Risk Factors and a Nomogram Model Establishment for Postoperative Delirium in Elderly Patients Undergoing Arthroplasty Surgery: A Single-Center Retrospective Study

Abstract

Objective. To explore the related risk factors of postoperative delirium (POD) after hip or knee arthroplasty in elderly orthopedic patients and the predictive value of related risk factors. Material and Methods. In total, 309 patients (≥60 years) who received knee and hip arthroplasty between January 2017 and May 2020 were consecutively selected into the POD and nonpostoperative delirium (NPOD) groups. Group bias was eliminated through propensity score matching. Univariate and multivariable logistic analysis was used to determine the risk factors for POD. The nomogram was made by R. Results. 58 patients were included in each group after propensity score matching; multivariable analysis demonstrated that LDH ( OR = 4.364 , P = 0.017 ), CHE ( OR = 4.640 , P = 0.004 ), Cystatin C ( OR = 5.283 , P = 0.006 ), arrhythmia ( OR = 5.253 , P = 0.002 ), and operation duration ( OR = 1.017 , P = 0.050 ) were independent risk factors of POD. LDH, CHE, Cystatin C, and arrhythmia were used to construct a nomogram to predict the POD. The nomogram was well calibrated and had moderate discriminative ability ( AUC = 0.821 , 95% CI: 0.760~0.883). Decision curve analysis demonstrated that the nomogram was clinically useful. Conclusions. Our study revealed that arrhythmia, operation duration, the increase of lactate dehydrogenase and Cystatin C, and the decrease of cholinesterase were reliable factors for predicting postoperative delirium after elderly hip and knee arthroplasty. Meanwhile, the nomogram we developed can assist the clinician to filtrate potential patients with postoperative delirium.

Lower preoperative serum uric acid level may be a risk factor for postoperative delirium in older patients undergoing hip fracture surgery: a matched retrospective case-control study

“…We also found that increased duration of surgery and NLR could serve as independent risk factors for POD in older patients undergoing hip fracture surgery. Numerous studies have established operation time as a risk factor for POD in various types of surgical procedures [ 26 , 27 ]. There is no doubt that longer surgery duration indicates a more complex procedure with severe surgical stimulation and may result in greater acute inflammatory responses [ 27 , 28 ].…”

“…Numerous studies have established operation time as a risk factor for POD in various types of surgical procedures [ 26 , 27 ]. There is no doubt that longer surgery duration indicates a more complex procedure with severe surgical stimulation and may result in greater acute inflammatory responses [ 27 , 28 ]. The neuroinflammatory hypothesis is a widely accepted explanation for the development of POD [ 29 ].…”

Incidence and associated factors of delirium after primary total joint arthroplasty in elderly patients: A systematic review and meta-analysis

“…The outcome of methodology quality assessment was as follows: 3 studies [6,8,15] scored 9, 12 studies [1,7,10,[16][17][18][19][20][21][22][23][24] scored 8, and 8 studies [25][26][27][28][29][30][31][32] scored 7.…”

Comparison of logistic regression and machine learning methods for predicting postoperative delirium in elderly patients: A retrospective study

“…But the AUC (area under the ROC curve) of them did not perform well. Besides, some models have been developed to predict delirium using logistic regression or machine learning methods for different surgeries 14 , 15 , 16 , 17 . A nomogram model with logistic regression was usually used to predict POD in a particular type of surgery in small‐scale cohorts 16 , 17 .…”

“…Besides, some models have been developed to predict delirium using logistic regression or machine learning methods for different surgeries 14 , 15 , 16 , 17 . A nomogram model with logistic regression was usually used to predict POD in a particular type of surgery in small‐scale cohorts 16 , 17 . Machine learning methods can be optimally applied when data are abundant 18 , and they have been widely used in various clinical domains to predict events of interest 15 , 19 , 20 .…”