Table 5.
Risk of anastomotic leakage confirmed at reoperation based on multivariate logistic regression analysis
| Variable | P | Odds ratio (95% CI) |
|---|---|---|
| NSAID use and type | <0.001 | — |
| None | — | 1 |
| Ibuprofen | 0.18 | 1.54 (0.82 to 2.86) |
| Diclofenac | <0.001 | 7.16 (3.82 to 13.4) |
| Intraoperative transfusion | <0.001 | 7.00 (4.51 to 10.9) |
| Rectal v colonic resection | <0.001 | 2.26 (1.62 to 3.15) |
| Male sex | 0.003 | 1.65 (1.18 to 2.30) |
| Surgical centre* | 0.003 | — |
| 1 | — | 1 |
| 2 | 0.021 | 0.09 (0.01 to 0.69) |
| 3 | 0.98 | 0.99 (0.51 to 1.91) |
| 4 | 0.77 | 1.09 (0.62 to 1.91) |
| 5 | 0.21 | 1.50 (0.80 to 2.77) |
| 6 | 0.007 | 2.10 (1.23 to 3.58) |
Model included the variables from univariate analyses (table 3) with P<0.1; and the significant interaction between NSAID use and drug type, and intraoperative transfusion (table 4).
*Centres numbered from 1 to 6.