What It Means to Be Evidence-based
By the Exercise Geeks
When contemplating evidence-based practice – regardless of whether it’s a physical therapy or strength and conditioning setting – we always reference the principles set forth by Dr. David Sackett. The late Canadian-American medical doctor was a pioneer in the field of evidence-based medicine. He deconstructed the practice into a triad of factors:
- Collect and appraise the best available clinical evidence.
- Integrate this evidence with clinical expertise.
- Apply this evidence in the context of patient/client values and expectations.
To better understand the role of each factor in the decision-making process, let’s dive deeper into each one individually.
1. Collect and appraise the best available clinical evidence.
For many, this factor is the only factor that is considered when discussing science-based practice. Clinical evidence refers to systematic reviews, randomized controlled trials, and other less robust studies published in peer-reviewed journals and online databases.
There’s no doubt that this factor is vital to the process of continued learning in our respective fields, and it is a worthwhile effort to promote practices based on as much of this information as possible. When considering this factor in the triad, it’s important to appreciate what “high-quality evidence” means. Just because an article is published in a peer-reviewed journal doesn’t mean it’s sound evidence on the topic.
The following is the hierarchy of evidence, from most robust to least:
- Meta-analyses: a statistical approach to combine the results from multiple studies in an effort to increase statistical power (over individual studies), improve estimates of the size of the effect and/or to resolve uncertainty when reports disagree.
- Systematic Reviews: a research study that exhaustively collects and evaluates all relevant studies based on predetermined selection criteria. Researchers use methods that are determined before they begin to frame one or more questions, then they find and analyze the studies that relate to that question.
- Critically Appraised Topics (CATs): a short summary of the best available evidence, created to answer a specific clinical question. A CAT looks like a short, rigorous version of a systematic review.
- Critically Appraised Individual Articles: authors of these articles evaluate and synopsize individual research studies.
- Randomized Controlled Trials: a type of scientific experiment where the people being studied are randomly allocated to the different treatments under study. The RCT is often considered the gold standard for a clinical trial.
- Cohort Studies: a type of longitudinal study that involves identification of two groups (cohorts) of individuals, one that received an exposure of interest and one that did not. These cohorts are then followed forward and outcomes are assessed over a specified period of time. (i.e. Following smokers vs. nonsmokers for a 20 year period and assessing rates of lung cancer between the two groups.)
- Case Controlled studies: epidemiological, observational studies thatcompare individuals who have an outcome of interest (cases) with those who do not have the outcome (controls). Case controlled studies look back retrospectively to compare how frequently the exposure to a risk factor is present in each group to determine the relationship between the risk factor and the outcome of interest. (i.e. Examining a group of individuals that have developed lung cancer and then looking back retrospectively to see how many of those individuals were smokers vs. nonsmokers.)
- Case studies/case series (uncontrolled): similar to case controlled studies except when looking back retrospectively there is no comparison to a group of individuals that did not have the outcome of interest (controls). (i.e. Examining whether physical activity decreases pain in an individual or series of individuals that has fibromyalgia. The specific intervention of physical activity would need to be compared to no physical activity or no change in activity in order for the study to be controlled.)
- Expert opinion: this category of evidence often gets placed into the hierarchy, but it is difficult to make a claim that it is really “evidence” at all. Moreover, when discussing the factors set forth by Sackett, expert opinion is better discussed when speaking about the next topic in the triad – clinical expertise.
Recent clinical examples from physical therapist Chris Leib demonstrates how current scientific evidence can be vital in giving a client the best possible care:
I’ll admit: I hate open chain lower body strengthening exercises. Examples of these exercises include machine-based seated leg extensions, seated or prone leg curls, the multi-hip device, seated hip abduction/adduction, and straight leg raises, just to name a few. My bias against this type of exercises comes mostly from clinical experience in seeing very little direct benefit to my patients/clients when attempting to improve a functional movement in which they are limited. Another way to say this is that clinical experience has shown me that there are much better ways to spend my time with patients/clients.
Another intervention that I’ve found practically useless in my years of experience as a therapist is therapeutic ultrasound. This treatment modality uses sound waves to “theoretically” improve circulation to subcutaneous tissues in an attempt to aid in their healing process. As I said, I have not seen much noticeable success with this treatment option in the past decade. Moreover, the research has generally not been kind regarding its effectiveness in treating pain and improving function (more on this topic later).
The problem with getting overly passionate about any topic (especially to the point of using the word “hate”) is that your judgment can become clouded and you can miss potential benefits of interventions that you have long biased your mind against.
This point became abundantly clear a few weeks ago when a client who is status-post his second ACL reconstruction asked me a question. His question pertained to methods for facilitating the maturation process of his surgical graft. My initial inclination was to separate that question from my scope of practice and refer it to his surgeon. However, I figured I would see what the research said on the matter and see if there was anything that could be done from a physical therapy or fitness standpoint.
I’m really glad I took this step. After performing a Google Scholar and PubMed search, I came across two valuable studies that changed my perspective on training individuals through this particular recovery process. The first was a 2002 study by Yamakoto et al. in Arthroscopy that discussed the influence of mechanical stress on ACL graft healing. After examining the findings of this study, one major piece of information caught my attention — and also chipped away at my negative bias towards open chain lower body exercises.
The study concluded that “tensile stress enhances the healing process of tendon-bone junctions.” Using clinical reasoning, I deduced that indeed open chain exercises are ideal for creating this type of stress. Maybe there is something to these exercises after all. It appears that at the point of the recovery process where my client currently finds himself, open chain exercises are necessary for the overall resiliency of the graft so that more high level activities can be undertaken with less risk of re-injury. Thank you, science, for this new information and a gentle demonstration of the dangers of overcommitting to dogmatic principles.
A second study dealing with a soft tissue graft repair made me swallow more prideful notions. In this case, it was in reference to therapeutic ultrasound. The referenced study was done in 2007 by Walsh and colleagues and was also published in Arthroscopy. This controlled animal trial looked at the effects of ultrasound on tendon-bone healing in an intra-articular sheep knee model. It concluded that the application of low-intensity pulsed ultrasound appears to improve healing at the tendon-bone interface for soft tissue grafts.
Although it is difficult to make definitive conclusions on how an animal study relates to human beings, this study does allow for the consideration of a generally risk-free, easy-to-apply technique that has potential for substantially decreasing chance for re-injury in the late stages of ACL reconstruction rehabilitation. Without this information, I would not have considered this an option for management of similar conditions with my patients, therefore possibly leaving greater potential for re-injury on the table.
Hopefully, the examples above make it quite clear how important it is to be diligent in appraising the scientific literature; do not simply rely on your clinical opinions. With that being said, we will now look at the importance of clinical experience and where it fits into the grand scheme of evidence-based practice.
2. Integrate this evidence with clinical expertise.
This factor can be quite subjective. After all, how does one define who is an expert and who is not? Are there a certain number of years that one must work in the field to be considered an expert? Are there specific credentials that are necessary to earn this distinction?
There are no cut-and-dry answers to these questions, and unfortunately, that doesn’t make for good science. However, the subjectivity of the often differing opinions that many professionals within the same field hold often sparks innovation and change in ways that no well-designed study ever could. Clinical experience and inter-professional discussion is where new and progressive questions are developed to be tested in the laboratory later on.
With more time working “in the trenches,” you begin to develop “a body of norms,” or an improved ability to recognize common patterns. Certain professionals develop this art more quickly than others, but few would debate that it is not a common denominator for those considered “experts” in their respective fields. This skill of pattern recognition allows professionals to develop systems that can be assessed with each client or patient during every encounter. Moreover, as professionals require less effort figuring out clinical presentations that once took up a majority of their time, more focus is able to be allocated towards clinical reasoning.
Clinical reasoning is the process by which professionals collect cues, access information, come to an understanding of a patient problem or situation, plan and implement interventions, evaluate outcomes, and reflect on and learn from the full experience. It has been demonstrated that the clinical reasoning process is dependent upon a professional’s attitude, philosophical perspective, and preconceptions – not simply on the number of years the individual has accumulated (Scheffer & Rubenfeld 2000; McCarthy 2003).
Clinical reasoning has been conceptualized well by the Australian Learning and Teaching Counsel by using a clinical reasoning cycle. This cycle is as follows:
- Consider the patient situation: client demographics, medical diagnoses, reason for seeing you
- Collect cues/information
- Review any forms to indicate past medical/fitness history.
- Gather information from the client as to perceived limitations/chief complaints.
- Recall knowledge from your area of expertise that is pertinent to the client’s complaints.
- Process information
- Interpret the information to develop an understanding of the client complaint.
- Discriminate what information needs to be prioritized.
- Relate isolated information together by recognizing clinically meaningful patterns.
- Infer meaning of subjective and objective information gathered by generating deductions and contemplating differential alternatives to initial opinions.
- Match the current situation to past experiences of a similar nature.
- Predict the likelihood of a positive outcome.
- Identify the problem/issue: Combine objective information gathered and clinical inferences made to develop a definitive prioritization for intervention (i.e. dominant limitation in strength, power, endurance, mobility, stability, etc.).
- Establish goals: Develop realistic, easily measurable goals in order to assess progress over time.
- Take action: Perform an intervention.
- Evaluate: Re-assess subjective and objective information initially gathered (baselines) in order to identify change.
- Reflect on the process and new learning: Regardless of whether change made during the evaluation process is positive or negative, it’s an opportunity for learning. If positive change is made, reflect on the core issues of why that particular client had favorable results and contemplate other presentations that would also likely benefit from similar interventions. If negative change is made, reflect on what can be done differently. Is it a matter of subtly adjusting parameters such as frequency, intensity, or duration, or does the entire direction of intervention need to be adjusted?
As one can easily see, this clinical reasoning cycle in itself creates a bit of an ongoing “scientific study.” Basically, a good clinician treats every client as an individual dynamic case study. We use the word dynamic because there is a continuous assessment and adjustment of parameters and intervention techniques based on changing client response.
To explore this factor further, let’s look at a commonly seen clinical example:
Knee osteoarthritis (OA) can be defined as a degenerative “wear-and-tear” of the cartilage of the knee joints. This degradation of cartilage can occur either in the patellofemoral joint (kneecap on thigh) and/or tibiofemoral joint (thigh on lower leg).
Knee OA is seen in individuals of all ages (more common in those over 50 years old) and is one of the most common structural diagnoses in a physical therapy clinic. Over the past several years, research has demonstrated pretty clearly that squatting can be stressful to the knees of individuals with this diagnosis.
Squats have been examined in multiple studies in order to determine the effect on a variety of tissue stresses. Below is some specific research regarding the squat and its proposed danger for those with knee OA:
- Studies have demonstrated that squats increased loads placed on the tibiofemoral and patellofemoral knee joints (Raske and Norlin 2002; Escamilla et al. 2001) as well as on the posterior cruciate ligament (Toutoungi et al. 2000).
- These stresses have indeed been identified to increase injury risk, especially among individuals with a history of knee pathology (Toutoungi et al. 2000).
- The act of making soft tissue contact between the posterior leg and posterior thigh (calf and hamstring), as in an “ass-to-grass” squat, causes a shift of the axis of rotation away from the knee. This axis shift can create a shear force that further loads the cartilage of the tibiofemoral joint and may possibly lead to a worsening presentation of knee OA (Kreighbaum and Barthels 1996).
With the information at hand, it seems pretty clear-cut: those with knee OA should not be squatting. However, before jumping to a conclusion, several clinical questions must also be answered:
i. Does the patient actually have symptoms when squatting?
If not, then why should they not squat? Where do we draw the line? If we eliminated all exercises that created joint stress, we would not be doing much exercise at all. The body is a system of systems where the joints need to take stress in order for muscles to get stronger to protect those same joints.
ii. Is squatting necessary for the patient to achieve unrestricted physical function?
Not all squatting is heavy loaded back squatting. If we tell people not to squat, we are also limiting their options for getting up from the ground or negotiating objects at waist level. Squatting is an assessment of the ability to flex the hips, knees, and ankles while weightbearing. Seems pretty important – and likely worth training, obviously with respect to any increase in symptomatic presentation.
iii. Are all cases of knee OA the same?
Absolutely, positively not! No two cases of any diagnosis are ever the same. For some patients with knee OA, it hurts simply to get up from a chair. For others, there are only symptoms when running for miles.Moreover, many individuals that have been diagnosed with radiographic knee OA (cartilage degeneration on an X-ray) have no complaints whatsoever. So with this in mind, it seems pretty shortsighted to contraindicate a movement based on a diagnosis that has wildly different presentations from person to person.
iv. Does the patient have a good understanding of squatting mechanics?
