TableĀ 2.
Blind agreement by GP clinical experts* |
||||
---|---|---|---|---|
Gold standard | Records included | Respiratory consultations identified | Agreement if consultation is respiratory or not respiratory | Complete agreement for all respiratory classifications included in consultations |
Training set (set 1) | 1200 | 529 (0.44; 0.40 to 0.48) | 1139 (0.95; 0.93 to 0.97) | 1037 (0.86; 0.84 to 0.89) |
Test set (set 2) | 1200 | 556 (0.46; 0.42 to 0.51) | 1146 (0.96; 0.94 to 0.97) | 1060 (0.88; 0.85 to 0.91) |
Validation set (set 3) | 1200 | 553 (0.46; 0.42 to 0.51) | 1151 (0.96; 0.94to 0.98) | 1046 (0.87; 0.84 to 0.90) |
Data are n (proportion; 95% CI).
*Consensus was reached for all records with discordant classifications following initial independent coding.
GP, general practitioner.