The RECOVERY Trial tested the efficacy of four different drugs in the treatment of COVID-19 patients in the United Kingdom. In the New England Journal of Medicine, S. James Adelstein Professor of Health Care Policy (Biostatistics) Sharon-Lise Normand, PhD discusses the successes and shortfalls of this trial.
In this trial, 11,303 patients with COVID-19 were randomly assigned a treatment plan. They received one of four drugs- dexamethasone, hydroxychloroquine, lopinavir–ritonavir, or azithromycin- or underwent usual care. The treatment was further randomized with some patients receiving convalescent plasma. Some with progressive COVID-19 were randomly assigned to receive tocilizumab.
The RECOVERY Trial found that patients, especially those with progressive COVID-19 who had been put on ventilators, responded well to the dexamethasone treatment relative to usual care. Those treated with dexamethasone showed a 2.8% absolute mortality benefit over patients in the control group. Furthermore, ventilated patients with progressive COVID-10 showed an absolute mortality reduction of 12.1% after being treated with dexamethasone.
The success of the RECOVERY Trial showed that broad populations of patients, hospitals, and trial coordinators can be brought together quickly in times of great urgency. The use of routine health care data, national registry data, and a single use online follow-up form proved that minimal data collection can still provide meaningful outcomes. Perhaps the largest success of the trial was discovering that dexamethasone showed promise for reducing COVID-19 short-term mortality rates among ventilated patients compared to usual care.
However, the RECOVERY Trial did have some limitations. The decisions to add or remove therapies in such trials require clear stopping rules and, unfortunately, these rules were not well-characterized by the investigators. All four drug treatment group comparisons utilized the same usual care group. Repeated use of the same comparison group can provide results that may be simply due to chance unless some control measures are introduced. Finally, patients were not randomized within hospitals nor were they blinded. Normand suggests that randomization with the use of permuted blocks within hospitals could have helped maintain blinding and provided more opportunities to balance treatment groups within hospitals that could help reduce hospital-specific differences attributed to differences in usual care.
Despite the shortfalls, Normand praises the RECOVERY Trial investigators and participants.
“The results of this trial represent an important step in the fight against one aspect of the disease and undoubtedly will have an effect on practice,” Normand says, “However, the methodologic caveats raised here are important to other investigators who are developing and revising treatment protocols in hospitals and to the broader research community struggling to produce reliable results in an efficient way, even in the face of a pandemic.”