Pre-Term Deliveries Due to COVID-19 Could Be Avoided by Studying EHRs
Using electronic health record data to simulate drug trials for pregnant patients could one day offer a solution to the current practice of delivering babies, even if they are pre-term, if the mother contracts COVID-19, according to a position paper published in Nature Medicine.
Pregnant patients have typically been excluded from drug trials out of concern for fetal safety. And when human health is on the line, drug studies assessing fetal safety in animal models may be viewed as far from definitive. Due to sheer lack of data concerning implications for fetal and maternal safety, clinicians are often unsure about prescribing drugs to pregnant patients.
A position paper led by first author Anup Challa and senior author David Aronoff, MD, director, Division of Infectious Diseases at Vanderbilt University Medical Center, outlines how these deficits can lead to undertreatment of chronic and acute illness in pregnant patients, while also posing additional risk of adverse drug reactions.
"Pregnant patients are an especially vulnerable population, since exposure to many common drugs could harm their unborn children,” said Challa, principal investigator for MADRE and student worker, Vanderbilt Drug Repurposing System. “Further complicating this problem is the fact that pregnant patients are not allowed to enroll in clinical trials, given the ethical issues that a harmful drug reaction in their fetus could pose.
“These factors have significantly reduced quality of care for pregnant patients with COVID-19, as the current guidance to OB/Gyns is to deliver pregnant patients if they contract COVID-19 - even if the patients are significantly pre-term. This could significantly harm their fetuses, and recent case reports have shown that COVID-19 transmission occurs in utero if not treated in the mother,” he added.
Challa and Aronoff propose that using electronic health record data to emulate randomized controlled trials could offer an alternative to delivering all pregnant patients with COVID-19.
Used to compare treatments, these trials involve enrollment of subjects who undergo interventions carried out in real time. In contrast, target trials are a type of observational study that simulates the trial through retrospective analysis of existing clinical data.
“We could identify therapeutics that are safe for use in these patients by using high-powered algorithms like machine learning to learn from cases in which pregnant patients have been exposed to current therapeutic candidates and understand how much risk they pose to a fetus,” Aronoff said.
Vanderbilt University Medical Center has electronic health records for more than 2 million patients, which allows investigators to design and conduct “trials” that simulate not only a real trial’s treatment strategies (drug versus no drug) and outcomes, but also eligibility criteria and random assignment to treatment at baseline.
Such trials are arguably “the only ethical way to gather human drug exposure data for pregnant people on a significant scale and across all classes of drugs,” the authors said.
“If expanded, the Vanderbilt platform of target trials may offer an alternative to delivering all pregnant patients with COVID-19 by identifying therapeutics that are safe,” Challa said. “Without a good understanding of safety for these drugs, we might be missing opportunities to treat COVID-19 in pregnant patients."
In addition to Challa and Aronoff, co-authors include Robert Lavieri, Ph.D., Ethan Lippmann, Ph.D., Lisa Bastarache, MS, Jill Pulley, MBA, and Jeffery Goldstein, MD, Ph.D.