Commentary | Open | Open Peer Review | Published:
Pilot and feasibility studies in exercise, physical activity, or rehabilitation research
Pilot and Feasibility Studiesvolume 4, Article number: 137 (2018)
Clinical trials of physical activity and rehabilitation interventions can be challenging. Pilot or feasibility studies can be conducted prior to a definitive randomized controlled trial (RCT), to improve the chances of conducting a high-quality RCT of a physical activity intervention.
Physical activity interventions or trials present unique challenges at the population, intervention, comparator and outcome levels. At each level, we present guidance for researchers on the design considerations for pilot or feasibility studies of physical activity interventions. When it comes to defining study population, physical activity trials often exclude participants with certain health conditions or other characteristics (e.g., age, gender) because of uncertainty of the safety of the exercise intervention or presumed differences in responsiveness, at the expense of trial generalizability. A pilot trial could help investigators determine refined inclusion and exclusion criteria to balance safety, adequate recruitment, and generalizability. At the intervention level, because exercise can be a complex intervention, pilot trials allow investigators to evaluate participant adherence and instructor fidelity to the intervention and participant experience. At the comparator level, control group dissatisfaction and post-randomization drop-out can occur, because of the desire to be randomized to the exercise group, and the difficulty with blinding to group allocation; an active control or deception could be used. Finally, at the outcome level, there should be an emphasis on the pilot or feasibility outcomes such as recruitment rate, adherence to exercise, and retention or fidelity, than the efficacy of the exercise intervention.
Physical activity and rehabilitation researchers can use pilot and feasibility studies to enhance the rigor of future trials, while also publishing the results of their pilot work to move the field forward. Researchers in this field are encouraged to use published reporting guidelines for pilot and feasibility studies and to consider the challenges discussed in this paper.
Randomized controlled trials (RCTs) of exercise, rehabilitation or physical activity interventions (from here, collectively called physical activity interventions) are challenging and resource-intensive. Physical activity interventions often require supervision by exercise or rehabilitation professionals . Participants have to dedicate time and effort to participate and achieve a prescribed frequency, intensity, type, and duration of physical activity; adherence is a challenge and blinding to group allocation is often impossible [1, 2]. Outcomes include performance-based measures such as gait speed, aerobic capacity, muscular strength tests, or functional tests of balance and mobility, such as the Timed Up and Go or timed chair stands, thus requiring in-person follow-up and good retention . There is increasing emphasis on reducing bias and improving the quality of clinical trials in the physical activity realm, as demonstrated by the publication of reporting guidelines specific to such interventions [1, 2]. One way to do this is to work out the uncertainties of the trial design beforehand. A feasibility study can be used to determine whether a future main study can be done, while a pilot study is a subset of a feasibility study that resembles the main trial but with the overall aim to assess the feasibility rather than the effectiveness of a study protocol . Therefore pilot or feasibility studies can be conducted prior to a definitive RCT, to improve the chances of success in conducting a high-quality RCT of a physical activity intervention.
Reasons for conducting pilot or feasibility studies can fall under four categories: to inform process (e.g., feasibility of recruitment, retention, intervention adherence), to understand resource requirements (e.g., time and budget issues), to inform management (e.g., personnel challenges, data collection or organization), and to advance scientific inquiry (e.g., intervention safety, appropriate dose, potential treatment effect) . Inferences about efficacy cannot be made from a physical activity study that is underpowered due to unsuccessful recruitment and retention, and it will be difficult to observe an effect if there is poor adherence. Therefore, instead of investing in underpowered hypothesis-testing studies, conducting exploratory pilot or feasibility studies with clear a priori criteria for success can better inform future efforts. If a pilot study turned out not to be feasible it is not considered a failed pilot study but rather a success, since the team has avoided spending resources on a large study that will not succeed .
Many pilot and feasibility studies examining physical interventions do not have clear a priori criteria for success and often focus on efficacy outcomes of the intervention rather than feasibility [5,6,7,8,9]. Researchers with training in exercise, physical activity, or rehabilitation research may not always have received formal training in the design and conduct of clinical trials nor have expertise in how to design and interpret a pilot study. We discuss pilot and feasibility studies in the context of physical activity interventions with the overall aim of promoting the successful use of pilot and feasibility studies in physical activity research. Challenges unique to physical activity pilot/feasibility studies will be presented under the categories population, intervention, comparator, and outcome (PICO). Example challenges are summarized in Table 1, and the examples used are summarized in Table 2.
