ICPE Research Spotlight 2021: Part I: Findings from Aetion-FDA research collaboration to advance understanding of COVID-19 with RWE
This year at ICPE All Access, Aetion’s scientific research is represented across symposia, presentations, and posters, including some initial findings from our COVID-19 research with the U.S. Food and Drug Administration (FDA) and our continued efforts to advocate for best practices in pharmacoepidemiologic methods. In this two-part series, we break down the key takeaways from each presentation. This is part one; see our full list of posters and presentations here.
The rapidly evolving COVID-19 landscape increased the urgency with which health care decision makers needed near real-time evidence to inform decisions about the natural history of this new disease and the associated treatment patterns—including safety and effectiveness—in the real world. From this need came a number of collaborations across health care sectors—including regulators, biopharma, data providers, hospital systems, technology organizations, and more—to facilitate access to COVID-19 real-world data (RWD), and to streamline the generation of real-world evidence (RWE).
In May 2020, FDA and Aetion announced a research collaboration to use RWD to advance the understanding of and response to COVID-19. The research highlighted below—prepared for ICPE All Access 2021 and powered by Aetion Evidence Platform®—shares some initial findings from this collaborative effort, and features co-authors from FDA, Aetion, and one of our data partners, HealthVerity.
Use of an electronic health record (EHR) to inform an administrative data algorithm to categorize inpatient COVID-19 severity
Elizabeth M. Garry, PhD, MPH; Andrew Weckstein, BA; Kenneth Quinto, MD, MPH; Tamar Lasky, PhD, FISPE; Aloka Chakravarty, PhD, MStat; Sandy Leonard, MPH; Sarah Vititoe, MPH; Imaani Easthausen, MS; Jeremy A. Rassen, ScD, FISPE; Nicolle Gatto, PhD, MPH, FISPE
Among patients hospitalized with COVID-19, the level of respiratory support needed—meaning whether a patient requires no supplemental oxygen, supplemental or non-invasive oxygen, or invasive mechanical ventilation—emerged as a key indicator of disease severity. Understanding and controlling for this factor in RWD analyses can help reduce confounding when conducting studies to determine inpatient treatment effectiveness.
In this project, authors developed a COVID-19 severity score based on the level of respiratory support for use in inpatient RWD studies. They used additional data available in Optum EHR to improve upon the algorithm initially developed for HealthVerity administrative data.
In both the EHR and the administrative data sources, authors found that algorithms based solely on procedure-related codes substantially missed a clinical need for supplemental oxygen or noninvasive ventilation among hospitalized COVID-19 patients, a misclassification that was differential by treatment status. Such disease severity algorithms will allow for more specific categorization of lower and higher disease severity patients in future research.
Categorization of COVID-19 severity to determine mortality risk
Elizabeth M. Garry, PhD, MPH; Andrew Weckstein, BA; Kenneth Quinto, MD, MPH; Marie C. Bradley, PhD, MScPH, MPharm; Tamar Lasky, PhD, FISPE; Sandy Leonard, MPH; Sarah Vititoe, MPH; Imaani Easthausen, MS; Nicolle Gatto, PhD, MPH, FISPE
To further establish the utility of the COVID-19 severity score described in the previous presentation, authors aimed to estimate the mortality risk and rate in a cohort of patients hospitalized for COVID-19, and to demonstrate that patients with greater disease severity upon admission had an increased risk of mortality.
Authors observed a positive association between the algorithm-defined severity at admission and the 28-day mortality risk and rate. Although performance remains to be validated, this provides assurance that this algorithm may be used for confounding control or stratification in treatment effect studies.
