PATIENT-REPORTED CONTENT OF DIABETES TREATMENT DISCUSSIONS WITH PHYSICIANS AND LEVEL OF GLYCEMIC CONTROL

Purpose. Prior research shows that patients with diabetes face extended periods of time when recommended glycemic targets are not met and that medication intensification is often not evident until a patient’s A1c is ≥9%. We explore the relationship between glycemic control and patient-reported content of diabetes treatment discussions with their physician. Methods. Eligible patients are those from a cohort initiating oral mono-therapy in 2000-2005 per pharmaceutical claims and receiving care from an integrated delivery system in southeast Michigan. A mixed mail/telephone survey was administered 1/21/08 - 5/2/08 to an A1c stratified, random sample of cohort members with continual health insurance, and at least one visit with A1c testing in the 6 months preceding survey administration (N=1580). Surveys focused on patient-physician discussions in the prior six-month period. Generalized estimating equations are used to compare survey responders to non-responders and survey responses by A1c (≤7%, 7-9%, ≥9%).  Results. 1053 patients (72%) completed the survey.  Results are presented for the first n=787 surveys processed.  Survey respondents were older (mean 67 vs. 61 years, p<0.01), more likely to be white (60% vs. 52%, p<0.01), and to have lower A1c values (mean 7.3 vs. 7.5%, p<0.01).  Patient-reported physician discussions of diet/exercise and medication treatment differed significantly by A1c level. For example, 81% of those with A1c≥9% reported a diet/exercise-related discussion compared to 70% for those with A1c≤7% (p<0.01).  Similarly, 88% with A1c≥9% reported a medication discussion compared to 70% for those with A1c≤7% (p<0.01).  Among those reporting discussions, 91% with A1c≥9% reported their physician recommended changes in the diet/exercise compared to 89% for those with A1c 7-9% and 82% for those with A1c≤7% (p<0.01).  Likewise, 86% with A1c≥9% reported physician recommended medication changes compared to 68% for those with A1c 7-9% and 48% for those with A1c≤7% (p<0.01).  Patient-reported glycemic control targets also varied by A1c, with those with A1c≥9% significantly less likely to report a target <7% (27% vs. 48%, p<0.01) and significantly more likely to indicate their target was either ‘too low’ or ‘too high’ (12% vs 5%, p<0.01).  Conclusions. Patient reported discussions are consistent with findings that treatment intensifications often fail to occur until A1s are above recommended targets. Future research is needed to evaluate whether initiation of treatment discussions sooner would facilitate improved glycemic control.