.
£ Relative Predictive Values Test.
¶ Differences in Diagnostic Likelihood Ratio Test