Physician Satisfaction with EHRs: Finding Balance in a Digital World

Dr. Howard Landa, MD, CMIO Vice Chairman of AMDIS

As Vice Chairman of the Association of Medical Directors of Information Systems (AMDIS), Howard Landa frequently discusses with CMIOs the challenges and opportunities they face in leading informatics and analytics initiatives at their health systems.  We asked Dr. Landa to address physician satisfaction with EHRs and the emerging technologies that may help improve the value of these systems. 

While we have succeeded in providing healthcare systems with a plethora of health information technology pathways designed to improve the efficiency and quality of care, we have also managed to make patient care much more complex. At a recent AMDIS symposium on the evolving role of the CMIO in physician leadership, a discussion regarding the impact of the Electronic Health Record (EHR) on medical practice yielded interesting results.  When polled, many front-line clinicians will tell you that they have a love-hate relationship with the EHR.  They don’t want to give it up completely and return to the ‘old school’ days of paper, but the way it has impacted their workflows has simultaneously improved efficiency and caused a complete shift in the care paradigm.

What are the issues causing the physicians and clinicians to feel this way? One is the transition from Clinician created prose to standardized and structured documentation. EHR templates structure the documentation process by capturing data in drop-down boxes, checked or unchecked boxes and prefilled templates.  The templates provide consistent documentation methods, speed up documentation time and allow rapid analysis of large groups of patients via automatic analytics tools.  Yet they leave little room for interpretation regarding the nuances of the patient’s specific case.  In the analog world, a physician would dictate findings which would later be transcribed and become part of the record.  The dictation process allowed the clinician to share why certain decisions were made, the timing of the care process and “if/then” statements which could guide steps in clinical decision making.  As other care team members provided care to the patient, these aspects of the story were often useful in providing clinical context and understanding.  While the structured data provided by the usage of the EHR is very valuable from a data analytics side, we have lost the physician’s intellectual input into those notes making it difficult to “individualize” the patient.  One patient’s notes-over-time may be almost indistinguishable from each other, raising both cognitive and coding concerns.

So what can we do to swing the pendulum back toward the middle and find a balance between free prose “storytelling” and the high value of structured data?  Natural Language Processing (NLP) is one option. NLP is a practical application which can process free text that is entered or dictated into the record.  NLP enables physicians to extract the data in a more evolutionary and biologic way – i.e., the way a human brain would work as opposed to a computer.  When NLP is combined with nomenclature codification schemas such as SNOMED, structured data can be extracted incorporated into the patient’s record. Leveraging the “OpenNotes” patient engagement/empowerment movement and applying analytics to this data, clinicians will be able to improve the accuracy of the patient story and connect what the patient says with what the physician hears. The patient engagement with the process makes the requirement of clear and individualized information all the more important.

So to find balance and truly deliver the promise of EHRs, NLP and other tools will allow the care team the ability to read the notes and extract data while keeping track of the nuances of patient care due to the specificity of prose.  We will also be able to ask the systems to categorize, store, prioritize, codify, and place structured extracted data where it can be redelivered when needed while still retaining the physician’s intellectual thought processes.  Health systems can then use the findings to improve weaknesses in clinical documentation or care processes, further stratify population health risks and outcomes, provide input into clinical research and more.

Dr. Howard Landa, MD is an accomplished CMIO with a proven record of leveraging innovation, technical expertise and operational knowhow to deliver HIT solutions.  Industry recognition including 20 years as Vice-Chair of the Association of Medical Directors of Information Systems; two years as the chairman of the HIMSS Physician Community, and recipient of  Modern Healthcare’s Top 25 Medical Informaticists Award in 2010, 2011 and 2012.