Maximising the impact of qualitative research in feasibility studies for randomised controlled trials: guidance for researchers
- First Online:
- Cite this article as:
- O’Cathain, A., Hoddinott, P., Lewin, S. et al. Pilot Feasibility Stud (2015) 1: 32. doi:10.1186/s40814-015-0026-y
- 8.9k Downloads
Feasibility studies are increasingly undertaken in preparation for randomised controlled trials in order to explore uncertainties and enable trialists to optimise the intervention or the conduct of the trial. Qualitative research can be used to examine and address key uncertainties prior to a full trial. We present guidance that researchers, research funders and reviewers may wish to consider when assessing or undertaking qualitative research within feasibility studies for randomised controlled trials. The guidance consists of 16 items within five domains: research questions, data collection, analysis, teamwork and reporting. Appropriate and well conducted qualitative research can make an important contribution to feasibility studies for randomised controlled trials. This guidance may help researchers to consider the full range of contributions that qualitative research can make in relation to their particular trial. The guidance may also help researchers and others to reflect on the utility of such qualitative research in practice, so that trial teams can decide when and how best to use these approaches in future studies.
KeywordsRandomised controlled trial Feasibility studies Pilot studies Qualitative methods Guidance
Medical Research Council
National Institute for Health Research
The United Kingdom Medical Research Council (UK MRC) guidance on the development and evaluation of complex interventions recommends an early phase of assessing feasibility prior to a full evaluation . In this feasibility and pilot phase, researchers can identify and address problems which might undermine the acceptability and delivery of the intervention or the conduct of the evaluation. When the outcome evaluation is a randomised controlled trial, this feasibility phase increases the chances of researchers evaluating the optimum intervention using the most appropriate and efficient recruitment practices and trial design. Alternatively, at the feasibility phase, researchers may identify fundamental problems with the intervention or trial conduct and return to the development phase rather than proceed to a full trial. The feasibility phase thus has the potential to ensure that money is not wasted on an expensive trial which produces a null result due to problems with recruitment, retention or delivery of the intervention .
Feasibility studies for randomised controlled trials can draw on a range of methods. Some feasibility studies use quantitative methods only. For example, researchers concerned about whether they could recruit to a trial, and whether the intervention was acceptable to health professionals and patients, undertook a pilot trial with outcomes related to recruitment and surveys to measure the acceptability of the intervention . Increasingly, qualitative or mixed methods are being used within feasibility studies for randomised controlled trials. A review of 296 journal articles reporting the use of qualitative research with trials published between 2008 and 2010 identified that 28 % of articles reported qualitative research undertaken prior to the full trial . Qualitative research was not only undertaken with trials of complex interventions but was also used with trials of drugs and devices where researchers recognised the complexity of the patient group receiving the intervention or the complexity of the environment in which the trial was to be undertaken . Yet, there is little guidance available on how to use qualitative methods within feasibility studies for trials. Here, we offer guidance in order to help researchers maximise the opportunities of this endeavour.
Getting the language right: feasibility studies, pilot studies and pilot trials
Before offering guidance on using qualitative methods at the feasibility phase of a trial, we first need to be clear about the meaning of the term ‘feasibility study’ because the language used to describe the preparatory phase for a trial is inconsistent . These types of studies can be called feasibility or pilot studies, with researchers making no clear distinction between the two when reporting their studies in journal articles . The MRC guidance for developing and evaluating complex interventions describes this as the ‘feasibility and piloting’ stage. The UK funding body, the National Institute for Health Research (NIHR), offers definitions of feasibility and pilot studies, distinguishing between the two . A feasibility study is undertaken to address the question ‘can the planned evaluation be done?’. In contrast, pilot studies are miniature versions of the main study. In the case of a randomised controlled trial, the pilot study is a pilot trial. A feasibility study for a randomised controlled trial does not necessarily involve a pilot randomised controlled trial  but may do so, and indeed, some researchers have described their studies as a ‘feasibility study and pilot trial’ in the titles of their journal articles . Other terms may be used to describe a feasibility study for a trial, for example a ‘formative’ study as part of ‘evidence-based planning’  or an exploratory pilot study  or a process evaluation with a pilot trial . In this guidance, we use the term ‘feasibility study’ to describe any study that addresses the question of whether the planned evaluation trial can be done regardless of the labels other researchers might use.
