Progress Report: Knee-Deep in Theories, Workflows, and Software Design

You are probably wondering why there have been so few blog posts in the last few months.   Well, it’s because I ran into a classic chicken-egg problem.   With the best of intentions and an optimistic outlook, I set out to write a handbook on clinical workflow analysis and workflow application development. The plan was for a 15-chapter book with the final chapter being a workflow application.   The problem appeared in chapters 9-12. These chapters were to include methods and concepts taken from my workflow research over the past few years.   However, try as I might, every version of the Chapter 9 that I tried to write eventually came to a halt because it became clear that more background information was required than could reasonably fit in a practical handbook.

For years, I have been wrestling with the idea of a theory that would bring together clinical work, clinical workflow analysis, and clinical care software design.   That research (working title, A Theory of Clinical Work and Clinical Systems) has been completed. I decided against taking the time to write a complete account of the theory with a literature review, etc. and instead chose to add relevant parts to the handbook. When I started the handbook, explaining the theory in a few pages, then moving on to the practical aspects of analyzing workflows and creating applications seemed reasonable.   Regrettably, every attempt to do that fell on its face.   There were too many times when I would have had to ask the reader to accept a concept or term without knowing its complete origin or why I thought it superior to other approaches.   No answer appeared.

Here, the chicken-egg problem arose.   I was gung-ho to do the handbook, but did not want to take the time to write a theory paper–after all, I am in business, not academia. Of course, it was quite clear that publishing the handbook without having provided an accounting of the methods and concepts being expounded would have been a disservice to handbook readers. I was stuck in an infinite loop of not wanting to write the paper and being eager to write the book, and the book depended on the paper. I wasted weeks on this, and finally giving up two weeks ago, decided that the theory paper—sigh—had to come first.

Things would have been bad enough if only the book were affected. Oh, to be so lucky.   I was in the middle of a tutorial series on Clinical Workflow Center (a patient engagement workflow project) and was about to start the clinical app tutorial on EHR Science. Both came to a halt because I realized the concepts that would eventually appear in each would depend on elements of the theory as well.  You see, the theory has changed my thinking about clinical workflows and software design to the point it is now normalized for me, thus what makes perfectly good sense in my bubble would be foreign to blog readers. No way around the blog dependencies has appeared either.   So, as you can see, everything came to a halt. Unfortunately, there is more to the story…

The theory developed over a number of years, making significant strides after I discovered the workflow research of Wil van der Aalst. It began as a means of guiding the design of clinical software, but with the influence of van der Aalst’s research, it morphed into a way of looking at clinical work, creating another problem: two seemingly independent theories with significant overlaps.   For a while I was comfortable with the overlaps, but eventually they seemed inelegant.   Of course, this meant taking time to see why they overlapped, and if that was necessarily so or simply bad framing.   As it turns out, it was bad framing. Where does this leave me?   With everything being held up by the theory paper.

The good news (I choose to look at it this way) is that version 3.0 of the outline for the theory paper is just about complete, except it will be longer than a paper, probably (hopefully) fewer than 100 pages.   The literature review is nearly complete. Having regularly scanned the literature over the years (many of papers and books are listed on EHR Science and Clinical Workflow Center resource pages) has helped immensely.   Already, five technical reviewers have agreed to read the theory paper/monograph, so it is nice to know others are interested in this work.

I am not looking forward to writing a 100-page document with citations, but I now accept it as being the only mentally healthful way to proceed. On the bright side, between answering critics, further explaining the theory, and demonstrating its practical applications, I will not have to dream up blog topics for years to come.

So, dear reader, if you are waiting for the handbook, it will appear quickly, once the monograph is done.   If the theory monograph is giving you the tingles, expect it in a few months. As for my next blog post, anywhere—who knows…


Leave a Reply

Your email address will not be published. Required fields are marked *