This is all too common. A patient says that it hurts when they squat. The patient then proceeds to demonstrate an imitation of Bambi on ice, which consequently results in knee pain. Yet, when you have them repeat the squat with a few simple instructions and an added heel lift, suddenly the patient is able to sit into that squat absolutely pain-free. Is this an OA problem or just a poor movement pattern?
v. Does the patient have the range of motion available to squat?
The squat calls for near full range of motion at the hips, knees, and ankles when performed to its fullest expression. If any of these ranges are limited, squatting will be nearly impossible without drastic compensations. In individuals with knee OA, these compensations often lead to excessive irritation to the already degenerated knee cartilage and is at least a contributing factor to their knee pain.
Oftentimes once mobility limitations are improved or external supports (i.e heel lifts) are utilized to adjust for mobility limitations, pain-free squatting can be resumed. In these cases, the degenerated cartilage did not change but nevertheless the patient’s ability to squat is no longer a problem. Again, is this an OA problem or a mobility problem?
Using the previous example to make a broader point, clinical experience shows that structural diagnosis should never be the only factor considered when deciding which movements are appropriate. Instead, an individualized movement assessment will always be more valuable for making these judgements.
Research of the manner described above should not be utilized to contraindicate any particular activity altogether. Instead, research should be used to give clinicians more information to make informed decisions. A skilled clinician does not use research to eliminate options for movement, but instead uses it to get people to perform a wider array of activity in a safer manner.
3. Apply this evidence in the context of patient/client values and expectations.
The importance of this factor in generating positive patient/client outcomes cannot be overstated. Specifically when dealing with patients in pain, there is this idea that there is a certain way a patient is “supposed to present” after a structural diagnosis has been made. Any veering from these theoretical presentations are deemed the fault of the patient’s psychological dealings with the situation, not a part of the overall scope of his or her condition.
The truth is that regardless of their diagnoses, no two people EVER present exactly the same. Recognizing and understanding how to work with different patient perceptions with regards to their condition proves to be more important than identifying an exact structural cause to injury.
Of course, physical therapists and strength and conditioning professionals don’t have the means to make change to structural damage; that’s the job of a surgeon. However, therapists and trainers are at the forefront of improving an individual’s physical function and movement quality. The first step in this process is often to decrease an individual’s fear of what they perceive are “dangerous” movements/activities. 
We can circle back to therapeutic ultrasound for an example from Chris’s clinical practice where patient values actually trumped current evidence and clinical experience:
It has long been recognized through scientific review that there is little evidence that active therapeutic ultrasound is more effective than placebo ultrasound for treating people with pain or a range of musculoskeletal injuries. However, a very specific case comes to mind with a patient suffering from chronic medial ankle pain.
The patient’s symptoms were so severe that walking for any longer than two minutes “flared up” her pain to the point where she was unable to walk for the rest of the day and often for several days afterwards. As mentioned, this was a chronic problem and she had received prior treatment for this issue in the past.
As part of that prior treatment, she had received therapeutic ultrasound and genuinely believed she had received benefit from the intervention in the past. When she initially mentioned this to me, I have to admit that I was skeptical of commencing its implementation due to my knowledge of the lack of supporting evidence. However, my mind quickly turned to the factors laid out by Sackett, and I was able to identify this situation as one where patient values needed to be respected.
I weighed the situation and came to the conclusion that it was unlikely there would be any harm in applying the treatment. The worst case scenario would be that it did nothing to help her, and we could move forward with other treatment options. I further reasoned that this intervention need not be performed in isolation, but rather in conjunction with other treatment approaches that I deemed more supported by the evidence.
Once implemented, it was immediately apparent that the addition of the ultrasound treatment was a game-changer in the progress of the patient’s function. The patient seemed to view the use of the ultrasound treatment as a sort of safety net, giving her the courage to perform exercise specifically geared to improve on her functional limitations at increasing repetitions and loads. These progressions were accomplished with minimal exacerbation of her condition. Her walking tolerance rapidly progressed from less than two minutes to ten minutes and any increase in pain during the session was mitigated by the therapeutic ultrasound treatment at the end of the session.
It’s impossible to determine whether the above patient genuinely received physiological benefit from the treatment or whether the act of applying the treatment decreased feelings of “danger” towards physical activity that she had previously linked to increasing symptoms. Clinically, however, an exact answer to this question is not important, as it is likely that different patients/clients will gain benefit from different interventions for different reasons. If a clinician fails to look past “the best current evidence” and appreciate the values of the patient, the best overall outcome for that patient is not being considered.
