Methods to Account for the Impact of COVID-19 in Real-World Retrospective Studies of Economic Outcomes: A Targeted Literature Review

OBJECTIVES: The emergence of the coronavirus disease 2019 (COVID-19) drastically changed patient diagnosis, treatment, healthcare activity and behaviors. It is critical that real-world (RW) research using data from 2020 onwards appropriately accounts for potential confounding from the COVID-19 pandemic; however, there are no accepted guidelines for adjustment.

METHODS: A targeted literature review (TLR) was conducted in OVID for retrospective studies published between January 2022 and June 2023 that used claims or EMR databases and evaluated economic outcomes. Studies that were not of a primarily retrospective design, those that specifically evaluated COVID as an outcome, or that were non-English language were excluded. Two reviewers screened abstracts identified through the search; a third reviewer was engaged in case of discordance. Results were extracted for all eligible articles with study data ending December 2019 or later.

RESULTS: In total, 111 studies were identified as relevant and reviewed. N=53 studies ended their study period between December 2019 and March 2020 and of those, only five stated their rationale of avoiding changes in healthcare activity and associated costs due to the pandemic. N=58 studies had study periods that included data following March 2020 and of those, 11 studies acknowledged the impact of COVID-19 on their findings, while only one study conducted specific analysis to account for the pandemic. This sole study removed data from the COVID-19 period in their sensitivity analysis.

CONCLUSIONS: Overall, most retrospective studies of economic outcomes did not mention or account for the COVID-19 pandemic. In the limited studies that referenced COVID-19, only one study utilized specific methodology to mitigate potential confounding related to the pandemic.