Challenges to physical activity research
Pilot or feasibility studies can be used to determine how many participants are potentially eligible and how many are excluded and why, while also considering current physical activity levels and providing realistic screening to recruitment ratios in the population of interest, so that when planning a larger trial, researchers have a good estimate of how long it will take or if they need to recruit more sites or broaden their inclusion criteria.
Inclusion and exclusion criteria
Considerations specific to physical activity interventions include determining the inclusion/exclusion criteria related to safe participation at the required frequency, intensity, duration, and type of physical activity and how to exclude those who are already participating in a comparable physical activity intervention. A basic criterion is to exclude individuals who meet any of the American College of Sports Medicine (ACSM) relative or absolute contraindications to exercise . Participant screening should include current physical activity status; signs and symptoms of cardiovascular, metabolic, or renal disease; and capacity to complete prescribed physical activity given the level of supervision . For example, the risk of exercise-related cardiovascular events is greater among inactive participants who engage in physical activity at vigorous intensities [10,11,12]. Screening could also include tools like the “Get Active Questionnaire” . Researchers often err on the side of caution and create eligibility criteria for physical activity trials that are too restrictive, or not generalizable to real world practice. For instance, an exercise-based rehabilitation intervention study comprised of participants with intensive care unit-acquired weakness had strict eligibility criteria, and this resulted in a low ratio of included to excluded participants . A pilot or feasibility study could have been used as a first step to refine the inclusion and exclusion criteria for the physical activity intervention. If the proposed definitive trial is a multicenter trial, one should evaluate feasibility of recruitment at multiple centers as opposed to assessing feasibility at a single site.
Baseline physical activity level
Physical activity trials often aim to recruit individuals who are physically inactive, but individuals who express interest may already participate in some activity. A pilot or feasibility trial could evaluate the characteristics of those who express interest, to determine potential selection bias, including baseline physical activity level. One approach is to consider what the “active ingredients” of the intervention are and exclude those already participating in activities that contain those active ingredients. For example, if one is examining the efficacy of a moderate-high resistance training intervention on muscle strength, should a person who does yoga and walks every day be eligible? For multicomponent interventions, the screening becomes more complex. Our recent trial of home exercise in women with vertebral fractures included balance training, posture training and resistance exercises, and daily aerobic physical activity; we excluded individuals with exercise participation ≥ 3 times per week that addressed ≥ 2 of the domains . Based on the inclusion criteria, individuals who walked daily were eligible because they were only participating in one of the domains (i.e., daily aerobic physical activity), and a key focus of the intervention was resistance and balance training. Researchers should use a physical activity screening tool validated for the population of interest, e.g., International Physical Activity Questionnaire, the Rapid Assessment of Physical Activity, or the Global Physical Activity Questionnaire.
A unique challenge in physical activity trials that can be evaluated in pilot/feasibility studies is feasibility of intervention implementation including fidelity of intervention delivery, assessment of intervention adherence, and consideration of factors that might influence adherence.
Fidelity of intervention delivery
There are two types of treatment fidelity that can be evaluated in pilot/feasibility studies: participant fidelity and instructor fidelity. Participant fidelity refers to whether the participant completes the prescribed activities or exercises as required by the protocol (e.g., the right frequency, intensity, duration, and type), and instructor fidelity refers to whether the activities or exercises are prescribed, tailored, or demonstrated in accordance with the protocol. In a study by Patten et al. ACSM-certified trainers delivered the exercise interventions after 6 h of training on the study protocol ; when 15% of the sessions were randomly audited, coach adherence to the training manual was 95%, representing high instructor fidelity . Suttanon et al. involved the caregivers in a home-based exercise program designed for older adults with Alzheimer’s disease . Some caregivers indicated that they had competing demands or had health conditions that prevented them from continuously supporting the exercise program. A pilot trial could record modifications made (or that need to be made) to the intervention to inform a future trial.
Another consideration is that physical activity interventions require that participants not only understand the study but also to complete assessments or to adhere to the exercise intervention, particularly when participants are required to understand and execute exercises with proper form while unsupervised [5, 8]. A pilot study might consider testing various methods for evaluating informed consent or willingness to participate in study activities. A simple way to evaluate whether the participants fully understand or can complete tasks or exercises might include describing the study and asking the participant to repeat what they are being asked to do in their own words. If the focus of the trial is explanatory, it is desirable to only enroll participants who are likely to adhere to the intervention, one could consider a “faintness of heart” period [15, 16]. For example, potential participants could be asked to complete one exercise or task daily and record it in an exercise adherence log for 2 weeks. Those who fail to return the log would lose eligibility, and those who return a completed log could then be randomized. Or, adherence could be a key outcome of a pilot trial.
