December 20, 2021

HOMERuN Collaborative: HOMERuN Diagnostic Error Projects

The Hospital Medicine Reengineering Network (HOMERuN) is a national network of Hospital Medicine investigators at 12 academic medical centers (AMCs) and 50 affiliated sites.

Organizers and Facilitators: Gopi Astik, MD, Molly Kantor, MD, Andrew Auerbach, MD, MPH, Jeffrey Schnipper, MD, MPH

Utility of Predictive Systems in Diagnostic Error (UPSIDE) Study Background: 3-year AHRQ-funded project, now entering year 3. Project Goals:  To determine the incidence of diagnostic errors among patients who die in hospital or are transferred to the ICU two days or more after admission, which factors contribute to risks of diagnostic errors, and to use risk estimates to understand highest priority areas for improvement. Training: We trained more than 100 faculty from 28 sites. Chart Reviews: Trained reviewers looked at the clinical factors, used SaferDX (adapted for inpatient setting) to define errors, assigned a level of harm due to the error, and then applied a process fault framework to determine underlying causes of diagnostic error.  Case Selection: We gathered a list of medical patients who died or went to ICU (n=26,915), and sites use that list to randomly select cases until 100 cases were reviewed.  Diagnostic errors are common:   26.1% of deaths or ICU transfers were associated with a diagnostic error.  Diagnostic errors are harmful:   Among patients who died and had an error, in 29% of those cases the error probably caused the death.

"I find the data very humbling that 25%-30% of our ICU transfers were unexpected deaths that had a diagnostic error. It shows me how many we are clearly missing because we didn’t know about that. Our hospital has a process for mortality review and you can refer other cases. I would say there’s almost no way we’re picking up on all those, otherwise we’d be hearing about this all the time."

Diagnostic Errors in COVID-19 Persons of Interest Background: Moore-funded project, now in analytic phase.  Project Goals: 1) To determine the rates of diagnostic errors and underlying causes of diagnostic errors among patients admitted for suspected COVID-19 infection; 2) To determine risk factors associated with death and ICU transfer among patients admitted for suspected COVID-19 infection.   Training and Chart Review Process: Similar to UPSIDE.  Case Selection: Random selection of 5 cases/month from list of patients admitted with “PUI” status (e.g., COVID test pending/suspicion for disease; n=258). Diagnostic errors are still fairly common in PUI patients: The mean diagnostic error rate was 14%.  The diagnostic error rate seems to have risen over the summer of 2020: A period of time of a lot of changes in health care for COVID patients.

"This would be good work to integrate into a Patient Safety / Quality Improvement curriculum where residents can do this abstraction, which could be revealing and help the system."

Causes of Diagnostic Errors in UPSIDE and COVID-19 Persons of Interest

Common DEER Factors Among UPSIDE Patients and PUI with Diagnostic Errors

Highlights: Top 3 for ICU transfers or deaths in terms of prevalence: clinical assessment, testing problems, and patient follow-up and monitoring. Top 3 for PUI patients: clinical assessment, testing problems, and history taking problems.

Ideas for Next Steps   Apply this methodology prospectively or do recent retrospective reviews.  We should think about how we can catch these errors while provider remembers details of the case or while it is happening.  Role for senior advisors to review charts flagged and give advice.  Diagnostic timeouts or checklist upon ICU transfer. Thresholds for ICU transfer have risen at some sites, so would need to figure out how to account for that.  Culture change. Acknowledging diagnostic uncertainty may lead to safer care.  Interest in studies of:  Intersection of diagnostic error in vulnerable populations. Patients with COPD/CHF and other diseases that may be victims of premature diagnostic closure.  Unnecessary antibiotics for pneumonia, potential sign of anchoring on pneumonia.  Challenges   Most health systems don't have a program in place to look at diagnostic errors. Most are outcomes-based programs like mortality reviews, some looking at ICU transfers.  Feedback to providers. We need to think about how to do this feasibly and provide thoughtful feedback without being punitive.   These reviews can be time-consuming. A potential solution would be to use residents on quality or safety rotations to do them.

"Next step I hope to see is looking into what are the causes of these diagnostic errors identified, which are hard to measure. The statistical issues are substantial."

The RELIANCE Study: Roflumilast or Azithromycin to Prevent COPD Exacerbations

Purpose of RELIANCE: Both roflumilast and azithromycin have been shown to reduce the risk of COPD exacerbations compared to placebo. However, there has not been a head-to-head comparison of these two FDA-approved medications. RELIANCE is intended to support hospital efforts to reduce the risk of all-cause hospitalization and premature deaths in individuals with COPD.  RELIANCE is Seeking Hospitalists: Hospitalists are critical in the development of post-discharge care plans and medications used by people with COPD. We found from preliminary work that identification of people with COPD while they are hospitalized is an efficient recruitment method for RELIANCE. Benefits and Compensation: $500/year honorarium for being a community partner (paid after registration) plus $200 per patient paid after submission of baseline form. Option to participate in clinical roundtables with COPD thought leaders. Contribute topic ideas for future grant proposals or publications related to hospitalist care. Community Partners will not be investigators / authors, but will be acknowledged in the RELIANCE publication.  We will have a webinar on the RELIANCE Study on January 11, 2022 at 1pm PT. Please reach out to Tiffany.Lee@ucsf.edu if you would like to attend or if you'd like more information about RELIANCE.

Key Takeaways Diagnostic errors are common in hospitalized patients, and may be a major driver of inpatient deaths.  Most health systems don't have a program in place to look at diagnostic errors, and face cultural and practical barriers to make that leap. To begin to screen and give feedback will require leadership buy in, thought for how feedback can be given to providers without being punitive, and streamlining of the review process.

Our next meeting will be on January 14, 2021. The Medical Education Working Group will hold focus group discussions.

Check out the HOMERuN COVID-19 Knowledge Base for more details. If you would like to join the HOMERuN Collaborative calls, please reach out to Tiffany.Lee@ucsf.edu.