Preoperative total lymphocyte count was not associated with adverse postoperative events following elective shoulder arthroplasty

Abstract

BackgroundWith the heightened focus on lowering adverse postoperative events through preoperative patient optimization, malnutrition may be a modifiable risk factor that could be addressed and lead to improved postoperative outcomes. However, an understanding of the association between malnutrition status and adverse postoperative events following shoulder arthroplasty is lacking. We sought to evaluate the association between preoperative malnutrition status, identified via the total serum lymphocyte count, and adverse postoperative events following shoulder arthroplasty. MethodsWe conducted a cohort study using data from Kaiser Permanente's Shoulder Arthroplasty Registry. Patients who underwent elective primary shoulder arthroplasty were identified (2005-2016). Patients with a preoperative total lymphocyte count <1500 cells/mm3 were defined as malnourished. A competing risk Cox multistate model was used to evaluate the association between malnutrition and revision or death with stratification by age. Multiple logistic regression was used to evaluate 5-year deep infection, 90-day emergency department visit, 90-day readmission, and 90-day venous thromboembolism. Receiver operating characteristic curves were generated for discrete total lymphocyte counts to determine whether a threshold exists to identify patients at higher risk of adverse postoperative events. ResultsThe final study cohort comprised 6956 shoulder arthroplasty patients, with 2133 (30.7%) identified as malnourished. No difference in septic or aseptic revision risks was observed when comparing patients with and without malnutrition; however, malnourished patients had a higher risk for death, regardless of age (<60 years: hazard ratio = 2.25, 95% confidence interval = 1.07-4.73; ≥60 years: hazard ratio = 1.47, 95% confidence interval = 1.28-1.70). We did not observe a difference in likelihood of deep infection, emergency department visit, readmission, or venous thromboembolism. Receiver operating characteristic curve analysis suggested preoperative total lymphocyte count performed poorly at differentiating adverse events. ConclusionOnly a higher mortality risk was observed to be associated with total lymphocyte count-defined malnutrition in patients undergoing elective shoulder arthroplasty. When instead looking at total lymphocyte count continuously, an optimal threshold for discriminating risk of adverse postoperative events could not be identified. Further study is needed to identify an appropriate indicator of malnutrition in shoulder arthroplasty patients and if this indicator can be modified to improve patient status and quality of care. Level of EvidenceLevel III; Retrospective Cohort Comparative Study