Consensus Between Radiologists, Specialists in Internal Medicine, and AI Software on Chest X-Rays in a Hospital-at-Home Service: Prospective Observational Study
2024

Agreement on Chest X-Rays in Home Hospitalization

Sample size: 260 publication Evidence: moderate

Author Information

Author(s): Eitan Grossbard, Yehonatan Marziano, Adam Sharabi, Eliyahu Abutbul, Aya Berman, Reut Kassif-Lerner, Galia Barkai, Hila Hakim, Gad Segal

Primary Institution: Chaim Sheba Medical Center, Faculty of Medicine, Tel-Aviv University

Hypothesis

How do interpretations of chest x-rays by radiologists, internal medicine specialists, and AI compare in a hospital-at-home setting?

Conclusion

Chest x-rays are crucial in home hospitalization, showing substantial agreement between specialists and radiologists, but AI interpretation needs further validation.

Supporting Evidence

  • 300 chest x-rays were performed at home for 260 patients.
  • 95.3% of x-rays were interpreted by internal medicine specialists.
  • Substantial agreement (κ=0.65) was found between internal medicine specialists and radiologists.

Takeaway

Doctors and AI looked at chest x-rays for patients at home, and they mostly agreed on what they saw, but the AI still needs to get better.

Methodology

A prospective observational study analyzing chest x-ray interpretations by different modalities in home hospitalization.

Potential Biases

Potential interpretation bias due to the limited number of physicians and specific AI algorithm used.

Limitations

The study was single-centered with a limited number of patients and used specific technology, which may introduce bias.

Participant Demographics

Median age of 74 years, with 55% women and 71% having cardiovascular disease.

Digital Object Identifier (DOI)

10.2196/55916

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication