Diagnostic error (failure to establish an accurate and timely explanation of the patient’s health problems or
communicate that explanation to the patient) is estimated to harm an estimated 249,900 hospitalized US adults
every year. Most errors are multifactorial, and a majority have cognitive causes. Clinician awareness of these
cognitive causes is vital for preventing future diagnostic errors. However, clinicians often do not have an
accurate awareness of their own diagnostic performance. Specifically, their metacognitive calibration
(alignment between their diagnostic performance and confidence in that performance) may be misaligned. To
help clinicians gain insight into their own diagnostic performance, we developed “Calibrate Dx: A Resource to
Improve Diagnostic Decisions,” which aims to improve calibration by leveraging the ideas behind self-regulated
learning theory and guiding clinicians through a series of steps to generate feedback about their own
diagnostic decision-making using self-assessment and peer debriefing. Pilot work suggests that clinicians think
calibrating their diagnostic performance is “critical” and a “moral and ethical responsibility,” but they previously
lacked a “structured way of doing [it].” Additionally, pilot participants emphasized a need to facilitate adoption
and use of Calibrate Dx in clinical practice. We will use the Safer Dx Framework (which approaches quality
improvement by considering the complex, sociotechnical system in which diagnosis occurs) to systematically
identify barriers and facilitators to Calibrate Dx use and create a guide, enabling further adoption and scale.
Then, we will evaluate whether Calibrate Dx use leads to improved calibration and changes in attitudes or
behaviors related to diagnostic safety in hospital medicine. We will use Kirkpatrick’s Framework for Evaluating
Educational Interventions to guide this work. Our Specific Aims are: 1) Create methodology for evaluating
calibration in clinical practice. Experts will help develop methods for measuring calibration in real, clinical
practice in the hospital setting. Then, the measurement methods will be piloted at two participating healthcare
institutions. 2) Develop implementation guide to account for sociotechnical barriers and facilitators while using
Calibrate Dx by conducting semi-structured interviews across two health care institutions. Interviews will be
conducted with key informants including hospital leadership, patient safety officers, safety and legal staff, and
physician-hospitalists to identify barriers and facilitators to use of Calibrate Dx for learning and improvement.
Strategies to avoid barriers and harness facilitators will be developed to serve as additional guidance for
Calibrate Dx use. 3a) Assess whether use of Calibrate Dx leads to improved calibration. We will recruit
hospitalists from two health care institutions to use Calibrate Dx. We will assess if resource use improves
calibration over time. We will also examine whether clinician-level factors relate to improved calibration. 3b)
Determine whether use of Calibrate Dx affects other attitudes and behaviors related to diagnostic safety (e.g.,
safety attitudes, humility, changes in practice) using a combination of surveys and structured interviews.