| Reference | Brennan et al32 | O’Neil et al23 28 | Wilson et al3 | Thomas et al2 | Vincent et al33 | Davis et al24 30 | Baker et al31 | Sari et al29 | Median percentage (interquartile range) |
| No. of records | 30 121 | 3141 | 14 179 | 14 700 | 1014 | 6579 | 3745 | 1006 | – |
| No. of patients with at least one adverse event | 1133 (3.8) | 237* (7.5) | 2353 (16.6) | 475 (3.2) | 110 (10.8) | 850 (12.9) | 255 (6.8) | 110 (10.9) | 9.2 (4.6 to 12.4) |
| No. of adverse events (if >1 adverse event per patient) | – | – | – | – | 119 (11.7) | – | 289 (7.7) | 136 (13.5) | 11.7 (7.7 to 13.5) |
| No. of preventable adverse events | – | 103* (43.5) | 1205 (51.2) | – | 57 (47.9) | 315 (37.1) | 106 (41.6) | – | 43.5 (39.4 to 49.6) |
| Outcome | |||||||||
| No or minor disability† | 644 (56.8) | – | 1073 (45.6) | 253 (53.3) | 73 (66.4) | 524 (61.6) | 161 (55.7) | – | 56.3 (51.4 to 62.8) |
| Temporary disability‡ | 187 (16.5) | – | 702 (29.8) | 150 (31.6) | 21 (19.1) | 162 (19.0) | 36 (12.5) | – | 19.1 (15.5 to 30.3) |
| Permanent disability§ | 74 (6.5) | – | 315 (13.4) | 40 (8.4) | 7 (6.4) | 87 (10.2) | 15 (5.2) | – | 7.0 (6.1 to 11.0) |
| Death | 154 (13.6) | – | 112 (4.8) | 31 (6.6) | 9 (8.2) | 38 (4.5) | 46 (15.9) | – | 7.4 (4.7 to 14.2) |
| Unknown | 75 (6.6) | – | 151 (6.4) | – | – | 40 (4.7) | 31 (10.7) | – | 6.5 (5.1 to 9.7) |
Numbers (percentages) except last column. Interquartile range = 25th to 75th percentile. Brennan: no. of records, number of AEs and percentages of outcomes were given. Percentage of AEs and numbers of outcomes were calculated. No number or percentage was given for preventability. No. of AEs with negligence (failure to meet standards reasonably to be expected) was 280 (24.7%). O’Neil: two different strategies were used. No. of records, number and percentage of AEs per strategy and percentage of preventability per strategy were given. Total number and percentage of AEs and total number and percentage of preventable AEs were calculated. Bates: no. of records, number of AEs and percentage of AEs were given. Wilson: no. of records, number and percentage of AEs, percentage of preventability and numbers were given. No. of preventable AEs and percentages of outcomes were calculated. Thomas: no. of records, number of AEs and percentages of outcomes were given. Percentage of AEs and numbers of outcomes were calculated. No number or percentage was given for preventability. Percentage of AEs with negligence (failure to meet standards reasonably to be expected) was 32.6 for Utah and 27.5 for Colorado. Vincent: no. of records, number and percentage of AEs, number and percentage of preventability and numbers and percentages of outcomes were given. Davis: no. of records, number and percentage of AEs, number and percentage of preventability and percentages of outcomes were given. Numbers of outcomes were calculated. Baker: no. of records, number of AEs, number and percentage of preventability and numbers and percentages of outcomes were given. Percentage of AEs was calculated. Sari: two different strategies were used, numbers and percentages in table are combined results. No. of records and number of AEs were given. Percentage of AEs was calculated.
*O’Neil et al used two different strategies. Bates et al evaluated the same population using the same methods for retrospective record review and reported another incidence. The numbers and percentages given here represent the mean of the combined results from the publication by O’Neil et al and the results by Bates et al. Idem for the preventability numbers.
†No disability or disability resolved within 1 month; contains Thomas’ categories of “emotional,” “insignificant” and “minor temporary” and Baker’s categories of “none” and “minimal.”
‡Contains Brennan’s categories of “disability resolved within 1–6 months” and “within 6–12 months,” Thomas’ category of “major temporary” and Baker’s categories of “impairment resolved within 1–6 months” and “within 6–12 months.”
§Contains Thomas’ categories of “minor permanent,” “significant permanent,” “major permanent” and “grave.”