Residency and fellowship are an exciting, extremely busy and a foundational period of one’s career. Having a large number of high quality experiences that strengthen one’s career foundation during this period increases the chances of long-term career success. Conducting research in residency can be a fulfilling effort with effective planning, time management and good mentorship. In this discussion, I will share some basic concepts and thoughts, along with a few examples. I have previously discussed many of these concepts and thoughts as a faculty expert on the NEJM 360 Resident Forum last month.

Prework:

To achieve research success during residency or fellowship, some prework is necessary before beginning the actual research project(s). The prework includes taking stock of one’s career goals and interests and abilities, a survey of one’s work environment for potential mentors and potential research projects, and identifying a research theme and mentor(s). There is no need to do this alone. One may do this prework with trusted buddies, faculty members, and program director and potential mentors.

1. What is my specialty and track? Examples of dedicated tracks are the physician scientist, quality improvement, and a subspecialty track like the infectious diseases fellowship healthcare epidemiology track.

2. What are my short-term (the duration of residency/ fellowship), intermediate-term (residency/fellowship plus 5-10 years), and long-term goals (duration of one’s physician career)? It is okay to know the general direction for the intermediate and long term, but it is helpful to have crisper goals for the short-term.

3. What do I hope to gain from conducting research in residency? E.g., research publication(s) to help gain entry into a specialty fellowship of interest, build on research path already begun in medical school, foundation for a long-term academic career, try out research in order to figure out whether a career that includes research is a right path for you.

4. What research question(s) am I interested in? What are some potential mentors working on? Are the questions aligned with my career goals and interests? It is okay to be open to work on a research question that a mentor is working on. Sometimes, one may choose to work on clinical case reports, case series, or projects related to system science, aka, quality improvement or healthcare delivery improvement, instead of conducting a research study, but these principles still apply. You don’t need to come up with the research question yourself. You can fall in love with a mentor’s research question or a question that someone else suggests. It is important that you and your mentor see value in pursuing that question, and know how the research adds new knowledge to what is already known in the field. FINER is a good acronym to use. The question must be Feasible (answerable with proper design of a research project), Interesting, Novel, Ethical and Relevant. My own fellowship mentor told me something that I have repeated to every mentee who has worked with me. “Make sure you love the research question sufficiently. The love needs to last at least until you finish the research study and get the manuscript accepted for publication. Ideally, interest in the line of research will last a lifetime.”

5. Do I want to work on a project from start to finish? Am I okay with contributing to a portion of a research project?

6. What skills do I currently have (e.g., knowledge of study design, excel, access database, biostatistics skills)? Do I have basic understanding of the process of conduction a research study from start to finish? Am I trained in research methods?

7. Once there is clarity on the above questions, it is important to identify a primary mentor, and perhaps additional mentor(s) based on your learning needs and the project needs. What do I want to learn from the mentor(s)? Do my mentor and I have sufficient team chemistry? Mentoring is a two-way relationship that is a joy and privilege, but not an entitlement. Guidance on how to identify a good mentor is included in the suggested additional reading at the end of this discussion.

Steps of Conducting a Research Project:

Once the above pre-work is done, the steps of conducting a research project are less complicated. My areas of strength are observational clinical research and clinical quality improvement project – so, I will describe the steps here. Making a realistic and detailed project plan with timelines, and skillful execution are necessary for success. The steps are:

1. Design a tangible research project around the research question that you have identified with your mentor(s) and assemble the research team. There should be a perfect alignment between the hypotheses, specific aims and the research methods. The methods should stand by themselves regardless of whether the answer to the research question is something we expected to find or not. It is good to have an idea of the end product in mind. Discuss an outline of the future manuscript with your mentor(s) and the research team, while keeping it flexible and open to change until the research study is done. Pay attention to potential tables and figures. These steps will help with focus. Identify three destination journals that might be interested in publishing your work, and also identify a couple conferences where you might present this work.

2. Set up frequent meetings with your mentor(s) to review progress, troubleshoot and discuss literature on the topic. Remember that there is no such thing as a stupid question.

3. Obtain IRB approval or if it is quality improvement study, obtain a letter of non-research or exemption from review. I have previously discussed the differences between QI work and QI research.

4. Set up data collection tools and a variable dictionary. Even the most sophisticated data analyses cannot overcome any flaw(s) in the data collection stage. So, time and effort invested in this stage is worth every bit. REDCap is a great tool for collecting data, if you have access to it. Data collection should be consistent throughout. Remember that you can combine data elements and categories in future, but not divide them further. E.g., if you count the number of days, you can calculate number of weeks later on, but may not be possible the other way.

5. Format the data and make it analysis ready. Work with someone trained in statistics for this step. Some find this step daunting. Even if you do not have grant funding to pay for biostatistics help, including a biostatistician on the team as a research partner and collaborator early on, especially at the study design phase will prevent future problems. Some using secondary data sets (i.e., data not specifically collected for the purpose of answering your research question; data registries and patient medical records are great examples) might find that data need to be reformatted and reoriented to help answer the research question at hand.   

6. Analyze data. Do this step by yourself if you are sufficiently trained. Your mentor might do this for you if he/ she has training. Or give the analysis ready data to a biostatistician.

7. Discuss interpretation of the data analyses and summarize findings.

8. Present/ Publish. Do seek guidance on how to make effective poster and podium presentations, and how to write effective research papers.

Suggested additional reading:

Introduction to Research in Residency: https://resident360.nejm.org/content_items/1841

It’s a Story, Not a Study: Writing an Effective Research Paper. Lingard, Lorelei PhD; Watling, Chris MD, PhD. Academic Medicine: December 2016 – Volume 91 – Issue 12 – p e12. doi: 10.1097/ACM.0000000000001389

Suggested additional reading on mentorship, finding good mentors, and being a good mentee:

PC: http://www.clipartpanda.com/clipart_images/do-some-research-clipart-59353453