The need for guidance on using qualitative methods in feasibility studies for randomised controlled trials
With the use of qualitative research in feasibility studies for randomised controlled trials becoming increasingly common, guidance on how to do this would be useful to both researchers and those commissioning and reviewing this research. Guidance is available or emerging in areas related to feasibility studies for trials. Guidance exists for undertaking quantitative pilot studies [13, 14], and a Consolidated Standards of Reporting Trials (CONSORT) statement for reporting feasibility studies (rather than undertaking them) is under development . UK MRC guidance has recently been developed for process evaluations undertaken alongside randomised controlled trials . This new guidance recommends that, in most cases, it is useful to use both qualitative and quantitative methods concurrently with a pilot or full trial. It also states that as feasibility studies will usually aim to refine understanding of how the intervention works, and facilitate ongoing adaptation of intervention and evaluation design in preparation for a full trial, qualitative data will likely be of particular value at this stage. However, that guidance does not address in any depth issues specific to the use of qualitative research during the feasibility phase of a trial. There is also guidance for writing proposals for using qualitative research with trials  and reporting qualitative research undertaken with trials . However, the feasibility phase of a trial is unique in that it may involve the ongoing adaptation of plans for conducting the trial and of the intervention in preparation for the full trial. Therefore, our guidance complements recent and upcoming guidance by focusing on the role of qualitative research specifically rather than the overall feasibility study and by addressing the iterative nature of research that may occur in feasibility studies for trials.
The focus of the guidance
A feasibility study may or may not include a pilot randomised controlled trial.
The feasibility phase follows the development phase of an intervention, in which qualitative methods may also be used . Although there may be overlap between the development of the intervention and the feasibility phase of the trial, this guidance assumes that an intervention has been developed, but that it might need further modification, including assessment of its practicability in the health care setting.
Qualitative methods can be used alone or in conjunction with quantitative methods, such as modelling and surveys, in the feasibility phase .
The definition of qualitative research is the explicit use of both qualitative data collection and analysis methods. This is distinguished from trialists’ reflective reports on the problems that they encountered in running a feasibility study and from the use of methods that may draw on qualitative approaches but do not meet our definition. For example, some researchers report using ‘observation’ and ‘field notes’ but show no evidence of qualitative data collection or analysis in their article and do not label these as qualitative research . Reflective practice by trialists and intervention deliverers is important for learning about trial conduct but is not the focus of the guidance presented here.
The guidance focuses on maximising the opportunities of qualitative research by presenting options, rather than delineating required actions. This is based on the understanding that the strengths of qualitative research are its flexibility and responsiveness to emerging issues from the field.
The guidance may be used by researchers when writing proposals and undertaking or reporting qualitative research within feasibility studies. If the feasibility study includes a pilot randomised controlled trial, reporting should follow the CONSORT statement that is currently under development .
Processes used to develop the guidance
This guidance is based on the experience of the authors of this paper. The authors came together in a workshop to write this guidance after meeting to discuss a study of how to maximise the value of qualitative research with randomised controlled trials which had been undertaken by two of the authors of this guidance (AOC, KJT) [4, 5]. That study involved undertaking a systematic mapping review of journal articles reporting qualitative research undertaken with randomised controlled trials and interviews with qualitative researchers and trialists; some of these articles are referenced to illustrate points made. Towards the end of this study, the UK MRC Hubs for Trials Methodology Research funded a conference to disseminate the findings from this study and a related 1-day workshop to develop guidance for using qualitative research with trials. The nine workshop members, all of whom are authors of this guidance, were identified for their experience in using qualitative research with trials. One member had also published a review of the use of qualitative research alongside trials of complex interventions .
The workshop focused on feasibility studies because these were identified as an underdeveloped aspect of trial methodology. The workshop members put forward items for the guidance, based on their experience and expert knowledge. Discussion took place about the importance of items and the different viewpoints within each item. Draft guidance was produced by AOC after the workshop. Subsequent development of the guidance was undertaken by email correspondence and meetings between sub-groups of the workshop membership. A draft of the guidance was then presented at a meeting of an MRC Methodology Hub for researchers with experience in undertaking qualitative research in feasibility studies for trials. Attendees viewed the guidance as helpful, and further insights emerged from this process, particularly around the analysis domain of the guidance.
