Clinical Risk and Outpatient Therapy Utilization for COVID-19 in the Medicare Population

Abstract

Importance: Multiple therapies are available for outpatient treatment of COVID-19 that are highly effective at preventing hospitalization and mortality. Although racial and socioeconomic disparities in use of these therapies have been documented, limited evidence exists on what factors explain differences in use and the potential public health relevance of these differences.

Objective: To assess COVID-19 outpatient treatment utilization in the Medicare population and simulate the potential outcome of allocating treatment according to patient risk for severe COVID-19.

Design, setting, and participants: This cross-sectional study included patients enrolled in Medicare in 2022 across the US, identified with 100% Medicare fee-for-service claims.

Main outcomes and measures: The primary outcome was any COVID-19 outpatient therapy utilization. Secondary outcomes included COVID-19 testing, ambulatory visits, and hospitalization. Differences in outcomes were estimated based on patient demographics, treatment contraindications, and a composite risk score for mortality after COVID-19 based on demographics and comorbidities. A simulation of reallocating COVID-19 treatment, particularly with nirmatrelvir, to those at high risk of severe disease was performed, and the potential COVID-19 hospitalizations and mortality outcomes were assessed.

Results: In 2022, 6.0% of 20 026 910 beneficiaries received outpatient COVID-19 treatment, 40.5% of which had no associated COVID-19 diagnosis within 10 days. Patients with higher risk for severe disease received less outpatient treatment, such as 6.4% of those aged 65 to 69 years compared with 4.9% of those 90 years and older (adjusted odds ratio [aOR], 0.64 [95% CI, 0.62-0.65]) and 6.4% of White patients compared with 3.0% of Black patients (aOR, 0.56 [95% CI, 0.54-0.58]). In the highest COVID-19 severity risk quintile, 2.6% were hospitalized for COVID-19 and 4.9% received outpatient treatment, compared with 0.2% and 7.5% in the lowest quintile. These patterns were similar among patients with a documented COVID-19 diagnosis, those with no claims for vaccination, and patients who are insured with Medicare Advantage. Differences were not explained by variable COVID-19 testing, ambulatory visits, or treatment contraindications. Reallocation of 2022 outpatient COVID-19 treatment, particularly with nirmatrelvir, based on risk for severe COVID-19 would have averted 16 503 COVID-19 deaths (16.3%) in the sample.

Conclusion: In this cross-sectional study, outpatient COVID-19 treatment was disproportionately accessed by beneficiaries at lower risk for severe infection, undermining its potential public health benefit. Undertreatment was not driven by lack of clinical access or treatment contraindications.

Figure 1.. Monthly Counts and Proportion of Infusions or Prescriptions of Outpatient COVID-19 Therapies Delivered to Fee-for-Service Medicare

A, The left panel shows the monthly volume of prescription claims (for oral medications) or infusion claims (for intravenous medications) of outpatient COVID-19 therapies from November 2020 to December 2022 for all Medicare beneficiaries, regardless of enrollment. The main analytic sample in subsequent exhibits focuses on continuously enrolled beneficiaries in 2022 alone. B, the graph shows the proportion of all prescriptions or infusions in a given month in 1 of 4 categories (monoclonal antibodies, molnupiravir, nirmatrelvir, or remdesivir). Prior to late 2021, only monoclonal antibodies were available for outpatient treatment of COVID-19.

Figure 2.. Adjusted Odds of Receiving Any Outpatient COVID-19 Treatment by Selected Patient Characteristics

Forest plot showing adjusted odds ratios (aORs) for a beneficiary-level logistic regression model with the outcome of receipt of any COVID-19 treatment in 2022, adjusting for all characteristics shown in Table 1 with robust standard errors clustered at the hospital referral region. Characteristics not shown for visualization: original reason for Medicare eligibility and telemedicine use. ^aOther race includes beneficiaries with American Indian or Alaska Native; unknown; or other race and ethnicity. These are the 3 options for self-identified race in the Medicare enrollment file besides Asian and Pacific Islander, Black, Hispanic, and non-Hispanic White. Less than 1% of beneficiaries are coded as unknown race.

Figure 3.. Rates of Outpatient COVID-19 Therapy and Inpatient Admissions by COVID-19 Severity Risk and Nirmatrelvir Contraindications

The figure shows rates of any outpatient COVID-19 treatment, nirmatrelvir treatment, and inpatient admission for Medicare beneficiaries in 2022 stratified by their COVID-19 severity risk quintile. This score serves as a measure of the aggregate observable risk across dozens of characteristics that a clinician may consider in deciding to prescribe COVID-19 therapy. Scores were based on estimated coefficients from a linear regression model of mortality following COVID-19 (within 21 days) in 2021, using each beneficiary’s characteristics in 2022 (eMethods 2 in Supplement 1). Beneficiaries were assigned to COVID-19 severity risk quintiles.