Skip to main content
. 2024 Aug 27;24:584. doi: 10.1186/s12888-024-06022-5

Table 4.

Validation of algorithm results

# % confirmed
Excluded clozapine treatment periods
 Total number of periods 425
 Number of periods randomly selected for validation (%) 90 (21.2%)
 Number of confirmed periodsa 76 84.4%
Clozapine treatment periods with identified start-date
 Total number of periods 1,902
 Number of periods randomly selected for validation (%) 211 (11.1%)
 Number of confirmed periods 210 99.5%
 Start-date accuracyb
  Highly accurate (range of up to 15 days) 202 96.2%
  Accurate (range of 15–30 days) 3 1.4%
  Partially accurate (range of 30–60 days) 2 1.0%
  Inaccurate (range of more than 60 days) 3 1.4%
 End-date accuracyb
  Highly accurate (range of up to 15 days) 67 71.3%
  Accurate (range of 15–30 days) 13 13.8%
  Partially accurate (range of 30–60 days) 13 13.8%
  Inaccurate (range of more than 60 days) 1 1.1%
  No end-date to detectc 116 (55.2%)
Clozapine treatment periods with undetermined start-date
 Total number of periods 885
 Number of periods randomly selected for validation (%) 114 (12.9%)
 Number of confirmed periods 114 100%

aConfirmed excluded periods are treatment periods found to be "empty", meaning no clozapine was prescribed or total length of treatment was shorter than 7 days

bValidated for periods with identified start-date only

cEnd-date could not be detected as patients were either transferred outside of SLaM or reached the end of the study observation window