Ju JW, Jang HS, Lee M, Lee HJ, Kwon W, Jang JY. Early postoperative fever as a predictor of pancreatic fistula after pancreaticoduodenectomy: a single-center retrospective observational study.
BMC Surg 2024;
24:229. [PMID:
39134979 PMCID:
PMC11318233 DOI:
10.1186/s12893-024-02521-0]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/03/2024] [Accepted: 07/30/2024] [Indexed: 08/15/2024] Open
Abstract
BACKGROUND
The connection between early postoperative fever and clinically relevant postoperative pancreatic fistula (CR-POPF) after pancreaticoduodenectomy remains unclear. This study aimed to investigate this association and assess the predictive value of early postoperative fever for CR-POPF.
METHODS
This retrospective observational study included adult patients who underwent pancreaticoduodenectomy at a tertiary teaching hospital between 2007 and 2019. Patients were categorized into those with early postoperative fever (≥ 38 °C in the first 48 h after surgery) and those without early postoperative fever groups. Weighted logistic regression analysis using stabilized inverse probability of treatment weighting (sIPTW) and multivariable logistic analysis were performed. The c-statistics of the receiver operating characteristic curves were calculated to evaluate the impact on the predictive power of adding early postoperative fever to previously identified predictors of CR-POPF.
RESULTS
Of the 1997 patients analyzed, 909 (45.1%) developed early postoperative fever. The overall incidence of CR-POPF among all the patients was 14.3%, with an incidence of 19.5% in the early postoperative fever group and 9.9% in the group without early postoperative fever. Early postoperative fever was significantly associated with a higher risk of CR-POPF after sIPTW (adjusted odds ratio [OR], 1.73; 95% confidence interval [CI], 1.34-2.22; P < 0.001) and multivariable logistic regression analysis (adjusted OR, 1.88; 95% CI, 1.42-2.49; P < 0.001). The c-statistics for the models with and without early postoperative fever were 0.76 (95% CI, 0.73-0.79) and 0.75 (95% CI, 0.72-0.78), respectively, showing a significant difference between the two (difference, 0.02; 95% CI, 0.00-0.03; DeLong's test, P = 0.005).
CONCLUSIONS
Early postoperative fever is a significant but not highly discriminative predictor of CR-POPF after pancreaticoduodenectomy. However, its widespread occurrence limits its applicability as a predictive marker.
Collapse