A Retrospective Review of Healthcare Disparities among Economic Subgroups with Hematologic Malignancies

Background: In 2007, one in five Americans reported delayed or inadequate health care. Of those, 15.5 million had no insurance coverage. The number of underinsured patients increased to 25.2 million in 2007. Furthermore, a recent study showed that Non-Hodgkin’s lymphoma patients with Medicaid or no insurance presented at a later stage. The mission of our institution, MetroHealth Medical Center, a county hospital, is to provide excellent health care to all individuals, regardless of their ability to pay. A significant number of patients who are uninsured or underinsured are treated at this institution, as well as patients with adequate insurance coverage. Along with financial and social staff to assist these patients in obtaining access to Medicare and Medicaid benefits, MetroHealth employs a unique rating system to assess patients’ financial needs in a timely manner and provide assistance with these needs. Our primary objective was to evaluate whether a lack of adequate health insurance significantly delayed the time to diagnosis and/or treatment of patients with hematologic malignancies.

Methods: Patients with hematologic malignancies, diagnosed between 1999 and 2007, were identified using the computerized Tumor Registry Board at our institution. We retrospectively compared the diagnostic interval (time from presentation to a medical provider, with a symptom or finding clearly related to malignancy, until diagnosis) and the treatment interval (time from diagnosis to treatment) in patients with or without insurance. Patients with hematologic malignancies who had an indication for treatment at the time of diagnosis were included in the study, however, patients requiring emergent treatment at the time of presentation were excluded (as financial evaluation would not be required for such treatment). We also considered covariates of age, sex, race and patient compliance (as measured by the number of “no show” appointments in the diagnostic and treatment intervals).

Results: A total of 206 patients were identified who met the inclusion criteria and for whom the relevant time intervals and covariates could be determined. Kaplan-Meier curves were constructed to compare subgroups based on insurance status. We found no difference in the

• diagnostic or

• treatment intervals for insured vs uninsured patient (a: 18 vs 20 days, p = 0.7071; b: 20 vs 28 days, p = 0.2433), or when the patients were grouped by type of insurance (a: p = 0.6000, b: p = 0.2196).

Cox proportional hazards models were used to evaluate the effect of the covariates on the diagnostic and treatment intervals. A significant correlation was found between patient non-compliance and both diagnostic (p = 0.0007, HR = 0.8184) and treatment intervals (p = 0.0499, HR = 0.8841). Hazard ratios should be interpreted as the relative decrease in the probability of diagnosis (or treatment) within any time interval, given one additional “no show” appointment. The treatment interval was also found to be significantly correlated with race (p = 0.0050, HR = 0.6884). Otherwise the time intervals were independent of the tested covariates.

Conclusion: The financial evaluation and assistance system employed at MetroHealth Medical Center allows patients to receive rapid access to needed health care. No delays in time to diagnosis or treatment, related to insurance status, were observed in patients with hematologic malignancies seen at this institution over a 9 year period. In a time when national health care reform is being contemplated, this system may serve as a useful model for health care access to individuals with varying financial means. Of note, a small disparity in treatment interval was seen related to race. This may be due to correlation between race and type of malignancy, but is being investigated further.