Every day, scientists conduct important research that brings real value to the medical field. And the most impactful efforts typically begin with a clear question that will add a new perspective to the current knowledge that exists, as well as a well-designed study format, explains Jamie Robertson, PhD, MPH, director of Innovation in Surgical Education at Brigham and Women’s Hospital, associate director of Harvard Medical School’s Global Clinical Scholars Research Training program and co-director for the Foundations of Clinical Research program. 

For instance, perhaps you notice that your patients who go running regularly have lower blood pressure than their counterparts. But, is running directly related to reduced hypertension rates, or could these patients also be making other lifestyle choices that are impacting the equation? Fully understanding the relationship can be valuable to help you provide good advice to your other patients to help them control or prevent hypertension. Yet in order to find out the connection, you’ll need to come up with a solid research question and conduct a study that will help you to understand whether running is related to blood pressure.  

Step 1: Develop a Targeted Question 

To develop a meaningful research question, it’s important to narrow your interest to a very specific area.  

“Often people get excited and want to study a big topic,” Robertson says. This could be something like the factors related to hypertension among adults of all ages. “But you can actually do your best work when you have a concise and focused question,” she points out. Narrow in on a specific subset of the population and make sure that what you want to answer is complex and not too obvious. An example might be looking at the impact that running three times a week has on the blood pressure levels of middle-age patients with no other confounding health factors. 

She points out that when you develop such a question, it’s important that you make sure that the data you seek to collect will be adding a new perspective, rather than just repeating information that already exists.  

“Think about why a medical student or resident should care about your research. It should not be something that they can already find in the medical literature or that they can answer just using common sense,” she adds. 

Step 2: Include the Right Elements 

Once you have an idea of what you want to tackle in your research, Robertson suggests using a method called PICO to flesh out the research question and all the nuances involved and determine how best to go about answering it.

Each letter of PICO stands for a different topic area, she says.  

• P=Population: Who are the people being studied? In this case, it might be runners between the ages of 35 and 55 in good health. 
• I=Intervention: What is the action that you want to explore? This could be the impact of running on hypertension. 
• C=Control: What is the other group that you will use in order to compare your results? You could use a group of people in the same age range with similar health status overall but who do not run.
• O=Outcome: What are the results? This would be the blood pressure levels.  

Using this method to define all of the elements can be essential in helping you focus your efforts and determine how to tackle the research study most effectively. 

Step 3: Explore Study Design Formats 

The next step is selecting the study format you want to use to gather your data. “People often ask me what the best study design is to use for their work. But there is no one right answer,” Robertson says. “We tend to think randomized clinical trials have the highest level of evidence. But that is not the right study design in all cases, and it may not even be ethical, or feasible, or cost effective, in some situations,” she stresses. Therefore, you’ll want to explore different options to see what fits best for your circumstances.  

Here are some common study designs that you might want to consider, along with ideas of how each one might work to answer the question about running and hypertension: 

• Randomized control trial: If you want to determine the relationship between running and blood pressure, using a randomized clinical trial can minimize the confounding factors and provide level 1 evidence while controlling bias. Yet, this method will be expensive to perform and will require an extended study period. There is also the question of whether participants will be compliant over the long term—in this case, continuing to run three times a week.  
• Prospective cohort study: A prospective cohort study would enable you to follow one group that runs and one group that does not and see how their blood pressure responds. No intervention would be needed. The drawbacks with this method, however, include that it takes a long time to see the results, and it would be hard to know if the runners will continue to run over time. Non-compliance would negate the results. 
• Case control study: “For this type of study, I’d go into the records and find patients diagnosed with hypertension and patients who don’t have it. I would pick groups with details that match except for this diagnosis. Then, I would look back and see their exposure to running,” she says. You could interview participants and find out how many times they ran five years ago, and also look for their exercise data using fitness tracking apps to get this information. The key is that for a case control study, you will select participants based on their diagnoses or health status and then look at their exercise history to see if patterns exist. 
• Retro study: “For a retro study, I would look at fitness records from 10 years ago [perhaps from a fitness tracker or app] and divide participants into two groups—those who ran three times a week and those who did not run. Then, I would go forward in time to the present and see if they now have hypertension,” she says. 
• Prospective case control study: Another less common option is a prospective case control study: “In this case, you would wait for people to get the disease or outcome of interest and then enroll them in your study. So, I would hang out in my clinic and wait for people to be diagnosed and then enroll them instead of getting them from the medical records,” she says. 

Contributing to the Field 

Robertson points out that when you consider all of these aspects to help you design and implement research studies to answer well thought out and informative questions, you can help to advance your research career in exciting directions, while also adding a significant body of knowledge to the field that may lead to improved outcomes.

Written by Lisa D. Ellis