Development of a core outcome set for cardiac rehabilitation in patients with myocardial infarction: a study protocol


Myocardial infarction (MI) cause severe outcomes for patients, and the risks of all-cause death and cardiovascular events are 30% higher in patients after MI than in the general population at both 1–3 years and 3–5 years after an MI event.1 Although mortality due to MI has decreased due to the widespread implementation of percutaneous coronary interventions and other therapies, improving the health and well-being of discharged patients to allow them to partake in normal social life remains an insufficiently achieved goal.2 3

Cardiac rehabilitation (CR) is an evidence-based intervention that includes complete medical, psychosocial and clinical outcome assessments, exercise training, cardiovascular risk factors management, patient education and behavioural guidance.4–7 CR is important throughout the comprehensive intervention after an MI event and has life-long implications.8 CR can reduce mortality and improve cardiac function, exercise tolerance and quality of life for patients after an MI.9–13 Moreover, CR significantly contributes to the secondary prevention of MI,14 but its emphasis extends beyond mere physical recuperation to encompass a comprehensive approach aimed at enhancing patients’ quality of life and emotional well-being.

The heterogeneity in the reported outcomes of clinical trials focusing on CR for patients with MI pose a challenge to researchers conducting systematic reviews because synthesising the available data becomes difficult and many variables end up being excluded from analyses.15–17 Developing a core outcome set (COS) may address these issues.

A COS is a consensus-based minimum set of outcomes that should be measured and reported in all clinical trials focusing on specific study areas.18 In addition, using a COS may reduce potential selective outcome reporting bias in clinical trials.

After searching the Core Outcome Measures in Effectiveness Trials (COMET) database, we found only one COS for MI ( The scope of that COS is for clinical trials of traditional Chinese and Western medicine, patients are participants through semistructured interviews, but they do not contribute to the consensus process.19 20

The scope of our research differs from that of the registered COS study. Our COS will focus on CR and we will consider any type of intervention. In addition, stakeholders and patients will be invited to collaborate in the consensus process, we will consider patient-centred outcomes for inclusion in the final COS.

Our aim with this study is to provide a standardised set of outcomes that should be measured in all clinical trials of CR on patients with MI. We used the Core Outcome Set Standards for Protocol Items, as outlined in the COS-STAP Statement21 to report this protocol (online supplemental file 1).

Supplemental material

The scope of this COS includes the following items:

  1. Population: patients with MI, aged 18 years or older.

  2. Interventions: CR.

  3. Settings: the COS will be useful for conducting randomised controlled trials.


We will use mixed methods to develop the final COS. These methods have been recommended by the COMET Handbook and they have been used in previous COS investigations.16 17 20–23

The study is planned to commence in June 2023 and is anticipated to conclude in June 2025.


Steering committee

We will form a national steering committee to support the development of the COS. The steering committee will comprise six experts representing different disciplines, including two clinicians, two clinical trialists, a nurse and an evidence-based medicine methodologist. The steering committee will review and confirm the study protocol, identify the preliminary outcomes list and participate in the Delphi survey and consensus meeting to promote the COS development.


We will invite four groups of stakeholders to participate in the development of the COS including patients or their representatives, COS users (clinical trialists), health professionals (clinicians and nurse), and methodologists.


The study will be conducted in five stages (figure 1):

  1. Conducting a systematic review to form a preliminary list of candidate outcomes.

  2. Carrying out patients’ semistructured interviews to add potential new outcomes.

  3. Completing two rounds of Delphi surveys for COS users, health professionals, and methodologists.

  4. Completing patient questionnaire surveys.

  5. Holding consensus meetings for patients and other stakeholders to determine the final COS.

Figure 1
Figure 1

Flowchart of core outcome set development for cardiac rehabilitation after myocardial infarction.

Stage 1: systematic review

Search strategy

We will search six databases, including the China National Knowledge Infrastructure, Wanfang Data Knowledge Service Platform, China Biomedical Literature Service System, PubMed, Cochrane Library and Embase. The search will include trials published between January 2013 and April 2023. In addition, clinical trials registered in China Clinical Trial Registry and between 2013 and 2023 will also be searched. The online supplemental file 2 outlines the search strategy for English databases. Box 1 lists the inclusion and exclusion criteria for published articles.

Supplemental material

Box 1

Systematic review inclusion and exclusion criteria for published literature

Inclusion criteria.

  • Study population: patients with MI aged 18 years or older.

  • Interventions: CR.

  • Study types: randomised controlled trials.

  • Languages: Chinese and English.

  • Follow-up lengths of at least 8 weeks.

Exclusion criteria

  • The outcomes cannot be extracted.

  • Full text cannot be obtained, or repeated published study.

According to a foreign guideline specifying a consensus on maintaining records for at least eight weeks, studies require a follow-up period of at least eight weeks.7

Literature screening and data extraction

Two investigators (ZG and XZ) will read the titles and abstracts of the published articles and protocols for screening, and they will select the articles for inclusion after retrieving and assessing full texts or protocols. The same investigators (ZG and XZ) will independently evaluate and extract the data from the selected articles. Disagreements will be resolved through discussion or consultation with a third investigator (RQ).

The data extracted will include the first author’s and the applicant’s name, publication or registration year, title, country where the study was conducted, number of participants and their ages, interventions, follow-up lengths, outcomes, outcome definitions, tools used to measure outcomes, measurement timepoints (intervention/follow-up length) and others. After extraction, we will cross check the data and will solve inconsistencies through discussion or consultation with the third investigator (RQ). RQ will double-check 10% of the data extracted by ZG and XZ.

Merging outcomes and grouping under outcome domains

Two researchers (ZG and XZ) will merge the outcomes and independently group them under outcome domains to develop a preliminary long list of outcomes. They will use published methods20 24 encompassing the following steps:

  1. Extract the outcomes verbatim ‘word-for-word’ from the study methods or results.

  2. Extract data according to ‘what’, ‘when’ and ‘how’ from the verbatim outcomes.

  3. Rewrite outcomes that similar and correspond to equal definitions.

  4. Translate English outcomes into Chinese according to the terminology formulated by the National Science and Technology Terminology Committee. In the absence of a set relevant term, two researchers will agree on an appropriate term.

  5. Extract composite outcomes as individual ones.

  6. Drop outcomes without a definition or measurement instruments that are difficult to merge with another outcome.

  7. Group outcomes into different outcome domains according to the Williamson/Clarke outcome taxonomy.25

  8. Classify outcomes according to specific interventions and follow-up lengths.

Stage 2: patients’ semistructured interviews

We will conduct semistructured interviews for patients who have experience on MI to record their perspectives and identify potential new outcomes. The interviews will be conducted with two objectives in mind: one is to help the patients understand the meaning of medical terms, the other is to give the patients sufficient time to think deeply about the outcomes that are important to them.

Participant selection

Inclusion criteria

Adult patients with MI who have undergone CR within the 5 years prior to the interview.

Exclusion criteria

  1. Patients with acute MI status.

  2. Patients with a combination of other serious illnesses who are too ill to participate in the study, and patients with a severe mental health disorder that prevents them from communicating properly.


According to literature, 30 patients are sufficient to achieve saturation.26 Thus, we will recruit at least 30 patients, using a simple sampling strategy. We will recruit eligible patients from the Cardiology Department of Dongzhimen Hospital, Beijing University of Chinese Medicine who agrees to participate and sign informed consent.

Recruitment and data collection

Researchers (ZG and XZ, who have been trained in qualitative research methods) will approach potential participants at the inpatient or outpatient ward of Cardiology in Dongzhimen Hospital, Beijing University of Chinese Medicine. The researchers will explain the study to the patients making sure they understand the aim of the study and think about outcomes that are important to them. The researchers will give separate written information sheets to potential eligible patients. If the patients agree to participate in the interview, they will be asked to sign an informed consent form. Then, the semistructured interviews will be conducted. During the interviews, the researchers will collect socioeconomic, demographic and other information about the patients. The audio of the interviews will be recorded. Patients will be able to make a request to withdraw from the study at any time.

Topic guide

Table 1 shows the outline of the semistructured interviews. The questions in the table will be translated into Chinese. The researchers should spend approximately 20–30 min talking with each patient.

Table 1

Questions in the semistructured interviews for patients

Data analysis

We will collect and analyse the data concurrently, and the interview recordings will be transcribed verbatim and imported into qualitative analysis software. We will analyse the data on the basis of a framework methodology27 used in published COS studies.21 After analysing the data, we will identify the outcomes that are important to patients; then, the two investigators will determine whether these outcomes are new. The third investigator (RQ) will double-check the results.

After data extraction, two researchers (ZG and XZ) will combine and group new outcomes into the long list of outcomes. The complete list of potential outcomes will be submitted to the steering committee before developing the questionnaires.

Stage 3: Delphi survey

Stakeholder selection

Four stakeholder groups will be invited to participate in the Delphi survey: clinicians, clinical trialists, nurses and evidence-based medicine methodologists in cardiovascular disease areas.

