JAMIA Open — Journal / Study Design Concept
What JAMIA Open Is
JAMIA Open (Journal of the American Medical Informatics Association Open) is a peer-reviewed, open-access journal published by the BMJ Group in partnership with the American Medical Informatics Association (AMIA). It focuses on applied health informatics research — the science of how information technology is used in healthcare delivery.
It is considered a credible, peer-reviewed source for clinical informatics research, including quality improvement studies on digital health tools like AI scribes.
Why JAMIA Open Is Relevant Here
The Tierney et al. study (2025) was published in JAMIA Open. This is the KUMC quality improvement study that provides the 81% / 77% / 73% / 67% / 64% satisfaction figures for Abridge.
While JAMIA Open is not as high-impact as JAMA, it is a specialty informatics journal — its peer reviewers are domain experts in health informatics, making the methodology more likely to be scrutinized for informatics-specific concerns (study design, measurement validity, bias).
Important Study Design Notes
The Tierney et al. KUMC study has several methodological nuances worth understanding:
1. Quality improvement study, not an RCT It was designed as an internal quality improvement initiative, not a randomized controlled trial. This means:
- No randomized assignment
- No control group
- Results are observational/directional, not causal
2. Non-identical pre/post surveys The pre-implementation survey (8 items) and post-implementation survey (11 items) had different questions. Only 2 items were directly comparable:
- “I find my current documentation workflow easy to use” → “Abridge has made my documentation workflow easy to use”
- “I usually complete the note before the next patient visit” → “With Abridge, I could complete the note before the next patient visit”
The 81%, 77%, 73%, etc. figures come from post-implementation survey responses only — the percentage of respondents who agreed or strongly agreed with each statement after using Abridge. They are not computed pre/post changes.
3. Selection bias Early adopters of AI documentation tools may have been more favorable toward technology and less burned out at baseline — limiting generalizability.
4. Single institution KUMC is one institution — academic, Midwestern, early adopter culture.
The 2 Statistically Computed Pre/Post Comparisons
These are the only two findings with computed statistical comparisons:
| Finding | Result |
|---|---|
| Documentation workflow easy to use | 6.91× higher odds post-intervention (P<.001) |
| Note completed before next patient | 4.95× higher odds post-intervention (P<.001) |
Related