Completed Projects Algorithm
Algorithm to Identify Clinically Significant Arthritis with Functional Limitation
The project goal is to explore using available administrative data to identify patients with clinically significant arthritis and functional limitation in a managed care setting and target them for possible intervention. This group with limitation is of particular interest because they are most likely to be receptive to interventions to improve their situation. Because there may be many patients with mild disease who may be getting by with self-medication and not feel ready or motivated to engage in any further interventions, identifying those who may more readily adopt any interventions because of more severe disease (pain, dysfunction, limitations, inability to engage in valued activities) would help target intervention outreach and make a more useful algorithm to health care systems and HMOs.
This project is a follow-up to a completed case definition validation study done at Fallon Clinic between November 2001 and October 2002. The studied classified 389 patients as having clinically significant arthritis or not and functional limitation. This effort uses the study data to evaluate the performance of newly developed algorithms using data from computerized administrative records at Fallon Clinic to identify patients with clinically significant arthritis and functional limitation.
The program objective was to develop an algorithm to easily and accurately identify patients with clinically significant arthritis and functional limitation in managed care settings using administrative data.
Abstracts, Publications, Presentations
Harrold LR, Li W, Yood RA, Fuller J, Gorwitz JH. Identification of patients with arthritis and functional limitations using administrative data. J Pub Health Management and Practice 2008 Sep-Oct;14(5):487–497.
The algorithms developed were considered to be accurate enough to identify patients for general disease management interventions and to screen for recruitment into clinical trials, but not accurate enough by themselves for epidemilogic and health services research. The authors felt arthritis is a more difficult target for such administrative algorithms because most encounters are in primary care (less information than with specialists). treatment generally doesn't involve prescription medications (which are easier to track), and arthritis chronicity means that diagnostic and treatment procedures may have occurred prior to the 12 month window of administrative data examined.
Leslie Harrold, M.D., M.P.H.
University of Massachusetts Medical School
55 Lake Avenue North
Worcester, MA 01655