Smith, Steven A.

Objective To develop and validate a model of incident type 2 diabetes based solely on administrative data. Data Sources/Study Setting Optum Labs Data Warehouse (OLDW), a national commercial administrative dataset. Study Design HealthImpact model was developed and internally validated using nested case–control study design; n = 473,049 in training cohort and n = 303,025 in internal validation cohort. HealthImpact was externally validated in 2,000,000 adults followed prospectively for 3 years. Only adults ≥18 years were included. Data Collection/Extraction Methods Patients with incident diabetes were identified using HEDIS rules. Control subjects were sampled from patients without diabetes. Medical and pharmacy claims data collected over 3 years prior to index date were used to build the model variables. Principal Findings HealthImpact, scored 0–100, has 48 variables with c‐statistic 0.80815. We identified HealthImpact threshold of 90 as identifying patients at high risk of incident diabetes. HealthImpact had excellent discrimination in external validation cohort (c‐statistic 0.8171). The sensitivity, specificity, positive predictive value, and negative predictive value of HealthImpact >90 for new diagnosis of diabetes within 3 years were 32.35, 94.92, 22.25, and 96.90 percent, respectively. Conclusions HealthImpact is an efficient and effective method of risk stratification for incident diabetes that is not predicated on patient‐provided information or laboratory tests.

Background: Carbohydrate counting may improve glycemic control in hospitalized cardiology patients by providing individualized insulin doses tailored to meal consumption. The purpose of this study was to compare glycemic outcomes with mealtime insulin dosed by carbohydrate counting versus fixed dosing in the inpatient setting. Materials and Methods: This single-center retrospective cohort study included 225 adult medical cardiology patients who received mealtime, basal, and correction-scale insulin concurrently for at least 72 h and up to 7 days in the interval March 1, 2010–November 7, 2013. Mealtime insulin was dosed by carbohydrate counting or with fixed doses determined prior to meal intake. An inpatient diabetes consult service was responsible for insulin management. Exclusion criteria included receipt of an insulin infusion. The primary end point compared mean daily postprandial glucose values, whereas secondary end points included comparison of preprandial glucose values and mean daily rates of hypoglycemia. Results: Mean postprandial glucose level on Day 7 was 204 and 183 mg/dL in the carbohydrate counting and fixed mealtime dose groups, respectively (unadjusted P =0.04, adjusted P =0.12). There were no statistical differences between groups on Days 2–6. Greater rates of preprandial hypoglycemia were observed in the carbohydrate counting cohort on Day 5 (8.6% vs. 1.5%, P =0.02), Day 6 (1.7% vs. 0%, P =0.01), and Day 7 (7.1% vs. 0%, P =0.008). No differences in postprandial hypoglycemia were seen. Conclusions: Mealtime insulin dosing by carbohydrate counting was associated with similar glycemic outcomes as fixed mealtime insulin dosing, except for a greater incidence of preprandial hypoglycemia. Additional comparative studies that include hospital outcomes are needed.