Table 2.
Comparisons of different Symptom Pattern variants based on 100 splits for the adult, child, and neonate modules, including use of health care experience information
| Adult module: | |||||
|---|---|---|---|---|---|
| Cluster | Single/Multiple | Symptom | Prior | CSMF accuracy (95% uncertainty interval [UI]) | Chance-corrected concordance (%) (95% UI) |
| 10 | Single | Top 40 | Uniform | 0.726 (0.714, 0.737) | 47.8 (47.4, 48.2) |
| 10 | Single | All | Uniform | 0.703 (0.687, 0.718) | 45.6 (44.9, 46.3) |
| 10 | Single | Top 40 | King-Lu | 0.653 (0.640, 0.672) | 42.6 (42.1, 43.4) |
| 10 | Single | All | King-Lu | 0.311 (0.291, 0.349) | 18.4 (17.4, 20.3) |
| 10 | Multiple | All | Uniform | 0.714 (0.697, 0.721) | 46.1 (45.7, 46.5) |
| 10 | Multiple | All | King-Lu | 0.708 (0.696, 0.719) | 46.0 (45.6, 46.6) |
| 1 | Single | Top 40 | Uniform | 0.668 (0.652, 0.681) | 42.7 (42.2, 43.0) |
| 1 | Single | All | Uniform | 0.632 (0.620, 0.643) | 40.3 (39.8, 40.5) |
| 1 | Single | Top 40 | King-Lu | 0.163 (0.147, 0.212) | 9.3 (8.3, 11.0) |
| 1 | Single | All | King-Lu | 0.043 (0.031, 0.057) | 0.4 (0.0, 0.9) |
| 1 | Multiple | All | Uniform | 0.651 (0.636, 0.665) | 39.2 (38.4, 39.4) |
| 1 | Multiple | All | King-Lu | 0.646 (0.630, 0.664) | 38.6 (38.1, 39.2) |
| Child module: | |||||
| Cluster | Single/Multiple | Symptom | Prior | CSMF accuracy (95% UI) | Chance-corrected concordance (%) (95% UI) |
| 10 | Single | Top 40 | Uniform | 0.718 (0.699, 0.738) | 45.2 (44.4, 46.2) |
| 10 | Single | All | Uniform | 0.740 (0.727, 0.757) | 50.9 (50.1, 51.8) |
| 10 | Single | Top 40 | King-Lu | 0.633 (0.617, 0.666) | 40.4 (39.2, 40.9) |
| 10 | Single | All | King-Lu | 0.469 (0.453, 0.516) | 36.8 (35.4, 38.0) |
| 10 | Multiple | All | Uniform | 0.749 (0.736, 0.766) | 51.8 (50.7, 52.9) |
| 10 | Multiple | All | King-Lu | 0.759 (0.745, 0.771) | 52.1 (51.5, 53.0) |
| 1 | Single | Top 40 | Uniform | 0.696 (0.676, 0.715) | 44.4 (44.0, 45.5) |
| 1 | Single | All | Uniform | 0.705 (0.692, 0.727) | 46.9 (45.6, 47.5) |
| 1 | Single | Top 40 | King-Lu | 0.263 (0.228, 0.280) | 16.6 (14.2, 17.7) |
| 1 | Single | All | King-Lu | 0.125 (0.104, 0.161) | 3.6 (2.3, 4.5) |
| 1 | Multiple | All | Uniform | 0.716 (0.701, 0.733) | 47.9 (46.5, 48.7) |
| 1 | Multiple | All | King-Lu | 0.723 (0.705, 0.741) | 47.9 (47.1, 48.6) |
| Neonate module: | |||||
| Cluster | Single/Multiple | Symptom | Prior | CSMF accuracy (95% UI) | Chance-corrected concordance (%) (95% UI) |
| 10 | Single | Top 40 | Uniform | 0.748 (0.730, 0.766) | 29.7 (28.7, 30.6) |
| 10 | Single | All | Uniform | 0.741 (0.720, 0.787) | 31.7 (31.2, 33.0) |
| 10 | Single | Top 40 | King-Lu | 0.679 (0.647, 0.704) | 27.9 (25.9, 28.5) |
| 10 | Single | All | King-Lu | 0.603 (0.553, 0.624) | 19.1 (18.0, 21.8) |
| 10 | Multiple | All | Uniform | 0.732 (0.712, 0.745) | 34.1 (32.8, 35.5) |
| 10 | Multiple | All | King-Lu | 0.736 (0.711, 0.752) | 33.6 (32.9, 35.5) |
| 1 | Single | Top 40 | Uniform | 0.663 (0.634, 0.691) | 28.8 (27.4, 29.6) |
| 1 | Single | All | Uniform | 0.604 (0.571, 0.639) | 26.4 (25.2, 27.6) |
| 1 | Single | Top 40 | King-Lu | 0.425 (0.391, 0.462) | 10.0 (9.2, 11.6) |
| 1 | Single | All | King-Lu | 0.363 (0.325, 0.384) | 0.0 (0.0, 0.5) |
| 1 | Multiple | All | Uniform | 0.564 (0.550, 0.580) | 29.5 (27.7, 30.4) |
| 1 | Multiple | All | King-Lu | 0.565 (0.541, 0.591) | 29.4 (27.8, 30.8) |