Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

Measuring Care Continuity

A Comparison of Claims-based Methods

Pollack, Craig E., MD, MHS; Hussey, Peter S., PhD; Rudin, Robert S., PhD; Fox, D. Steven, MD; Lai, Julie, MPH; Schneider, Eric C., MD

doi: 10.1097/MLR.0000000000000018
Applied Methods

Background: Assessing care continuity is important in evaluating the impact of health care reform and changes to health care delivery. Multiple measures of care continuity have been developed for use with claims data.

Objective: This study examined whether alternative continuity measures provide distinct assessments of coordination within predefined episodes of care.

Research Design and Subjects: This was a retrospective cohort study using 2008–2009 claims files for a national 5% sample of beneficiaries with congestive heart failure, chronic obstructive pulmonary disease, and diabetes mellitus.

Measures: Correlations among 4 measures of care continuity—the Bice-Boxerman Continuity of Care Index, Herfindahl Index, usual provider of care, and Sequential Continuity of Care Index—were derived at the provider- and practice-levels.

Results: Across the 3 conditions, results on 4 claims-based care coordination measures were highly correlated at the provider-level (Pearson correlation coefficient r=0.87–0.98) and practice-level (r=0.75–0.98). Correlation of the results was also high for the same measures between the provider- and practice-levels (r=0.65–0.92).

Conclusions: Claims-based care continuity measures are all highly correlated with one another within episodes of care.

*Johns Hopkins School of Medicine

Johns Hopkins Bloomberg School of Public Health, Baltimore, MD

RAND Corporation, Santa Monica, CA

§Division of General Medicine and Primary Care, Brigham and Women’s Hospital

Harvard Medical School

Harvard School of Public Health, Boston, MA

This paper was presented at the 2013 Society of General Internal Medicine annual meeting on April 25, 2016.

Supported by the Aetna Foundation. C.E.P. supported by the National Cancer Institute (NCI) and Office of Behavioral and Social Sciences (K07 CA151910).

The authors declare no conflict of interest.

Reprints: Craig E. Pollack, MD, MHS, Johns Hopkins School of Medicine, 2024 E. Monument Street, 2-521, Baltimore, MD 21287. E-mail:

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.