Related Experiment Video
Updated: Apr 20, 2026
Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
The ICD-10 Charlson Comorbidity Index predicted mortality but not resource utilization following hip fracture
Barbara Toson1, Lara A Harvey1, Jacqueline C T Close2
1Falls and Injury Prevention Group, Neuroscience Research Australia, University of New South Wales, Barker Street, Randwick, NSW 2031, Australia.
Insights
The Charlson Comorbidity Index (CCI) effectively predicts mortality in hip fracture patients but not resource utilization. Using individual conditions and the Quan algorithm is recommended for better prediction in administrative data.
Area of Science:
- Geriatric Medicine
- Health Services Research
- Epidemiology
Background:
- Hip fractures represent a significant health burden in older adults.
- Accurate prediction of outcomes is crucial for resource allocation and patient management.
- The Charlson Comorbidity Index (CCI) is a widely used comorbidity measure, but its performance in hip fracture populations using administrative data requires evaluation.
Purpose of the Study:
- To assess the performance of the Charlson Comorbidity Index (CCI) in predicting mortality, 30-day readmission, and length of stay (LOS) in hip fracture patients.
- To compare the effectiveness of different International Classification of Diseases, 10th Revision (ICD-10) coding algorithms for comorbidity assessment.
- To determine optimal modeling strategies for predicting outcomes in this population.
Main Methods:
- Utilized linked hospitalization and death data for 47,698 New South Wales residents aged 65+ admitted for hip fracture.
- Ascertained comorbidities using ICD-10 coding algorithms (Sundararajan 2004, Quan 2005).
- Employed regression models, assessing area under the receiver operating curve (AUC) and Akaike information criterion for model fit and discrimination.
Main Results:
- Both algorithms demonstrated acceptable discrimination for in-hospital, 30-day, and 1-year mortality (AUC 0.69-0.76).
- Predictive ability for 30-day readmission (AUC 0.54-0.57) and LOS (adjusted R(2) 0.007-0.045) was poor.
- The Quan algorithm showed better model fit; modeling individual conditions outperformed weighted scores. A 1-year lookback improved 1-year mortality prediction.
Conclusions:
- The CCI is a valid tool for predicting mortality in hip fracture patients but not for resource utilization (LOS, readmission).
- The Quan algorithm is preferable to the Sundararajan algorithm for comorbidity assessment in this context.
- Modeling individual conditions rather than using categorized weighted scores is recommended for improved predictive accuracy.
Objectives:
To evaluate the performance of the Charlson Comorbidity Index (CCI) in the prediction of mortality, 30-day readmission, and length of stay (LOS) in a hip fracture population using algorithms designed for use in International Classification of Diseases, 10th Revision (ICD-10)--coded administrative data sets.
Study Design And Setting:
Hospitalization and death data for 47,698 New South Wales residents aged 65 years and over, admitted for hip fracture, were linked. Comorbidities were ascertained using ICD-10 coding algorithms developed by Sundararajan (2004) and Quan (2005). Regression models were fitted, and area under the receiver operating curve (AUC) and Akaike information criterion were assessed.
Results:
Both algorithms had acceptable discrimination in predicting in-hospital (AUC, 0.72-0.76), 30-day (0.72-0.75), and 1-year mortality (0.69-0.75) but poor ability to predict 30-day readmission (0.54-0.57) or LOS (adjusted R(2), 0.007-0.045). The Quan algorithm provided better model fit than the Sundararajan algorithm. Models incorporating comorbidities as individual variables performed better than the Charlson weighted or updated Quan weighted score. Including a 1-year lookback period increased predictive ability for 1-year mortality only.
Conclusion:
The CCI is a valid tool for predicting mortality but not resource utilization after hip fracture. We recommend the use of the Quan algorithm rather than Sundararajan algorithm and to model individual conditions rather than categorized weighted scores.
More Related Videos
09:51The Transition to an Anterior-Based Muscle Sparing Approach Improves Early Postoperative Function but is Associated with a Learning Curve
Published on: September 7, 2022
06:17Augmented Reality Navigation-Guided Core Decompression for Osteonecrosis of Femoral Head
Published on: April 12, 2022