Assessment of intervention adherence
A pilot or feasibility trial is ideal for assessing whether participants will adhere to an intervention with predefined criteria for success that are aligned with a hypothesized minimum dose. The criteria for success could be defined based on a dose response study, or prior work that informs what is realistic or what might be a minimum effective dose. If adherence is a primary feasibility outcome, the methods for assessing adherence must be rigorous, a goal that is easy to achieve in supervised interventions, but less so in unsupervised ones. To monitor exercise adherence for the latter, participants are often asked to complete a self-reported exercise diary, which is subject to social desirability bias and non-response bias. For example, Suttanon et al. indicated that only two thirds of monthly exercise logs were completed by the end of the study . Completing calendars are considered burdensome by participants and could lead to study withdrawal . Activity monitors can also be used to assess physical activity levels but have limited utility in interventions that do not involve walking or running; they do not capture resistance training, cycling, balance training, or swimming, for example. A pilot/feasibility trial could also consider the feasibility of compliance with physical activity monitors. For example, Patten et al. found that the majority of the participants did not wear the device provided . Since participants and those delivering interventions may not be blind to group allocation, blinding assessors becomes more important in physical activity trials to reduce detection bias. Physical activity trials often include physical activity outcomes or adherence, or other outcomes (e.g., exercise self-efficacy) that when assessed, could reveal group allocation. Therefore, it would be wise to have separate assessors not blind to group allocation evaluate these outcomes so as to ensure that assessors examining non-exercise related outcomes remain blind to group allocation.
Factors influencing intervention adherence
A pilot or feasibility study should also explore factors that influence adherence, including intervention type and setting, participant satisfaction or experience measures, or conduct exit interviews to gather input on their overall experience with the study, or the intervention. The method of delivery, group versus individual exercise program, as well as the setting, such as community setting or at home, can influence adherence to the physical activity intervention . Home-based exercise interventions do not involve the planning of transportation or scheduling difficulties, but unsupervised or intermittently supervised interventions may compromise adherence or intervention fidelity. Intermittent supervision by means of home visits or telephone calls may help but add cost (e.g., time, travel), complexity, and variability in terms of intervention fidelity. The number of home visits or telephone calls should be determined when designing the exercise intervention, and the proximity of persons delivering the intervention to the participants should be considered. A cross-sectional study examined exercise adherence to strength training in older women and showed that those who were older, had greater physical activity level, better perceived overall health, and were supervised by a well-trained professional were more likely to adhere to their exercise regimen . To enhance adherence, some studies include exercise-related behavior interventions that involve goal setting, time management, and overcoming barriers [18, 19]. Behavior intervention strategies, however, do not always improve exercise retention or adherence . Nevertheless, to ensure effective counseling, the strategies employed should be standardized and completed by trained personnel . When it comes to physical activity interventions, there is substantial scope for assessing feasibility in advance of a large trial, including intervention fidelity, participant experience and adherence.
An ideal comparison group is a placebo, such that the participants are blinded to whether they are in the active treatment group. However, a placebo control is difficult or near impossible for physical activity interventions, resulting in the potential for performance bias (i.e., that knowledge of the intervention received will influence outcomes). Further, what often attracts potential participants is the opportunity to be in the exercise group. A feasibility study should assess contamination or post-randomization drop-out rates in the comparison group relative to the intervention group. As an example, Barker et al. experienced a 23% drop-out rate in the control group, which was higher than anticipated, versus a 14% drop-out rate in the exercise group . Participants lost interest after being allocated to the control group, and it was the most commonly reported reason for study withdrawal . Individuals randomized to control who really want to start exercise may seek an exercise program outside of the study; a pilot study could monitor whether this is an issue. Compensatory strategies might include an attention control or an active control that is attractive to participants, with or without deception (e.g., tell participants the study is comparing two exercise interventions). For instance, Giangregorio et al. and Suttanon et al. made sure that the control group received the same number and duration of home visits and phone calls [9, 14]. Patten et al. made sure that all participants received identical evidence-based smoking cessation counseling and added exercise counseling in the intervention group . Studies of challenging balance interventions could provide seated or lying exercises as the comparator . Therefore, an alternative exercise intervention can be selected for the comparator group as long as it does not have an effect on the outcome of interest. A feasibility study could assess participant adherence to or acceptance of the attention/active control, and whether it prevented contamination or post-randomization drop out.
