Prognostic Nutritional Index (PNI) as a potential predictor and intervention target for perioperative ischemic stroke: a retrospective cohort study.

BACKGROUND

The prognostic nutritional index (PNI) is a nutritional indicator and predictor of various diseases. However it is unclear whether PNI can be a predictor of perioperative ischemic stroke. This study aims to evaluate the association of the preoperative PNI and ischemic stroke in patients undergoing non-cardiac surgery.

METHODS

The retrospective cohort study included patients who underwent noncardiac surgery between January 2008 and August 2019. The patients were divided into PNI ≥ 38.8 and PNI < 38.8 groups according to the cut-off value of PNI. Univariate and multivariate logistic regression analyses were performed to explore the association between PNI and perioperative ischemic stroke. Subsequently, propensity score matching (PSM) analysis was performed to eliminate the confounding factors of covariates and further validate the results. Subgroup analyses were completed to assess the predictive utility of PNI for perioperative ischemic stroke in different groups.

RESULTS

Amongst 221,542 hospitalized patients enrolled, 485 (0.22%) experienced an ischemic stroke within 30 days of the surgery, 22.1% of patients were malnourished according to PNI < 38.8, and the occurrence of perioperative ischemic stroke was 0.34% (169/49055) in the PNI < 38.8 group. PNI < 38.8 was significantly associated with an increased incidence of perioperative ischemic stroke whether in univariate logistic regression analysis (OR = 1.884, 95% CI: 1.559-2.267, P < 0.001) or multivariate logistic regression analysis (OR = 1.306, 95% CI: 1.061-1.602, P = 0.011). After PSM analysis, the ORs of PNI < 38.8 group were 1.250 (95% CI: 1.000-1.556, P = 0.050) and 1.357 (95% CI: 1.077-1.704, P = 0.009) in univariate logistic regression analysis and multivariate logistic regression analysis respectively. The subgroup analysis indicated that reduced PNI was significantly associated to an increased risk of perioperative ischemic stroke in patients over 65 years old, ASA II, not taking aspirin before surgery, without a history of stroke, who had neurosurgery, non-emergency surgery, and were admitted to ICU after surgery.

CONCLUSIONS

Our study indicates that low preoperative PNI is significantly associated with a higher incidence of ischemic stroke in patients undergoing non-cardiac surgery. Preoperative PNI, as a preoperative nutritional status evaluation index, is an independent risk factor useful to predict perioperative ischemic stroke risk, which could be used as an intervenable preoperative clinical biochemical index to reduce the incidence of perioperative ischemic stroke.