Comparing comorbidity measures for predicting mortality and hospitalization in three population-based cohorts
2011

Comparing Comorbidity Measures for Predicting Health Outcomes

Sample size: 732416 publication 10 minutes Evidence: moderate

Author Information

Author(s): Jacqueline M Quail, Lisa M Lix, Beliz Acan Osman, Gary F Teare

Primary Institution: Saskatchewan Health Quality Council

Hypothesis

Which comorbidity measure is optimal for predicting mortality and hospitalization in different populations?

Conclusion

The optimal comorbidity measure depends on the health outcome and not on the disease characteristics of the study population.

Supporting Evidence

  • The Elixhauser index was the best predictor of mortality.
  • The number of diagnoses was the best predictor of hospitalization.
  • Results were consistent across different populations.
  • Models in chronic disease cohorts had poorer performance than the general population cohort.
  • Age-restricted cohorts showed greater changes in predictive performance.

Takeaway

This study looked at different ways to measure health problems in people to see which method best predicts if they will die or need to go to the hospital. It found that the best method depends on what health issue you are looking at.

Methodology

The study used administrative health data from Saskatchewan to create cohorts and assessed predictive performance of five comorbidity measures for mortality and hospitalization outcomes.

Potential Biases

Potential misclassification of comorbid conditions due to coding inaccuracies.

Limitations

Misclassification bias may occur due to inaccurate diagnosis coding, and the study only used one year of data for comorbidity measures.

Participant Demographics

The study included Saskatchewan residents aged 20 and older, with specific cohorts for diabetes and osteoporosis.

Statistical Information

Confidence Interval

95% CI: 0.886, 0.892

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1472-6963-11-146

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication