I recently attended my institution's student research day. As a part of our entry-level Doctor of Physical Therapy curriculum, students take a series of evidence-based practice courses and continue to apply concepts learned in these courses throughout the curriculum. For one course, Critical Inquiry, students work in small groups mentored by a faculty member and a physical therapist who is practicing in our health care system to develop a PICO (Patient, Intervention, Comparison, Outcome) question and perform a systematic review to answer the question. Students then present the findings of their systematic review at student research day.

There were a variety of systematic review topics, ranging from inspiratory/expiratory muscle training for individuals with spinal cord injury to modified constraint-induced movement therapy of the lower extremity for people with stroke. As I virtually went from poster to poster, there was one consistent theme that emerged: movement and activity-based interventions improve outcomes at the body structure function and the activity levels. Exactly what type of movement/activity-based intervention is most efficacious and what is the optimal dose, however, were not clear. Another interesting issue that emerged was that the sample sizes in the studies included in the systematic reviews were all relatively small. Very few studies had sample sizes greater than 50.

Clinicians are eager to implement the latest research findings to improve the lives of our patients. The gaps in the research literature related to what specific type of movement intervention and the exact dosage, however, challenge the ability of clinicians to translate research findings safely and optimally to practice. A recent review by Stinear and colleagues1 of large rehabilitation trials (all studies had n > 100) for people with stroke found that motor function improves as a result of rehabilitation interventions, but that in most cases there was no difference between the comparison and experimental groups. This is also true regarding rehabilitation trials involving people with Parkinson disease.2

Although the amount and quality of rehabilitation research have grown since I started practicing, there are still very few large, phase 3 rehabilitation trials to guide practice. Most rehabilitation research trials are smaller phase 1 or phase 2 studies. A recent systematic review of motor rehabilitation trials in people with stroke found that in 1410 trials published between 1972 and 2018, the mean sample size was 46.4 (SD = 60.8) and a mean PEDro score of 6.1 (SD = 1.5).3 Although an important guide to practice based on a synthesis of the best, current evidence, the recent clinical practice guideline to improve locomotor capability in people with stroke, traumatic brain injury, and spinal cord injury that recommends the use of moderate- to high-intensity walking training (mean sample size = 42, SD = 36) and virtual reality (mean sample size = 36, SD = 21) is based on studies with relatively small sample sizes.4 These sample sizes are smaller than in many other fields. For example, the median trial sample size from a systematic review that evaluated the use and definition of perioperative outcomes in people undergoing cardiac surgery was 351 (range, 57-4752).5

Rehabilitation researchers who work to determine the efficacy and effectiveness of physical therapy interventions face many unique challenges when performing clinical trials. The complex psychosocial factors that people with neurological disorders face make it difficult to tease out the impact of the many different ingredients that are a part of rehabilitation practice. To better guide physical therapy practice, there is a need for researchers to provide detailed information on the dosage of both the experimental and comparison interventions and to implement trials with larger sample sizes. Although we have improved on our use of outcome measures of clinical capacity, we need to measure what matters to patients (that is a topic for another editorial). Different models of clinical trials such as the single-arm, open-label methodology used by Gracies and colleagues6 in this issue can provide important insight to guide clinical practice and an opportunity to address some of the aforementioned challenges.

Despite the challenges faced by both clinicians and researchers who work with people with neurological health conditions, it is gratifying to know that the movement and activity-based interventions that we do have a positive impact on our patients' lives. The fun part is that there is still plenty of exciting work ahead.

REFERENCES