Table 4. Comparison of the performance criteria obtained during the collaborative study for the different methods (first stage of evaluation).
Methods | Analytical specificity ASP (%) ad | Diagnostic sensitivity DSE (%) bd | Diagnostic specificity DSP (%) cd | Significant variation between results produced by the method and theoretically expected results |
---|---|---|---|---|
M1 | (H1) 89.6 BC | (H1) 89.5 BCD | (H1) 89.7 AB | Yes for all criteria (p < 0.001) |
(83.0–93.8) | (83.0–93.8) | (83.1–93.9) | ||
(H2) 84.8 CDEF | (H2) 85.2 DE | (H2) 84.1 B | ||
(77.5–90.0) | (78.0–90.4) | (76.8–89.5) | ||
M2 | (H1) 87.5 BCDE | (H1) 84.4 DE | (H1) 92.6 AB | Yes for all criteria |
(79.9–92.5) | (76.4–90.1) | (86.1–96.2) | (H1) p < 0.001 for ASP and DSE, p = 0.006 for DSP | |
(H2) 83.7 DEF | (H2) 80.0 E | (H2) 89.8 AB | (H2) p < 0.001 for all criteria | |
(75.6–89.5) | (71.5–86.5) | (82.7–94.2) | ||
Ma | (H1) 88.3 BCDE | (H1) 84.0 DE | (H1) 95.6 AB | No for DSP ((H1) p = 0.494, (H2) p = 0.242), |
(75.8–94.8) | (70.7–91.9) | (85.2–98.8) | yes for ASP and DSE (p < 0.001) | |
(H2) 85.8 BCDEF | (H2) 81.3 DE | (H2) 93.3 AB | ||
(72.8–93.2) | (67.7–90.1) | (82.1–97.7) | ||
M3 | (H1) 79.8 EF | (H1) 86.7 CDE | (H1) 68.3 CD | Yes for all criteria (p < 0.001) |
(68.3–87.8) | (76.1–93.0) | (56.0–78.4) | ||
(H2) 77.4 F | (H2) 86.7 CDE | (H2) 61.9 D | ||
(65.7–86.0) | (76.1–93.0) | (49.6–72.9) | ||
M4 | (H1) 95.8 A | (H1) 96.7 A | (H1) 94.4 A | (H1) No for DSE (p = 0.060) and DSP (p = 0.059), |
(89.5–98.4) | (90.7–98.9) | (87.6–97.6) | yes for ASP (p = 0.002) | |
(H2) 92.5 AB | (H2) 94.7 AB | (H2) 88.9 AB | (H2) Yes for all criteria (p = 0.007 for DSE, p = 0.002 for DSP | |
(85.2–96.4) | (87.9–97.7) | (80.7–93.9) | and p < 0.001 for ASP) | |
M5 | (H1) 91.7 AB | (H1) 93.3 ABC | (H1) 88.9 AB | Yes for all criteria |
(82.8–96.3) | (84.4–97.3) | (78.8–94.5) | (H1) p = 0.014 for DSE, p = 0.013 for DSP and p < 0.001 for ASP | |
(H2) 88.7 BCD | (H2) 91.4 ABCD | (H2) 84.1 BC | (H2) p = 0.003 for DSE, p = 0.001 for DSP and p < 0.001 for ASP | |
(78.6–94.4) | (81.9–96.2) | (73.2–91.1) | ||
M6 | (H1) 95.8 A | (H1) 96.3 A | (H1) 95.1 A | (H1) No for DSE (p = 0.060) and DSP (p = 0.120), |
(89.0–98.5) | (89.7–98.7) | (88.0–98.1) | yes for ASP (p = 0.004) | |
(H2) 89.4 BCD | (H2) 92.6 ABC | (H2) 84 BC | (H2) Yes for all criteria (p = 0.002 for DSE, p = 0.001 for DSP | |
(80.8–94.4) | (84.8–96.6) | (74.5–90.4) | and p < 0.001 for ASP) |
a Analytical specificity (95% confidence interval): ability of the method to detect the target when it is present in the sample and to fail to detect the target when it is not present in the sample. Values followed by the same letter in a column are not significantly different (p = 0.05) according to Fisher’s exact test (two tailed).
b Diagnostic sensitivity (95% confidence interval): ability of the method to detect the target when it is present in the sample. Values followed by the same letter in a column are not significantly different (p = 0.05) according to Fisher’s exact test (two tailed).
c Diagnostic specificity (95% confidence interval): ability of the method to fail to detect the target when it is not present in the sample. Values followed by the same letter in a column are not significantly different (p = 0.05) according to Fisher’s exact test (two tailed).
d For each criterion, we present data derived from the two scenarios described in the Materials & Methods section for the interpretation of indeterminate results.