Performance of 5 algorithms in identifying prescription opioid–related deaths from the Statistics Canada Vital Statistics Death Database in 2010
| Algorithm | No. of deaths | Sensitivity (95% CI) | Specificity (95% CI) | Positive predictive value* (95% CI) | κ (95% CI) | AUC (95% CI) | LR (95% CI) | ||
|---|---|---|---|---|---|---|---|---|---|
| Coroner data | Vital statistics data | Positive | Negative | ||||||
| 1 | 542 | 425 | 0.72 (0.68–0.75) | 0.99 (0.99–0.99) | 0.91 (0.88–0.94) | 0.80 (0.77–0.83) | 0.86 (0.84–0.88) | 1701 (1108–2295) | 0.28 (0.24–0.33) |
| 2 | 542 | 447 | 0.75 (0.71–0.78) | 0.99 (0.99–0.99) | 0.90 (0.88–0.93) | 0.82 (0.79–0.84) | 0.87 (0.85–0.89) | 1525 (1032–2018) | 0.25 (0.22–0.29) |
| 3 | 542 | 437 | 0.72 (0.68–0.76) | 0.99 (0.99–0.99) | 0.89 (0.86–0.92) | 0.80 (0.77–0.82) | 0.86 (0.84–0.88) | 1347 (929–1764) | 0.28 (0.24–0.32) |
| 4 | 542 | 466 | 0.75 (0.71–0.79) | 0.99 (0.99–0.99) | 0.87 (0.84–0.90) | 0.81 (0.78–0.83) | 0.87 (0.86–0.89) | 1119 (810–1429) | 0.25 (0.21–0.29) |
| 5 | 542 | 692 | 0.89 (0.87–0.92) | 0.99 (0.99–0.99) | 0.70 (0.67–0.74) | 0.78 (0.76–0.81) | 0.94 (0.93–0.96) | 380 (324–436) | 0.11 (0.08–0.13) |
Note: AUC = area under the receiver operating characteristic curve, CI = confidence interval, LR = likelihood ratio.
↵* Negative predictive value = 0.99 for all algorithms.