New multimorbidity score outperforms Charlson Comorbidity Index
Click Here to Manage Email Alerts
The Cambridge Multimorbidity Score — a new method of measuring multimorbidity — outperformed the currently used Charlson Comorbidity Index across all outcomes, according to research published in CMAJ.
“The score provides a useful means of quantifying multimorbidity, which will be of use for risk stratification and to inform health service design and delivery,” Rupert A. Payne, PhD, MRCGP, FRCPE, FBPhS, a consultant senior lecturer at the Centre for Academic Primary Care in Bristol Medical School at the University of Bristol, United Kingdom, told Healio Primary Care.
“The score has several advantages over existing measures,” he continued. “For example, it uses contemporary data, it predicts a range of outcomes — primary utilization, hospitalization, as well as the more usual death, and also an ‘average’ measure that captures all three of these outcomes — and includes a wider range of clinical conditions than some more limited measures.”
Payne and colleagues developed and tested the Cambridge Multimorbidity Score using data from 148 primary care practices across the United Kingdom. Specifically, they modelled the association between 37 morbidities and three outcomes at 1 year and at 5 years: primary care consultations, unplanned hospitalizations and death.
The models that included all 37 conditions were “acceptable predictors” of primary care consultations (C-index = 0.732; 95% CI, 0.731-0.734), unplanned hospitalizations (C-index 0.742; 95% CI, 0.737-0.747) and death at 1 year (C-index = 0.912; 95% CI, 0.905-0.918).
Simplified models that focused on only 20 of the most common conditions that had largest effect on patients performed almost as well as the more complex models, the researchers said. The simplified models had similarly predicted 5-year outcomes of death (C-index = 0.889; 95% CI, 0.885-0.892) and primary care consultations (C-index = 0.735; 95% CI, 0.734-0.736), but researchers found that they did not perform as well predicting unplanned hospitalizations (C-index = 0.708; 95% CI, 0.705-0.712).
A general-outcome multimorbidity score — which the researchers said they developed by averaging the standardized weights of the separate outcome scores — performed similarly to outcome-specific models.
Compared with the Charlson Comorbidity Index, Payne and colleagues’ models performed significantly better for predicting primary care consultations (C-index = 0.691; 95% CI, 0.69-0.693) and unplanned hospitalizations (C-index = 0.703; 95% CI, 0.697-0.709). The models performed similarly in predicting death (C-index = 0.907; 95% CI, 0.9-0.914).
Payne said that although the score was developed in the U.K., it uses population-based data and is therefore likely to be generalizable to U.S. populations. However, he acknowledged that others will likely wish to study the model in specific patient groups, health care settings, and in other populations before it is widely implemented.
“This tool will be of use to people involved in planning health services and for deciding who should potentially be prioritized for tailored multimorbidity care,” Payne explained. “It's not specifically designed for use with an individual patient; rather, for deciding which patients within a particular population — for example, those people registered with a particular family practice — should be targeted with interventions designed to optimize care for multimorbidity.” – by Erin Michael
Disclosure: Payne reports no relevant financial disclosures.