Differences in diabetes management of nursing home patients based on functional and cognitive status

To describe practice patterns regarding diabetes management among nursing home (NH) physicians and to identify variation in this practice based on patient characteristics.

Mailed survey.

Nursing home physicians from the American Medical Directors Association (AMDA) Foundation Long-Term Care Research Network (n = 142), as well as other members of AMDA who were Certified Medical Directors (CMD) (n = 68) and members who were not CMD certified (n = 45). Response rates to the survey were 51%, 33%, and 23%, respectively.

Physician and facility characteristics were queried. Responses to 12 items pertaining to diabetes management and 5 items pertaining to use of specific oral diabetes medications were evaluated in the context of 3 different patient profiles that reflected different combinations of functional and cognitive impairment. Responses were based on the physicians' perception of how they manage diabetes under these specified patient profiles.

Responses from members of the Research Network indicated highly significant variability (P < .01) between the 3 patient profiles for all of the 12 management items. Ordering a special diet, monitoring lipid panel, and ordering routine ophthalmology was less likely for the patient profile with both functional and cognitive impairment (P < .01). These differences among the patient profiles for these 3 interventions were present in the responses from all 3 categories of physicians (Research Network, CMD, and non-CMD members of AMDA). There was no statistically significant variability among the 3 patient profiles for any of the 3 physician groups regarding the likelihood of using of any of the 5 classes of oral diabetic medications. Non-CMD physicians were more likely to have less NH experience; otherwise, there were no differences among the 3 physician groups.

Nursing home physicians appear to alter the approach to diabetes management based on the functional and/or cognitive status of the patient. This was particularly true for those physicians who were members of the AMDA Foundation Research Network. These findings have implications for initiatives designed to guide clinical practice as well as efforts by regulatory bodies to evaluate appropriate care. Further research is needed to measure the actual impact of different approaches to diabetes management on relevant outcomes in this population.