Y Bar-Dayan1,2, I Beer3, M Boaz4,5, Z Landau2,6, D Jakubowicz1,2, J Wainstein1,2. 1. Diabetes Unit, Wolfson Medical Center, Holon, Israel. 2. Sackler Faculty of Medicine, Tel Aviv, Israel. 3. Department of Medicine, Wolfson Medical Center, Holon, Israel. 4. Epidemiology and Research Unit, Wolfson Medical Center, Holon, Israel. 5. Ariel University, Ariel, Israel. 6. Pediatric Diabetes Unit, Wolfson Medical Center, Holon, Israel.
Abstract
AIMS: Stress hyperglycaemia during hospitalisation may be the first sign of diabetes mellitus (DM). Most hospitals routinely measure blood glucose, which may enable early diagnosis. This study measured the prevalence of hyperglycaemia in hospitalised adults with no history of diabetes, and whether the discharge summary recommended work-up. METHODS: Files with at least one random blood glucose (RBG) sample were included and reviewed for specific discharge recommendations concerning elevated blood glucose. Hyperglycaemia was defined as serum glucose > 200 mg/dl. Length of stay, in-hospital mortality and 3-year mortality were examined. RESULTS: Among 5274 discharged patients, 1479 had DM. They were older and had a higher incidence of cerebrovascular risk factors. Among 3714 patients without known DM, 211 (5.7%) had at least one RBG > 200 mg/dl. Of these patients, 31 died and 24 left against medical advice. Of the remaining 156, 25(16%) files included instructions to the family physician. These patients were younger, more overweight and less frequently diagnosed with dementia or other mental illness. Patients with RBG > 200 mg/dl had prolonged hospital stay (6.5 ± 5.3 vs. 4.0 ± 4.8; p < 0.001). In-hospital mortality and 3-year mortality were increased by 5.1 and 1.89, respectively (p < 0.001 for both parameters) compared to those without RBG ≤ 200 mg/dl. RBG > 200 mg/dl emerged as a significant, independent predictor of prolonged hospital stay and death. CONCLUSIONS: Random blood glucose > 200 mg/dl is common in medical departments and is associated with increased in-hospital and 3-year out-hospital mortality.
AIMS: Stress hyperglycaemia during hospitalisation may be the first sign of diabetes mellitus (DM). Most hospitals routinely measure blood glucose, which may enable early diagnosis. This study measured the prevalence of hyperglycaemia in hospitalised adults with no history of diabetes, and whether the discharge summary recommended work-up. METHODS: Files with at least one random blood glucose (RBG) sample were included and reviewed for specific discharge recommendations concerning elevated blood glucose. Hyperglycaemia was defined as serum glucose > 200 mg/dl. Length of stay, in-hospital mortality and 3-year mortality were examined. RESULTS: Among 5274 discharged patients, 1479 had DM. They were older and had a higher incidence of cerebrovascular risk factors. Among 3714 patients without known DM, 211 (5.7%) had at least one RBG > 200 mg/dl. Of these patients, 31 died and 24 left against medical advice. Of the remaining 156, 25(16%) files included instructions to the family physician. These patients were younger, more overweight and less frequently diagnosed with dementia or other mental illness. Patients with RBG > 200 mg/dl had prolonged hospital stay (6.5 ± 5.3 vs. 4.0 ± 4.8; p < 0.001). In-hospital mortality and 3-year mortality were increased by 5.1 and 1.89, respectively (p < 0.001 for both parameters) compared to those without RBG ≤ 200 mg/dl. RBG > 200 mg/dl emerged as a significant, independent predictor of prolonged hospital stay and death. CONCLUSIONS: Random blood glucose > 200 mg/dl is common in medical departments and is associated with increased in-hospital and 3-year out-hospital mortality.