HIT Think

Will Electronic Records Cure Health Care?

Register now

Until we are totally confident we know how to design and deploy EHRs in a manner that will substantially improve health care, why would we want to proliferate these expensive systems?  The thinking is that EHR interoperability will solve health care's crisis. But ask yourself:  Whenever you've received inadequate care, what was the root cause? Was it (1) because your doctor couldn't access a medical record that was in some other doctor's office? Was it (2) because your doctor did not have access to the clinical knowledge that would have led to accurate diagnosis and/or effective treatment? Or was it (3) because medical science, itself, just does not know enough?

Of those three causes for suboptimal healthcare, I believe the first one (lack of EHR interoperability) is actually the least impacting. For most clinical episodes, the treating physician is not truly handicapped by not being able to see what’s in some other physician’s record of your prior care. The second one seems to be considerably more instrumental. No physician can learn all she or he needs to learn, remember all that was learned, and apply it effectively during a brief clinical encounter. So we should clearly enable access to whatever is currently known by medical science, by providing computer-retrievable knowledge at the point of care. Not to do so is just plain foolish … or professionally arrogant.

The third cause, in my opinion, is actually the most significant deficiency in health care. Medical science just does not know enough. The reason for this is that health care does not learn from its own experiences. No one is retrospectively analyzing all the clinical encounters every day, to determine the early signs of what eventually become definitive diagnoses. No one is evaluating what treatments actually work best for various conditions, and under what circumstances. Medical science only moves forward via controlled clinical studies, which are too targeted and expensive to be our only strategy for advancing the science. We need to mine the data on real-life clinical encounters--nationwide. If you doubt this assertion, think about hormone-replacement therapy.  The message here is that data interoperability, attained through a standardized clinical vocabulary, is more critical than operational interoperability.

Once we have determined, through data analyses (while controlling for potentially confounding variables), how to diagnose and treat more effectively, we must convert that learning into a "clinical guidance system", operational at the point of care. We would monitor outcomes, assuming we can figure out how to measure them, so that the system can be empirically enhanced--thereby establishing continuous quality improvement (CQI) for healthcare. That, along with systemization of health care delivery, via processes like triage and rational incentives, is the only way that we can prevent the current crisis from turning into an apocalypse.

We need to conduct pilots of alternative EHR approaches, rigorously analyzing both the financial and clinical outcomes--so that we can learn what truly works best. The point-and-click documentation requirement of most existing EHRs has ironically been demonstrated to decrease the productivity of physicians. That is the last thing we need … particularly if there are no offsetting benefits derived from improved quality and value. Let’s figure out how to do it right:  How to make data entry physician-friendly and highly efficient. Let’s bring the best minds together to design and evaluate these systems, which will determine the future of our nation’s healthcare. Let's not throw money at this devastating problem until we know for sure it will buy the cure.

Joe Weber is chief marketing officer at Valadoc LLC, which trains medical coordinators to document physician/patient visits to improve documentation in electronic health records and enhance care coordination. He can be reached at joeweber@alum.mit.edu.

For reprint and licensing requests for this article, click here.