The incidence, causes, risk factors, and predictive nomogram of 30-day unplanned reoperation after spinal surgery: a multi-center retrospective study.
TL;DR
The newly developed nomogram exhibits good discrimination, favourable calibration, and superior net clinical benefit, enabling accurate individualized prediction of short-term unplanned reoperation.
Abstract
Background
This study aims to elucidate the demographic characteristics, incidence, causes, and risk factors of unplanned reoperations post-spinal surgery (URPS) within 30 days, and to develop and validate a nomogram-based prediction model.
Methods
We retrospectively analysed data from patients who underwent spinal surgery at three tertiary hospitals between January 2013 and January 2023, with a particular focus on those requiring unplanned reoperation within 30 days postoperatively. Demographic characteristics, incidence rates, and temporal patterns of URPS were assessed. A detailed review of causes and presenting symptoms across different postoperative intervals was also performed. Furthermore, we investigated risk factors for URPS and constructed a predictive nomogram.
Results
Among 77,026 individuals included, 476 cases of URPS within 30 days were identified, comprising 258 males (54.2%) and 218 females (45.8%). Patient ages ranged from 4 to 84 years, with a mean of 48.98 ± 14.75 years. The overall incidence of URPS was 0.62%, showing a gradual annual increase. The leading causes were surgical site infection (SSI) and symptomatic epidural haematoma (SEH), which predominantly manifested as signs of surgical infection, radicular pain, and motor paralysis. The acute phase (postoperative days 2-10) represented the most critical period for symptom presentation, whereas SEH occurred mainly during the hyper-acute phase (days 0-1). Independent risk factors for URPS included drinking (p = 0.027, OR: 1.36, 95% CI: 1.040-1.790), diabetes mellitus (DM) (p < 0.001, OR: 2.056, 95% CI: 1.580-2.676), length of hospital stay (LOS) ≥ 7 days (p < 0.001, OR: 2.269, 95% CI: 1.566-3.288), first operation time (FOT) ≥ 3 h (p = 0.008, OR: 1.396, 95% CI: 1.093-1.783), drainage volume (DV) ≥ 300 ml (p < 0.001, OR: 5.156, 95% CI: 3.854-6.899), and disease type (DT) (p < 0.001, OR: 1.571, 95% CI: 1.382-1.787). The predictive nomogram demonstrated good discriminative ability, with an area under the receiver operating characteristic curve (AUC) of 0.752 (95% CI: 0.726-0.778). Calibration curves showed excellent agreement between predicted and observed probabilities (Hosmer-Lemeshow test, p = 0.892, Brier score = 0.185, R2 = 0.232), and decision curve analysis (DCA) confirmed net clinical benefit across a wide range of threshold probabilities. Patients who experienced URPS had significantly higher nomogram scores than those who did not (p < 0.001), validating the model's risk-stratification capacity.
Conclusions
This study delineates the incidence, causes, and independent risk factors for URPS within 30 days. The newly developed nomogram exhibits good discrimination, favourable calibration, and superior net clinical benefit, enabling accurate individualized prediction of short-term unplanned reoperation. This predictive tool can assist clinicians in preoperative risk stratification, identification of high-risk patients, and design of tailored intervention strategies, thereby potentially reducing the incidence of unplanned reoperations and improving postoperative outcomes.