When the major EHRs in the market are being installed in hospitals, there is little latitude for making significant changes to them on the basis of input from physicians or nurses. These systems are so complex that any possible customization is rather minor. For example, physicians and nurses can provide input about clinical order sets (see: EMR-Based Order Sets as a Locus of Control of Hospital-Based Physicians; Evidence-Based Order Sets Deployed at the University of Kansas Hospital). By way of defense of EHR vendors, they all have had (or claim to have had) significant physician and nurse input during the alpha and beta phases of their systems. And, of course, all EHR vendors will certainly implement rapid changes to their systems when critical patient safety problems are pointed out to them. A recent article discussed the need for the involvement of physicians and nurses in EHRs designs and improvement, particularly with regard to modules used in Emergency Departments (EDs) (see: Clinician Involvement Critical in EHR Design, Improvement). Below is an excerpt from it:
To ensure patient safety and quality of care, front-line clinician involvement is critical in the design and deployment of electronic health record (EHR) systems, as reported in an article published...in the Annals of Emergency Medicine. Efforts to improve EHR systems are constantly evolving....The researchers represent a joint committee [of members] of the American College of Emergency Physicians, who combined efforts to analyze the benefits and potential threats to patient safety that can result from EHR choices. Some vendors prohibit users from sharing hazards even in the academic literature. Although EHR systems that are marketed and sold to hospitals and healthcare providers are certified by the Certification Commission for Health Information Technology, that process is not focused on safety issues....and no mechanism exists to allow or encourage system users to address ongoing safety issues or provide feedback. "In fact, some vendors prohibit users from sharing hazards even in the academic literature," they write. Emergency Department (ED) characteristics, including rapid turnover, frequent transitions of care, continual interruptions, large patient volume, and unfamiliar patients, make EDs "particularly error prone" when using EHR systems or ED information systems (EDISs) to reduce medical errors and standardize care, the authors write. The report describes 4 potential "fictitious yet realistic clinical scenarios" in which choice of EDIS can make a difference and offers 7 recommendations to improve patient safety when using EDIS.
The issue of patient safety with regard to any clinical information system, be it an LIS, RIS, or EHR, is critical. There now seems to be a growing chorus of system users as well as governmental and regulatory bodies that are beginning to discuss patient safety issues as they relate to EHRs (see: Joint Commission puts focus on EHR, patient safety; HHS Releases Health IT Safety Plan; How Often are EHRs Placing Patients at Risk?). The committee members quoted in the article above are correct in highlighting EHR vendor's contract "gag clauses" as a major barrier to EHR improvements in patient safety (see: EHR vendor contracts and the gag clause).
I would draw an analogy here between adverse reactions to drugs and EHR errors that threaten patient safety. Reporting the former to the FDA is mandatory and information about these problems is widely reported. EHR errors, when reported to the vendor and by way of contrast, may result in a quiet discussion between the vendor and the reporting hospital client with contractual restrictions on public disclosure. Even public disclosure of the specific language in EHR vendor gag clause contractual language is prohibited. Here's an important quote from the IEEE Spectrum article referenced above:
Part of the reason for the lack of safety-related information is that, unlike medical devices, reporting of problems with electronic health record systems to the US Federal Drug Administration is not mandatory. In addition, it is often difficult to identify whether an “EHR problem” is due to a simple human error, a design flaw that encourages human errors to occur, or some specific software or other type of system-related error. Further, even if there was a reliable means to identify and classify these issues, EHR vendors frequently include “hold harmless clauses” in their contracts that help them escape responsibility for errors or defects in their software, as well as nondisclosure or confidentiality clauses that can prevent EHR users from disclosing the problems they encountered with a vendor’s software.
For me, the difficulty in assessing whether patient harm is attributable to human or software error is an insufficient reason for not reporting such perceived problems to a federal agency. Gag clauses in contracts regarding disclosure of EHR patient safety issues should also be prohibited.