We will contact potential stakeholders from the lists of members of the Clinical Research Method of Cardiovascular Disease of Professional Committee of Chinese Association of Integrative Medicine, the China Research Institute of China Information Association for Traditional Chinese Medicine and Pharmacy and the Clinical Research Platform for Chinese—Western Medicine. Clinicians and nurses will also be identified from the employees at the Dongzhimen Hospital, Beijing University of Chinese Medicine and the Zhongnan Hospital of Wuhan University. Finally, additional researchers will also be identified from the lists of authors of the included clinical trials or protocols in the systematic review.

Inclusion criteria for stakeholders

The inclusion criteria for stakeholders are the following:

  1. All the stakeholders will have a bachelor’s or higher educational level degree.

  2. All the stakeholders will have at least 1 year of work experience.

  3. The clinicians and nurses will have experience working in tertiary hospitals.

  4. Researchers will have participated in the design, administration, management or statistical analysis of cardiovascular-related clinical trials, or they will have conducted a systematic review of cardiovascular-related fields over the 10 years prior to the study start time.

  5. Stakeholders will be selected from any geographical region in China.


To the best of our knowledge, no standard sample size calculation method for Delphi surveys exists. We plan to use the snowball method to expand the survey’s sample size. Participants will be asked to forward the questionnaire they receive to their collaborators meeting the inclusion criteria. The sample size for each stakeholder group will be at least 20.

We will design questionnaires for professionals including all outcomes.

Round 1 of the Delphi survey

The questionnaire in the first round of the Delphi survey will include the introduction, outcomes and open-ended questions. The outcomes will include both clinician-reported and patient-reported outcomes.

In the questionnaire, we will use a 9-point Likert scoring system. A score of ‘1–3’ will indicate that ‘the outcome is not important for the COS’, a score of ‘4–6’ will indicate that ‘the outcome is important but not critical for the COS’ and a score of ‘7–9’ will indicate that ‘the outcome is critical for the COS’. The participants will be asked to score the importance of outcomes according to their own experience. In addition, there will be an ‘unclear’ option to select in cases when the participants cannot determine the importance of a given outcome.

At the end of the first round of the Delphi survey, there will be an open-ended question: what outcomes do you think are important but were not listed in this questionnaire?

Before we distribute the first round Delphi survey, we will send the questionnaire to the steering committee for evaluation. We will revise the questionnaire on the basis of the steering committee’s comments. After that, we will send the survey to different stakeholders by email or Wechat. The survey will be accessible to participants for 2–3 weeks.

The analysis of the first Delphi round will include the number of questionnaires completed, the number completed by each stakeholder group, the rating results for each outcome, and the rating results for each data domain by each stakeholder group. We will include all outcomes considered critically important by more than 50% of the stakeholders into the second round Delphi interview. If participants provide potential new outcomes, two researchers (ZG and XZ) will identify them as new. The third researcher (RQ) will decide whether to include the outcome.

Round 2 of the Delphi survey

We will develop a second Delphi round according to the results of the first round of Delphi survey. The new outcomes will be presented at the end of the questionnaire. Each stakeholders’ score distribution will be reported. Round 2 of the Delphi survey will be sent to all the stakeholders who complete the first Delphi round, and the stakeholders will be asked to rescore the outcomes. The stakeholders will have 2–3 weeks to complete the survey. If the response rate is less than 80% at the end of the third week, we will extend the Delphi survey deadline or will invite other eligible experts to participate in the survey.

The analysis for the second round of the Delphi survey will encompass assessing the response rate and identifying potential attrition bias. We will detect attrition bias by calculating the mean score of each outcome for participants of each stakeholder group whether they completed the two rounds of questionnaires or not. During the consensus meeting, we will list the outcomes and indicate whether they achieve consensus.

We will determine whether outcomes get to be included in the COS as follows:

  1. Consensus that an outcome should be included in the COS: more than 70% of participants scoring the outcome as 7–9 and less than 15% of participants scoring it as 1–3.

  2. Consensus that an outcome should be excluded from the COS: 50% or less of participants scoring the outcome as 7–9.

  3. No consensus: all other scoring scenarios.

Stage 4: patient survey

We will invite an MI patient with an undergraduate degree to participate in the development of the questionnaire. The outcomes in the questionnaire will be described in simple and understandable language. Three other patients will review the questionnaire. Patients who agree to participate in the survey will sign an informed consent form. A paper or Web-based questionnaire will be sent to patients, and they will be asked to score the importance of outcomes. We will modify the labels of the 1–9 Likert scale for patients, instead a traffic light system will be used to score each outcome as published in another COS study.28 We will label scores 1–3 as ‘not important’, scores 4–6 as ‘important’ and scores 7–9 as ‘very important’. The patients will have the chance to choose ‘unclear’ if they are not sure of the importance of a given outcome. The patients will be allowed to complete the questionnaire with the help of the researchers. At the end of the questionnaire, there will be an open-ended question: which other outcomes do you think are important but were not included in the questionnaire?