Guidance for using qualitative research in feasibility studies for trials
Aspects of the feasibility study
Issues to consider
1. Research questions
a. When designing the feasibility study, consider the wide range of questions. Then, consider those best addressed by qualitative research.
b. Prioritise the initial questions by identifying key uncertainties, whilst allowing for the possibility of emergent questions.
c. Consider the often overlooked questions, such as ‘what is considered to be usual care?’.
2. Design and data collection
a. Consider the range of qualitative methods that might be used to address the key feasibility questions, including dynamic or iterative approaches which allow learning from early qualitative research findings to be implemented before further qualitative research is undertaken as part of the feasibility study.
b. Select from a range of appropriate qualitative methods to address the feasibility questions and provide a rationale for the choices made; non-participant observation may be an important consideration.
c. Pay attention to diversity when sampling participants, groups, sites and stage of intervention.
d. Appreciate the difference between qualitative research and public and patient involvement.
a. Consider timing of analysis which might be in stages in a dynamic approach.
b. Many different approaches to analysis can be used, including framework, thematic and grounded theory-informed analysis.
c. Data can cover a breadth of issues, but the analysis may focus on a few key issues.
a. Have a qualitative researcher as part of the feasibility study design team.
b. Consider relationships between the qualitative researchers and the wider feasibility study team.
c. Consider who will make changes to the intervention or trial conduct.
a. Publish feasibility studies where possible because they help other researchers consider the feasibility of similar interventions or trials.
b. Describe the qualitative analysis and findings in detail.
c. Be explicit about the learning for a future trial or a similar body of interventions or generic learning for trials.
- 1.Research questions
When designing the feasibility study, consider the full range of questions that could be addressed. Then, consider those best addressed by qualitative research.
Questions that qualitative researcha can address in a feasibility study for a randomised controlled trial
Category of question
Examples of possible questions
Intervention content and delivery
To what extent does the planned intervention need to be refined or adapted to make it more acceptable to users or more relevant or useful to the specific context in which it is delivered?
Consider the different aspects of the intervention and which are fixed and flexible. The intervention may be different in practice from the planned intervention and may need to be documented so it can be delivered consistently in the full trial.
Mechanisms of action
How might the intervention be working? How might it produce the outcomes important to the trial? Data collected to address these questions may be interpreted in relation to the theory upon which the intervention is based or may help to develop new theory.
Perceived value, benefits, harms or unintended consequences of the intervention
What value do service providers and intervention users place on the intervention and the outcomes it plans to deliver? What benefits and harms do they feel they have experienced from the intervention so that these can be measured in the full trial?
Acceptability of intervention in principle
Are service users or health care providers unhappy with any aspect of the content or delivery of the intervention?
Feasibility and acceptability of intervention in practice
What are service users or health care providers’ views of the implementation of the intervention? Has implementation varied by setting? Are there any important intervention-context interactions? Should implementation be tailored by setting?
Fidelity, reach and dose of intervention
Is the right amount of the intervention getting to the right recipients in the right way? Do those delivering the intervention and/or receiving it adhere to the planned intervention? If not, what are the reasons for this? What are the limits of acceptable tailoring of the intervention?
Trial design, conduct and processes
Recruitment and retention
How do the planned recruitment practices work in the field? Do recruitment practices need to be improved to increase recruitment rates and levels of informed consent? If so, how? Are the trial participants willing to be randomised? Are clinicians willing to recruit patients, or are they uncomfortable? Are there ways in which trial procedures could be improved to increase retention rates?
Diversity of participants
Are the planned recruitment practices likely to result in recruitment of the desired range of participants for the trial? If not, how might recruitment practices be improved?
How is the planned trial communication implemented by recruiters and received by participants? How can trial communication be improved to ensure recruiters understand patients’ views about participating in the trial?
Acceptability of the trial in principle
Is the trial design acceptable to patients, recruiters and service providers in principle?
Acceptability of the trial in practice
Is the trial design acceptable to patients, recruiters and service providers in practice, or are there ways in which participants try to alter the procedures?
Are the informed consent procedures appropriate and acceptable to likely trial participants?
