Abstract
Background/Aims
Type 2 diabetes is one of the nation’s most prevalent chronic diseases, and the rate is increasing every year. In 2006 a diabetes system of care using a nine-component all-or-none bundle (diabetes bundle) of measures was introduced to Geisinger primary care physicians (PCP). We evaluated how the diabetes bundle impacted the medical cost of diabetes.
Methods
A cohort of diabetes patients between 2005 and 2008 was extracted from Geisinger Health Plan (GHP) claims, among them ~45% of members were Geisinger PCP patients, and thus under diabetes bundle management, and ~55% were not, and thus taken as controls. After applying inclusion and exclusion criteria, 2189 patients in Bundle group and 3554 patients in non-Bundle control group were used for analysis. A subgroup of 3102 Medicare beneficiaries (1862 were Geisinger patients, and 1240 were not) was also analyzed. Generalized linear models were applied to determine the association between cost and Bundle/control groups, adjusting for comorbidities and patients’ characteristics. The estimations of cost attributed to Bundle year by year were performed using difference-in-difference method. Separate analysis of outpatient cost and inpatient cost was also performed. Marginal effect for each comorbidities and characteristics was estimated by recycling prediction.
Results
After adjusting for age, sex, and comorbidities, the average costs attributed to Bundle were −$1569, $20 and $589 dollars for Medicare beneficiaries from 2006 to 2008 (positive means cost reduction, i.e. saving money). And they were −$470, −$251 and $104 for overall patients. Bundle reduces cost for patients with comorbidities, especially for diabetes-related complications, for example, patients with chronic kidney disease (CKD) saved $427 in 2008 than those without CKD. Reduction of inpatient cost accounted for most of saving, the average inpatient costs attributed to Bundle were $117, $340 and $720 from 2006 to 2008.
Conclusions
Patients under diabetes bundle management were beneficial in term of cost, especially for the long term cost, the benefit could be observed as early as 2 years after bundle implementation, and the reduction of inpatient cost account for most of reduced cost.
Keywords: All-or-non diabetes bundle, Cost reduction