Data analysis

We will retain all outcomes voted by more than 70% of participants in the list for the consensus meeting. Outcomes voted by less than 50% of participants will be excluded from the list. We will retain outcomes with votes from 50% to 70% of participants into the consensus meeting list of non-consensus results. We expect at least 30 patients to participate in the survey.

We will show the results of the survey for patients during the second round of the Delphi survey.

Stage 5: consensus meeting

We will hold separate in-person consensus meetings for professionals and patients.

We will invite at least 15 eligible professionals (including those in the steering committee) from different stakeholder groups to attend the consensus meeting.

The consensus meeting inclusion criteria for professionals are the following:

  1. Stakeholders will have a master’s degree or higher degree.

  2. Stakeholders will have at least 5 years of related work experience.

  3. Clinicians and nurses will have tertiary hospital working experience.

  4. Clinical researchers will have participated in the design, conduct, management or statistical analysis of cardiovascular-related clinical trials, or conducted a systematic review on cardiovascular-related fields during the 10 years prior to the study start.

  5. Stakeholders will be selected from any geographical region.

  6. Stakeholders will be allowed to have participated in the Delphi survey.

We will report the results of the second round of the Delphi survey at the consensus meeting. The outcomes will be presented according to specific interventions and follow-up lengths, because previous research has shown that this approach may improve the adoption of COSs by Chinese trialists.29

We will invite at least 10 patients to participate in the meeting. The inclusion and exclusion criteria for patients partaking in the consensus meeting are the same as those for their participation in the semistructured interviews.

The participants of the consensus meeting will have the chance to discuss any outcome they want after being informed of the results of the patient survey. Then, they will be asked to vote anonymously for the outcomes that should be included in the COS. We will include the outcomes voted on by at least 70% of participants in the COS. For patient-reported outcomes, those voted for by both professionals and patients will be included in the final COS. The final COS will recommend core outcomes for specific interventions and different length of follow-up.

Patient and public involvement

Patients will be involved in the development of the questionnaire. In addition, patients or their representatives (such as their family members or caregivers) will be invited to participate in the semistructured interviews, questionnaire survey, and consensus meeting.


A variety of CR therapies for patients with MI have attracted increasing attention among healthcare researchers and clinicians. Developing a COS is necessary to improve outcome reporting.

Previous COS investigations have compared traditional Chinese medicine and Western medicine for treating patients with MI. For this study, we will focus on CR for patients with MI. We intentionally excluded other populations benefiting from CR, such as patients with heart failure or atrial fibrillation, to maintain a narrowly defined population and facilitate achieving a consensus. We chose this approach after researching the application of COSs in clinical trials in China30 and finding that a broader scope COSs may not get used consistently by Chinese trialists. A potentially more effective strategy to enhance applicability and promote widespread use might be to develop separate CR COSs across different diseases, and then to generate a Meta-COS.

We will conduct this COS study only in China. Therefore, the dissemination of the resulting COS in other countries (especially in high-income countries) may be limited.

During the study, we will invite different stakeholders to participate, including clinicians, clinical trialists, nurses, evidence-based medicine methodologists and patients. However, the patients will not participate in the Delphi survey because medical taxonomy is difficult for patients without a medical education background. Instead, the patients will be invited to take part in a survey designed to be easily understandable and written in plain language. In addition, to ensure patients’ opinions are incorporated into the final COS, we will invite them to participate in a consensus meeting.

For the study, we will focus on ‘what to measure’. The outcome measures, especially those for patient-reported outcomes,29 should be searched, assessed and agreed on by different stakeholders according to the guideline on ‘how to select outcome measurement instruments for outcomes included in a COS’.31 Thus, we plan to conduct a new study to recommend ‘how to measure’ core outcomes, after the final COS is developed.

Study registration and status

Our current data analysis pertains primarily to the preliminary systematic review. We have not yet carried out any crucial components of the study, including patient interviews, the Delphi survey, or the consensus meetings.

Ethics and dissemination

The study has obtained ethical approval by the Ethics Committee of Dongzhimen Hospital, Beijing University of Chinese Medicine (2022DZMEC-349). We will obtain informed consent from patients who participate in the semistructured interviews, questionnaire survey and consensus meeting. Informed consent for the other participants is unnecessary.

When the final COS is completed, we will report it using the COS-STAR. We will publish this research in a journal and report the results at national and international conferences. In addition, we will disseminate our findings on the website of China Information Association for Traditional Chinese Medicine and Pharmacy Clinical Research Information Association ( We will also send the publication to researchers who participated in the Delphi surveys and the consensus meeting for their use in their research.

This post was originally published on