Understanding Prediabetes in a Medicare Advantage Population Using Data Adaptive Techniques.

The objective was to identify individuals with undiagnosed prediabetes from administrative data using adaptive techniques. The data source was a national Medicare Advantage Prescription Drug (MAPD) plan administrative data set. A retrospective, cross-sectional study developed and evaluated data adaptive logistic regression, decision tree, neural network, and ensemble predictive models for metabolic syndrome and prediabetes using 3 mutually exclusive cohorts (N = 279,903). The misclassification rate (MCR), average squared error (ASE), c-statistics, sensitivity (SN), and false positive (FP) rates were compared to select the final predictive models. MAPD individuals with continuous enrollment from 2013 to 2014 were included. Metabolic syndrome and prediabetes were defined using clinical guidelines, diagnosis, and laboratory data. A total of 512 variables identified through subject matter expertise in addition to utilizing all data available were evaluated for the modeling. The ensemble model demonstrated better discrimination (c-statistics, MCR, and ASE of 0.83, 0.24, and 0.16, respectively), high SN, and low FP rate in predicting metabolic syndrome than the individual data adaptive modeling techniques. Logistic regression demonstrated better discrimination (c-statistics, MCR, and ASE of 0.67, 0.13, and 0.11 respectively), high SN, and low FP rate in predicting prediabetes than the other adaptive modeling techniques or ensemble methods. The scored data predicted prediabetes in 44% of the MAPD population, which is comparable to 2005-2006 National Health and Nutrition Examination Survey prediabetes rates of 41%. The logistic regression model demonstrated good performance in predicting undiagnosed prediabetes in MAPD individuals.

Predictors of all-cause 30 day readmission among Medicare patients with type 2 diabetes.

OBJECTIVE: Readmission is costly among patients with type 2 diabetes (T2DM) in Medicare Advantage Prescription Drug Plans; identifying high-risk patients is necessary for targeting reduction programs. The objective of this study was to develop a claims-based algorithm to predict all-cause 30 day readmission among patients with T2DM. METHODS: This study used administrative data from 1 January 2012 through 31 January 2014. The cohort included hospitalized T2DM patients, aged 18-90 with ≥12 months' continuous enrollment before an unplanned hospital admission and ≥1 month of enrollment post-discharge, excluding patients in long-term care >30 days pre-index. Multivariate logistic regression predicted the likelihood of readmission following hospitalization in 2013. The analytic file was randomly split into training and test datasets to build and validate the model. Candidate variables included physician and patient demographics, baseline clinical conditions, and healthcare utilization metrics. Clinical conditions were classified using the Healthcare Cost and Utilization Project clinical classification system for ICD-9-CM. RESULTS: Of 63,237 individuals, 17.1% experienced a readmission. Of nearly 200 candidate variables, 14 were predictors of readmission, including total cumulative number of days for inpatient stays and the number of emergency department visits in the baseline period. Male gender, older age, and certain comorbidities were associated with higher likelihood of readmission. The final model demonstrated good discriminant ability (c-statistic = 0.82). CONCLUSIONS: This study provided evidence that certain patient characteristics and healthcare utilization are predictive of readmission. An algorithm with good discriminant ability was developed which could be used to target readmission reduction programs. Physician gender, specialty, and ownership status did not appear to influence the likelihood of readmission.

A Historical Retrospective Analysis of the Impact of Diabetes Quality Measure Attainment on Outcomes in Medicare Advantage Members.

The objective of this study was to assess achievement of 4 diabetes mellitus (DM)-related quality measures (QMs) and examine the relationship between QM attainment, concurrent health care costs, and DM complications over 1 year by conducting a retrospective analysis of claims data for Medicare Advantage Prescription Drug plan members with DM. Claims and member-level quality data were used to assess QM achievement, concurrent health care costs, and presence of new or worsening DM complications during the QM year. Multivariable regression models were used to examine the relationship between QM achievement and outcome measures controlling for potentially confounding baseline characteristics. QM attainment rates ranged from 54.2% for DM Treatment measure to 83.4% for Cholesterol Screening measure. Odds of new or worsening complications were greater for members who did not meet the Blood Sugar Controlled performance goal (odds ratio [OR]: 1.12, P < 0.001), DM Treatment goal (OR: 1.40, P < 0.001), or Cholesterol Screening goal (OR: 1.32, P < 0.001). Failure to attain the DM Medication Adherence goal was associated with lower odds of new or worsening complications (OR: 0.94, P < 0.001). In the regression models, all-cause health care costs were greater for members who achieved the Blood Sugar Controlled quality goal (P < 0.001), but lower for members who attained DM Treatment (P < 0.001) and low-density lipoprotein Cholesterol Screening goals (P < 0.001). There was no statistically significant relationship between attaining the DM Medication Adherence measure and all-cause costs. Achievement rates for individual QMs varied across the study population and relationships between QM attainment, health care costs, and DM complications during the QM measurement year were mixed.

