Model performed better than Charlson Comorbidity Index for predicting consultations, admissions
MONDAY, Feb. 3, 2020 (HealthDay News) — The Cambridge Multimorbidity Score is robust for predicting key outcomes, according to a study published online Feb. 3 in CMAJ, the journal of the Canadian Medical Association.
Rupert A. Payne, M.B.Ch.B., Ph.D., from the University of Bristol in the United Kingdom, and colleagues modelled the association between 37 morbidities and three care outcomes (primary care consultation, unplanned hospital admission, and death) at one and five years. A total of 300,000 development and 150,000 validation samples were extracted from the U.K. Clinical Practice Research Datalink. A general outcome multimorbidity score was constructed; performance was compared to that of the Charlson Comorbidity Index.
The researchers found that models including all 37 conditions were acceptable predictors of general practitioner consultations, unplanned hospital admission, and death at one year (C-indices, 0.732, 0.742, and 0.912, respectively). Similar predictive ability was seen with reduction of models to the 20 conditions with the greatest combined prevalence/weight (C-indices, 0.727, 0.738, and 0.910, respectively). For consultations and death, models also predicted five-year outcomes (C-indices, 0.735 and 0.889, respectively); performance was worse for admissions (C-index, 0.708). For consultations and admissions, these models performed better than the Charlson Comorbidity Index (C-indices, 0.691 and 0.703, respectively), while performance for mortality was similar (C-index, 0.907).
“These scores may be of considerable value for policy development and health care priority-setting, providing accurate, easy-to-implement ways of optimizing health care delivery to an aging population with multiple illnesses,” Payne said in a statement.
Copyright © 2020 HealthDay. All rights reserved.