Comparing clinical and economic characteristics between commercially-insured patients with diabetic neuropathy and demographically-matched diabetic controls.

ABSTRACT

OBJECTIVE: To examine medical conditions associated with diabetic neuropathy (DN) and to identify drivers of healthcare charges and utilization using administrative claims. METHODS: The study examined commercially-insured under-age-65 individuals with 24 months continuous enrollment in a large US national health plan. DN patients were identified by having at least one claim with a DN diagnosis between July 2004 and June 2005. A demographically-matched control cohort of patients with diabetes but no DN (10 1 ratio) was constructed using propensity scoring. Overall illness burden via a comprehensive disease classification, year 2 (July 2005 through June 2006) distribution of charges, and reasons for inpatient admissions and emergency room (ER) visits were compared between DN patients and diabetic controls. Multivariate regressions were used to assess the marginal contribution of DN to healthcare charges and utilization, and the most common reasons for ER and inpatient admissions, controlling for differences in overall illness burden. RESULTS: Both DN patients (n = 8655) and diabetic controls (n = 86 550) had a mean age of 51 years, and 46% were female. Compared with controls, DN patients had more comorbid medical conditions (9.7 vs. 6.8, p < 0.05) and higher total healthcare charges. Controlling for differences in overall illness burden, DN patients had significantly more hospital days (0.67), more ER (0.09), physician office (0.62), and outpatient hospital visits (2.87), and higher total healthcare charges ($5696) than controls (all p < 0.05), with majority of the difference in charges from inpatient service ($3975, p < 0.05). Patients with DN were also far more likely to be hospitalized (ketoacidosis, neurological manifestation, heart disease, skin infection) or have an ER encounter (amputation) for diabetes-related complications. Due to the use of a retrospective claims database, limitations of this analysis include a lack of formal diagnostic testing of patients, inability to measure factors such as disease duration and severity that are not captured in such databases, and the possible lack of generalization from this group of patients with diabetes to other populations. CONCLUSIONS: DN patients had significantly more comorbid medical conditions and higher healthcare charges and utilization than age- and sex-matched diabetic controls. Controlling for differences in overall illness burden, DN patients incurred more ER visits and inpatient admissions.