Feasibility outcomes could include recruitment rate [5,6,7,8, 14], consent rate , adherence to the physical activity intervention [5,6,7,8, 14], study retention [7, 8, 14], adverse events [5, 6, 8, 9], and participant experience or satisfaction [5, 7, 8, 14]. The decision to move from a pilot/feasibility study to a full trial should be based on the feasibility objectives and not the secondary outcome measures. Sometimes, emphasis is placed on efficacy outcomes for which the pilot/feasibility trial was not designed or powered to assess. Outcomes related to the efficacy of the intervention can be measured, but should be treated as exploratory, and not hypothesis-testing. However, one could explore the feasibility of measuring the outcomes that will be used to assess efficacy in a future trial, e.g., potential for ceiling or floor effects, the time taken to complete assessments, and extent to which data is missing. Intention to treat analyses require imputation of missing data, but sometimes data are not missing at random e.g., because of difficulties completing performance-based measures, or higher drop-out rate in controls because of dissatisfaction with group allocation. All pilot or feasibility studies should have a priori criteria for success related to the primary feasibility questions, to inform unbiased conclusions as to whether it is feasible for the pilot study to become a main study . Unfortunately, however, many pilot and feasibility studies of physical activity interventions fail to specify criteria for success (Table 2), and instead make vague claims about high completion rates, good attendance, an absence of adverse events, or strong patient acceptability. In summary, pilot or feasibility outcomes should be measuring whether a future trial is realistic or feasible. Pilot or feasibility trials should not be used to provide a conclusive estimate as to whether an intervention is effective.
Reporting of pilot and feasibility trials of physical activity interventions
High-quality pilot and feasibility trials have the potential to advance the future success of clinical trials in the physical activity realm. They form the foundation for exceptional grant proposals to secure funds for large trials, as they can increase reviewer confidence regarding the team’s capacity to do the work. Further, they can still contribute to academic productivity if they are published. Although pilot or feasibility studies are informative, they are not often published due to emphasis on efficacy outcomes or they lack clear feasibility objectives and do not specify criteria for success (Table 2) . In addition, there are misconceptions regarding what is considered a pilot study. For example, studies performed with limited funding, in a short time frame, or with restricted resources, are not reasons a study should be called a pilot . Given the complex nature of physical activity interventions, pilot and feasibility studies in this field should be more abundant. Grant review panels and journals, however, will not recognize the value of pilot studies without rigor in design, conduct, and reporting. Accordingly, the Consolidated Standards of Reporting Trials (CONSORT) group has now created extensions specific to the reporting of pilot and feasibility trials  and of non-pharmacological interventions . Adhering to the CONSORT 26-item checklist can improve reporting and can also help in designing and evaluating future trials . Journals are increasingly becoming receptive to high quality pilot trials, and the journal Pilot and Feasibility Studies is an ideal destination.
Pilot and feasibility studies of physical activity interventions can inform the design and conduct of large, definitive trials by testing how the trial processes work together and by evaluating feasibility of recruitment, retention, adherence, intervention delivery, outcome assessment, or any other aspect of the trial that could compromise its success. Feasibility issues unique to physical activity trials include the challenge of blinding of group allocation, retaining the comparison group, participant and instructor fidelity, and being able to recruit people not already active enough, but interested enough in becoming active to sign up for the study. A pilot trial should have clear objectives and a priori criteria for success and should be designed, conducted, and reported using the same standards as any high-quality randomized controlled trial. Overall, pilot and feasibility studies can ensure that scarce research money and researcher efforts are being invested in large trials that can provide definitive answers to important research questions.
American College of Sports Medicine
Consolidated standards of reporting trials
Population, intervention, comparator, and outcome
Randomized controlled trials
Slade SC, Dionne CE, Underwood M, Buchbinder R. Consensus on exercise reporting template (CERT): explanation and elaboration statement. Br J Sports Med. 2016;50(23):1428–37.
Eldridge SM, Chan CL, Campbell MJ, Bond CM, Hopewell S, Thabane L, et al. CONSORT 2010 statement: extension to randomised pilot and feasibility trials. Pilot Feasibility Stud [Internet]. 2016;355:1–29. Available from: http://pilotfeasibilitystudies.biomedcentral.com/articles/10.1186/s40814-016-0105-8
Eldridge SM, Lancaster GA, Campbell MJ, Thabane L, Hopewell S, Coleman CL, et al. Defining feasibility and pilot studies in preparation for randomised controlled trials: development of a conceptual framework. PLoS One. 2016;11(3):1–22.
