| Study | Setting | Admissions with ⩾1 AE*% (95% CI) | AEs that were preventable†% (95% CI) | AEs that were associated with death% (95% CI) | Admissions associated with potentially preventable death (%) |
| Present study | 21 Dutch hospitals, 7926 hospital admissions (2004) | 5.7 (5.1 to 6.4) | 39.6 | 7.6 | 0.12 |
| Andres et al19 | 24 Spanish hospitals, 5624 hospital admissions (2005) | 8.4 (7.7 to 9.1) | 42.6 | 4.4 | 0.19 ** |
| Michel et al8 | 7 hospitals in France, 778 hospital admissions (2002) | 14.5 | 27.6 | NR | NR |
| Schioler et al10 | 17 hospitals in Denmark, 1097 hospital admissions (2001) | 9.0 | 40.4 | 4.9 | 0.178§ |
| Baker et al2 | 20 Canadian hospitals, 3745 hospital admissions (2000) | 7.5 (5.7 to 9.3) | 36.9 | 20.8 | 0.66 ¶ |
| Vincent et al12 | 2 hospitals in London, England, 1014 hospital admissions (1998) | 10.8 | 48 | 8 | 0.41§ |
| Davis et al4 5 | 13 hospitals in New Zealand, 6579 hospital admissions (1998) | 12.9 | 37.1 | 4.5 | 0.28 ** |
| Thomas et al11 | 28 hospitals in Utah and Colorado, 14 700 hospital admissions (1992) | 2.9 (0.2)*** | 50 | 6.6 | 0.13 ** |
| Wilson et al13 | 28 hospitals in New South Wales and South Australia, 14 179 hospital admissions (1992) | 16.6 (15.2 to 17.9) | 51.2 | 4.9 | 0.55 ** |
| Brennan et al1 | 51 hospitals in New York, 30 195 hospital admissions (1984) | 3.7 (3.2 to 4.2) | 27.6‡ | 13.6 | 0.26 ** |
| Leape et al3 |
NR, not reported
*Most of the studies used causation score ⩾4, except Davis et al45 and Wilson et al.13 They used causation score ⩾2. Not all studies reported 95% CIs.
†All studies reported on high preventable AEs (preventability score ⩾4).
‡ Measured negligence instead of preventability.
¶Reported.2
§Estimated by calculation: proportion of AEs × proportion of preventable AEs × proportion that contributed to death.
**Estimated by calculation: proportion of AEs × proportion of preventable AEs that contributed to death.
***SD.