To investigate the risk factors of surgical site infection (SSI) after posterior lumbar interbody fusion (PLIF), and construct a quantitative risk prediction model to guide clinical application.

Total of 1 333 patients admitted to Haian People’s Hospital from January 2019 to January 2023 for elective PLIF were selected, and were divided into SSI group (44 cases) and non SSI group (1 289 cases) according to whether SSI occurred within 30 days after surgery. Clinical samples of patients in SSI group were collected for pathogen isolation and identification, and the risk factors of SSI were screened by univariate and multivariate Logistic regression analysis, respectively, and a nomogram model was constructed. The predictive efficiency was analyzed by receiver operating curve (ROC).

The incidence of SSI was 3.3% (44/1 333). Total of 38 strains of pathogens were isolated, including 16 strains of Klebsiella pneumoniae, 13 strains of Escherichia coli, 7 strains of Pseudomonas aeruginosa and 2 other strains. In SSI group, the age of patients was older, urinary tract infection and multi-segment surgery (≥ 2 segments) were more, subcutaneous fat thickness, lumbar multifidus muscle (LMM) fat infiltration, operation time, drainage time were longer and the number of drainage tubes were larger, with significant differences (all P < 0.05). Univariate Logistic regression analysis showed that SSI was associated with age ≥ 60 years old, multi-segment surgery (≥ 2 segments), subcutaneous fat thickness (≥ 14.0 mm), LMM fat infiltration (≥ 22.5 %), operation time (≥ 190.0 min) and drainage time (≥ 2.0 d) (all P < 0.05). Multivariate Logistic regression analysis showed that age ≥ 60 years old, multi-segment surgery (≥ 2 segments), subcutaneous fat thickness (≥ 14.0 mm), LMM fat infiltration (≥ 22.5 %), operation time (≥ 190.0 min) and drainage time (≥ 2.0 d) were all risk factors for SSI (all P < 0.05). ROC analysis showed that AUC value of nomogram in predicting SSI was significantly higher than that of Logistic regression model (0.845 vs. 0.769: Z = 6.325, P < 0.001).

PLIF has a certain risk of SSI, body weight distribution indicators such as subcutaneous fat thickness and LMM fat infiltration may be important potential risk factors for SSI. The establishment of quantitative nomogram prediction model is helpful to guide clinical for early identification of SSI high-risk groups, and carry out targeted intervention to improve clinical prognosis.