Table 3. Multinomial logistic regression analysis (N=760).
| Complications, OR (95% CI) | ||
|---|---|---|
| Minor (n=259) | Major (n=31) | |
| Unadjusted | ||
| Superomedial pedicle | 1.00 | 1.00 |
| Inferior pedicle | 1.93 (1.37–2.72) | 0.73 (0.29–1.82) |
| Superior pedicle | 0.58 (0.33–1.02) | – |
| Model 1 | ||
| Superomedial pedicle | 1.00 | 1.00 |
| Inferior pedicle | 2.03 (1.43–2.89) | 0.77 (0.30–1.94) |
| Superior pedicle | 0.66 (0.37–1.16) | – |
| Model 2 | ||
| Superomedial pedicle | 1.00 | 1.00 |
| Inferior pedicle | 1.92 (1.35–2.72) | 0.73 (0.29–1.83) |
| Superior pedicle | 0.76 (0.43–1.35) | – |
| Model 1+2 | ||
| Superomedial pedicle | 1.00 | 1.00 |
| Inferior pedicle | 2.03 (1.41–2.91) | 0.75 (0.29–1.92) |
| Superior pedicle | 0.85 (0.47–1.53) | – |
Reference for minor and major complications was no complications (n=470). Differences between complication groups were tested using multinomial logistic regression analysis. Model 1 was adjusted for age, body mass index, and comorbidity (cardiovascular disease, asthma/COPD, DM, other). Model 2 was adjusted for operative details [laterality, resected tissue (resg upper quartile), drain, operating surgeon]. N, number of participants; n, number of participants in the subgroup; OR, odds ratio; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus.