Open AccessA nomogram for accurately predicting the surgical site infection following transforaminal lumbar interbody fusion in type 2 diabetes patients, based on glycemic variability
Author(s) -
Liu Hang,
Zhang Wei,
Hu Qiang,
Liu Lei,
Xie Zhiyang,
Xu Yuzhu,
Jing Genyang,
Wang Yuntao
Publication year - 2023
Publication title -
international wound journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.867
H-Index - 63
eISSN - 1742-481X
pISSN - 1742-4801
DOI - 10.1111/iwj.13948
Subject(s) - medicine , nomogram , perioperative , medical record , logistic regression , glycemic , surgical site infection , diabetes mellitus , spinal fusion , surgery , cohort , univariate analysis , type 2 diabetes , complication , retrospective cohort study , incidence (geometry) , multivariate analysis , physics , optics , endocrinology
Surgical site infection (SSI) is a common and serious complication of transforaminal lumbar interbody fusion (TLIF), and the occurrence of SSI usually leads to prolonged hospitalisation, increased medical costs, poor prognosis, and even death. The objectives of this study were to compare the incidence of SSI in patients with type 2 diabetes, investigate the correlation between perioperative glycemic variability and postoperative SSI, and develop a nomogram model to predict the risk of SSI. This study retrospectively analysed 339 patients with type 2 diabetes who underwent TLIF in the spinal surgery department of the Affiliated Zhongda Hospital of Southeast University from January 2018 to September 2021. The medical records of all patients were collected, and postoperative infection cases were determined according to the diagnostic criteria of surgical site infection. The risk factors for postoperative SSI were analysed by univariate and multivariate logistic regression. And Nomogram prediction model was established and validated. The nomogram incorporated seven independent predictors. Preoperative FPG‐CV was the most important independent risk predictor of SSI, followed by preoperative MFBG, duration of drain placement, postoperative FPG‐CV, preoperative blood glucose control scheme, duration of diabetes >5 years, and the number of fused vertebrae ≥2. The nomogram showed good diagnostic accuracy for the SS of both the training cohort and the validation cohort (AUC = 0.915 and AUC = 0.890). The calibration curves for the two cohorts both showed optimal agreement between nomogram prediction and actual observation. In conclusion, preoperative and postoperative glycemic variability is closely related to the occurrence of SSI. We developed and validated a nomogram to accurately predict the risk of SSI after TLIF surgery. It's helpful for spinal surgeons to formulate reasonable treatment plans and prevention strategies for type 2 diabetes patients.