Table 3.
Estimates of logistic regression for binary outcome model of surgical mortality.
| No. of survivors | No. of non-survivors | Adjusted OR (95% CI) | p value | |
|---|---|---|---|---|
| Nationality | ||||
| Israeli | 853/872 (97.8%) | 19/872 (2.2%) | 1.0 [ref.] | |
| Palestinian | 199/207 (96.1%) | 8/207 (3.9%) | 1.52 (0.64 to 3.63) | 0.346 |
| STS complexity score | ||||
| 1 | 349/351 (99.4%) | 2/351 (0.6%) | 1.0 [ref.] | |
| 2 | 360/362 (99.4%) | 2/362 (0.6%) | 0.94 (0.13 to 6.69) | 0.946 |
| 3 | 123/127 (96.9%) | 4/127 (3.1%) | 5.42 (0.98 to 30.05) | 0.053 |
| 4 | 186/198 (93.9%) | 12/198 (6.1%) | 10.61 (2.34 to 48.22) | 0.002 |
| 5 | 34/41 (82.9%) | 7/41 (17.1%) | 34.99 (6.98 to 175.42) | < 0.001 |
| Constant | 0.005 | < 0.001 |
Abbreviations: OR, odds ratio; CI, confidence interval; STS, Society of Thoracic Surgeons.
Notes: The model also included patient's sex and natural logarithm of age (days) at surgery – omitted because of low statistical significance. The Hosmer and Lemeshow test for goodness of fit yielded p = 0.87.