Clinical and economic outcomes of pharmacological stress tests in patients with a history of COVID‐19

Abstract Background Despite millions of COVID‐19 cases in the United States, it remains unknown whether a history of COVID‐19 infection impacts the safety of pharmacologic myocardial perfusion imaging stress testing (pharmacologic MPI). Hypothesis The aim of this study was to assess if a prior COVID‐19 infection was associated with a higher risk of complications during and following pharmacologic MPI testing. Methods This retrospective cohort analysis included 179 803 adults (≥18 years) from the PharMetrics® Plus claims database who underwent pharmacologic MPI between March 1, 2020 and February 28, 2021. Patients with a history of COVID‐19 infection (COVID‐19 group) were compared with propensity‐score matched no‐COVID‐19 history group for reversal agent use, 30‐day resource use, and post‐MPI cardiac events/procedures. Results The most commonly used stress agent was regadenoson (91.7%). The COVID‐19 group (n = 6372; 3.5%) had slightly higher: reversal agent use (difference 1.13% [95% confidence interval [CI]: 0.33, 1.92]), all‐cause costs (difference USD $128 [95% CI: $73–$181]), and office visits (81.5% vs. 77.0%) than the no‐COVID‐19 group. Prior COVID‐19 infection did not appear to impact subsequent cardiac events/procedures. Conclusions COVID‐19 history was associated with slightly higher reversal agent use, all‐cause costs, and office visits after pharmacologic MPI; however, the differences were not clinically meaningful. Concerns for use of stress agents in patients with prior COVID‐19 do not appear to be warranted.


| INTRODUCTION
Over the last 2.5 years, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), or COVID-19, infections affected several million individuals with a broad range of complications and outcomes, ranging from short-lived effects with no or mild symptoms, to prolonged hospitalizations, severe complications, or death. 1,2 In addition to their baseline risk, COVID-19 survivors may have a heightened risk of cardiac complications and may be referred for pharmacologic myocardial perfusion imaging stress tests (pharmacologic MPI) for symptoms related, or not, to their prior COVID-19 infection. [3][4][5][6][7][8] Depending on the local setting, COVID-19 incidence, and availability, the use of pharmacologic MPI may be recommended instead of an exercise stress test to reduce droplet exposure risk and thus avoid the potential spread of COVID-19. 9 Although lab closures in response to local and national recommendations early in the pandemic resulted in reduced cardiac diagnostic testing procedures, volumes of procedures in the United States and Canada in 2021 slightly surpassed prepandemic levels. 10 However, it remains unknown whether a prior COVID-19 infection impacts the safety of pharmacologic MPI.
Guidelines recommend regadenoson as the preferred coronary vasodilator agent for pharmacologic MPI during the pandemic 9 due to its shorter infusion time (10 seconds) compared with several minutes for adenosine and dipyridamole, which helps to minimize contact time between health professionals and patients. [11][12][13] In the event of serious or intolerable adverse effects during pharmacologic MPI, reversal agents such as caffeine and the adenosine receptor antagonist (aminophylline) can be administered to reverse the effects of the vasodilator agent. [14][15][16][17] In the regadenoson phase 3 trials of 2015 patients undergoing pharmacologic MPI (median [range] age: 66  years; primarily White males), 3% of patients receiving regadenoson (n = 46/1337) and 2% of those receiving adenosine (n = 12/678) received reversal agents. 11,18 Recently, Hasnie et al. reported that 1 out of 15 patients (6.7%) who recently recovered from COVID-19 used aminophylline. 19 The aim of this study was to assess if a prior COVID-19 infection was associated with a higher risk of complications during and following pharmacologic MPI testing. Specifically, the effect of a prior history of COVID-19 infection on reversal agent use, subsequent healthcare visits, all-cause costs, and cardiac events/procedures following pharmacologic MPI was examined.

| Study design and data source
This retrospective cohort analysis used the IQVIA PharMetrics ® Plus claims database, a longitudinal health-plan database of medical and pharmacy claims, including 200 million enrollees from US national and subnational health plans and self-insured employer groups.
The study was conducted in compliance with all national requirements for non-interventional studies using deidentified data. The data from IQVIA PharMetrics ® Plus is permitted to be in research and publications; informed consent, ethics committee approval, and Institutional Review Board approval were not necessary for this study.

