RESUMO
OBJECTIVE: To determine the role of hypoglycemia, hyperglycemia or the combination of both as independent risk factors for falls in a hospital population. Secondary objectives included evaluation of other risk factors for falling and their relationships with glucose levels. RESEARCH DESIGN AND METHODS: Retrospective cohort study over 2 years on hospitalized subjects (N = 57411) analyzing in-hospital-falls and capillary glucose values. Bivariate analysis (χ2 test) and multivariate analysis (logistic regression) were performed to test for correlation of glucose values, age, sex, Charlson index, service of care, diagnosis at discharge and diabetes treatment with risk of in-hospital-falls. RESULTS: The comparison of patients who experienced a fall (fall population) with the non-fall population suggested that: glucose determinations were significantly more frequent in the fall population (OR 3.45; CI 2.98-3.99; p < 0.0001); values of glucose below 70 mg/dl and over 200 mg/dl were significantly associated to falls during hospitalization (OR 1.76; CI 1.42-2.19; p < 0.001) as compared to glycemic values between 70 and 200 mg/dl; diabetes treatment was significantly correlated to risk of fall (OR 2.97; CI 2.54-3.49; p < 0.001); the frequency of glycemia measurements below 70 mg/dl and over 200 mg/dl in the same subject was significantly associated to falls during hospitalization (OR 1.01; CI 1.01-1.02; p < 0.001). CONCLUSION: Hypoglycemia and hyperglycemia during hospital stays are correlated with an increased risk for falls in the hospitalized population. Presence of diabetes, use of insulin or glucose variability could potentially constitute risk factors for falls inside the hospital as well.