Development of a postoperative delirium risk scoring tool using data from the Australian and New Zealand Hip Fracture Registry: an analysis of 6672 patients 2017-2018

Abstract

Background and purpose: This study aimed to determine the incidence, predictors of postoperative delirium and develop a post-surgery delirium risk scoring tool.

Patients and methods: A total of 6672 hip fracture patients with documented assessment for delirium were analyzed from the Australia and New Zealand Hip Fracture Registry between June 2017 and December 2018.Thirty-six variables for the prediction of delirium using univariate and multivariate logistic regression were assessed. The models were assessed for diagnostic accuracy using C-statistic and calibration using Hosmer-Lemeshow goodness-of-fit test. A Delirium Risk Score was developed based on the regression coefficients.

Results: Delirium developed in 2599/6672 (39.0%) hip fracture patients. Seven independent predictors of delirium were identified; age above 80 years (OR=1.6 CI 1.4-1.9; p=0.001), male (OR=1.3 CI 1.1-1.5; p=0.007), absent pre-operative cognitive assessment (OR=1.5 CI 1.3-1.9; p=0.001), impaired pre-operative cognitive state (OR=1.7 CI 1.3 -2.1; p=0.001), surgery delay (OR=1.7 CI 1.2-2.5; p=0.002) and mobilisation day 1 post-surgery (OR=1.9 CI 1.4-2.6; p=0.001). The C-statistics for the training and validation datasets were 0.74 and 0.75, respectively. Calibration was good (χ2=35.72 (9); p<0.001). The Delirium Risk Score for patients ranged from 0 to 42 in the validation data and when used alone as a risk predictor, had similar levels of diagnostic accuracy (C-statistic=0.742) indicating its potential for use as a stand-alone risk scoring tool.

Conclusion: We have designed and validated a delirium risk score for predicting delirium following surgery for a hip fracture using seven predicting factors. This could assist clinicians in identifying high risk patients requiring higher levels of observation and post-surgical care.

Keywords: Delirium; hip fractures; multivariate; risk score.