Table 5.
Severity of potential harm from clinical prescribing errors by study period.
| Potential harma severity | Control—Paper | Intervention—First 70 days of eMM | Intervention—One-year post-eMM | |||
|---|---|---|---|---|---|---|
| n | % (95% CI) | n | % (95% CI) | n | % (95% CI) | |
| 1: No/minimal harm | 382 | 27.3% (24.5–30.1) | 797 | 27.1% (25.2–29.1) | 281 | 26.9% (23.7–30.3) |
| 2: Temporary harm requiring monitoring | 664 | 47.5% (44.7–50.3) | 1333 | 45.3% (43.4–47.3) | 463 | 44.4% (41.1–47.7) |
| 3: Temporary harm requiring intervention | 299 | 21.4% (18.6–24.2) | 686 | 23.3% (21.4–25.3) | 266 | 25.5% (22.2–28.8) |
| 4: Permanent harm requiring intervention | 48 | 3.4% (0.6–6.3) | 112 | 3.8% (1.9–5.8) | 28 | 2.7% (0–6.0) |
| 5: Potential death | 6 | 0.4% (0–3.3) | 14 | 0.5% (0.0–2.5) | 5 | 0.5% (0.0–3.8) |
| Total of prescribing errors with a score ≥3 | 353 | 25.2% (23.0–27.6) | 812 | 27.6% (26.0–29.3) | 299 | 28.7% (25.9–31.5) |
aDefinitions provided in Supplementary Table 6. Confidence intervals (CIs) for the potential harm categories are 95% simultaneous multinomial CIs. For the total, CIs are exact (Clopper–Pearson) binomial 95% CIs.