Systems Thinking

I recently read an article titled “Cowboys and Pit Crews” by one of my favorite writers, Atul Gawande.  I was turned on to Gawande’s writing, specifically the books Better and Complications by son Derek.  He is in his fourth year of medical school and apparently they assign some of his writing to the students.  Gawande writes insightfully particularly around policy and process issues in medicine with the objective of improving the overall system – training, practice, etc.

The main point of the article is that medicine has evolved from a field where individual doctors “could hold all the key information patients needed in their heads and manage everything required themselves” to one where clearly that’s not possible.  Anyone who has experienced even something as common as a torn ACL can think about the number of people who take care of them from internist, radiologist and orthopedic surgeons to the MRI tech, nurses and physical therapists – easily more than a dozen.

Success then results as much (or more) from the system working properly as from the individual e.g. the surgeon performing well.  The places that have the best results for, say cataract surgery or ACL replacement are well run systems that, of course, have competent surgeon members. Having grown up in a family of physicians (both parents, 2 siblings, many cousins etc.) suddenly what they do seemed to more and more like what I and members of my team here at SugarSync do.

In managing a system, the first thing you need to do is know your goals and measurement criteria then set up measurement systems to know if you are meeting those goals.  Our VP of Engineering Brock LaPorte popped into my head as I thought about systems engineering as this is his primary mode of operation.  Sure – he can solve individual technical problems but his main contribution is making sure the systems – hardware, software and people, function well.    That always starts with instrumentation – without that data you can’t improve the system.    This is part of the engineering discipline both academically and in practice.  As I’ve chatted with my two sons in medical school – this is notably absent in their training.  It might be part of advanced degree programs in public health but not medicine itself.

Any experienced, successful manager of a major internet service builds in instrumentation capabilities from the start.  They needn’t be expensive – at SugarSync most of our measurement tools are open source – built and contributed over the years by practitioners.    As much as possible is captured in the logs – even data that may not be analyzed at the time – for future use.  With the digitization of health care records – more data should be available for measurement and analysis to allow us to have better functioning systems.  Hopefully medical educators will include these key skills into the training.  I believe there is an opportunity for more cross-pollination with industries such as internet services that do this well.  The stakes are high both in terms of health and dollars.

2 thoughts on “Systems Thinking

  1. Well written, but I think you are missing some insights into “real medical practice” The digital age has certainly been embraced in an overall sense by the field of medicine. It was decades ago that computers were introduced into radiology to produce the CT scan…..a computer used to take apart radiographic body-slices …. and reconstruct the images to create anatomically correct images of the body….and revelation of normal or pathological findings.

    Ar some point in the future….probably many years away….electronic medical records (EMR), also referred to as EHR (electronic health records) will have real meaning in terms of making an individual’s health information available to “systems” anywhere. And, at that point in the future….the information contained in those data banks will have meaningful use. That is certainly not the case at present. There are multiple companies that have created the programs to create a product called the EMR. The big problem is that basically non of these systems can talk to each other….and absent that capability…..all we can say is that at a given location (hospital,physician’s office,ambulatory surgical facilty) the patient’s record has been digitized. Bottom line….there is still a very long way to go. The underlying problem here does not lie with the medical school.

    My two grandsons in medical school….both in their clinical years….are missing the summation of all the training. The clinical years are broken down into training in the various disciplines …… and so you see a clump of trees here, and another clump of trees there…..but you don’t see the entire forest for what it is.
    You study the fragments…such as medicine…surgery…ob/gyn….psychiatry, etc, but at the level of advancing medical student it is difficult to tie them together into one “system” as you portray it. But it does not take an advanced degree in public health to do it.

    The integrated system you are looking for is called “a well trained, concerned, and caring primary care physician”. It is the good family physician, or internist, or pediatrician that embodies the “system” you are looking for. These individuals gather the information…..the data if you will….to diagnose….treat…and usually cure the patient. The processor, if you will, is not a product of Intel…but the working of a brain taught to gather relevant information….compile the information….analytically arrive at a diagnosis….and produce the treatment and the cure.

    I love computers……I love the digital age….I embrace the technology, and always am eager to see what lies around the corner. But make no mistake…..the systems that we have in medicine are good….they work. I don’t believe the intricate, differential thinking we in medicine….particularly primary care….will ever be replaced with machines. They can help juggling data, retrieve archived information easily…..but they will always function along side us…never instead of us.




    • I agree that the primary care physician is the ultimate leader/manager of the system and that this human role is not replaceable with computers. I do believe, however, that there is an opportunity to better instrument the systems with more complete data.

Comments are closed.