| Study population
The study population consisted of patients aged ≥18 years who underwent pharmacologic MPI between March 1, 2020 and

| Outcome definitions
Reversal agent use was defined by procedure codes or National Drug Code on the index date. Among those requiring reversal agent use, a prespecified diagnosis was sought as a potential indication of reversal agent use. [14][15][16][17]  The rate of reversal agent use and proportion with each type of healthcare visit were compared between the matched groups. The 95% confidence intervals (CIs) for the difference in frequencies were estimated based on t distribution, and odds ratios (ORs) were estimated using logistic models. The impact on all-cause costs was estimated as median differences between the matched pairs and 95% confidence limits (CLs) for the median difference were estimated by distribution-free CLs for quantiles. Chronological trends in pharmacologic MPI use during the study period were assessed by taking potential changes in the size of the database itself into account (e.g., by leaving insurance programs).
The monthly pharmacologic MPI rate/10 000 beneficiaries were estimated by dividing the number of monthly indexes pharmacologic MPI from March 2020 to February 2021 by the monthly total number of beneficiaries in the database and multiplying by 10 000.
The monthly reversal agent use rate by geographic region was estimated among pharmacologic MPI patients with known geographic regions. Reversal agent use rate (%) was estimated per geographical region and was calculated as a proportion of pharmacologic MPI with reversal-agent use among total pharmacologic MPI in each region.
The statistical analysis plan was developed and iteratively revised by the authors. Statistical analyses were performed using SAS software, Version 9.04.

| Patients
In total, 179 803 patients with complete data underwent pharmacologic MPI and were included in the study (Supporting Information: Figure S1). Age ranged from 52 to 69 years, and 64% (N = 115 668) of patients were aged <65 years (Table 1)

| Reversal agent use
In total, before matching, reversal agents were used in 13 890 patients (7.7%). Reversal agents used were aminophylline (98.0%) and caffeine (2.0%); theophylline was not used. The most common diagnoses given on the date of reversal agent use were chest pain (54.6%) and arrhythmia (13.8%).
Reversal agent use was higher in patients with a history of COVID-19 infection than in those without a history of COVID-19 infection (8.8% vs. 7.7%; difference 1.1% [95% CI: 0.4, 1.8]). Among patients aged <65 years, reversal agent use was higher among those with a COVID-19 history than those without a COVID-19 history (8.9% vs. 7.3%); while among patients aged ≥65 years, reversal agent use was similar in those with or without a history of COVID-19 infection (8.4% vs. 8.3%). In general, reversal agent use appeared to be higher in women (9.4%) than in men (6.3%).
Among patients with a history of COVID-19 infection, reversal agent use was lower in patients who had experienced severe COVID-19 than in those who had experienced nonsevere COVID-19 (8.4% vs. 9.0%). Use of reversal agents was higher among patients with recent COVID-19 infection (≤2 weeks between COVID-19 diagnosis and pharmacologic MPI; 11.0%) compared to those with pharmacologic MPI greater than 2 weeks after COVID-19 diagnosis (8.7%). There was no significant change over time with longer time since the COVID-19 diagnosis ( Figure 1A). In a multivariate model, factors associated with reversal agent use included: a history of COVID-19 infection, age ≥65 years, female sex, use of dipyridamole as a pharmacological stress agent, comorbidities (myocardial infarction, congestive heart failure, cerebrovascular disease, dementia, chronic pulmonary disease, hemiplegia or paraplegia, and renal disease), and MPI place of service in an outpatient hospital or other settings (vs. physicians' office) ( Figure 1B).

| All-cause costs and healthcare visits
All-cause costs were minimally higher among those with versus those without a history of COVID-19 infection (

| Cardiac outcomes after pharmacologic MPI
Among patients enrolled by December 2020, there was no difference in cumulative incidence of cardiac events or procedures between patients with a history of COVID-19 infection and those without during the 90-day post-pharmacologic MPI period (Figure 2A-C).
There was no difference in HRs between the two groups in cardiac angiography without revascularization

| Chronological trends in pharmacologic MPI and geographical variations in reversal agent use
Overall rates of pharmacologic MPI use from March 2020 to February 2021 were similar between the Northeast, Midwest, South, and West regions ( Figure 2D). Slightly lower pharmacologic MPI rates were observed between March and May 2020 (2.5/ 10 000 beneficiaries), with a dip in April 2020 (1.5/10 000 beneficiaries) but a recovery in June 2020 (3.4/10 000 beneficiaries). The reversal agent use rate against total pharmacologic MPI was also slightly lower in April 2020 (7.2%) but rather stable during the study period (7.2%-8.2%).
Reversal agent use generally remained unchanged over the entire period ( Figure 2E).

| CONCLUSIONS
In a large claims database of patients undergoing pharmacologic MPI, a prior COVID-19 infection was associated with slightly higher reversal agent use, all-cause costs, and office visits; however, the differences were not clinically meaningful. Concerns for use of stress agents in patients with prior COVID-19 do not appear to be warranted.

AUTHOR CONTRIBUTIONS
David Walker and Tomomi Kimura conceived and developed the study. Hicham Skali, David Walker, Jeanette Jiang, and Tomomi Kimura provided acquisition, analysis, or interpretation of data.
All authors critically revised each draft of the manuscript and reviewed and approved the final version.

ACKNOWLEDGMENTS
The authors would like to thank Robson Barbosa Lima for his contribution to the study. This study was funded by Astellas Pharma Global Development Inc. Medical writing support for the outline and the first draft was provided by Andrea Bothwell, BSc, on behalf of Oxford PharmaGenesis Inc., with funding provided by Astellas.