Table 5. Results of analytical sensitivity (ASE), repeatability (DA), reproducibility (CO) and concordance odds ratio (COR) obtained during the collaborative study for the different methods.
| Number of results and criteria | Results obtained for each method | |||||||
|---|---|---|---|---|---|---|---|---|
| M1 | M2 | Maa | M3 | M4 | M5 | M6 | ||
| Number of positive results per /number of results (per dilution level and per method) and p-value (exact binomial test) and significance with the theoretical detection level of 95%cd | D1b | 67/75 | 32/75 | 23/30 | 58/75 | 62/75 | 75/75 | 75/75 |
| (0.034 S*) | (< 0.001 S***) | (< 0.001 S***) | (< 0.001 S***) | (< 0.001 S***) | (1.000 NS) | (1.000 NS) | ||
| D2 | 68/75 | 29/75 | 26/30 | 66/75 | 74/75 | 75/75 | 72/75 | |
| (0.081 NS) | (< 0.001 S***) | (0.061 NS) | (0.012 S*) | (0.979 NS) | (1.000 NS) | (0.730 NS) | ||
| D3 | 56/75 | 25/75 | 24/30 | 70/75 | 72/75 | 73/75 | 68/75 | |
| (< 0.001 S***) | (< 0.001 S***) | (< 0.001 S***) | 0.321 NS | 0.730 NS | 0.894 NS | 0.081 NS | ||
| D4 | 38/75 | 20/75 | 21/30 | 65/75 | 62/75 | 69/75 | 63/75 | |
| (< 0.001 S***) | (< 0.001 S***) | (< 0.001 S***) | 0.004 S** | (< 0.001 S***) | 0.172 NS | (< 0.001 S***) | ||
| D5 | 48/75 | 15/75 | 22/30 | 50/75 | 55/75 | 65/75 | 53/75 | |
| (< 0.001 S***) | (< .001 S***) | (< 0.001 S***) | (< 0.001 S***) | (< 0.001 S***) | (0.004 S**) | (< 0.001 S***) | ||
| Overall number of results per method | 375 | 375 | 150 | 375 | 375 | 375 | 375 | |
| Overall number of positive results per method | 277 | 121 | 116 | 309 | 325 | 357 | 331 | |
| Analytical sensitivity « overall ASE » (%)e | 73.9 B | 32.3 C | 77.3 AB | 82.4 AB | 86.7 AB | 95.2 A | 88.3 AB | |
| (69.2–78.1) | (27.7–37.2) | (70.0–83.3) | (78.2–85.9) | (82.9–89.7) | (92.5–96.9) | (84.6–91.1) | ||
| Repeatability DA (%)f | 81.7 | 92.5 | 91.0 | 88.1 | 90.6 | 94.9 | 88.5 | |
| (77.9–87.5) | (88.7–96.5) | (89.7–92.4) | (84.2–94.1) | (85.5–96.5) | (89.7–100) | (81.5–97.1) | ||
| Reproducibility CO (%)g | 71.9 | 52.1 | 57.9 | 72.5 | 82.9 | 93.0 | 86.4 | |
| (64.7–79.6) | (42.7–59.7) | (47.7–68.3) | (64.6–80.3) | (75.3–92.0) | (86.0–100) | (78.4–95.7) | ||
| Odds ratio CORh | 1.74 | 11.41 | 7.36 | 2.79 | 1.99 | 1.39 | 1.22 | |
| (1.46–2.07) | (9.18–14.18) | (4.89–11.08) | (2.30–3.39) | (1.60–2.48) | (1.03–1.88) | (0.98–1.51) | ||
| p-value (Fisher) for significance in performance between laboratories c | 2.06e-05 S*** | 4.09e-48 S*** | NAa | 1.54e-09 S*** | 8.24e-04 S*** | 0.08 NS | 0.60 NS | |
aValues for method Ma were calculated from the results of 2 laboratories (vs 5 for other methods). Consequently, some calculations were not implemented (NA value).
bD1: dilution 1.0·10−1 (= samples A1-B1-C1); D2: dilution 1.0·10−2 (= samples A2-B2-C2), D3: dilution 3.3·10−3 (= samples A3-B3-C3); D4: dilution 1.1·10−3 (= samples A4-B4-C4); D5: dilution 3.7·10−4 (= samples A5-B5-C5)
c NS: not significant (p ≥ 0.05); S*: 0.01 ≤ p < 0.05; S**: 0.001 ≤ p < 0.01; S***: p < 0.001
dThe shaded cells indicate a statistical significance of variation with the theoretical detection level of 95%
eOverall analytical sensitivity (95% confidence interval): is the overall probability of detection calculated per method as the ratio between positive results and the number of results obtained per method for this second stage of the evaluation. Values followed by the same letter in a column are not significantly different (p = 0.05) according to Fisher’s exact test (two tailed)
fRepeatability (bootstrap confidence interval): is the probability of finding the same result from two identical test portions analyzed in the same laboratory, under repeatability conditions
gReproducibility (bootstrap confidence interval): is the percentage chance of finding the same result for two identical samples analyzed in two different laboratories
hThe concordance odds ratio (95% confidence interval) was calculated as the ratio DA x (1—CO)/CO x (1- DA). The larger the odds ratio (above 1), the more predominant interlaboratory variation