Thabane L, Ma J, Chu R, Cheng J, Ismaila A, Rios LP, et al. A tutorial on pilot studies: the what, why and how. BMC Med Res Methodol [Internet]. 2010;10(1):1–10. Available from: http://bmcmedresmethodol.biomedcentral.com/articles/10.1186/1471-2288-10-1
Connolly B, Thompson A, Douiri A, Moxham J, Hart N. Exercise-based rehabilitation after hospital discharge for survivors of critical illness with intensive care unit-acquired weakness: a pilot feasibility trial. J Crit Care [Internet]. 2015;30(3):589–98. Available from: https://doi.org/10.1016/j.jcrc.2015.02.002
Granger CL, Chao C, McDonald CF, Berney S, Denehy L. Safety and feasibility of an exercise intervention for patients following lung resection: a pilot randomized controlled trial. Integr Cancer Ther [Internet]. 2013;12(3):213–24. Available from: http://journals.sagepub.com/doi/10.1177/1534735412450461
Patten CA, Bronars CA, Douglas KSV, Ussher MH, Levine JA, Tye SJ, et al. Supervised, vigorous intensity exercise intervention for depressed female smokers: a pilot study. Nicotine Tob Res. 2017;19(1):77–86.
Barker AL, Talevski J, Bohensky MA, Brand CA, Cameron PA, Morello RT. Feasibility of Pilates exercise to decrease falls risk: a pilot randomized controlled trial in community-dwelling older people. Clin Rehabil [Internet]. 2016;30(10):984–96. Available from: http://journals.sagepub.com/doi/10.1177/0269215515606197
Suttanon P, Hill KD, Said CM, Williams SB, Byrne KN, LoGiudice D, et al. Feasibility, safety and preliminary evidence of the effectiveness of a home-based exercise programme for older people with Alzheimer’s disease: a pilot randomized controlled trial. Clin Rehabil [Internet]. 2013;27(5):427–38. Available from: http://journals.sagepub.com/doi/10.1177/0269215512460877
Riebe D, Franklin BA, Thompson PD, Garber CE, Whitfield GP, Magal M, et al. Updating ACSM’s recommendations for exercise preparticipation health screening. Med Sci Sports Exerc. 2015;47(11):2473–9.
Franklin BA, McCullough PA. Cardiorespiratory fitness : an independent and additive marker of risk stratification. Mayo Clin Proc. 2009;84(9):776–9.
United States Department of Health and Human Services. Physical Activity Guidelines Advisory Committee Report, 2008 [Internet]. 2008. Available from: https://health.gov/paguidelines/Report/pdf/CommitteeReport.pdf. Accessed 23 June 2018.
Canadian Society for Exercise Physiology. Get Active Questionnaire [Internet]. 2017. Available from: http://www.csep.ca/en/publications/get-active-questionnaire. Accessed 22 Dec 2017.
Giangregorio LM, Thabane L, Adachi JD, Ashe MC, Bleakney RR, Braun EA, et al. Build better bones with exercise: protocol for a feasibility study of a multicenter randomized controlled trial of 12 months of home exercise in women with a vertebral fracture. Phys Ther [Internet]. 2014;94(9):1337–52. Available from: https://academic.oup.com/ptj/article-lookup/doi/10.2522/ptj.20130625
Veterans Administration Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension: results in patients with diastolic blood pressures averaging 115 through 129 mm Hg. JAMA. 1967;202(11):1028–34.
Sackett DL. Explanatory and pragmatic clinical trials: a primer and application to a recent asthma trial. Pol Arch Med Wewn. 2011;121(7–8):259–63.
Seguin RA, Economos CD, Palombo R, Hyatt R, Kuder J, Nelson ME. Strength training and older women: a cross-sectional study examining factors related to exercise adherence. J Aging Phys Act. 2010;18(2):201–18.
Ackermann RT, Deyo ÃRA, LoGerfo JP. Prompting primary providers to increase community exercise referrals for older adults: a randomized trial. J Am Geriatr Soc. 2005;53:283–89.
Petrella RJ, Koval JJ, Cunningham DA, Paterson DH. Can primary care doctors prescribe exercise to improve fitness? The step test exercise prescription (STEP) project. Am J Prev Med. 2003;24(4):316–22.
Cox KL, Burke V, Beilin LJ, Derbyshire AJ, Grove JR, Blanksby BA, et al. Short and long-term adherence to swimming and walking programs in older women - the sedentary women exercise adherence trial (SWEAT 2). Prev Med (Baltim). 2008;46(6):511–7.
Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee IM, et al. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc. 2011;43(7):1334–59.
Boutron I, Altman DG, Moher D, Schulz KF, Ravaud P. CONSORT statement for randomized trials of nonpharmacologic Treatments: a 2017 update and a CONSORT extension for nonpharmacologic trial abstracts. Ann Intern Med. 2018;167:40–47.
The authors would like to thank Gillian Lancaster and Lehana Thabane for their help in the preparation of this manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.