To create self-assessment tools for return to play in post-COVID-19 athletes, extensive literature regarding post-COVID-19 complications, safe return to physical activity, and exercise guidelines were reviewed to create the questionnaires. The literature was searched using PubMed and Scopus search engines. The terms used were ‘Covid-19’ OR ‘COVID’ OR ‘coronavirus’ OR ‘SARS-Cov-2’ AND ‘return to play’ OR ‘RTP’ OR ‘return to sport’ OR ‘RTS’ OR ‘athletes’ OR ‘Recommendation’ OR ‘Guidelines’ OR ‘consideration’ AND ‘post-infection’ OR ‘after infection.’ The search was conducted between September 1, 2022 and December 1, 2022.

Based on an extensive search, we divided the questionnaire domains into three categories: (1) to identify high-risk factors for developing severe post-COVID-19 complications, (2) to identify symptoms of severe post-COVID complications, and (3) to identify severe physical limitations. Each group comprised two sports physicians, one family medicine specialist with extensive experience in managing COVID-19 in the community, and one public health physician. Proposed questionnaires are provided in the supplementary files (Supplements 1, 2).

Our target sample included athletes who contracted COVID-19 and were followed-up for post-COVID-19 at the National Sports Institute between January 1, 2022 and May 10, 2023. An initial estimated sample size was 20:1, with a target sample size of 220 based on 11 items. A total of 312 athletes contracted COVID-19 and were followed-up at the National Sports Institute between January 1, 2022 and May 10, 2023. We used recall methods and asked the athletes to recall their post-COVID-19 symptoms from the latest COVID-19 infection between January 1, 2022 and May 10, 2023. This method was used because we predicted a low prevalence of COVID-19 in 2023. Therefore, a prospective study was not feasible. We chose January 1, 2022 because we would like to minimize the recall bias to one and a half years, and May 10, 2023 because that was the last date before our ethical approval.

Of the 312 athletes, 33 had two episodes of COVID-19 infections during the study period. Therefore, we selected 279 athletes using WhatsApp. However, only 161 athletes (57.7% of 279) responded and agreed to participate in the study. Eighteen percent of the athletes (29 of 161) had difficulty understanding English language, and 12 were aged <16 years old; thus, they were excluded from the study. Therefore, only 120 athletes participated in the first round of answering the questionnaires (criterion validity) using Google Forms (Google LLC, Mountain View, CA, USA), which were then blasted using WhatsApp (WhatsApp Inc., Mountain View, CA, USA)—we are using the WhatsApp message to contact all athletes. Subsequently, the first 50 athletes were selected to answer a second round of questionnaire for reliability testing. Nonetheless, only 32 athletes responded, despite three reminders given through WhatsApp. Phone call was not made to respect the athletes’ privacy and their decision not to participate in the second questionnaire.

Files of 120 athletes were traced from the medical records of the National Sports Institute, and pre-participation health screening forms were extracted along with ECG, chest radiograph, and return-to-play duration. The following health screening forms were included: (1) demographic data; age, gender, and sports participation; (2) anthropometric data: height, weight, body mass index; (3) vital signs: blood pressure, resting heart rate, saturation of peripheral oxygen; (4) ECG report; and (5) chest X-ray report.

A pre-participation evaluation form, along with the ECG and chest X-ray reports, were available in electronic medical records and re-examined by a single examiner. The following categories were used to compare the findings with those of the Physical Activity Readiness Questionnaires for post-COVID-19 athletes (PAR-COVID-Q): all criteria should be fulfilled to classify patients as having no severe post-COVID-19 complications and safely return to sports category. (1) Athletes with absence of ongoing symptoms of cardiovascular disease, extreme fatigue or reduced performance, and no history of admission or referral post-COVID-19 were considered not having post-COVID-19 complications. (2) Athletes with normal ECG or ECG characteristics of athlete’s heart based on international criteria for ECG interpretations in athletes were considered not having post-COVID-19 complications. None of the athletes who completed the questionnaire had abnormal ECG findings. (3) Chest radiography without pulmonary fibrosis, collapse, or consolidation indicated no post-COVID-19 complications.

For content validation, internal reviewers reviewed the initial questionnaires consisting of four experts: (1) associate professor (Dr. Nur Amani) from primary care who was highly involved in the management of post-COVID in primary care settings; (2) associate professor (Dr. Nur Farizah) from the public health department and experts in COVID; (3) associate professor (Dr. Nahar) from sports medicine and researchers; and (4) Dr. Munawwar who was a sports physician working closely in the post-COVID clinic at the National Sports Institute, Malaysia. The initial questionnaires were revised, and preliminary questionnaires were developed, consisting of 11 items. Several suggestions and advice are provided regarding the wordings and terminologies used. A second version of the amended questionnaire was sent for content validation to external reviewers consisting of one consultant sports cardiologist, one consultant infectious disease physician, one exercise physiologist expert, one respiratory physician, one sports physician who was highly involved in post-COVID management in athletes, and one athlete who had a history of COVID-19 infection, using the Likert scale, as shown in Supplement 2. Eleven items remained, with changes in item 2(e) being replaced with chest tightness to reduce performance, and re-validated by the expert review. The final expert reviewer rated the amended questionnaire as adequately relevant, clear, simple, and without any doubt. The final set of questionnaires is shown in Figure 1. Overall validity indices are presented in Table 1.

