COVID Patient ‘Lottery’ Ensured Scarce Drugs Divided Equitably

A weighted “lottery” system designed to allocate scarce COVID-19 treatment to patients most in need worked as intended from an equity and population health standpoint, a researcher said.

Across nearly two dozen hospitals, the system resulted in 61% of the allotted remdesivir (Veklury) being allocated to either essential workers or patients from disadvantaged neighborhoods, reported Douglas White, MD, of the University of Pittsburgh School of Medicine.

Per design, the weighted lottery gave 25% higher chances to COVID-19 patients who were essential workers or who lived in disadvantaged areas based on Area Deprivation Index (ADI) scores.

It also reduced the chance of allocation by 50% for patients whose death was expected within a year due to an end-stage condition or those who required mechanical ventilation or extracorporeal membrane oxygenation (ECMO). White noted this was after data emerged that severe COVID patients had lower likelihood of benefit with remdesivir.

The weighted lottery was developed when there was a 3-week remdesivir shortage during spring 2020 across 21 hospitals in Pennsylvania, he explained in a presentation at the American Thoracic Society virtual meeting.

The researchers wanted to see if the weighted lottery was feasible, and to examine whether therapeutics were allocated to promote equity and population health outcomes.

“This is a major improvement over the first-come, first-served approach that many hospitals have used; that approach is likely to worsen disparities for those with access-to-care barriers,” said White in a statement.

White detailed how the daily lottery worked with a variety of stakeholders, including clinicians, hospital operations experts, and diversity and inclusion experts, as well as the Commonwealth of Pennsylvania Ethical Allocation Committee. A centralized search of the electronic health record identified patients with supplemental oxygen requirements, and then the local hospital allocation team conferred with clinicians and made the referral.

All patients met clinical eligibility criteria for remdesivir emergency use authorization (EUA). The lottery was conducted daily in a central location in Pittsburgh using a random number generator.

Overall, 93 patients were eligible for remdesivir, with a mean age of 68, with about an equal gender split. About 70% of patients were white. There were 44% of patients from disadvantaged neighborhoods (defined as an ADI score of 8 or higher), and 20% who were essential workers. A little less than 10% were expected to die within the year, and 17% were receiving mechanical ventilation or ECMO.

Of these, 59% of patients were allocated to remdesivir, though seven patients refused treatment, as all 55 courses of remdesivir were allocated to patients in 21 hospitals over the 3-week period. The authors noted during this shortage, the general population’s chances of receiving remdesivir ranged from 28% to 88%.

The team outlined the protocol with instructions for other hospitals wanting to design and implement a similar lottery system. White advocated the efficiency of conducting one lottery involving many regional hospitals versus each hospital doing their own allocation.

“It was our goal from the outset to develop a framework and process that any hospital can use,” White said. “This approach can simultaneously accomplish fair allocation and rapid learning, because the lottery creates a natural experiment in which some patients receive the scarce drug while others do not.”