2022 Clinical Workflow Flexibility Challenges in the EHR: Defining the Problem (Part Three)

In our two previous articles, we described an “as-is” exercise with the EHRA workgroup and the HIMSS Physician Committee members where one issue became the focus: the flexibility of workflows. This series of blog posts reviews the results of the exercise and describes a path forward. So far, this series has covered Learning the EHR, Personalizing the EHR and Understanding the Patient. In this article we discuss the fourth and final topic that was raised: Providing Care.

 

Providing Care

Providing medical care for the patient necessitates “last mile” flexibility that allows clinicians to jump in and out of the expected workflow. It has been demonstrated that a defined “workflow” that physicians agreed upon during development most commonly erodes after the first several steps. (4,5) The ability to step back into the workflow after a short detour needs to be readily available in an intuitive and simple manner. There should also be flexibility in how the EHR is used, both to conform to an individual’s working style as well as to respond to the demands of a clinical situation. Along these lines, but not limited to them, is the ability to have more than one patient record open with safeguards that ensure the correct data is always entered into the right patient’s chart.

 

Two major facets that impact direct patient care is a lack of clarity on what the important data is and how to surface that information for users as they work with a patient. Every health record is a means of communication: between patient and provider; between provider and provider; and between laboratory, radiology and providers. It also serves as a record of interactions with important archival and legal implications. Thus, critical or salient information needs to be conveyed in an appropriate and visible manner. What was noted, and not noted, what was included in an evaluation and not included, and what was acted upon and who acted on it needs to be easily identified in the record. Furthermore, if appropriate action was not taken, the responsible party should be easily identified. This information must be prominently and reliably identifiable, and the process flexible enough, that it is not an additional chore for the provider.

 

Where Do We Go?

Each of these areas represent a vast domain to explore, and our collaboration will continue to do so. Despite the complexity established by this exercise, there are solutions that are beginning to emerge. Many of these problem areas circle around the long trail of interactions that are needed to create a clear picture of what happened with a patient. Automation in the form of machine learning is already being explored for these purposes and has a genuine chance to move the needle. Building trust around this automation, and in particular highlighting clinically relevant aspects of a patient record, is going to be an essential first step to more deeply integrating this technology. Crowdsourcing of user personalization and surfacing helpful tips, workflows and settings across “similar” users within an EHR vendor ecosystem is another area that seems to have merit. There are good examples from consumer software regarding the sharing of software settings en masse that can offer good lessons (i.e. video game “mods”, email rule algorithms). The HIMSS Physician Committee and EHRA Clinician Experience Workgroup will be taking further steps to explore solution spaces for these areas of focus, and welcome input from the broader clinical community.

References:

1. Holman, G. T., Beasley, J. W., Karsh, B. T., Stone, J. A., Smith, P. D., & Wetterneck, T. B. (2016). The myth of standardized workflow in primary care. J Am Med Inform Assoc, 23(1), 29-37. doi:10.1093/jamia/ocv107
2. Beasley, J. W., Wetterneck, T. B., Temte, J., Lapin, J. A., Smith, P., Rivera-Rodriguez, A. J., & Karsh, B.-T. (2011). Information Chaos in Primary Care: Implications for Physician Performance and Patient Safety. The Journal of the American Board of Family Medicine, 24(6), 745-751. doi:10.3122/jabfm.2011.06.100255