≡ Menu

Decision Management and reforming hospitals


The McKinsey Quarterly had a nice piece on Reforming hospitals with IT investment that contained a great paragraph:

Combined with clinical-decision-support (CDS) tools that give physicians best-practice guidelines for medical procedures and with stricter coding classifications, electronic health records not only broaden access to medical information but also serve as a forcing agent to spur the adoption of standard operating procedures and best medical practice.

While the bulk of the article was on Electronic Medical Records/Electronic Health Records I appreciated this emphasis on the role of decision making in delivering the benefits of electronic medical records. Too many articles on EMR/EHR systems assert that simply storing, managing and regurgitating information about patients will improve results. Given the pressure on medical staff to act rapidly in emergencies and to minimize time spent with patients for cost reasons, simply having more information is insufficient. Medical staff need active systems, systems that will make decisions (“this test is not necessary”, “this procedure should be substituted for that one”, “this drug is contraindicated for this patient”) if that information is to be put to work.

I have blogged about the role of decision management in effectively taking medical records online before. We will not get the value we need out of these systems unless we combine integrated, available medical records with best practice and analytics-based treatment recommendations, problem identification and drug interaction/dosage monitoring and more. Healthcare decision management.


Comments on this entry are closed.

  • Tim Richardson August 23, 2010, 8:00 am

    Have you seen many articles with published logical rules – I’ve read articles that show the RESULTS of the published rule (eg: …a 20% decrease in improper/inappropriate antibiotic ordering…) but most of the articles I’ve seen do not publish the actual rule (eg: If-then statements that match signs/symptoms to decisions).

    I’m in the position of recommending modifications to an EMR for an orthopedics physical therapy practice and I’m trying to find published examples of the actual rules that I can hand to the software guy who will write them into the program.

    Thank you,

    Tim Richardson, PT

    • James Taylor August 23, 2010, 8:16 am

      Thanks for the comment. Probably the easiest place to find some examples is on Corticon’s RulesWorld site. Check out the healthcare section. You will need to register (it’s free) but you should find something useful.
      Let me know if not

  • Nick van Goor August 23, 2010, 12:50 pm

    Hi James,

    Last week I visited with a friend of mine who works for a healthcare system here in Minnesota. He showed me their “brand new” EMR system, which went live late last year. I was shocked at how archaic and un-userfriendly the system was. In fact, he told me that since the new system was installed, provider productivity had gone down 15%. This is a system that was five years in the making (which might be one of the problems).

    I’ve become more and more fascinated with decisioning in medicine. I’m in the process of reading “How doctors think,” by Jerome Groopman, which is quite enlightening and provides insight into how we might help doctors use analytics. I think it isn’t that analytics is distrusted by doctors, it’s just that traditional analytics doesn’t fundamentally solve their problem: the cost of a false negative is catastrophic.

    I wrote a blog-entry today on some other thoughts I had on decisioning in healthcare, and would value your opinion.



  • Mark Eastwood August 25, 2010, 10:22 am

    A few years ago I worked with a larger player in the space of electronic health records/hospital information systems and was shocked to learn that while they embed a rule engine, they don’t really use it. A few of their clients have since gone on to make use, but at the time only a very few.

    To me the idea that decision automation could alert a doctor or other hospital staff of potential problems with a patient or to prevent errors relating to patient care is a big win. We’re not replacing doctors or nurses, we’re helping them do a better job. Seeing an alert that patient XYZ has had a change in some lab result by greater than x% could be life threatening. Helping the medical team to react more quickly or creating some kind of “are you sure you want to give this patient more of drug X” might just save a life. How many kids die each year because they were given the wrong dose of medicine? (Too many I’m sure.)

    Diagnostics is another area that’s very difficult; yes I do watch the TV show House. I can imagine decision support systems for doctors that would support diagnosis, prompt them to run a particular test or examine this result or ask this question.

    Lastly I know that there are publications listing the current care standards for many conditions and diseases. What if this could be tracked, match up a doctor’s performance against care standards and create a “report card”. It would point out a need for professional education or changes in their current practices. Hospitals and insurance companies might use the data to compensate doctors for “better grades”. Decision support systems could help doctors perform at high levels of proficiency even with large patient loads.