Identifying co-morbidity in surgical patients using administrative data with the Royal College of Surgeons Charlson Score

TLDR

Surgical outcomes are influenced by co‑morbidity. The Royal College of Surgeons Co‑morbidity Consensus Group was convened to improve existing instruments that identify co‑morbidity in ICD‑10 administrative data. The RCS Charlson Score was developed with a coding philosophy that enhances international transferability and avoids misclassifying complications as co‑morbidity, and was validated in English Hospital Episode Statistics data for AAA repair, aortic valve replacement, total hip replacement and transurethral prostate resection. Patients with co‑morbidity were older, more likely to be admitted as emergencies, stayed longer, required more augmented care, and had higher in‑hospital and 1‑year mortality, and multivariable prognostic models incorporating the RCS Charlson Score outperformed models based only on age, sex, admission method and prior emergency admissions.

Abstract

Abstract Background Surgical outcomes are influenced by co-morbidity. The Royal College of Surgeons (RCS) Co-morbidity Consensus Group was convened to improve existing instruments that identify co-morbidity in International Classification of Diseases tenth revision administrative data. Methods The RCS Charlson Score was developed using a coding philosophy that enhances international transferability and avoids misclassifying complications as co-morbidity. The score was validated in English Hospital Episode Statistics data for abdominal aortic aneurysm (AAA) repair, aortic valve replacement, total hip replacement and transurethral prostate resection. Results With exception of AAA, patients with co-morbidity were older and more likely to be admitted as an emergency than those without. All patients with co-morbidity stayed longer in hospital, required more augmented care, and had higher in-hospital and 1-year mortality rates. Multivariable prognostic models incorporating the RCS Charlson Score had better discriminatory power than those that relied only on age, sex, admission method (elective or emergency) and number of emergency admissions in the preceding year. Conclusion The RCS Charlson Score identifies co-morbidity in surgical patients in England at least as well as existing instruments. Given its explicit coding philosophy, it may be used as a co-morbidity scoring instrument for international comparisons.