# Safety II Together: Coupling teaming science with patient engagement and health information transparency to coproduce diagnostic excellence

> **NIH AHRQ R18** · UNIVERSITY OF TEXAS HLTH SCI CTR HOUSTON · 2024 · $999,346

## Abstract

Ambulatory diagnostic error is a vexing problem because events unfold across time with shifting team
members. Accurate and timely diagnosis relies on keeping patients and health care professionals (HCPs) on
the same page, both in and between clinic visits. Experts urge shared mental models (SMMs) of the diagnostic
process (DxP), because research shows their central role for effective team communication and coordination.
Yet little is known about how patients and HCPs develop and sustain SMMs, especially in the dynamic setting
of the DxP. While we have many examples of diagnostic errors in Safety-I research, there are few models of
“what good looks like” using a Safety-II lens. Since diagnostic excellence relies on factors and interactions
beyond any single HCP, progress will require identifying specific teaming behaviors that help patients and
HCPs to coproduce safety at known DxP risk points. We will close this foundational knowledge gap. We will
then develop and disseminate tools to support patients, care partners, interpreters and clinicians as active
contributors to the DxP through 3 partnering national organizations. In our current AHRQ grant, we learned that
patients have unique insights about the DxP, including identification of “diagnostic blindspots” – events or
conditions pertaining to DxP safety that HCPs or systems may not observe. These include events occurring
between visits, omissions, and misalignments stemming from lack of a SMM between patients and HCPs.
Interestingly, online access to visit notes (“open notes”) – now nearly universal through 2021 federal legislation
– may help development of SMMs because patients can gain context on HCP perspectives and potential gaps
in their understanding by reading notes. One sobering finding was that the most common contributing factor to
patient-reported diagnostic error was not feeling heard. Based on these data, we designed an online tool –
OurDX – to elicit patient priorities, histories, and potential concerns. We found that patients with limited English
proficiency (LEP) were 5x as likely to report not feeling heard, and older patients with poorer health were more
likely to identify a diagnostic blindspot. Teams that work need new approaches to share information, to listen,
and to respond, especially since harnessing the unique knowledge of each member (including early
identification of otherwise undetected blindspots) can increase system resilience. Our diagnostic center of
excellence – Safety2gether – is based on innovative and disciplined coupling of teaming science with patient
engagement to coproduce diagnostic safety, with a focus on safety-II principles and priority populations: LEP
and elderly patients with chronic illness and their care partners. Safety2gether will partner with patients, train
learners, define future research questions, foster collaborations, and share multistakeholder tools through 3
national organizations to both learn from, and serve as a resource for, othe...

## Key facts

- **NIH application ID:** 10928222
- **Project number:** 5R18HS029362-03
- **Recipient organization:** UNIVERSITY OF TEXAS HLTH SCI CTR HOUSTON
- **Principal Investigator:** SIGALL BELL
- **Activity code:** R18 (R01, R21, SBIR, etc.)
- **Funding institute:** AHRQ
- **Fiscal year:** 2024
- **Award amount:** $999,346
- **Award type:** 5
- **Project period:** 2022-09-30 → 2026-09-29

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10928222

## Citation

> US National Institutes of Health, RePORTER application 10928222, Safety II Together: Coupling teaming science with patient engagement and health information transparency to coproduce diagnostic excellence (5R18HS029362-03). Retrieved via AI Analytics 2026-05-27 from https://api.ai-analytics.org/grant/nih/10928222. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*