Table 2 Reporting of essential items applicable to both derivation and validation studies
Characteristic | Number (%) N = 100 |
|---|---|
General information | |
 Rationale provided for the derivation or validation of the model | |
  Yes | 66 (66.0) |
 Source of data | |
  Prospective cohort | 60 (60.0) |
  Retrospective cohort—registry data | 17 (17.0) |
  Retrospective cohort—data from past studies | 12 (12.0) |
  Retrospective cohort—chart review | 10 (10.0) |
  Unclear or not reported | 1 (1.0) |
 Inclusion criteria (definition used for selection) | |
  TIA—Classical time-based definition | 92 (92.0) |
  TIA—Tissue-based definition | 5 (5.0) |
  Minor Stroke—Classical time-based definition | 8 (8.0) |
  Minor Stroke—Tissue-based definition | 1 (1.0) |
  Unclear or not reported | 7 (7.0) |
 Time-based versus tissue-based inclusion criteria | |
  Time-based TIA/minor stroke | 88 (88.0) |
  Tissue-based TIA/minor stroke | 5 (5.0) |
  Unclear or not reported | 7 (7.0) |
 TIA versus minor stroke inclusion criteria | |
  TIA only | 85 (85.0) |
  Both TIA and minor stroke | 8 (8.0) |
  Unclear or not reported | 7 (7.0) |
 Definition used for outcome of stroke | |
  Clinical definition | 73 (73.0) |
  Tissue-based definition | 7 (7.0) |
  Unclear or not reported | 20 (20.0) |
 Time of outcome prediction | |
  2-day | 28 (28.0) |
  3-day | 2 (2.0) |
  7-day | 66 (66.0) |
  14-day | 2 (2.0) |
  28-day | 4 (4.0) |
  30-day | 16 (16.0) |
  90-day | 73 (73.0) |
  Additional outcome periods considered in study | 8 (8.0) |
   1-year | 2 (2.0) |
   3-year | 4 (4.0) |
   14-year | 2 (2.0) |
 Recruitment method | |
  Consecutive participants | 50 (50.0) |
  Nonconsecutive sample | 1 (1.0) |
  Unclear or not reported | 49 (49.0) |
 Number of sites | |
  1 | 41 (41.0) |
  2 | 2 (2.0) |
  3 | 3 (3.0) |
  4 | 0 (0.0) |
  5–9 | 13 (13.0) |
  10+ | 38 (38.0) |
  Unclear or not reported | 3 (3.0) |
 Setting | |
  Tertiary | 48 (48.0) |
  Community | 10 (10.0) |
  Tertiary and community | 7 (7.0) |
  Unclear or not reported | 35 (35.0) |
 Location | |
  Urban | 50 (50.0) |
  Rural | 0 (0.0) |
  Urban and rural | 8 (8.0) |
  Unclear or not reported | 42 (42.0) |
 Study dates were provided | |
  Yes | 95 (95.0) |
Outcome to be predicted | |
 Same outcome definition (and method of measurement) used in all patients | |
  Yes | 44 (44.0) |
 Explicitly stated that outcome was assessed without knowledge of the candidate predictors (i.e. blinded) | |
  Yes | 24 (24.0) |
Candidate predictors | |
 All measurement predictors defined with information on how to measure | |
  Yes | 63 (63.0) |
 All measurement predictors defined in terms of when to measure (e.g. pre-hospital, blood pressure measurement time, etc) | |
  Yes | 47 (47.0) |
 Explicitly stated that predictors were assessed blinded for outcome, and for each other | |
  Yes | 12 (12.0) |
Sample size | |
 Sample size justification provided (practical justification such as using existing study cohort or an RCT data is considered justified) | |
  Yes | 49 (49.0) |
 Flow of participants provided through a flow diagram |  |
  Yes | 21 (21.0) |
Missing data | |
 Number of participants with any missing value (on any of predictors and outcomes) reported | |
  Yes | 34 (34.0) |
 Number of participants with missing data for each predictor reported | |
  Yes | 28 (28.0) |
 Number of participants lost to follow-up reported* | |
  Yes | 17 (27.9) |
  No, unclear or not reported | 44 (72.1) |
  Not applicable (retrospective cohort) | 39 (39.0) |
 Handling of missing data* | |
  Complete case analysis | 14 (15.1) |
  Predictor with missing values omitted | 0 (0.0) |
  Single imputation | 0 (0.0) |
  Multiple imputation | 1 (1.1) |
  Not handled, unclear, or not reported | 78 (83.9) |
 Not applicable (there were no missing data) | 7 (7.0) |
Model performance | |
 Method used for calibration | |
  Calibration plot | 0 (0.0) |
  Calibration slope | 0 (0.0) |
  Hosmer-Lemeshow test | 5 (5.0) |
  Unclear or not reported | 95 (95.0) |
 Method used for discrimination | |
  C-statistic/AUC-ROC | 79 (79.0) |
  D-statistic | 0 (0.0) |
  Log-rank | 2 (2.0) |
  Unclear or not reported | 19 (19.0) |
 Classification measures reported | |
  Sensitivity | 28 (28.0) |
  Specificity | 28 (28.0) |
  Predictive values | 11 (11.0) |
  Net reclassification improvement | 13 (13.0) |
  Unclear or not reported | 60 (60.0) |
 A priori cut points used for the classification measures | |
  Yes | 25 (25.0) |
 Confidence intervals (CIs) provided for the performance measures | |
  Yes | 65 (65.0) |