Early treatment revisions by addition or switch for type 2 diabetes: impact on glycemic control, diabetic complications, and healthcare costs.

BACKGROUND: The study examined the prevalence of early treatment revisions after glycosylated hemoglobin (HbA1c) ≥9.0% (75 mmol/mol) and estimated the impact of early treatment revisions on glycemic control, diabetic complications, and costs. RESEARCH DESIGN AND METHODS: A retrospective cohort study of administrative claims data of plan members with type 2 diabetes and HbA1c ≥9.0% (75 mmol/mol) was completed. Treatment revision was identified as treatment addition or switch. Glycemic control was measured as HbA1c during 6-12 months following the first qualifying HbA1c ≥9.0% (75 mmol/mol) laboratory result. Complications severity (via Diabetes Complication Severity Index (DCSI)) and costs were measured after 12, 24, and 36 months. Unadjusted comparisons and multivariable models were used to examine the relationship between early treatment revision (within 90 days of HbA1c) and outcomes after controlling for potentially confounding factors measured during a 12-month baseline period. RESULTS: 8463 participants were included with a mean baseline HbA1c of 10.2% (75 mmol/mol). Early treatment revision was associated with greater reduction in HbA1c at 6-12 months (-2.10% vs -1.87%; p<0.001). No significant relationship was observed between early treatment revision and DCSI at 12, 24, or 36 months (p=0.931, p=0.332, and p=0.418). Total costs, medical costs, and pharmacy costs at 12, 24, or 36 months were greater for the early treatment revision group compared with the delayed treatment revision group (all p<0.05). CONCLUSIONS: The findings suggest that in patients with type 2 diabetes mellitus, treatment revision within 90 days of finding an HbA1c ≥9.0% is associated with a greater level of near-term glycemic control and higher cost. The impact on end points such as diabetic complications may not be realized over relatively short time frames.

Type 2 Diabetes Mellitus Patients' Healthcare Costs Related to Inpatient Hospitalizations: A Retrospective Administrative Claims Database Study.

INTRODUCTION: Diabetes-related healthcare costs are increasing in the United States, with inpatient hospitalization the largest component of medical expenditures. The aims of this study were to characterize hospitalized type 2 diabetes mellitus (T2DM) patients, understand the relationship between hospitalization and healthcare costs, and explore treatment modification after inpatient hospitalization. METHODS: A retrospective cohort analysis of Humana Medicare Advantage and commercial members with T2DM was conducted. T2DM members were identified and assigned to three groups: (1) inpatient hospitalization (IPH) without a 30-day readmit (IPH group); (2) IPH with a 30-day readmission (IPH readmission group); and, (3) matched non-IPH group. Demographics, clinical characteristics, comorbidities and healthcare costs were measured based on enrollment data and claims. Descriptive statistics were used and the relationship between IPH and costs was assessed using generalized linear models. RESULTS: A total of 15,555 IPH patients, 1757 IPH readmission patients, and 17,312 matched non-IPH patients were included in the study. The IPH readmission group had the highest adjusted mean all-cause total costs ($76,806), followed by the IPH group ($42,011), and the non-IPH group ($9624). A similar trend was observed for adjusted all-cause mean medical and pharmacy costs. DM-related total healthcare costs were highest for the IPH readmission group ($13,714), followed by the IPH group ($7477), and non-IPH group ($1620). While overall therapy modification (discontinuation, addition, switch) was low, T2DM patients with an IPH (with or without a readmission) had greater rates of therapy modification relative to the non-IPH patients. CONCLUSION: Adjusted all-cause and DM-related total costs were greatest for IPH readmission patients. Rates of treatment modification within 10 days of discharge after IPH were generally low. Identifying T2DM patients at high risk of readmission and employing methods to decrease that risk during the index hospitalization could have a significant impact on health system costs. FUNDING: Novo Nordisk.