As I sat finishing up after a busy clinic day, I received an email tip describing a new feature in our EMR. As I read this new tip, I was struck by one main thought: We’re still so far from building – or even talking about providing – tools that truly improve care.
The tip described how I could now add annual reminders for an upper endoscopy(EGD). An annual EGD can be very important for a tiny subset of patients, but there is little consensus on whom those patients are, and little evidence about how frequently to apply the procedure1. There are, however, significant costs and real potential harms.
My immediate thought was that this feature’s most likely effect would be over-utilization of an invasive procedure and increased revenue for the hospital, accompanied by a decrease in overall value of the care we are providing. Most importantly, it would expose some patients to increased potential harms.
Perhaps mitigating this was the cognitive load required to use the feature: I would need a two-page tip sheet describing all the things I should consider before adding the reminder into a record. It struck me that Apple would never have allowed this feature to be released.
I want us all to do better for patients, whether in high-tech changes to multi-million-dollar software, improvement projects, or simple changes in clinic. I therefore suggest a simple checklist of questions. If answered before building, and then released with the new change, this checklist could help to (1) improve the features built, and (2) improve adoption and (3) decrease burnout of health providers who wish to use those features to improve patient care.
The questions are:
- How does this new feature/change improve patient care?
- How will the benefit of this change be measured?
- What potential harms should we consider?
- What conflicts of interest should we be aware of and be willing to declare? (personal, corporate, or payor)
- What is the user experience of this new change?
- What alternatives were considered/should we consider to solve this problem?
- What would the ideal clinical workflow be, and what are the plans to move to that more ideal workflow?
- If there are limitations, why do they exist?
I tested the checklist by applying it to the tip I had received. I had initially found the tip demoralizing – a sign of a system that did not care about patient value or provider experience, but after applying the checklist I felt newly empowered. I knew why I had initially felt uncomfortable reading the tip – because of the potential harms and inherent conflicts. I now had responses in a format that I could share with the creators of the tip who most certainly had not intended these potential harms. And I could see a path (not easy but possible) of how the best intentions of the feature could be realized in a way that really could improve value to patients and better support their physicians.
I shared this review with those who created the tip. They thanked me and said they liked the idea of the checklist. It was clear that they were not aware of the controversy regarding the timing of ordering of EGDs. I have been able to meet with the consultant who had only the intention of making the care of patients safer with their work on this feature. We will work together to try to apply some of the lessons learned in the analysis.
I plan to use this in all the healthcare innovation projects that I participate in. I’ve set it up as an open-source checklist, and would welcome your contributions.
The original version of this was also published at KevinMD.
References
- Uptodate, “Barrett’s esophagus: Surveillance and management”, https://www.uptodate.com/contents/barretts-esophagus-surveillance-and-management, Accessed on 11 January 2019.
Images
- Created by the author
- Photo by Ian Espinosa on Unsplash, Unsplash license