Jay Rollins talks about the implications for IT that Electronic Medical Records could have and why the concept of Personal Medical Records may be better.


For months now the health care IT community has been busy trying to figure out where President Obama was going to spend money. There were all kinds of hints about Electronic Medical Records (EMR) and other technologies that help streamline the health care industry would be candidates. A couple of industry organizations also drafted proposals and lobbied hard for how best to spend money earmarked to build efficiency in health care.

Last week, a number was put out there. $19.2 Billion is being set aside for Health care IT. It is missing a couple of things such as specifics for long-term care and nursing homes, but for the most part, EMR seems to be a focus.

I am relatively new to the health care industry and started looking at EMR in earnest. Early adopters will likely be rewarded similarly to the E-Prescribe effort that was launched last year. Basically, doctors who used an electronic prescription system would get a 2% reimbursement bonus from medicare during the first year of the program. That percentage goes down year over year until it starts penalizing doctors who do not adopt the system (i.e., -1% reimbursement deduction if E-Prescribe is not being used). Assumptions are that early adopters of EMR will see something similar. With some larger health care companies, 2% is a pretty big number.

The challenge is having to be an early adopter. And that requires a different resource set from an IT department standpoint. Additional staff to deal with bugs, process reengineering efforts, changing standards, etc. And although many software providers have been involved in developing EMR software for several years now, there will be a rush of new products to the market that can claim they can do everything. Then you also get legacy applications that try to tweak something here or there to be “EMR compliant” or some other nonsense.

What worries me is the concept of EMR in general. Basically, you have disparate data stored by various health care providers: The family doctor, the hospital you stayed in to have your gall bladder removed, the lab that did all your blood tests when you had some infection years ago, the pharmacy, the two different insurance companies you have had in the past three years, the GI specialist, the rehabilitation facility you stayed in after your car accident…well, you get the picture. This data, specific to particular health events in your life are all over the place. Privacy rules provide kind of a limited power of attorney so that the hospital can talk to the doctor’s office and get the lab work, but only as it relates to relevant health information for the situation.

How is this worrisome? I try to equate this to data warehouse architectures. One of the the key concepts of developing this architecture is the idea of a record master. A single source of truth that brings all of the contact points into one master record and makes decisions as to what change happened first or what changes are allowed or prohibited. Did they have the mumps shot before the measles shot? Does that have an impact on the symptoms the patient is going through right now? I mean, if there are timing dependencies and business rules dealing with a customer record in a retail database, there surely is the same situation in a health record.

After researching this further, I came across the concept of the Personal Medical Record (PMR) which is maintained by the user. All the health information, family history information, previous prescriptions or health conditions are all kept in a single record. This PMR concept is really the Patient Master. No more filling out the same form over and over again regarding allergies, hospital stays, current prescriptions, etc. The patient decides who gets what information when, but at the same time, there is a huge side benefit. You get to mine the data and glean intelligence from it. Professionals, especially doctors, can make better decisions with more information. Not more data, but more information.

As I learn more and more about EMR and all the integration challenges, I am becoming convinced that the PMR approach is what is needed. The integration costs and the bugs that are going to have to be worked out are going to be large and many. EMR from a standards perspective is more mature and better documented so Washington can feel more comfortable backing EMR, but I fear they are backing the wrong horse.

Not all will be lost. Much of the infrastructure put in place for EMR can likely be used in a PMR publish-subscribe model, but if PMR is done first, the cost can potentially be 10x (SWAG) less. We will have to see what happens. There are a lot of smart people working on EMR. I just hope they are not succumbing to Group Think and enabling another $10,000 toilet seat Washington spending scenario.