Recapping the Evidence-based Approach
Being an evidence-based fitness professional can be broadly described as using scientifically backed methods to assess, design, and implement programs to clients. However, problems arise when we start requiring every technique or method we utilize to be backed by a wealth of original scientific research. Therefore, as we discussed above, we must integrate several factors into our assessment of clients, our designing of their programs, and overall training methods and philosophies.
- Search the current literature.
- Evaluate original research studies, while searching for trends and limitations.
- Read reviews and meta-analyses pertaining to the subject, while understanding what studies were included and the scope of the reviews.
- See where the current scientifically-backed body of knowledge sits.
- Pull from personal experience and/or from the experiences of trusted colleagues.
- Pull from your own experiences with various patients, clients, and athletes.
- Speak to trusted colleagues in the field, which is why it is extremely important to maintain excellent relationships with coaches, trainers, physical therapists, researchers, and others in the allied health professions. Understanding one’s scope and limitations is critical, and having trusted colleagues is vital to being evidence-based.
- Take personal experiences along with colleagues’, and put together a list of key points, tips, and thoughts.
- Merge the literature and experience with scientific basis and practical application.
- Take current understanding of the literature, trends, the limitations of the research, personal and colleagues’ experiences in order to come up with points of agreement and disagreement.
- Use the points of agreements as a foundation and continue to evaluate the points of disagreements.
- Implement certain techniques and methods even if there is no current research to support them, provided there is a scientific basis to support their use.
- Implement approaches and evaluate.
- Remember that each client is different, and just because somethingshould work with them, it may, in fact, not. Therefore, constantly evaluate methods, look for trends of what works with various populations, and pay attention to client/athlete feedback.
- Science is dynamic and ever-changing. Always look for new evidence, changing trends, and feedback from colleagues.
Being evidence-based is about more than just bridging science with practical application. Being evidence-based is about understanding what we know and using it as a basis, while learning from our own experiences and those around us. It is a dynamic, ever-changing, and cyclic process, which allows us all to grow as practitioners.
- Escamilla, R.F., et al. (2001). Effects of technique variations on knee biomechanics during the squat and leg press. Medicine & Science in Sports & Exercise, 33 (9), 1552–66.
- Gatchel, R.J., Peng, Y.B., Peters, M.L., Fuchs, P.N., & Turk, D.C. (2007). The biopsychosocial approach to chronic pain: Scientific advances and future directions. Psychological Bulletin, 133(4), 581-624.
- Heidari, B. (2011). Knee osteoarthritis prevalence, risk factor, pathogenesis and features: Part 1. Caspian Journal of Internal Medicine, 2 (2), 205-212.
- Leyland K.M., Hart D.J., Javaid M.K., Judge A., et al. (2012). The natural history of radiographic knee osteoarthritis: a fourteen-year population-based cohort study. Arthritis & Rheumatology ,64(7), 2243-51.
- Raske, A., & Norlin, R. (2002). Injury incidence and prevalence among elite weight and power lifters. The American Journal of Sports Medicine, 30 (2), 248–56.
- Robertson, V.J., & Baker, K.G. (2010). A review of therapeutic ultrasound: Effectiveness studies. Physical Therapy, 81, 1339-1350.
- Toutoungi, D.E., et al. 2000. Cruciate ligament forces in the human knee during rehabilitation exercises. Clinical Biomechanics, 15 (3), 176–87.
- Walsh, W.R., Stephens, P., Vizesi, F., Bruce, W., Huckle, J., & Yu, Y. (2007). Effects of Low-Intensity Pulsed Ultrasound on Tendon–Bone Healing in an Intra-articular Sheep Knee Model. Arthroscopy: The Journal of Arthroscopic & Related Surgery, 23(2), 197-204.
- Yamakado, K., Kitaoka, K., Yamada, H., Hashiba, K., Nakamura, R., & Tomita, K. (2002). The influence of mechanical stress on graft healing in a bone tunnel [Abstract]. Arthroscopy: The Journal of Arthroscopic & Related Surgery,18(1), 82-90.
1. To support this notion, recent research has demonstrated evidence that pain is the product of a “biopsychosocial model” that shares contributing factors from physiological structural damage, psychological catastrophizing, and socioeconomic means for treatment and education. This fairly well-studied model makes it quite clear that all pain is at least partially “in an individual’s head” and not recognizing that fact is disregarding current science.