Use of Cardiopulmonary Exercise Testing to Evaluate Long COVID-19 Symptoms in Adults

Key Points Question Is exercise capacity reduced more than 3 months after SARS-CoV-2 infection among those with long COVID-19 (LC) symptoms compared with recovered individuals without symptoms, and what patterns of limitations on cardiopulmonary exercise testing (CPET) are common? Findings In this systematic review and meta-analysis of 38 studies comprising 2160 participants, exercise capacity was reduced by 4.9 mL/kg/min among individuals with symptoms consistent with LC compared with individuals without symptoms more than 3 months after SARS-CoV-2 infection. Findings among individuals with exertional intolerance suggest that deconditioning, dysfunctional breathing, chronotropic incompetence, and abnormal peripheral oxygen extraction and/or use may contribute to reduced exercise capacity. Meaning These findings suggest that CPET may provide insight into the mechanisms for reduced exercise capacity among individuals with LC.

Condition being studied: post-acute sequelae of COVID-19, also known as Long COVID, which according to the WHO definition is >3 months after acute infection with SARS-CoV-2.
Inclusion criteria: all studies of adults with confirmed COVID-19 at least 3 months after onset that include cardiopulmonary exercise testing with measurement of peak VO 2 published since 2020 will be included. Baseline cardiopulmonary exercise testing from interventional or randomized controlled trials will also be included if they meet the other inclusion criteria.
Exclusion criteria: studies of children, studies of conditions other than COVID-19/SARS-CoV-2, studies in the acute or early post-acute phase (<3 months after infection), review articles, case reports.
Intervention/exposure: Cardiopulmonary exercise testing, which includes measurement of metabolic gases with either treadmill or cycle ergometer exercise.
Participants/population: We are interested in all adults with COVID without respect to hospitalization status or severity of acute illness.
Inclusion/Exclusion criteria: adults with confirmed COVID-19 at least 3 months after onset that include cardiopulmonary exercise testing with measurement of peak VO 2 will be included. We excluded studies of children, studies of conditions other than COVID-19/SARS-CoV-2, studies in the acute phase (<3 months after infection).
Comparators/control: We will include case series without controls, as well as studies with healthy controls, control participants with unexplained dyspnea, or that compare those who have fully recovered from COVID compared to those reporting ongoing symptoms.
Types of studies to be included: We will include observational studies including case series, cross-sectional studies, case-control studies, and cohort studies. We will also include randomized trials of interventions, in which case we will use baseline CPET data. We will exclude case reports and review articles.
Context: We will include studies that include the full spectrum of COVID-19; specifically, we will not restrict to only studies of those requiring ICU or hospitalization during acute infection.
Main Outcomes: The primary outcome will be peak VO 2 (in ml/kg/min and % predicted). If meta-analysis is possible, studies that do not include this measure will be excluded from meta-analysis. We will report the difference in peak VO 2 between those with and without COVID and among those with COVID between those with and without post-acute sequelae.
Additional outcomes: Additional outcomes will include the proportion with exercise limitation <80 or 85% of predicted (different studies use different cutoffs), difference in exercise capacity between those with and without cardiopulmonary symptoms (absolute and relative difference with 95% confidence intervals and p value), common features among those with limitations (i.e., reduced oxygen pulse pressure, reduced chronotropic response). We will likely report these effect measures in odds-ratios as we expect that many of the studies may be case-control studies.

