Electronic Health Reporter

Guest post by Jonathan Bertman, founder and president of Amazing Charts.

The big news out of HIMSS13 was no surprise to those of us who work in the electronic health record (EHR) industry. The results of a two-year (2010-2012) survey from American EHR revealed that user satisfaction levels with EHRs are dropping in multiple areas – very bad news for EHR vendors.

The results didn’t surprise me at all. I talk to hundreds of EHR users each year about their concerns, and have divided the sources of their satisfaction into three broad areas. Users talk about their dissatisfaction with unusable systems because of poor design; unaffordable prices; and the inherent unfairness of the purchase process, often requiring multi-year contracts committing clinicians to a system they have never had a chance to use in the real world.

In multiple surveys over the past decade, users have given low marks for the usability of even basic EHR functions like order entry and results reporting.[1] More complex EHR functionality, such as clinical decision support, receives even lower grades from the clinicians I talk to at continuing medical education conferences around the country. But all of us agree that clinical decision support is vitally important because across most domains in medicine, practice has lagged behind knowledge by at least several years. Clinical decision support can make us smarter clinicians by bring global medical knowledge to the local point of delivery to facilitate improved patient outcomes.

Clinical decision support has many different definitions, but I like to explain it as evidence- and rules-based recommendations delivered to the clinician in the exam room when seeing the patient. Broad examples include:

• Contradiction alerts to let a clinician know that a particular medication or treatment is not recommended for a particular patient given their medical history,
• Suggestions to help a clinician come up with ideas as to how to manage or treat a patient’s condition,
• Automated data interpretation to help a clinician determine whether a fluctuation in readings is within or outside of a typical patient.

Reasons for dissatisfaction with EHR-based clinical decision support systems (CDSSs) range from the presentation of irrelevant information, to problems with workflow integration and lack of intelligent information filtering. Today, however, the opportunity exists to build an intelligent EHR system that delivers relevant clinical decision support information when and where it is needed. The basic concept is simple: use big data analytics of patient health information from EHR charts to deliver smarter decision support for clinicians.

Here’s one scenario for how this could work: a clinician opts-in to give his/her EHR company permission to access, de-identify and analyze her patient data. The data set is then compared with data sets from thousands of other practices. Based on population, demographics, public health trends, and more, we could build a profile of the clinician’s patient panel. The clinical decision support system could then be customized and optimized to handle the particular gaps in care found in his or her practice, such as the percentage of patients with diabetes who do not meet current HbA1C target goals. The information and recommendations delivered during patient care would be more relevant to the individual practice, and the“knowledgebase” of the clinical decision support system would be continuously updated with latest global medical knowledge to ensure further relevance.

As long as we ensure that personal health information and other practice data are never used without full disclosure, transparency, and permission, then smarter EHR-based clinical decision support systems will be a boon to both healthcare providers and their patients. If the goal is to help clinicians make good choices at the point-of-care, this will be one the best investments we can make in EHR systems.