Frequent Outpatient Contact and Decreasing Medication Affordability in Patients With Diabetes From 1997 to 2004
- Richard W. Grant, MD, MPH1,
- Ellen P. McCarthy, PHD2,
- Daniel E. Singer, MD1 and
- James B. Meigs, MD, MPH1
- 1General Medicine Units, Massachusetts General Hospital, Boston, Massachusetts
- 2Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
- Address correspondence and reprint requests to Richard W. Grant, MD, MPH, 50-9 Staniford St., Boston, MA 02114. E-mail: rgrant{at}partners.org
Despite significant changes in treatment standards for diabetes, there has been little improvement in overall rates of metabolic control (1–5). In recent years, diabetes medical regimen complexity has also increased markedly, with the proportion of U.S. patients taking five or more prescription medicines nearly doubling (from 18 to 30%) from 1994 to 2003 (6). Given the persistent difficulty in achieving ideal metabolic control and the increasing complexity of medical care, further insight into current diabetes outpatient management is required.
RESEARCH DESIGN AND METHODS
We used data from the National Health Interview Survey (NHIS) to examine U.S. annual trends in diabetes care from 1997 to 2004. The NHIS is an annual nationwide in-person household survey conducted by the Census Bureau that includes an in-depth adult health care access and utilization survey. Sampled adults were asked about demographic and socioeconomic information, diagnosed conditions, self-reported ability to afford prescription medications in the prior year, perceived barriers to obtaining clinic appointments (e.g., long waits for appointments, difficulty making appointments, difficulty reaching the clinic by phone), and frequency of outpatient clinical contact (excluding overnight hospitalizations, telephone calls, and emergency room, dental, or home visits). Response rates over this 8-year period varied from 80.8 to 90.3%.
We limited our analyses to adults who self-identified as having a diagnosis of nongestational diabetes. We created a single independent variable for year (numbered 1–8) and used logistic regression for dichotomous and linear …
