TABLE 4.
Results and calculated specificities of the commercial methods for the detection of anti-measles IgM antibodies evaluated with the non-measles sera panel (n = 187), by subseta
| Method | No. of positive or equivocal results (specificity, %) |
95% CI of specificity | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Chikungunya (n = 4) | Dengue (n = 34) | Parvovirus B19 (n = 35) | Roseola (n = 38) | Rubella (n = 36) | Zika (n = 3) | Unknownb (n = 37) | Total (n = 187) | ||
| Enzygnost | 0 (100) | 0 (100) | 7 (80.0) | 2 (94.7) | 0 (100) | 0 (100) | 0 (100) | 9 (95.2) | 90.8–97.6 |
| Euroimmun | 1 (75) | 0 (100) | 6 (82.9) | 1 (97.4) | 0 (100) | 0 (100) | 0 (100) | 8 (95.7) | 91.4–98.0 |
| Euroimmun Nucleoprotein | 0 (100) | 0 (100) | 1 (97.1) | 0 (100) | 0 (100) | 0 (100) | 1 (97.3) | 2 (98.9) | 95.8–99.8 |
| IBL | 0 (100) | 2 (94.1) | 0 (100) | 0 (100) | 3 (91.7) | 1 (66.7) | 1 (97.3) | 7 (96.3) | 92.1–98.3 |
| LIAISON XL | 0 (100) | 0 (100) | 1 (97.1) | 0 (100) | 0 (100) | 0 (100) | 0 (100) | 1 (99.5d) | 96.6–100 |
| Microimmune | 0 (100) | 1 (97.1) | 0 (100) | 2 (94.7) | 0 (100) | 1 (66.7) | 1 (97.3) | 5 (97.3) | 93.5–99.0 |
| NovaLisa | 0 (100) | 2 (94.1) | 15 (57.1) | 2 (94.7) | 0 (100) | 1 (66.7) | 1 (97.3) | 21 (88.8d) | 83.1–92.8 |
| Serion (activity calculator)c | 0 (100) | 0 (100) | 16 (54.3) | 2 (94.7) | 0 (100) | 2 (33.3) | 1 (97.3) | 21 (88.8d) | 83.1–92.8 |
| Serion (OD range)c | 0 (100) | 0 (100) | 15 (57.1) | 2 (94.7) | 0 (100) | 2 (33.3) | 1 (97.3) | 20 (89.3d) | 83.7–93.2 |
| Serion (special case formula)c | 1 (75) | 0 (100) | 17 (51.4) | 3 (92.1) | 0 (100) | 2 (33.3) | 2 (94.6) | 25 (86.6d) | 80.7–91.0 |
a
Specimens with equivocal results were counted as positive.
b
This panel of sera included fever/rash illness of unknown etiology.
c
Three methods of sample result determination, using the single set of optical density data from the test plates, were provided in the manufacturer’s IFU. All three methods were evaluated.
d
Significant difference (P < 0.05) between the most specific (LIAISON XL) and the least specific methods (Serion, all three result determination methods, and NovaLisa) based on nonoverlapping 95% confidence intervals.