Patient access to chronic medications during the Covid-19 pandemic: Evidence from a comprehensive dataset of US insurance claims

Abstract

Patient access and adherence to chronic medications is critical. In this work, we evaluate whether disruptions related to Covid-19 have affected new and existing patients' access to pharmacological therapies without interruption. We do so by performing a retrospective analysis on a dataset of 9.4 billion US prescription drug claims from 252 million patients from May, 2019 through August, 2020 (about 93% of prescriptions dispensed within those months). Using fixed effect (conditional likelihood) linear models, we evaluate continuity of care, how many days of supply patients received, and the likelihood of discontinuing therapy for drugs from classes with significant population health impacts. Findings indicate that more prescriptions were filled in March 2020 than in any prior month, followed by a significant drop in monthly dispensing. Compared to the pre-Covid era, a patient's likelihood of discontinuing some medications increased after the spread of Covid: norgestrel-ethinyl estradiol (hormonal contraceptive) discontinuation increased 0.62% (95% CI: 0.59% to 0.65%, p<0.001); dexmethylphenidate HCL (ADHD stimulant treatment) discontinuation increased 2.84% (95% CI: 2.79% to 2.89%, p<0.001); escitalopram oxalate (SSRI antidepressant) discontinuation increased 0.57% (95% CI: 0.561% to 0.578%, p<0.001); and haloperidol (antipsychotic) discontinuation increased 1.49% (95% CI: 1.41% to 1.57%, p<0.001). In contrast, the likelihood of discontinuing tacrolimus (immunosuppressant) decreased 0.15% (95% CI: 0.12% to 0.19%, p<0.001). The likelihood of discontinuing buprenorphine/naloxone (opioid addiction therapy) decreased 0.59% (95% CI: 0.55% to 0.62% decrease, p<0.001). We also observe a notable decline in new patients accessing these latter two therapies. Most US patients were able to access chronic medications during the early months of Covid-19, but still were more likely to discontinue their therapies than in previous months. Further, fewer than normal new patients started taking medications that may be vital to their care. Providers would do well to inquire about adherence and provide prompt, nonjudgmental, re-initiation of medications. From a policy perspective, opioid management programs seem to demonstrate a robust ability to manage existing patients in spite of disruption.

Fig 1. US prescription drug claims dispensed by month.

The figure shows the total number of approved pharmaceutical claims in the database; the red line indicates the ramp up of Covid-19 cases and response in the US. The trendline is based on data from May 2019 through February 2020 and represents a linear forecast for the period beyond February 2020. There were 12.05% fewer claims in August 2020 than in August 2019, likely due a combination of factors: stock up in March 2020, reduced demand (due to delayed elective procedures), and access challenges.

Fig 2. Total days of supply (DOS) dispensed by month across key drugs.

Total days of supply (DOS) dispensed by month across key drugs. March 2020 represents the peak demand for many drugs, though there are exceptions such as dexmethylphenidate HCL; this stimulant used to treat ADHD essentially saw the early onset of the summer decline when Covid-19 began to impact the US. The red line indicates the ramp up of Covid-19 cases and response in the US.

Fig 3. Proportion of total rejected for attempting to fill too early by month.

Proportion of claims rejected for attempting to refill too early. This is calculated by dividing the number of claims rejected with NCPDP Code 79 –Early Refill by the total number of claims for the drug. Pre-COVID means are calculated through February 2020. The red line indicates the ramp up of Covid-19 cases and response in the US.