Low predictive power of comorbidity indices identified for mortality after acute arthroplasty surgery undertaken for femoral neck fracture

Research output: Contribution to journalArticle

Abstract

Aims Our aim was to examine the Elixhauser and Charlson comorbidity indices, based on administrative data available before surgery, and to establish their predictive value for mortality for patients who underwent hip arthroplasty in the management of a femoral neck fracture. Patients and Methods We analyzed data from 42 354 patients from the Swedish Hip Arthroplasty Register between 2005 and 2012. Only the first operated hip was included for patients with bilateral arthroplasty. We obtained comorbidity data by linkage from the Swedish National Patient Register, as well as death dates from the national population register. We used univariable Cox regression models to predict mortality based on the comorbidity indices, as well as multivariable regression with age and gender. Predictive power was evaluated by a concordance index, ranging from 0.5 to 1 (with the higher value being the better predictive power). A concordance index less than 0.7 was considered poor. We used bootstrapping for internal validation of the results. Results The predictive power of mortality was poor for both the Elixhauser and Charlson comorbidity indices (concordance indices less than 0.7). The Charlson Comorbidity Index was superior to Elixhauser, and a model with age and gender was superior to both indices. Conclusion Preoperative comorbidity from administrative data did not predict mortality for patients with a hip fracture treated by arthroplasty. This was true even if association on group level existed.

Details

Authors
Organisations
External organisations
  • Swedish Hip Arthroplasty Register
  • University of Gothenburg
  • Prince Philip Hospital, Llanelli
Research areas and keywords

Subject classification (UKÄ) – MANDATORY

  • Orthopedics
  • Public Health, Global Health, Social Medicine and Epidemiology
Original languageEnglish
Pages (from-to)104-112
Number of pages9
JournalBone and Joint Journal
Volume101B
Issue number1
Publication statusPublished - 2019
Publication categoryResearch
Peer-reviewedYes