Search Strategy & Information Sources:
A comprehensive, electronic search strategy will be used to identify studies published since 2020 and indexed in PubMed, EMBASE, and Web of Science by a research librarian (PT) with extensive experience in systematic reviews. Unpublished abstracts from conference proceedings and indexed preprints will be included as part of our gray literature search. We will also review references from studies selected for data extraction. The search strategy will include terms and synonyms for the following: (COVID or SARS-CoV-2) AND ("cardiopulmonary exercise test*" OR (CPET or CPX or CPEX) OR exercise capacity OR VO2 OR anaerobic threshold). Searches will be tailored to each database depending on indexing terminology. Searches were conducted on December 20, 2021, and rerun prior to the final analysis on May 24, 2022; pre-prints were searched through June 9, 2022. Abstracts were reviewed for inclusion by two independent reviewers (MSD & KS); if there is disagreement after consensus discussion, a third reviewer will be consulted. All data extraction was done independently, in duplicate, using REDCap for data entry.
Gray literature plan: see search strategy for details; we will review conference abstracts, pre-prints, and references from studies that meet the inclusion criteria.
Data Extraction (Selection & Coding): Data including authors, title, date of study, location of study, sample size (including total with COVID, total with Cardiopulmonary Long COVID, and COVID-negative controls, if included), median time since acute infection and interquartile range, inclusion criteria (with particular attention to inclusion of hospitalized/ICU/ambulatory during acute illness and those with specific comorbidities or populations of interest), comparator group, exercise modality (treadmill or cycle ergometer), peak VO2 (in ml/kg/min and % predicted), proportion with exercise limitation <85% of predicted, difference in exercise capacity between those with and without cardiopulmonary symptoms (absolute and relative difference with 95% confidence intervals and p value), common features among those with limitations (i.e., reduced oxygen pulse pressure, reduced chronotropic response). If available, other cardiopulmonary parameters will be recorded including echocardiographic, pulmonary function tests, chest computed tomography, and cardiac magnetic resonance imaging.
Data Management: Studies identified through the searches will be managed using Covidence. Data extracted will be recorded using REDCap.
Quality Assessment: We will use Cochrane's Quality in Prognostic Studies (QUIPS) tool to assess for bias of included studies. We will assess study populations (especially choice of control groups), study attrition for noncross-sectional studies, peak VO 2 assessment quality, outcome measurement, study confounding, and statistical analysis and reporting. We will use Cochrane's Quality in Prognostic Studies (QUIPS) tool to assess for bias of included studies.
Data synthesis: Overall findings of each study will be summarized in a table. If possible, a meta-analysis will be performed to compare the peak VO2 among those with and without COVID. An odds ratio of having reduced exercise capacity may also be estimated if possible. Heterogeneity will be assessed using I 2 . The primary subgroup we plan to investigate is to compare peak VO2 (and the other explanatory variables for reduced exercise capacity) among those with and without PASC/Long COVID. If possible, we may also compare those with severe acute infection requiring hospitalization and/or ICU care with those who were asymptomatic or had mild acute infection. Lastly, we may compare the early post-acute period (3-6 months), medium term (6-12 months), and long term (>12 months). Analyses will be performed using STATA version 17.

Analysis of subgroups:
The primary subgroup we plan to investigate is to compare peak VO2 (and the other explanatory variables for reduced exercise capacity) among those with and without PASC/Long COVID. If possible, we may also compare those with severe acute infection requiring hospitalization and/or ICU care with those who were asymptomatic or had mild acute infection. Lastly, we may compare the early post-acute period (3-6 months), medium term (6-12 months), and long term (>12 months).

Risk of Bias/Quality Assessment:
Risk of bias will be assessed at both the study and the outcome level for each included study. Publication bias will be assessed using a Funnel Plot. The strength of the body of evidence will be assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) framework. Low risk The study sample represents the population of interest Moderate risk on key characteristics, sufficient to limit potential High risk bias of the observed relationship between PF and outcome

Study Attrition
Goal: To judge the risk of attrition bias (likelihood that relationship between PF and outcome are different for completing and non-completing participants).

number included __________________________________
proportion retained Study Attrition Summary Loss to follow-up (from baseline sample to study population analyzed) is not associated with key characteristics (i.e., the study data adequately represent the sample) sufficient to limit potential bias to the observed relationship between PF and outcome.
Overall comments on Study Attrition __________________________________

Prognostic Factor Measurement
Goal: To judge the risk of measurement bias related to how PF was measured (differential measurement of PF related to the level of outcome).

Definition of the PF (CPET)
A clear definition or description of CPET is provided (e.g., including exercise modality & protocol, stopping criteria, assessment of submaximal tests (ie RER, Borg, HR, double product), and clear specification of the method of measurement and classification of limitations) Valid and Reliable Measurement of PF (CPET) No Method of PF measurement is adequately valid and reliable to limit misclassification bias (e.g., may include relevant outside sources of information on measurement properties, also characteristics, such as blind measurement and limited reliance on recall) Especially how tests are interpreted, how anaerobic Overall comments on CPET quality __________________________________ No Partial Yes Unsure Partial Yes Unsure

Outcome Measurement
Goal: To judge the risk of bias related to the measurement of outcome (differential measurement of outcome related to the baseline level of PF).
Overall comments on assessment of PASC/Long COVID/symptoms __________________________________

Study Confounding
Goal: To judge the risk of bias due to confounding (i.e. the effect of PF is distorted by another factor that is related to PF and outcome).