Incidence of accidental falls and development of a fall risk prediction model among elderly patients with diabetes mellitus: A prospective cohort study.

Aims: To investigate the incidence of accidental falls and develop a fall risk prediction tool in elderly patients with diabetes mellitus.

Background: The risk of fall in elderly patients with diabetes is higher than that in the general elderly, there is fewer fall assessment tools for elderly patients with diabetes.

Design: A prospective cohort study.

Methods: Between June and September 2019, a total of 1007 elderly patients with diabetes were enrolled from a tertiary specialist diabetes hospital in Tianjin and were prospectively followed up for 6 months to determine outcomes of accidental falls through telephone. Demographic and diseases related factors were collected at baseline. Incidence of falls was investigated, and a nomogram was developed based on logistic regression model. SPSS 21.0 and R 3.6.3 were used to analyse the data. The article was reported in accordance with STROBE guidelines.

Results: Among 1007 elderly patients, 950 finished the follow-up. A total of 133 falls occurred in 93 patients during the follow-up period, with a fall rate of 9.79%. Diabetic peripheral neuropathy, walking aids, depression, fall history, fatigue and sex were independent predictors of accidental fall in diabetes elderly patients. The sensitivity and specificity of the predictive model were 73.12% and 52.63%, respectively, and a fall risk prediction nomogram was developed based on the regression model.

Conclusions: A nomogram including 6 easily available prediction factors (diabetic peripheral neuropathy, walking aids, depression, fall history within 1 year, fatigue, sex) was developed, and it can be used in safety management among Chinese elderly patients diagnosed with diabetes.

Relevance To Clinical Practice: Nomogram can be used to identify diabetic elderly patients at high risk of accidental falls.