For face validation, 11 athletes (1:10) had a direct interview session for 5–10 minutes regarding the PAR-COVID-Q. These athletes were national athletes who had been followed-up by the National Sports Institute for any reason and had contracted COVID-19. All the athletes agreed that the questionnaires were simple and easy to understand, less burdensome for follow-up, and easy to be self-administered. However, two athletes mentioned that some athletes in Malaysia might have problems with the Questionnaires in English language.

Absence of severe post-COVID-19 complications was characterized by a normal pre-participation evaluation form, normal ECG for athletes, and normal chest X-ray. The Spearman’s correlation coefficient (ρ) was calculated for the correlation between the total score of the questionnaire and presence of post-COVID complication, and ρ=1 suggested a significant correlation between both scales.

The first 50 respondents were asked to answer the questionnaires 7 days after their first response. However, only 32 responded, despite being given a reminder 3 times.

Data are presented as frequencies (number), proportions (%), means, and standard deviations. Validity index (VI) was calculated after analyzing content validity. The VI was calculated as the number of experts providing a score >4 on the Likert scale (scores, 1–5) divided by the total number of experts [18,19].

To test the criterion validity, the Spearman’s Pearson correlation was used between the total score of all items and post-COVID-19 complications. We expect that a score of 11 (11 yeses) correlates with the absence of post-COVID complications, and a score <11 correlates with the need for further medical evaluation and/or the presence of post-COVID complications. Questionnaires were considered valid if the P-value was <0.05.

To test for reliability, internal consistency was estimated using the alpha coefficient. Cronbach’s α of at least 0.70 was suggested to indicate adequate internal consistency. Cut-off points for intraclass correlation coefficient (ICC) analysis were based on the guidelines which state that an ICC below 0.40 is a poor reliability; between 0.40–0.59 is fair; between 0.60–0.74 is good; and between 0.75–1.00 is excellent. For test-retest reliability, the Spearman’s correlation was used, and a P-value of <0.05 was considered significant. This study was approved by the National Sports Institute of Malaysia (RE/A/001/2023-002/2023). Informed consent was obtained from all individual participants included in the study.

The questionnaire was used to screen for post-COVID-19 complications. In this questionnaire, there are three main domains covered to identify athletes at risk of severe complications and who should undergo medical screening by trained professionals: (1) a domain that identifies athletes with high-risk factors of developing severe complications covered by items one and five; (2) a domain that identifies symptoms at risk of developing severe COVID-19 complications covered by item 2 and its subdivision; and (3) a domain that identifies limitations of normal activity by items 3 and 4. These items were extensively reviewed by expert physicians with renowned experience in managing COVID-19 in Malaysia.

When testing for reliability, there was excellent internal consistency and perfect test-retest reliability in the 32 athletes. This demonstrated that our self-screening questionnaires were reproducible and highly consistent. When we examined the validity of the criterion, athletes with a score of 11 (11 yeses) successfully returned to sports, with normal health screening, ECG characteristics, and chest radiographs. Two of the athletes underwent further evaluation for post-COVID-19 complications. One of the two athletes who had chest pain post-COVID had normal cardiac enzyme levels, a normal echocardiogram finding with an ejection fraction of 67%, normal hemodynamic response, and ECG changes during the stress test exercise up to stage 5, and was cleared for return to play. These results demonstrate that the PAR-COVID-19 is a valid questionnaire that can act as a self-screening tool for athletes prior to their return to play after mild COVID-19.

This is the first valid and reliable questionnaire as a self-screening tool for athletes returning to play after mild COVID-19. The application of these screening tools should help ease the healthcare burden and reduce costs. The PAR-COVID-Q is an excellent self-assessment tool. These questionnaires took <5 minutes to complete, were straightforward, and easy to understand. With efficient self-assessment tools, it creates less of a burden for athletes to visit hospitals.

We believe that the questionnaire has excellent reliability, reflecting the uncomplicated nature of the questions. The PAR-COVID-Q can be administered with minimal investigator assistance, thereby avoiding potential observer bias.

However, we face many challenges and limitations when developing the PAR-COVID-Q. The prevalence of COVID-19 fluctuates worldwide. At the National Sports Institute in Malaysia, we observed a very low prevalence of COVID-19 among athletes, probably due to a lack of testing done these days. Thus, it is challenging to perform a prospective study because of the small sample size. Thus, this was a retrospective study. The main problem with retrospective studies is recall bias. To minimize bias, we limited the duration of our study to one and a half year starting from January 1, 2022. Nonetheless, when we compared athletes’ responses to their respective health screenings, the answers provided corresponded to the symptoms noted in the health screening. Even though the current prevalence of COVID-19 is low, we should expect an increase in the prevalence of COVID due to the swift evolution and mutation of the virus, which could lead to another pandemic. Other challenges included the struggle to obtain the target sample size. We targeted a 20:1 ratio as the optimal sample size for our study. However, we received a moderate response from the athletes (57.7%). This could be due to their busy schedules as national athletes, and many of the athletes trained outside the sports institute, making it difficult to reach them. As the questionnaires were delivered in English language, we observed that 18% of them had difficulty understanding English language, and most of them were from Malay ethnicity. Thus, we anticipate that there will be problems in delivering these self-assessment tools to non-English-fluent populations, and translation studies can be conducted to overcome such challenges. This study was not validated in children aged <16 years of age.

Nevertheless, the PAR-COVID Q is a valid, reliable, user-friendly, and low athlete burden screening tool for athletes aged >16 years to return to sports following mild COVID-19 infection. Athletes who answer all yes (11 yeses) can gradually return to sports safely. In contrast, athletes who answered <11 yeses should undergo a thorough medical assessment to ensure that there are no severe complications of COVID-19.