Health Insurance and Duration of Symptoms Prior to Emergency Department Visit: An Analysis of 19,850 Patients With Suspected COVID-19

ABSTRACT

Study Objectives: The COVID-19 pandemic has emphasized disparities in health outcomes across social and economic strata. The mechanisms of this relationship are poorly understood, but the length of time patients exhibit symptoms prior to getting tested for COVID-19 increases the opportunity for community transmission. We hypothesized that there is a relationship between insurance coverage and the duration of COVID-19 symptoms prior to seeking care at the emergency department (ED). Methods: A national, multi-institution (n=45 sites) registry collected information on ED visits in which patients were tested for suspected COVID-19. Demographics and clinical characteristics were summarized for the total cohort. Insurance was categorized into private (private or commercial), public (Medicare, Medicaid, or dual-eligible), worker’s compensation or unknown, or no health insurance. Negative binomial regression was used to analyze both the unadjusted and adjusted relationship between insurance and the time from symptom onset to ED presentation. Adjustments included age, sex, race, ethnicity, medical history, smoking status, drug use, and number of COVID symptoms. Results: Baseline demographic and clinical characteristics of included patients (n=19,850) are displayed in Table 1. The average time from symptom onset to ED presentation among patients with suspected COVID-19 was 5.4 days. In unadjusted analysis, patients with private insurance had significantly longer time of symptom onset prior to ED presentation than patients with public insurance (5.6 vs. 5.3 days, p=0.007). After multivariate adjustment, increased duration of symptoms prior to ED presentation was significantly associated with private insurance [rate ratio (RR) 1.07, 95% confidence interval (CI) 1.03 – 1.10] and no health insurance (RR 1.06, 95% CI 1.07 – 1.13) compared to public insurance (Figure 1). Patients residing in states with Medicaid expansion were not independently associated with the increased time to ED presentation (RR 1.03, 95% CI 1.00 – 1.07). Conclusion: Patients with private insurance or no insurance waited significantly longer to present to the ED. The extended duration of symptoms prior to presentation increases the opportunities for community transmission. The results from this study can be used by health systems to target the patients at increased risk for delayed ED presentation. [Formula presented]