Adaptation of trial conduct to local context
Will the planned trial procedures allow the trial to operate effectively in the proposed context? Do any changes need to be made to these procedures?
Impact of trial on staff, researchers, participants and the health system
Does this trial have any unanticipated negative impacts on recruiters, participants, other stakeholders and the health system? How can these impacts be minimised (e.g. workload involved in recruitment, numbers of measures undertaken)?
Patient and public involvement
How is patient and public involvement best achieved in the trial?
Breadth and selection of outcomes
Are outcomes important to service users selected for measurement in the full trial—both primary and secondary? Do some trial participants feel that they have experienced or noticed improvements in some outcomes that need to be included in the full trial?
Accuracy of measures
Are the process and outcome measures valid for this participant group?
Completion of measures
Can completion rates of measures be improved?
Development of measures
If validated measures do not exist for all the outcomes to be measured in the full trial, can they be developed in preparation for the trial?
Prioritise the questions for the qualitative research by identifying key uncertainties.
Consider often overlooked questions.
How do the intervention components and delivery processes work in the real world?
How does the choice of comparator affect the trial?
The focus of qualitative research undertaken with trials tends to be on the intervention, but qualitative research can also help to understand the control. Interventions can be compared with active controls or usual care, and there may be issues to explore regarding the comparability of an active control and the intervention or the extent to which the trial may change usual care . Such research may help the trial team to consider whether there is sufficient difference between the groups being compared in any trial. For instance, the planned intervention may not be that different from usual care in some settings and may need to be enhanced prior to use in the full trial. Differences between the intervention and usual care will have implications for the relative effectiveness of the intervention and the transferability of the trial findings to other contexts.
To what degree does equipoise exist?
- 2.Design and data collection
Consider the range of qualitative methods that might be used to address the key feasibility questions, including dynamic or iterative approaches which allow learning from early qualitative research findings to be implemented before further qualitative research is undertaken as part of the feasibility study.
Case study 1: Donovan and colleagues  undertook qualitative research within a feasibility study for a trial of prostate testing for cancer and treatment.
Research question: The authors are explicit in the introduction of the paper that the most important uncertainty for the full trial was whether participants would agree to randomisation. Therefore exploring this issue, and ways of improving recruitment, was key to decision making about the feasibility of a full trial.
Design and data collection: The qualitative research was a combination of indepth interviews with patients who had undergone the recruitment process and audiotape recordings of recruitment appointments with follow up interviews with recruiters. The data collection and analysis was ‘dynamic’ in that initial qualitative findings were acted on during the feasibility study and further qualitative research undertaken to check if improvements had occurred. The qualitative research showed that recruiters had difficulty discussing equipoise and presenting treatments equally. These findings were summarised and fed back to recruiters in training sessions. Changes were also made to the content and presentation of information in response to findings that patients misinterpreted the language used in the original trial information. Recruitment rates for the pilot trial were monitored over time, showing that they increased as these changes were made.
Analysis: Methods of constant comparison were used and references are given.
Reporting: The qualitative findings are reported in detail including quotes. The effect of the qualitative research on the full trial is clear in the abstract and the body of the paper. The recruitment rate increased during the pilot trial and three armed trial was identified as feasible.
Other approaches suitable for feasibility studies include iterative ‘rapid ethnographic assessment’ which has been used to adapt and tailor interventions to the different contexts in which the trial was planned . This approach applies a range of methods including participant observation, focus groups, interviews and social mapping . Other researchers have used ‘mixed methods formative research’ at the feasibility stage  and action research where potential participants and practitioners are actively involved in the research to assess the feasibility of an intervention and to ensure a good intervention-context fit [35, 36]. For instance, a participatory approach informed by the principles of action research was used to design, implement and evaluate the FEeding Support Team (FEST) intervention [35, 36].
Select from a range of appropriate qualitative methods to address the feasibility questions and provide a rationale for the choices made; non-participant observation may be an important consideration.
