Evaluating sources of bias in observational studies of angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker use during COVID-19: beyond confounding

Abstract

Concerns over ACE inhibitor or ARB use to treat hypertension during COVID-19 remain unresolved. Although studies using more robust methodologies provided some clarity, sources of bias persist and it remains critical to quickly address this question. In this review, we discuss pernicious sources of bias using a causal model framework, including time-varying confounder, collider, information, and time-dependent bias, in the context of recently published studies. We discuss causal inference methodologies that can address these issues, including causal diagrams, time-to-event analyses, sensitivity analyses, and marginal structural modeling. We discuss effect modification and we propose a role for causal mediation analysis to estimate indirect effects via mediating factors, especially components of the renin--angiotensin system. Thorough knowledge of these sources of bias and the appropriate methodologies to address them is crucial when evaluating observational studies to inform patient management decisions regarding whether ACE inhibitors or ARBs are associated with greater risk from COVID-19.

International Analysis of Electronic Health Records of Children and Youth Hospitalized With COVID-19 Infection in 6 Countries

“…However, values should be interpreted with some caution because tests are not obtained consistently on all patients and reflect physician decisions based on the patient’s clinical presentation and local health care workflows. 29 Evaluation of the number of patients with available laboratory tests indicated that the number tested dropped quickly during the first week of hospitalization. Nonetheless, our findings were consistent with clinical results presented in case series and meta-analyses, showing generally normal white blood cell counts and abnormally elevated inflammatory markers and coagulation tests.…”

Renin-angiotensin system blockers and COVID-19

“…In contrast, most observational studies which analyzed in-hospital exposure to RAS-blockers concluded in favor of a strong protective effect associated with treatment continuation [1,5]. However, among the myriad of observational studies published on RAS blockers and COVID-19 since the SARS-CoV-2 outbreak, many have suffered from important methodological limitations [6].…”

COVID-19 morbidity and mortality associated with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers use among 14,129 patients with hypertension from a US integrated healthcare system

“…Many epidemiologic studies have investigated the use of ACEIs or ARBs in the general population with COVID-19 infection [ 3 ] rather than patients with hypertension. This approach may introduce collider bias since hypertension increases the likelihood of ACEI or ARB use and a test for COVID-19 [ 4 ]. Furthermore, the findings have not always been consistent in the different populations studied.…”