Frailty and Diabetes in Older Hospitalized Adults: What Drives Outcomes?

Presentation Number: MON 612
Date of Presentation: April 3rd, 2017

Heather Theresa MacKenzie*1, Barna Tugwell2, Olga Theou2 and Kenneth Rockwood2
1Dalhousie University, Halifax, NS, CANADA, 2Dalhousie University, Halifax, NS, Canada


Background While diabetes is a common condition among older hospitalized adults, little is known about the outcomes of these patients compared to older patients without diabetes and the impact of frailty and blood glucose on these outcomes. We assessed outcomes in a population of older adults admitted to hospital, including length of stay (LOS), mortality, and glucose control. We wished to determine the impact of frailty and diabetes on hospital outcomes. Methods This study analyzed patients aged 65 or older admitted through the emergency department and consulted to internal medicine at a Canadian tertiary care hospital. Patients were assessed by a senior internist-geriatrician with a Comprehensive Geriatric Assessment (CGA) form, determining their Frailty Index (FI) and Clinical Frailty Scale (CFS), standardized validated frailty assessment instruments. Frailty was defined as a score of 5-9 on the CFS and is a continuous measure on FI. Outcomes such as mortality and LOS were obtained from a hospital administrative database. Admission glucose was defined as the fasting or random glucose test done within 1 day of the CGA. Results This study included 433 patients whose average age was 81.5 ± 8.2 years, 56.6% female. Mean LOS was 24.5 ± 38.1 days, and mortality rate 18.7%; 79.5% were frail. Of 433 patients, 32.1% had diabetes; LOS and mortality were not different between patients with and without diabetes. Patients with diabetes had higher scores on FI (0.50 ± 0.12 vs. 0.47 ± 0.14, p < 0.05) and CFS (86.9% vs. 76%, p < 0.05) than those without diabetes. In logistic regression modeling including frailty, age, sex, and diabetes status as independent risk factors, only frailty as measured by either the FI or the CFS was associated with an increased odds ratio for mortality. Frailty as measured by the FI was associated with a LOS >14 days, whereas diabetes status was not a significant predictor of LOS. Admission glucose was available for 386 patients. In patients with diabetes, mean glucose levels decreased with increasing frailty. The opposite trend was seen in the non-diabetes population, where mean glucose increased with increasing frailty. Mean glucose was 8.7 ± 3.5mmol/L in severely frail patients with diabetes compared with 11.1 ± 5.1 mmol/L in those who were not frail. In patients with diabetes, glucose was not predictive of death or LOS. Conclusions Patients with diabetes were more likely to be frail, but the level of frailty, rather than the presence of diabetes, was more predictive of hospital outcomes such as mortality and LOS. Patients with diabetes who were most frail tended to have the lowest mean glucose values on admission compared to nonfrail patients with diabetes. The practical impact of policies targeting higher blood glucose levels for severely frail individuals is not clear and requires further study.


Nothing to Disclose: HTM, BT, OT, KR