Is Epic! Epic?
A look into the short comings of modern day electronic health records and how the developers can turn things around
Looking through the history of medicine, there are few technological advances which stand out as truly transformative events. The discovery of diethyl ether as a general anesthetic in 1846 would take surgery from a hectic horror show full of screams to the operating theatre where increasingly complex and invasive techniques could be invented.[i] The discovery of Penicillium by Sir Alexander Fleming in 1928 would make small skin infections a treatable nuisance rather than a life-threatening conditioning.[ii] These discoveries and others like them have shaped the path of medicine and have disrupted entire fields of study. Flashing forward to present day, many have looked to electronic health records as medicine’s next disruptive technology.
The idea behind the electronic health record (EHR) is to collect a person’s health information over the entirety of their life and to store it in a searchable database. The end goal is to reduce medical errors, provide decision support tools, increase communication between a patient’s providers, and offer a platform for efficient research efforts.[iii] Although EHRs were originally developed in the 1960s, the uptake among physicians has moved at a turtle’s pace. In fact, in 2006 it was estimated that only 10 percent of physicians were utilizing an EHR. Despite this, the past decade has seen rapid growth in adoption of EHRs, driven in part by government legislation and in part by the increased usability of available software.[iv]
Though the increase in EHR utilization is promising, current EHRs have yet to fulfill their potential as a driver of better healthcare. Medicine at its essence is about human interactions and about a physician’s ability to connect with a patient and support them in health-related decisions. In many respects use of an EHR cheapens this relationship as physicians place a computer between themselves and the patient, often spending as much time clicking boxes as examining the patient. The available EHRs are clunky and non-intuitive resulting in wasted physician time and inaccurate recording of patient data. Likewise, they are designed with as much focus on billing as patient care limiting the utility of the information collected.
In terms of increasing communication between providers, there are currently over 450 EHR products utilized by physicians, few of which can communicate with each other. Thus, if a patient from one hospital system receives care at a neighboring hospital system, the healthcare providers remain reliant on the telephone and fax machine to receive the patient’s past medical record. This results in increased physician workload, ordering of duplicate studies, and a delay in patients receiving the care they require. This lack of communication between EHRs also detracts from the ability of researchers to efficiently aggregate patient data and to make potentially groundbreaking discoveries.
Despite the challenges outlined above, the potential to disrupt healthcare delivery remains for EHR developers. The first innovation required is a central database which links to all EHRs and enables the seamless transfer of patient information from one institution to another. In this model a patient’s health information is their property and will travel with them rather than being held hostage at their home institution. A central database will further allow for aggregation of patient data and more efficient/effective population level studies.
The second innovation required is development of patient-centric EHRs with integrated voice recognition and scribing capabilities. This will allow physicians to better record the information that is pertinent to patient care and to remove the computer as an obstacle between the physician and the patient. In addition, these future EHRs must have mechanism for detecting discrepancies in the data and highlighting errors. Current studies indicate that up to 60 percent of outpatient EHRs contain medication discrepancies demonstrating the need for integrated controls.[v] Finally, EHRs should move towards leveraging machine learning to support complex decision making. Thankfully, research is underway on this front as cancer centers collaborate with IBM Watson to determine optimal treatment protocols specific to the genetic composition of the cancer.[vi] With these innovations and more, EHRs may yet usher in a revolutionarily new era of healthcare delivery. (Word Count: 800)
[i] Atul Gawande, M.D., M.P.H. Two Hundred Years of Surgery. N Engl J Med 2012; 366:1716-1723May 3, 2012DOI: 10.1056/NEJMra1202392. Accessed November 14, 2016.
[ii] “History of Antibiotics.” Explorable: Think Outside the Box. https://explorable.com/history-of-antibiotics. Accessed November 14, 2016.
[iii] Haleh Ayatollahi, MSc, PhD, Nader Mirani, MSc, and Hamid Haghani, MSc. Electronic Health Records: What Are the Most Important Barriers? Perspect Health Inf Manag. 2014 Fall; Published online 2014 Oct 1. Accessed November 14, 2016.
[iv] Ryan Shay. EHR adoption rates: 19 must-see stats (blog). Practice Fusion Blog. Created January 1, 2016. http://www.practicefusion.com/blog/ehr-adoption-rates/. Accessed November 14, 2016.
[v] Linsky, A, Simon, SR. Medication discrepancies in integrated electronic health records. BMJ quality & safety. 2012;22(2):103-109.
[vi] Doyle-Lindrud S. Watson will see you now: a supercomputer to help clinicians make informed treatment decisions. Clin J Oncol Nurs. 2015 Feb;19(1):31-2. doi: 10.1188/15.CJON.31-32.
*Cover image* <https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcSTSLK9u4dtptivSlljhWrZ_F4xKnDkZ_dWncEjeYGE3uyjc-52og>
Student comments on Is Epic! Epic?
With physicians today making so many mistakes (perhaps more so than ever before) we really need to look to software and technology to see if there is any way to protect the patients who put their lives in the doctor’s care. Patient centric EHRs seem like exactly what the doctor ordered.
One thing I’ve heard about EHRs is they are very non-user-friendly from the doctors perspective and they aren’t flexible enough. Is Epic doing work to mitigate this? Is voice recognition enough to solve this? On the voice recognition side, are they teaming up with companies like Microsoft and Google to improve their voice recognition systems? So many questions to ask about such an important topic.
Like Harambe, Epic has it’s own legendary legacy to write. Hopefully it can crack the nut and make healthcare more efficient and safer for us all.
Thanks for the post – think this is a super interesting topic, especially following our IBM Watson discussion.
One significant hurdle to overcome will be patient privacy and other legal issues associated with storing and sharing data. As you mention, the healthcare system is incredibility fragmented and includes multiple stakeholders which has slowed the adoptions of EHR systems. International models may prove as helpful role models in countries that have national healthcare. While not directly applicable to the US, perhaps the solutions they are developing there may be somewhat transferable to overcome some of the difficulties you describe.