Table 2 Evaluation of the processing quality of concepts by type of variable on the test set
Chief complaint | Observation | Specialists' notes | Procedures | Prescriptions | Total | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
N* | % | 95% CI | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI | |
Correctly-processed concepts | 117 | 91.4 | 85.1-95.6 | 1,481 | 85.3 | 83.6-86.9 | 11 | 73.3 | 44.9-92.2 | 53 | 88.3 | 77.4-95.2 | 12 | 92.3 | 64.0-99.8 | 1,674 | 85.8 | 84.1-87.3 |
Missing code in the ECMT § | 0 | - | - | 6 | 0.3 | 0.1-0.8 | 0 | - | - | 0 | - | - | 0 | - | 6 | 0.3 | 0.1-0.7 | |
Missing code in UrgIndex filters | 0 | - | - | 16 | 0.9 | 0.5-1.5 | 0 | - | - | 0 | - | - | 1 | 7.7 | 0.2-36.0 | 17 | 0.9 | 0.5-1.4 |
Term not recognized by UrgIndex † | 6 | 4.7 | 1.7-9.9 | 176 | 10.1 | 8.8-11.7 | 1 | 6.7 | 0.2-31.9 | 7 | 11.7 | 4.8-22.6 | 0 | - | 190 | 9.7 | 8.5-11.1 | |
Negation not recognized by UrgIndex | 0 | - | - | 9 | 0.5 | 0.2-1.0 | 0 | - | - | 0 | - | - | 0 | - | 9 | 0.5 | 0.2-0.9 | |
Other UrgIndex error | 5 | 3.9 | 1.3-8.9 | 48 | 2.8 | 2.0-3.6 | 3 | 20 | 0.4-48.1 | 0 | - | - | 0 | - | 56 | 2.9 | 2.2-3.7 | |
Total number of concepts | 128 | 100 | - | 1,736 | 100 | - | 15 | 100 | - | 60 | 100 | - | 13 | 100 | - | 1,952 | 100 | - |