Opinion: Is the right EHR even out there yet?

An article I read today from Healthcare IT News talked about the problems yet another CIO faced when trying to select an EHR for his organization. The author of the article went on to pose the question, “Is the right EHR even out there yet?”

In one word: No.

Sure, there might be a really good EHR for small to medium pain management facilities, and a really good EHR for 1-3 physician family practices out there, but is there an EHR that is the right EHR for any given specialty or facility? No. You can’t place all the blame on the EHR vendors, though. With healthcare legislation playing a bigger role each day, I’m not surprised the “right” EHR isn’t out there yet.

Imagine you were trying to build the perfect EHR just 5 years ago. You probably thought you were coming up with all these awesome features to build into this perfect system: electronic order entry, letting doctors speak into microphones that transcribe their voices into text, a picture of the patient built into the system so you were sure you were documenting on the right patient chart, and ICD-9 codes built into the system! Maybe, for just a short moment, you had a nearly perfect EHR. But just a few years later, you learned that your EHR now had to track patient’s smoking status, keep an up-to-date problem list with current and past diagnoses, perform drug-allergy interaction checks, and a handful of other (Meaningful Use) measures. You spend lots of time and energy coding all these new features into your system, when lo and behold, a whole new list of things your must have are added. Now you’re coding in a feature that lets patient’s see their records (patient portal), and standardizing all your information into a format that can be sent to other organizations. Throw into the mix an entirely new diagnosis coding system (ICD-10) and some payment reform measures which must be considered, and now you’ve dug yourself quite a deep hole. Your once-perfect EHR is now a mess of data with checkboxes everywhere to cover all the bases. You haven’t had any time at all to consider usability so providers aren’t having to spend extra time finding all these fancy new checkboxes, and you’ve been throwing expensive new updates at them every few years.

I think its easy to see why the “right” EHR isn’t out there yet.

I do believe, however, that all these changes with regard to healthcare legislation will have a large impact on the number of EHRs able to maintain CCHIT certification, in turn reducing the number of options for CIOs and other decision makers in EHR selection (not saying that makes it any easier to make your selection…).

In reflection, this is something I hope to see in my professional career, the emergence of one or a select few EHRs that just, “have it all.” I would love to be part of a project like this; collaborating with an extensive team consisting of programmers, physicians, specialists, legislators, coders, and a dozen more healthcare role-players to research and start building the “right” EHR.


IBM’s Watson: Clinical Decision Making in Healthcare

I came across an article today that talked about IBM’s powerful thinking machine, Watson, and how “he” could be put to a practical use in the real world (aside from beating Jeopardy grand champions). One new possibility, which is already becoming a reality, is to have Watson help with clinical decision making in health care. According to IBM, only 20% of the knowledge physicians use to make their diagnosis and treatment decisions is evidenced-based. As a patient, that scares me. This means 80% of the physicians decision is based on intuition and prior experiences with somewhat similar cases. As a result of this, IBM states that 1 in 5 diagnoses are incorrect or incomplete. According to the Information Week article, this is the latest on Watson’s oncology learning diet:

Over the last year, Watson has been trained on more than 600,000 pieces of medical evidence and two million pages of text from 42 medical journals and clinical trials in the field of oncology. Sloan-Kettering has added details on 1,500 lung-cancer cases, training the technology to interpret physicians’ notes, lab results and clinical research on specialized treatments based on the genetics of tumors.

All this information has been added to Watson’s brain in just the last year. At this pace, Watson will soon have every published medical journal, clinical trial, and fragment of medical evidence stuffed into his mainframe.

IBM has developed an iPad app that can be used to find the best cancer treatment pathway for a given patient. Watson takes into account everything in the patient’s chart, and the physician can add new complications or conditions to the patient’s chart right within the app (they don’t even have to type it, Watson can comprehend your speech!). A short video on IBM’s website gives an amazing overview of just how powerful this app can be when diagnosing patients and selecting evidence-based treatment plans. Watson even gives a percentage indicating the level of confidence he has with each answer. Here is the video showing how the Watson iPad application works. I think it’s awesome, and in the future will be an extremely useful tool for providers to assist in clinical decision making.