Wednesday, January 20, 2016

Measuring Impact as Clinician-Scientist: Personal Insights

Over the past few months, I've been thinking about how to have impact through research as a MD. For PhDs, it's really no brainer because most of their time can be spent thinking about high yield research topics; their creativity and time are rewarded by grants and scientific papers.  It's not easy to be a great scientist but the incentives and path are mostly clear.

However, I think it's not clear for MD-scientists because research time comes at the expense of clinical productivity. Every day I spend doing research is one day loss generating revenue for the hospital...that translates to about $1000-2000 /day of direct salary loss and about 10-15x more in loss revenue for the hospital.

So, what happens to research? I accept that research is the cornerstone for creating new knowledge and innovation which give institutions a competitive edge.  Personally, I love clinical research and can spent hours and days mining data, recognizing new patterns, writing papers, etc. I did that for too many years. So, what happened? I fell behind. I was slower than my colleagues because I didn't read as much volume as them. Yes, I did published over 15 papers and gave over 20 talks over the past four years but who cares? How did my research actually have positive impact at the expense of my clinical practice? Every hour I spent working on a manuscript generates one hour of research productivity, but at the expense clinical time. What is the point? Why should I continue this when it's a lose-lose situation for me?

This cynicism has been the impetus for the evolution of my thought process of how to have impact as a radiologist-scientist without compromising clinical productivity or proficiency.  I think I finally have an answer. The first question I had was "How do you measure impact?". In the past, the measure had been number of papers I published. But, if I want to have maximum impact without losing clinical revenue, then publications are totally the wrong measure because I get no direct compensation for my time AND I lose out on clinical revenue for the hospital!! So,  I decided an alternate way  research can be impactful for clinician-scientist (without compromising clinical proficiency and productivity) is to adopt a business model.  Is the research valuable enough that people will pay for it?   

Developers will not waste time on writing software if nobody sees value and are not willing to pay for it! The entire business world understands this...and I think one way to be an effective radiologist-scientist is to adopt this core business principle. I know this is very cynical because some pure radiologist-scientist like one of my mentors believe we should measure impact by how happy our research makes us. That's a purist belief and I accept his perspective; but I think I'm more a pragmatist than a purist.

So, I've decided I'm going focus on research which can have potential commercial value, like developing software and tools that people will find value in and will pay for. I can still do research to quantify the incremental benefit of these tools and software and publish it. But, at the same time, the tools will meet people needs, the revenue generated can be invested further into research projects, so that the impact can be compounded and scaled at a logarithmic instead of a linear scale. By recently adopting this model, I've started to appreciate how much easier it is to draw developers into my research projects. It's a win win because I can attract super smart scientists into my area of medical research and together create something new and better than existing standards.

Having recently adopted this new approach, I feel like I can spend my time and talents in clinical work, which I love and is more high yield given my training; I am now as clinically productive (if not more) than my colleagues;  yet, I'm also able to be freed of the responsibility to be on the front line doing the day to day research work. I can collaborate with developers, students, and trainees to provide clinical insights and guidance to develop my research ideas with potential commercial value. Interestingly, this new approach has forced me to reach out to people with different background (i.e. data science, computer vision, etc) and as a result, has increased the quality of my research projects.

If the research has practical value, publications will automatically come.  I think the measure of impact should be how much value your target audience perceives the products of your research, rather than number of papers you can get from it. I think this perspective will save me and others like me from burnt out and in turn, will permit more practical,  high yield, feasible and sustainable careers  for clinician-scientists.

No comments:

Post a Comment