Change in Use of Glucose-Lowering Medications in Frail and Robust Residents Following Hospital Discharge to Long-Term Care Facilities

Objectives Hospital discharge to a long-term care facility (LTCF) often represents an opportunity to reassess care goals, treatment intensity, and medication use. We examined changes in glucose-lowering medication (GLM) dispensing following first hospital discharge to LTCFs and whether these changes differed by resident frailty status. 

Design Retrospective cohort study. 

Setting and Participants Adults aged ≥65 with type 2 diabetes (T2D) newly discharged to LTCFs in Victoria, Australia, from 2012 and 2018. 

Methods Using linked hospital and medication dispensing data, we compared age- and sex-adjusted prevalence of GLM use in the 90-day period prehospitalization and 90-day period postdischarge to LTCF using Poisson regression. Multivariable adjusted relative risks (aRRs) were estimated using generalized estimating equations. 

Results Among 19,704 individuals discharged to an LTCF (78.7% frail, 45.9% aged ≥85 years, 54.6% female), overall GLM use declined from 59.7% (95% CI 58.4-61.0) to 54.8% (95% CI 53.6-56.1) in frail residents, and 59.3% (95% CI 56.9-61.9) to 56.2% (95% CI 53.8-58.7) in robust residents. Changes in age- and sex-adjusted GLM prevalence among frail and robust residents were mainly attributable to decreased use of metformin and sulfonylureas and increased use of insulin. In multivariable-adjusted analyses, frailty was associated with lower likelihood of dispensing metformin monotherapy within 90 days postdischarge (aRR 0.91, 95% CI 0.85-0.97) and metformin plus sulfonylurea at 90 days (aRR 0.89, 95% CI 0.79-0.99). There were no statistically significant differences in aRRs for combination therapy with ≥2 GLMs. In sensitivity analyses, frailty was associated with greater likelihood of no GLM dispensing within 6 months (aRR 1.04, 95% CI 1.02-1.07), but not within 90 days postdischarge (aRR 1.02, 95% CI 1.00-1.05). 

Conclusions and Implications There are potential missed opportunities to reassess GLM regimens in the 90-day period postdischarge to LTCFs. Resident frailty status did not appear to be associated with meaningful T2D treatment deintensification. Further initiatives may be needed to promote postdischarge medication reviews to optimize diabetes care in this vulnerable population.