Researchers need to select from a range of qualitative methods including telephone and face-to-face interviews, focus groups, non-participant observation, paper/audio/video diaries, case notes kept by health professionals and discussions in online chat rooms and social media. Decisions on data collection and analysis methods should depend on the research questions posed and the context in which data will be collected. To date, feasibility studies for trials have often tended to rely solely upon interviews or focus group discussions with participants and intervention deliverers and have not drawn on the wider range of methods available [21, 22, 23, 24]. Researchers tend also to use focus groups and may do this because they think they are cheap and quick when in practice, they are challenging to both organise and analyse. Some researchers are explicit about why focus groups are the best approach for their study. For example, in a randomised trial on the use of diaphragms to prevent sexually transmitted infection, the research team conducted 12 focus groups with women before and after they received the intervention to consider its acceptability and feasibility. This data collection approach was justified on the basis that the researchers felt focus groups would generate more open discussion . However, focus groups may be problematic in a feasibility study because they tend towards consensus and can mask dissenting views, with the possibility of premature conceptual closure. It may also be the case that participants who are prepared to talk openly within a group setting may differ from the target population for a trial as, in general, focus groups tend to attract more educated and confident individuals .
Pay attention to diversity when sampling participants, groups, sites and stage of intervention.
All of the different approaches to sampling in qualitative research—such as purposive, key informant and snowballing—are relevant to feasibility studies. A particular challenge for sampling within the feasibility phase is the need to address the wide range of uncertainties about the full trial or the intervention within the resource limitations of the study.
It can be difficult to decide when enough has been learnt about the trial intervention or the conduct of the trial (or when data saturation has occurred) to recommend moving on to the full trial. Researchers will need to make pragmatic decisions on which emerging analysis themes warrant more data collection and where sufficient data are available. In practice, sample sizes for qualitative research in feasibility studies are usually small (typically between 5 and 20 individuals [10, 12, 22, 23, 24]). This may be reasonable, given that simulations suggest that 10 users will identify a minimum of 80 % of the problems with the technology during usability testing, and 20 users will identify 95 % of the problems . However, sample size will be dependent on the study; for example, there may be therapist effects to consider and a need to sample a range of patients using different therapists or a range of contexts.
Appreciate the difference between qualitative research and public and patient involvement.
Case study 2: Hind and colleagues  use qualitative research to explore the acceptability of computerised cognitive behavioural therapy for the treatment of depression in people with multiple sclerosis. This is undertaken in the context of a wider study which included a pilot randomised controlled trial.
Research question: In the introduction of the paper the authors reference previous research which identifies the importance of exploring whether an intervention engages specific target groups, and the importance of understanding the acceptability of computerised cognitive behaviour therapy.
Design and data collection: A patient representative was a member of the research team and was involved in the design and conduct of the study. Data collection for the qualitative study consisted of face-to-face semi-structured interviews with 17 patients who had used one of two computerised cognitive behaviour therapy packages. There was also brief weekly written feedback from patients and brief telephone interviews at the start of the intervention to identify immediate problems.
Analysis: Framework analysis was used and is referenced. A patient representative participated in the analysis of the data.
Reporting: Although not mentioned in the title of the paper, or the abstract, the authors are clear in the introduction that this paper that the qualitative research was undertaken in the same study as a pilot trial. The qualitative findings are described in detail using quotes from participants. The conclusions relate to the intervention - that computerised cognitive behaviour therapy packages would need to be adapted for people with chronic physical disease - but are not explicit about the implications for a full randomised controlled trial.
Consider the timing of analysis, which might be in stages in a dynamic approach.
Many different approaches to analysis can be used, including framework, thematic and grounded theory-informed analysis.
Data can cover a breadth of issues, but the analysis may focus on a few key issues.
Have a qualitative researcher as part of feasibility study design team.
Consider relationships between the qualitative researchers and the wider feasibility study team.
How the qualitative researchers interact with the wider feasibility study team is an important concern. If study participants view the qualitative researchers as closely aligned with the team delivering the intervention or conducting the pilot trial, then participants may feel less able to offer honest criticisms of the intervention or trial conduct. On the other hand, where qualitative researchers work too independently from the wider team, they may not develop a deep understanding of the needs of the trial and the implications of their findings for the trial.
Consider who will make changes to the intervention or trial conduct.
Publish feasibility studies where possible because they help other researchers to consider the feasibility of similar interventions or trials.
