Composite measures, which aggregate performance on individual measures into a summary score, are increasingly being used to evaluate facility performance. There is little understanding of the unique perspective that composite measures provide.
To examine whether high/low (ie, high or low) performers on a composite measures are also high/low performers on most of the individual measures that comprise the composite.
We used data from 2 previous studies, one involving 5 measures from 632 hospitals and one involving 28 measures from 112 Veterans Health Administration (VA) nursing homes; and new data on hospital readmissions for 3 conditions from 131 VA hospitals. To compare high/low performers on a composite to high/low performers on the component measures, we used 2-dimensional tables to categorize facilities into high/low performance on the composite and on the individual component measures.
In the first study, over a third of the 162 hospitals in the top quintile based on the composite were in the top quintile on at most 1 of the 5 individual measures. In the second study, over 40% of the 27 high-performing nursing homes on the composite were high performers on 8 or fewer of the 28 individual measures. In the third study, 20% of the 61 low performers on the composite were low performers on only 1 of the 3 individual measures.
Composite measures can identify as high/low performers facilities that perform “pretty well” (or “pretty poorly”) across many individual measures but may not be high/low performers on most of them.
*Center for Healthcare Organization and Implementation Research, VA Boston Healthcare System (152M)
†Questrom School of Business
‡Department of Surgery, Boston University School of Medicine
§School of Public Health, Boston University, Boston, MA
Supported by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development IIR 06-260.
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
The authors declare no conflict of interest.
Reprints: Michael Shwartz, PhD, Questrom School of Business, Boston University, 595 Commonwealth Avenue, Boston, MA 02215. E-mail: firstname.lastname@example.org.