We’ve all seen the spectrum of research that’s published nowadays, whether it’s the good, the bad or the straight-up irrelevant. Producing quality, clinically-significant research is the backbone of growth in any clinical field, which beckons the willing and qualified to begin producing studies that will make an actual difference.
After being heavily involved in all things research for over 30 years, Phil Page PhD, PT, ATC, CSCS, FACSM knows a thing or two about what it takes to compile an impactful study. In the first edition of the Journal of Performance Health, he shared his expertise in an article titled “Clinical Research 101.” Today, we’ll be featuring his tips and tricks to help polish your clinical researching skills so you can make worthwhile contributions to the clinical world!
The research process: it’s elementary
More often than not, the process you will follow throughout your clinical research is one that you learned in grade school: the good ‘ole scientific method. Not only must each component of the method be completed to perform an adequate study, but they must be done in the correct order:
- Research question
- Research design
“The research design must adequately address the research question, and the statistics must satisfy the design. Therefore, these components of the research process are all interrelated in order to answer the research question” (Page, 2016).
Got your research question and your hypothesis? Great! Before you barrel through the rest of the steps, let’s break them down to ensure that you’re performing each task with enough diligence and detail to produce a reliable report.
Choose the right research design
When deciding on your research design, you’ve got a couple of options. To start, you’re going to identify the type of study you’re conducting: descriptive (analytical) or experimental (clinical). Descriptive studies are built on existing variables that simply need observed, while experimental studies rely on manipulative variables.
Next, you’ll need to choose the level of evidence your study falls under. The Oxford Center for Evidence-Based Medicine (OCEBM) has provided classic levels of evidence from one through five. However, Dr. Page suggests modifying these five levels by discerning between descriptive or experimental designs:
Dr. Page notes, “The level may be graded up or down based on the quality of the study; however, in contrast, the level should not be based on effect size. A small effect size from a well-designed study may be clinically meaningful” (2016).
Concentrate on the methods behind your madness
Let’s say you’ve chosen a clinical research design. It’s time to buckle down and choose a research method: the most important component of a research report! This detail-driven section will provide several key pieces of information:
- Were the methods appropriate to answer the research question?
- Can they be reviewed, critiqued and analyzed?
- Do they allow others to replicate the study to validate and collect additional data?
- Can they be translated into clinical practice such as a clinical intervention?
- Is the study relevant to specific populations?
Need additional guidance? Remember to dip your clinical research design into some “PICO” to make sure you’re addressing four key components:
- Population: inclusion and exclusion criteria, as well as sample size
- Intervention: specific details that support replication
- Comparison: what the intervention is being compared to (or control)
- Outcomes: the outcome measures, including validity and reliability, as well as clinically important differences
Finally, identify the reporting guideline you fall under:
- CONSORT: Randomized clinical trials
- STROBE: Observational studies
- STARD: Diagnostic studies
- PRISMA: Systematic review or meta-analysis
Identify relevant statistics
Got your data? Good. Now it’s time to do a thorough analysis of the outcomes to extract the important information. “Traditional statistics don’t always provide important clinical interpretation. Statistical significance only indicates if the researcher made the correct statistical decision regarding a hypothesis. More clinically relevant and useful analysis should be included” (Page, 2016).
Make sure you’re including the following variables to report with full transparency and determine if the study provided statistical significance:
- Effect sizes: How large the difference is between groups
- Can be classified as trivial (<0.2), small (0.2-0.5), moderate (0.5-0.8) or large (>0.8).
- Confidence intervals: A range of estimates containing the true value within a population
- Minimal Clinically Important Differences (MCID): How clinically meaningful the differences between groups or after an intervention was
- Example: Dropping two points on a pain scale of 0-10 would be clinically meaningful, whereas a drop of one would not, even if it’s statistically significant.
- Statistical power analysis: A minimal number of subjects to detect a difference between groups, indicating that the study is adequately powered for statistical analysis
Interpret with intelligence
After you’ve done your statistical due diligence, it’s time to clinically interpret the results of your study. Here are some key points to cover:
There you have it: a well-structured research study! For more detail and how to structure your research paper, read the full article in the Journal of Performance Health. Have other research questions you need answered? Comment below and Dr. Page will be happy to share his wealth of knowledge!
This study was featured in an article originally published in the Journal of Performance Health written by Phil Page PhD, PT, ATC, CSCS, FACSM. Read the full article now or take a peek at the entire Journal here!
Page P 2016. Clinical Research 101 J Performance Health 1(1):55-59