Other researchers can learn from feasibility studies, and where this is likely to be the case, we recommend publishing them in peer-reviewed journal articles. Other researchers might be willing to take forward to full trial an intervention that the original researchers were unable or unwilling to take beyond the feasibility study. Or, other researchers might learn how to develop better interventions or trials within the same field or understand which qualitative methods are most fruitful in different contexts. Publishing what went wrong within a feasibility study can be as helpful as publishing what went right. Explicit description of how decisions were made about which research questions and uncertainties were prioritised may help others to understand how to make these types of decisions in their future feasibility studies.
Describe the qualitative analysis and findings in detail.
Be explicit about the learning for a future trial or a similar body of interventions or generic learning for trials.
Case study 3: Farquhar and colleagues  undertook a qualitative study to explore the feasibility, acceptability and appropriateness of a widely used quality of life instrument. This was undertaken within a pilot trial in preparation for a Phase III trial of a complex intervention for intractable breathlessness in patients with advanced chronic obstructive airways disease.
Research question: In the introduction of the paper the authors explain that it was important to explore the feasibility and acceptability of the instrument because it had not been used with this patient group.
Design and data collection: The instrument is administered in the context of an interview. Data collection for the qualitative study consisted of the audio-recordings of these interviews. 13 patients in the intervention and control arms of the pilot trial completed the instrument on 3-5 occasions each.
Analysis: Framework analysis was used and is referenced.
Reporting: Although not mentioned in the title of the paper, the authors are clear in the abstract that this qualitative research was undertaken in the context of a Phase II trial in preparation for a Phase III trial. The key words include ‘feasibility studies’. Within the methods section of the paper they state that the pilot trial was published elsewhere and give references so that readers can connect the different components of this study if required. The qualitative findings are described in detail using quotes from participants; they identify the difficulties participants had completing the instrument at different stages of the pilot trial. The authors state clearly in the conclusion section of the paper the implications of this work for both the full trial and for the future development of the instrument. The instrument was rejected for use in the full trial because of the difficulties identified.
Once a feasibility study is complete, researchers must make the difficult decision of whether to progress to the full trial or publish why a full trial cannot be undertaken. There is guidance on how to make this decision, which encourages the systematic identification and appraisal of problems and potential solutions and improves the transparency of decision-making processes . Too often, progression criteria are framed almost entirely in quantitative terms and it is unclear the extent to which qualitative data may or not play a direct role in informing the decision on whether to proceed to a full trial. For example, if researchers fall just short of a quantitative criterion, but have a sufficient qualitative understanding of why this happened and how to improve it, then it might be possible to proceed. Related to this, qualitative research may identify potential harms at the feasibility stage; the intervention could be modified to avoid these in the full trial, or a decision could be made not to proceed to a full trial even if progression criteria were met.
Exploring uncertainties before a full trial is underway can enable trialists to address problems or optimise the intervention or conduct of the trial. We present guidance that researchers, research funders and reviewers may wish to consider when assessing or undertaking qualitative research in feasibility studies. This guidance consists of 16 items framed around five research domains: research questions, data collection, analysis, teamwork and reporting. A strength of the guidance is that it is based on a combination of experiences from both published feasibility studies and researchers from eight universities in three countries. A limitation is that the guidance was not developed using consensus methods. The guidance is not meant as a straitjacket but as a way of helping researchers to reflect on their practice. A useful future exercise would be to develop worked examples of how research teams have used the guidance to plan and undertake their qualitative research within feasibility studies for trials. This would help to highlight the strengths and limitations of the guidance in different contexts. Using qualitative research with trials is still a developing area, and so, we present this guidance as a starting point for others to build on, as understanding of the importance of this vital stage of preparation for randomised controlled trials grows. Researchers may also wish to reflect on the utility of different qualitative methods and approaches within their studies to help other researchers make decisions about their future feasibility studies.
The workshop was funded by the MRC North West and ConDuCT Hubs for Trial Methodology. This work was undertaken with the support of the MRC ConDuCT-II Hub (Collaboration and innovation for Difficult and Complex randomised controlled Trials In Invasive procedures—MR/K025643/1). We would like to thank attendees at the ConDuCT-II Hub workshop on feasibility studies held at the University of Bristol in October 2014 who discussed and commented on a presentation of an earlier version of this guidance. JLD is a NIHR Senior Investigator. SL is supported by funding from the South African Medical Research Council.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.