The benefit of "tight glycemic control" in decreasing microvascular events over a 10 year period is based upon the UKPDS trial. An epidemiological follow-up demonstrates an association with improved microvascular and macrovascular complications over a 20 year period. However, a recent prospective cohort study suggests that the cardiovascular benefit of tighter glycemic control in more recent onset disease may be limited to individuals with a lower illness burden. The VADT and ACCORD studies demonstrate an increasing rate and risk of hypoglycemia in patients with a longer duration of diabetes. Such patients are still healthier than many veterans who have significant comorbid disease burdens alongside diabetes. Overall, the relative risks and benefits of tight glycemic control, especially with newer medications, have been understudied in patients with additional chronic complex diseases. The VHA needs to better understand the costs of secondary prevention (such as tight glycemic control) and diabetes complications (such as lower extremity amputations) among Veterans with diabetes and chronic comorbidities.
The goal of this pilot study was to develop cost estimates that can be incorporated into cost-effectiveness models that assess the comparative effectiveness of newer DM medications among veterans with diabetes and comorbidities. The study had 3 specific aims. Among veterans aged 25-64 years with newly diagnosed diabetes, the aims were:
Aim 1: Evaluate healthcare resource utilization and costs associated with the provision of diabetes-related outpatient care to veterans, with and without chronic complex illnesses. We evaluated differences in utilization and costs among patient groups according to the presence of concordant and discordant co-morbid chronic complex illnesses.
Aim 2: Evaluate healthcare utilization and costs associated with inpatient care for diabetes-related complications in veterans, with and without chronic complex illnesses. Complications included macrovascular, microvascular and metabolic complications.
Aim 3: Among veterans with discordant illnesses, we specifically explored inpatient and outpatient utilization and costs for veterans with mental illness and/or substance abuse as these patients may require more intensive care to achieve tight glycemic control.
We used the Diabetes Epidemiology Cohort (DEpiC), developed using VHA data for FY 2003 and FY 2004, to evaluate resource utilization in the inpatient and outpatient settings for veterans with recently diagnosed diabetes, using existing definitions of recent diagnosis and type II diabetes. Chronic comorbidity was categorized as either dominant (limited life expectancy and extensive medical treatment likely to minimize attention to diabetes), discordant (not related to diabetes in terms of pathophysiology, treatment plans, or self-management strategies), or concordant (related to diabetes with synergies in treatment plans and management). Categorization was based on ICD-9 and CPT codes, using a validated and established protocol developed within this research group. Costs were developed from Health Economics Resource Center average cost data, in order to develop estimates of costs of delivering diabetes care, managing diabetes complications in the outpatient setting and managing diabetes complications in the inpatient setting.
In FY 2004 we identified 53,513 veterans with incident DM. The population with DM carries a heavy burden of illness with 7.7% having microvascular illnesses, 24% having macrovascular disease, 24% discordant illness and 10% dominant illnesses.
The cost associated with treating patients with no comorbidities and new onset diabetes was $2,416 [Confidence Interval (CI) $2,088 - $2,203]; an incremental cost of $743. In terms of veterans with discordant disease only in FY2004, the cost associated with treating these patients with new onset diabetes was $6,867 [CI $6,645 - $7,087]; an incremental cost of $2,361. As expected, the Veteran's comorbidity profile is associated with higher incremental costs for new onset diabetes.
For veterans with concordant disease only in FY 2004. the cost associated with treating these patients with new onset diabetes was $5,338 [CI $5,070 - $5,605]; an incremental cost of $2,508. For those veterans with dominant disease in FY 2004, the cost associated with treating these patients with new onset diabetes was $19,907[CI $18,810- $21,000]; an incremental cost of $11,181. As expected, the Veteran's comorbidity profile is associated with higher incremental costs for new onset diabetes.
Next we estimated the costs of treating diabetes complications in the VHA. We assessed the costs associated with lower extremity amputations. The average direct inpatient cost of an amputation in FY2004 was $35,521 [CI $34,070 - $36,971]. Also, 25% of veterans with recent-onset diabetes had mental illness. In FY 2004, the cost associated with treating these patients was $12,398 [CI $18,810- $21,000] ; an incremental cost of $4,467.
We have demonstrated that healthcare costs in the VHA associated with diabetes and its complications vary based on comorbidity profile. The results from this pilot award will be used to develop of a full comparative effectiveness model focusing on Veterans with diabetes and incorporating the chronic comorbidity experienced by this patient population.
The expected significance of this pilot study is that results from the analysis can be used to project long-term costs of Veterans with incident diabetes and co-morbid conditions. This information could inform care coordination strategies within Primary Care or in collaboration with specialists.
None at this time.