Improving care in a resident practice for patients with diabetes

Curricular redesign and introduction of the Chronic Care Model in our residency clinic during 2005-2007 achieved limited success in glycemic (glycated hemoglobin level [A(1c)]), lipid (low-density lipoprotein fraction [LDL]), and blood pressure (BP) control for patients with diabetes.

Beginning in January 2008, ancillary staff performed previsit, protocol-driven reviews of medical records of patients with diabetes to identify those not at A(1c), LDL, and BP goals; inserted electronic prompts into the records regarding deficiencies; and obtained samples for A(1c) or lipid panel when needed. Faculty feedback regarding resident-specific panel reviews was added in May 2008, and point-of-care A(1c) testing was implemented in February 2009.

We conducted a 2-year retrospective study of all patients at our facility with diabetes mellitus, who had at least 1 visit during January to June 2008 (baseline) and 1 visit during July to December 2009 (follow-up). Measures included the most current A(1c), LDL, and BP results. Paired outcome results were compared using the McNemar χ(2) test.

A total of 522 patients with diabetes mellitus were seen during the baseline and follow-up periods, and 456 patients (87.4%) had paired A(1c) results, with A(1c) < 7.0% for 138 of 456 patients (30.3%) at baseline and 166 of 456 patients (36.4%) at follow-up (P  =  .011). For LDL, 460 patients (88.1%) had paired results, with LDL < 100 mg/dL for 225 of 460 patients (48.9%) at baseline and 262 of 460 patients (57.0%) at follow-up (P  =  .004). A total of 513 patients (98.3%) had paired BP results in which the BP < 130/80 mm Hg for 124 of 513 patients (24.2%) at baseline and for 188 of 513 patients (36.6%) at follow-up (P < .001). There were 421 patients (80.7%) with paired results for all 3 measures, with 17 of 421 patients (4.0%) at goal at baseline and 41 of 421 patients (9.7%) at goal at follow-up (P  =  .001).

The interventions resulted in statistically significant improvements in the proportion of patients with diabetes who attained goal for A(1c), LDL, and BP levels. Our redesign elements may be useful in enhancing resident education and in improving patient care.