Discussed in this post: “How Often Do EHRs Result in Patient Harm?” (AJN, News, March).
When we first had computers in the hospital—that is, while we still charted on paper but had quick online access to lab, radiology, and pathology results and could easily look up a patient’s prior admission history—it was wonderful. No more little lab slips floating all over the nurses’ station. No more unit-to-unit searches trying to figure out who last had custody of the patient’s X-ray films. (How could objects so large be so easily lost?)
A rocky transition to EHRs.
The transition to almost fully digital charting, on the other hand, has been pretty much a nightmare from the beginning. Nursing was rarely included in initial needs assessments. Many rollouts were chaotic, without additional staffing for the inevitable glitches that are bound to occur. Training of frontline clinical staff has been routinely minimal; we seem to be expected to pick up the many fine points of new software by some kind of digital osmosis.
That elusive clinician friendly EHR.
It’s very clear at this point that electronic health records (EHRs) were designed primarily for data collection and billing purposes. I have yet to see a system that could in any way be called clinician friendly. We are not communicating as clearly as we used to, and it’s a little tiring to keep hearing “But isn’t it great that we don’t have to decipher handwriting anymore!?” One advantage does not transform a labyrinth into a model.
Read more about this serious but still unacknowledged issue in “How Often Do EHRs Result in Patient Harm?” in this month’s AJN, which notes findings of a 2019 joint investigation by Kaiser Health News and Fortune magazine that found serious documented and potential patient safety issues traced to computer system errors:
“The authors describe systemic [computer] failures in transmission of physician orders, mix-ups in patient profiles that could lead to potentially catastrophic medication errors, and issues with ‘interface,’ where online systems used by pharmacies and laboratories for example, might fail to communicate with each other.”
Incredibly, there are no reporting requirements for near-misses or actual patient harm that has resulted from these computer issues.
Paper charting forces providers to be organized. It needs physical space for storage. Physical files lack of backups and its security is limited. They can be wiped out during a fire and inappropriate storage conditions. As a result, files are impossible to recover. Likewise, paper medical records are time-consuming and error-prone due to form layouts and handwriting misinterpretation. In other words, paper charting has an increased risk of human error and loss of data integrity.
On the other hand, electronic medical records (EHR) are part of information technology (IT). EHR allows for the collection and storage of accurate information, coordination of care, and reduction of paperwork and unnecessary tests and procedures. A study conducted by Allen et al. (2014) discusses that seventeen communities using health information technology infrastructure became leaders in building and strengthening their health IT to provide more effective care management (p. 150). IT has revolutionized health care and supports the nursing role in providing care. It also helps nurses to keep patient information in one place and access it when necessary. IT has many benefits such as management of staffing, locating medical profiles, and tracking patient treatments. A study suggests that IT has reduced mortality rates, medical errors, and operational costs, and it has increased patient safety and efficient care coordination (Tso-Ying et al., 2017, p. 917).
In my assessment, the EHR benefits outweigh the possible potential for harm. Likewise, evidence-based research is available to improve or fix any failure of the system. Input from health care providers interacting with IT should also be taken into consideration for future strategies in preventing patient harm form EHR.
Allen, A., Des Jardins, T. R., Heider, A., Kanger, C. R., Lobach, D. F., McWilliams, L., … Turske, S. A. (2014). Making it local: Beacon communities use health information technology to optimize care management. Population Health Management, 17(3), 149- 158. doi: 10.1089/pop.2013.0084
Tso-Ying, L., Gi-Tseng, S., Li-Tseng, K., & Mei-Ling, Y. (2017). The use of information technology to enhance patient safety and nursing efficiency. Technology & Health Care, 25(5), 917-928. doi: 10.3233/THC-170848
Another oft-unappreciated effect of the EHR is related to one important use of health records: legal and risk-management issues. Nurses brought to deposition to testify about their care are astonished when they’re asked, “Show me in here where you documented that.” They can’t find it. The printed records bear zero resemblance to the screens clinicians see and use. Even asking for something from the IT department isn’t straightforward. Sure, I can ask for “Any and all records for John Doe from the admission of 1/2/2020 to 1/6/2020,” but that won’t show me where the person was who accessed and read a given record entry (in the office? another floor? lab?), made changes in it or what changes were, or the various views with different dropdowns and defaults seen by lab, MD, therapy, nursing, dietary, etc. And there are many, many more aspects of an EHR in a hospital that most folks will never see. I once read 4000+ pages for a patient who was in the ICU for a time. There were 16 pp of systolic BPs. Some pages later, 14 pp of diastolic BPs. Somewhere else, 8 pp of mean arterial pressures. Good luck with trying to correlate that with assessment findings, SVR, UO, pressor rate changes, ABGs, MD calls, and new meds even if you can find them … and they wonder why my invoice shows such a huge bill for many hours of record review.
Another use of medical records is for teaching purposes. How would I use this record to teach a new ICU nurse about the interactions between inotropes, vasodilators, the importance of looking at systolic and diastolic and not just means? Not to mentino the rest of it.
And of course, the idea that once it’s in there, it’s there forever. I once had a case where a woman suffered a significant brain injury with hemiplegia; she had a ventriculoperitoneal shunt placed. Some months later she went to the hospital for a procedure to revise it. While they were in there, they found a small aneurysm, completely asymptomatic and benign-looking, so they clipped it. Days later the overworked and likely sleep-deprived R1 who had never seen her dictated her discharge note, and when I got it after three admissions to other facilities I read that she was hemiplegic because of the aneurysm. It took me quote some time sleuthing to discover that obviously, the R1 saw the op note and made that erroneous assumption.
My own EHR has me diagnosed with at least two signficant medical conditions, which would be good to know if either of them were accurate, and they are definitely not. I only discovered this by accident when transferring to a different PCP and transporting my own records by hand. What if somebody decided to treat an unconscious me for one of them? Or misdiagnosed me entirely because of a “preexisting” condition? Harm indeed.