Why Do We Miss The Point With EHR Research?
All in all, I think it’s safe to say that most scientific research is pretty solid. Yes, you can pick apart just about any study and point out where things could have been done differently or where the results could be interpreted slightly differently, but it’s rare that someone could [justifiably] argue that a published study should be considered invalid. Peer review, performed by other experts doing research in the same field does a reliable job of sifting out the wheat from the chaff.
Yet, it is with stunning number of researchers who deign to dip their toe into the EHR realm fail to understand how correlation of the use of a tool and enhanced quality outcomes may not necessarily prove causation. And those of us in the industry who attempt to interpret the research (perhaps moreso than the researchers) aren’t helping the matter either. There’s that issue of mechanism… To be fair, as far as researchers go, they are not the majority, but their work does seem to get an inappropriate amount of attention
Let’s take example 1 in a study funded by the ADA (Emphasis mine):
In the practices randomized for the diabetes management protocols, staff identified and referred 175 patients, who were eligible for intensified therapeutic management based on evidence-based guidelines. Access to the EHR was critical, as the CDEs [Certified Diabetes Educators] often left recommended prescriptions for doctors, who could act on them within a day.
And of course the the more telling statement of the study (Again, emphasis mine):
Dr Zgibor observed that patients receiving care from the CDEs administering evidence-based protocols were more likely to have medication adjusted quickly; changes took longer to occur for the usual care group. This is the hallmark of “clinical inertia,” and patients can suffer health effects during the time it takes for primary care to catch up with where doses need to be.
Given these two statements, why is it so hard to connect the dots between what’s really going on? First, let’s remove “Electronic” from EHR and ask if the the two statements still remain true? And of course they do. The fact that the patient’s record was electronic has absolutely nothing to do with the outcome.
Now let’s look at a study that actually proves the implementation of an EHR improved diabetic patient outcomes:
Mary Reed, research scientist at the Kaiser Permanente Northern California division of research, said the team looked at “three tightly linked steps” in the care of diabetic patients: whether or not the patient received medication treatment intensifications, whether or not he underwent follow-up testing, and whether or not a patient’s blood sugar or lipid levels were affected by the EHR.
“What Kaiser did, completely unrelated to the study, was a staggered roll out schedule for the EHR implementation,” said Reed. “They sequentially rolled out the EHR at one medical center after the other, across all 17.” This allowed Reed and her team to compare “pre and post” patient data at every medical center. “Because of the staggered nature, we were able to look at changes over time,” she said.
Therefore, in this study they already knew that treatment intensifications were what improved patient outcomes and and they could prove that using the EHR helped them identify individuals who needed treatment intensifications moreso than when they were working with paper records. [Here’s the study in full]
Yet, it’s studies like the former, and not the latter that EHR Vendors, HIMSS, News outlets, even HISTalk rally behind to show that EHRs are beneficial. I suppose it’s a good thing doctors aren’t buying it.
There are some themes emerging from the research though that are the point. Here are two items to ponder on and make the healthcare world a better place.
Communication:
Patient care improves in situations where the patient is being treated by multiple providers and those providers communicate effectively with each other and with the patient. This is not a new insight by any means. Yet, people are confusing the fact that we have really amazing communication technology in the world right now with what EHRs do. EHRs were designed to be documentation tools, not necessarily communication tools. Therefore, we need to design EHRs with an emphasis on communication instead of…whatever it is they think they’re emphasizing now. There are wonderful communication tools like drawn diagrams and graphs or even thoughtful visualizations of data sets that were completely lost in the transition from paper to digital. Why? Because no one knew they were creating communication tools.
Data Analysis:
As was shown in the Kaiser P. study above and numerous others, data analysis of large data sets can prove to be immensely helpful as long as we know what we’re asking of the data in the first place. The Big Data hype is cool and all, but humans are wonderfully good at finding patterns and relationships even though they might not actually exist. Collecting data about all we can is a noble endeavor. However, it doesn’t inherently make us better. As long as we have specific questions to ask of the data when we already know the mechanisms we need to enact, data analysis will help answer those questions. If you ask the internet to make you rich…that probably won’t happen. If you use the internet to find the opportunities you know will make you rich, your outcome will greatly improve.
Aaron Berdofe, MHI - Health Informatics 