International classification of disease clinical modification 9 modeling of a patient comorbidity score predicts incidence of perioperative complications in a nationwide inpatient sample assessment of complications in spine surgery

Abstract

Of background data: A patient comorbidity score (RCS) was developed from a prospective study of complications occurring in spine surgery patients.

Objective: To validate the RCS, we present an International Classification of Disease Clinical Modification (ICD-CM)-9 model of the score and correlate the score with complication incidence in a group of patients from the Nationwide Inpatient Sample database. We compare the predictive value of the score with the Charlson index.

Study design: We conducted a retrospective assessment of Nationwide Inpatient Sample patients undergoing cervical or thoracolumbar spine surgery for degenerative pathology from 2002 to 2009.

Methods: We generated an ICD-9-CM coding-based model of our prospectively derived RCS, categorizing diagnostic codes to represent relevant comorbidities. Multivariate models were constructed to eliminate the least significant variables. ICD-9-CM coding was also used to calculate a Charlson comorbidity score for each patient. The accuracy of the RCS was compared with the Charlson index through the use of a receiver-operating curve.

Results: A total of 352,535 patients undergoing 369,454 spine procedures for degenerative disease were gathered. Hypertension and hyperlipidemia were the most common comorbidities. Cervical procedures resulted in 8286 complications (4.50%), whereas thoracolumbar procedures produced 25,118 complications (13.55%). Increasing RCS correlated linearly with increasing complication incidence (odds ratio [OR] 1.11; 95% confidence interval [CI], 1.10-1.13; P<0.0001). Logistic regression revealed that neurological deficit, cardiac conditions, and drug or alcohol use had greatest association with complication occurrence. The Charlson index also correlated with complication occurrence in both cervical (OR 1.25; 95% CI, 1.23-1.27) and thoracolumbar (OR 1.11; 95% CI, 1.10-1.12) patient groups. Receiver-operating curve analysis allowed a comparison of accuracy of the indices by comparing predictive values. The RCS performed as well as the Charlson index in predicting complication occurrence in both cervical and thoracic spine patients.

Conclusions: ICD-9-based modeling validated that RCS correlates with complication occurrence. The RCS performed as well as the Charlson index in predicting risk of complication in spine patients.