MMHA 6601 Week 6 Adoption Resistance Post Example

Week 6 Discussion: Initial Post and Replies

Reading Physician Hesitation About Ambient Documentation as a List of Design Requirements

An ambient documentation tool records the spoken visit once the patient agrees, then writes a first version of the note that the physician checks, corrects and signs. In the composite primary care network I use for this course, a pilot offer went to forty physicians and only eleven accepted. Leaders described the other twenty-nine as resistant. I want to argue that their hesitation is better read as a set of design requirements, and that a rollout that treats it that way will reach more physicians than one that tries to overcome it.

The burden these tools address is real. When researchers timed office physicians minute by minute in four specialties, the computer and paperwork side of the day took close to twice as long as the conversations in the exam room (Sinsky et al., 2016). One large medical group that offered ambient scribes to its physicians reported that thousands adopted them within weeks, with users describing less time on notes after hours and more attention to patients during visits (Tierney et al., 2024). So the question is not whether the problem matters but why so many physicians in our network declined.

The reasons the hesitant physicians gave, collected in a short survey, fell into four groups. First, accuracy: several worried the draft would contain errors they would miss when tired, and that they would be signing notes they had not written. Second, privacy: some worried about patients' reactions to a recording device, especially in visits about mental health, substance use or intimate partner violence. Third, style: a few said their notes are how they think, and a drafted note would be someone else's thinking. Fourth, time: some expected that reviewing and correcting drafts would take as long as writing.

Each concern translates into a requirement. Accuracy becomes a requirement for drafts that mark uncertain content and a quality check on a sample of signed notes. Privacy becomes a requirement for easy opt-out by the patient at any point, a visible indicator when the tool is listening, and clear rules on how long audio is kept. Style becomes a requirement for templates physicians can shape. Time becomes a measurement requirement: track how long review actually takes and share the numbers.

This reading fits the technology acceptance model, in which perceived usefulness and perceived ease of use drive whether clinicians use a system (Holden & Karsh, 2010). The hesitant physicians are telling us exactly where usefulness and ease are in doubt. Treating that as resistance to be managed wastes the information.

Question for the class: when staff at your organization hesitated about a new tool, did leaders treat the reasons as requirements, and what changed as a result?

Reply to Keisha

Keisha, your example of nurses refusing a new bed-alarm system until the alarm volume could be adjusted is a perfect case of a concern that turned out to be a requirement. Once the vendor allowed per-room volume, use went up. I wonder whether the same is true for patient consent with ambient tools. In your experience, would a patient-facing sign in the room be enough, or would nurses want the physician to ask aloud each time? The answer would change the workflow a lot, and it seems like exactly the kind of question the hesitant physicians in my network were raising.

Reply to Tom

Tom, you made the fair point that some resistance is simply habit and that waiting for every physician to agree could stall the project. I agree that not every concern deserves a design change. One way to separate the two might be to ask whether a concern names a condition under which the physician would use the tool. "I would use it if I could see which parts it was unsure about" is a requirement. "I just do not like it" is not, at least not yet. In the technology acceptance research, early experience often shifts perceived ease of use, so offering a two-week trial with an easy exit might move the second group without changing the design at all. How did your organization handle physicians who would not even try?

References

Holden, R. J., & Karsh, B.-T. (2010). The technology acceptance model: Its past and its future in health care. Journal of Biomedical Informatics, 43(1), 159-172. https://doi.org/10.1016/j.jbi.2009.07.002

Sinsky, C., Colligan, L., Li, L., Prgomet, M., Reynolds, S., Goeders, L., Westbrook, J., Tutty, M., & Blike, G. (2016). Allocation of physician time in ambulatory practice: A time and motion study in 4 specialties. Annals of Internal Medicine, 165(11), 753-760. https://doi.org/10.7326/M16-0961

Tierney, A. A., Gayre, G., Hoberman, B., Mattern, B., Ballesca, M., Kipnis, P., Liu, V., & Lee, K. (2024). Ambient artificial intelligence scribes to alleviate the burden of clinical documentation. NEJM Catalyst Innovations in Care Delivery, 5(3). https://doi.org/10.1056/CAT.23.0404