IRBs and P-Values: Your Research Team

You’ve been called in for an emergency surgery, but you know the Institutional Review Board (IRB) documentation is due today…you also know that the person you hired to handle this isn’t as “on the ball” as you had hoped. These and other situations can arise in the time and resource-sensitive arena of research—especially if you don’t take care to hire the right team.

Dr. Jeffrey Katz, Director of the Orthopaedics and Arthritis Center for Outcomes Research at Brigham and Women’s Hospital, has experienced the gamut if not also run the gauntlet in practice of clinical research. He notes, “The principal investigator (PI) defines the research question, decides on the hypothesis, and develops a plan to test the hypothesis. But none of it is worth much if at the end of the day you don’t have a team that can carry it out. Members of the research team must have the appropriate level of experience, an enthusiasm for the work, and a broad skill set. And the value of good interpersonal skills should not be underestimated. The people you hire will be interacting not only with patients, but with staff members, all of whom may have different ways of communicating. Quantitative skills and a solid comprehension of the clinical problem you are trying to address are also critical to the success of the project.”

Dr. Katz, Professor of Medicine and Orthopaedic Surgery at Harvard Medical School, outlines the basics of a team. “Your group should include a project manager, at least one clinician, a statistician, a research assistant who is comfortable talking to patients, and a data manager.”

There are enough “unknowns” that crop up during a research project…you don’t want more surprises with the people you bring on board. “Ideally, you should hire someone who is a ‘known quantity, ’ i.e., someone who already has relationships—and experience—in the world of clinical research. That way there would not be a lot of surprises because you know their strengths and weaknesses. If that isn’t possible, turn to colleagues around the country to find reliable, talented people.”

As there is a logical progression to research, there is also a logical progression to formulating a research team. Dr. Katz: “Don’t, for example, hire the research assistant first because that person doesn’t have the broad view."

"You should initially search for a project manager (PM)—someone who is able to shift their thinking from the day to day details of the study protocol to the larger picture of clinical and policy issues. You want someone who can see the forest and the trees because you will be conferring with them about your strategic decision making."

"Additionally, the project manager should be personable and nurturing because they are often working with young people who assist with the study in various ways. While it is sometimes hard to know whether the applicant has the appropriate technical skills, you can contact someone they have already worked with to assess what they can and cannot handle.”

And make sure, recommends Dr. Katz, that the project manager is not uncomfortable with budgets. “Aside from the all important personality skills, the project manager needs the technical skills to help compile and manage budgets. It’s not necessary that they have a financial background, but they should have the ability to learn the financial side of things.”

“If you are part of a large institution there will likely be an in-house statistician whose expertise you can draw upon. If this isn’t possible, contact a local public health school to determine if one of their statisticians may be qualified for the project. That being said, technical qualifications are not enough. You need someone who has a genuine interest in clinical problems…they need to be interested in the application of the numbers they are working with.”

Dr. Kurt Spindler is Professor and Vice-Chairman of the Department of Orthopaedics and Rehabilitation at Vanderbilt University Medical School. An avid researcher, Dr. Spindler has also served as an ad hoc grant reviewer for NIH (National Institutes of Health) on many occasions. He states, “A pivotal team member is the research coordinator (project manager). Imagine if your project involves 80 surgeons at 40 sites. Each institution has its own IRB, so you must have a top notch person in order to handle the coordination of all of the related activities and documentation. It can be challenging to find someone with experience in this area.”

“Some research coordinators have Ph.D.s and some have Masters degrees; either way, the critical thing is that the person understand the concepts behind the project. Why are we conducting the study, what are the potential outcomes, what are the pitfalls, etc. In the best case scenario, the research coordinator will have some orthopedic knowledge because they will have to deal with inclusion and exclusion criteria, the clinical risks for patients, etc. This person also needs to be able to get quality data from the sites—and do so in a timely manner. Lastly, they must understand the basics of data management. In my experience, you’re lucky if it only takes 6 to 12 months to find such a person.”

Dr. Spindler puts the primary emphasis on hiring the appropriate research coordinator. He recommends,

When interviewing, ask questions along the lines of, ‘Have you helped write IRBs? What is your experience in interacting with investigators and patients? Have you collected data from statisticians?’

"The research coordinator—not the principal investigator—is the one interacting daily with all of these people. Get specifics on their prior involvement with data management. This is important because of all of the research steps and details involved. What if, for example, data from a research site is missing, or is not interpretable? This person needs to understand the checks and balances involved.”

And while Dr. Spindler isn’t advocating hiring pachyderms, he does suggest that your research coordinator have thick skin. “They have to deal with surgeons—enough said. In general, research coordinators often feel underappreciated for their many efforts. As for surgeons, they must realize that without this person the study can’t be done. And the research coordinators are often dealing with things that drive surgeons around the bend—like communicating with IRBs and completing all of their paperwork. Everyone involved in the project has pressures, but I encourage surgeons to be respectful of their fellow team members. A research coordinator is your right hand…don’t cut it off.”

Another veteran principal investigator is Dr. Christopher Bono, Chief of Orthopaedic Spine Service at Brigham and Women’s Hospital. He remarks, “If you are undertaking a clinical research project, the most critical thing is to have a clinical research coordinator who has some experience (but doesn’t necessarily have a medical background). This individual will be the point person for the study, and must be familiar with the scientific process and technical writing. They need to understand that to a great extent, the job can be tedious, with a lot of computer work and perhaps not as much patient contact as the person would like. I know of problems that have arisen when a research coordinator is brought on board, but misunderstands the job description and begins crafting their own project ideas. Finally, they must be unusually organized and compulsive when it comes to record keeping.”

“Red flags? The person you are interviewing tells you that the job of research coordinator is not difficult. Or they arrive late to your appointment and have a disheveled appearance. One thing you can do in the interview is to provide the person with a scenario and ask, ‘If you had to do this in a day, how would you do it?’”

In hiring a statistician, says Dr. Bono, be sure that the person hasn’t gotten lost in the data. “You need someone who not only has an intimate understanding of the numbers, but who comprehends the clinical goals of the study. Otherwise, that person will become distracted by the numbers."

"An example of a good statistician is someone who asks the PI, ‘What kind of a difference are you looking for in these two groups (one procedure versus another)? How much of a difference is important?’"

"Someone who is savvy in that way will know how to properly design and analyze your study. If, on the other hand, they only ask you about outcome measures and then say that you need X number of patients to achieve statistical significance, then that is someone who doesn’t understand the whole picture.”

Dr. Bono adds, “Certain projects require a research nurse, i.e., someone who conducts follow up patient examinations. This usually comes into play if the study involves a multicenter protocol for a specific exam or questionnaire (and, for example, every month each person has to undergo an exam). For this position you obviously need a ‘people person’ in order to encourage people to return each month…and they should be gently aggressive with regard to pursuing the participants.”

So to formulate a team whose parts click well together—and do so for the entirety of the project—make sure the individuals are completely clear about the job descriptions, ask the right questions, and talk to those who have gone before you.