Examining the evolving pandemic: Treatment patterns in patients hospitalized with COVID-19 by race and ethnicity
Sarah Vititoe, MPH; Imaani Easthausen, MS; Tamar Lasky, PhD, FISPE; Aloka Chakravarty, PhD, MStat; Marie C. Bradley PhD, MPharm, MScPH; Laura M. Roe, MMCi; Nicolle Gatto, PhD, MPH, FISPE; Andrew Weckstein, BA; Elizabeth M. Garry, PhD, MPH
The COVID-19 pandemic in the U.S. continues to highlight health disparities across racial and ethnic groups. While several studies have reported differences in COVID-19 hospitalizations and mortality by race and ethnicity, few have characterized the differences in treatment patterns among those hospitalized with COVID-19. In this study, authors used Optum de-identified COVID-19 EHR data to describe the differences in treatment patterns of hospitalized COVID-19 patients by race and ethnicity, controlling for factors such as baseline comorbidities, disease severity, and month of admission.
The findings suggest racial and ethnic differences in inpatient COVID-19 treatment patterns among the study cohort—for example, Black/African American patients were less likely to receive dexamethasone treatment than White patients. Additional research is needed to understand driving factors for these differential treatment patterns, as well as how differences in treatment patterns may have impacted outcomes in different racial and ethnic groups.
Temporal trends in medication utilization among hospitalized COVID-19 patients
Andrew Weckstein, BA; Sarah Vititoe, MPH; Donna Rivera, PharmD, MSc; Marie C. Bradley, PhD, MScPH, MPharm; Silvia Perez-Vilar, PharmD, PhD; Sandy Leonard, MPH; Elizabeth M. Garry, PhD, MPH; Nicolle M. Gatto, PhD, MPH, FISPE; Jeremy A. Rassen, ScD, FISPE
RWD can play an important role in illuminating changes in treatment patterns for hospitalized COVID-19 patients over time. Authors aimed to evaluate trends and shifts in medication usage for hospitalized COVID-19 patients in the U.S., overall and stratified by disease severity. The findings demonstrate a considerable shift in inpatient COVID-19 management from March 2020 through May 2021, highlighting an evolving treatment paradigm. For example, while azithromycin and hydroxychloroquine were used most frequently early in the pandemic, dexamethasone and remdesivir, respectively, later surpassed hydroxychloroquine as the most commonly used medications.
The research demonstrated how the thoughtful use of RWD can provide researchers, regulators, and drug manufacturers with an up-to-date understanding of large-scale health care trends, which can inform critical public health decisions. It also illustrates how the clinical management of diseases can change rapidly in response to new information, regulatory communications, and other real-world events: for example, the decline in hydroxychloroquine use following the publication of emerging scientific data and FDA regulatory actions, and the spike in dexamethasone use following the positive RECOVERY trial results.
Temporal changes in the baseline patient characteristics of COVID-19 patients
Margaret Okobi, MPH; Patricia Prince, MPH; Laura M. Roe, MMCi; Donna Rivera, PharmD, MSc; Kenneth Quinto, MD, MPH; Sandy Leonard, MPH; Elizabeth M. Garry, PhD, MPH; Andrew Weckstein, BA; Melanie Wang, MPH, MBA; Nicolle Gatto, PhD, MPH, FISPE
This research aimed to characterize the evolving presentation of COVID-19 as the pandemic progressed. The authors summarized the temporal changes in patient characteristics, including comorbidities and medication use, during the baseline period prior to COVID-19 diagnosis.
The research team used HealthVerity’s Chargemaster and medical and pharmacy claims data for the analyses. One key element of their work involved exploring ways to identify COVID-19 patients in claims and Chargemaster data, for example, by analyzing which ICD-10 diagnosis codes and lab codes would best capture patients with COVID-19.
This study found that the average documented COVID-19 patient changed remarkably over the course of the pandemic, a change that can likely be attributed to increased access to testing. The study also found that metabolic syndrome—which includes a range of conditions such as cardiovascular disease, hypertension, diabetes, and obesity—remained the most common set of comorbidities amongst COVID-19 patients. The authors suggested that access to health care resources may significantly impact the patients seen in RWD.