CODEX Digest - 1.29.26

1) Artificial intelligence large language models improve patient comprehension of radiologist magnetic resonance imaging reports. (subscription required)

Berzolla E, Gosnell GG, Chen L, et al. Arthroscopy. 2025;41(11):4607-4614.e4.

Communication barriers in test results can limit patients' understanding and participation in decisions. This randomized controlled trial finds that AI-generated knee MRI reports increased comprehension but decreased trust, highlighting the importance of physician involvement.

2) Ambient Scribe Technology in Simulated Patient Encounters Across Specialties.

Brunner J, Morrissey S, Stevens EM, et al. JAMA Netw Open. 2026;9(1):e2552870.

Ambient scribes can be a promising tool to improve record keeping during clinician patient encounters, but challenges exist. This research letter examines AI scribe records from four specialties and finds concerns related to ambient scribe-generated physical exam findings, patient history accuracy, and diagnostic conclusions.

3) Characterizing interpretive and diagnostic competency development in pathology training. (subscription required)

Brunyé TT, Shucard H, Eguchi MM, et al. Mod Pathol. 2025;38(12):100802.

Diagnostic expertise development is an essential goal in pathology residency. This longitudinal study tracks residents from various programs and different stages of their training to evaluate how experience influences their diagnostic process. Training improved feature description, image review efficiency and diagnostic accuracy but did not consistently enhance visual search or attention allocation.

4) Assessing the clinical relevance of point-of-care ultrasound for hospitalists: influence on clinical reasoning and decision-making.

Dias Souza M, Rahhal H, Tibério IFLC. PLoS One. 2025;20(12):e0338202.

Image analysis at the point-of-care can assist in confirming diagnosis through rapid access to physical condition data. This single-center Brazilian study with limited access to traditional radiologic imaging surveyed how hospitalists' diagnostic hypotheses, clinical practice, and diagnostic confidence changed after using POCUS.

5) Depressed or delirious? Overdiagnosis of depression in medically hospitalized patients. (subscription required)   

Howland M, Thompson N, Mahajan A, et al. J Psychosom Res. 2025;201:112493.

Misdiagnosing psychological issues can lead to stigma, inappropriate treatment, and poor communication. This retrospective chart review study identifies strengths and gaps in generalist recognition of depression and delirium in hospitalized patients across age groups, notes the risk of diagnostic bias from medication history, and underscores the need for a broad differential diagnosis to prevent misdiagnosis.

6) What goes wrong in out-of-hours primary care in Denmark? A compensation claim analysis using the healthcare complaints analysis tool.

Køpfli ML, Huibers L, Brinck CES, et al. BMJ Qual Saf. Epub 2025 Dec 19.

After-hours primary care is a setting where diagnostic delays may lead to serious harm. This restrospecitive cohort study finds that compensation claims analysis provides value as a monitoring and factor identification strategy that surfaces concerns reducing the safety of out-of-hours primary care.

7) Health inequalities and trends in heart failure diagnosis in primary care in England, 2000-21: a national retrospective cohort data-linkage study.

Lawson CA, Ali MR, McCann GP, et al. Lancet Prim Care. 2025;1(6):100060.

Delayed heart failure diagnosis leads to missed interventions and emergency care. This UK study finds that while most patients show symptoms in primary care, timely diagnosis is limited, especially for women, sicker individuals, and individuals with socioeonomic deprivation. Diagnosing during emergency admissions worsens outcomes and increases inequities.

8) An ethnographic study of diagnosis of physical illness in people with mental health conditions in the emergency department.

Liberati E, Jayaprakash A, Lindsay R, et al. Soc Sci Med. 2026;391:118927.

Patients with mental health conditions can experience barriers to effective care. This English ethnographic study documents a sociotechnical set of factors contributing to inequities in the diagnosis of physical illness for people with mental health conditions presenting to an emergency department. Biases, stigma, spacial, human factors engineering and other influences are discussed.

​​​​​​

9) Diagnostic follow-up of positive results on low-dose computed tomography screening in the Medicare population.

Pinsky PF, Silvestri G, Osarogiagbon R, et al. Ann Intern Med. Epub 2025 Dec 16.

Poor follow-up after positive cancer screenings can undermine diagnostic excellence. This retrospective cohort study reviews Medicare patients' adherence to guidelines following positive screening results, analyzing factors linked to guideline-recommended care and rates of diagnostic procedures by lung cancer diagnosis status. About one third received less intensive testing than recommended.

10) Projected cancer screening and outcomes under the 2025 federal Medicaid eligibility restrictions. (subscription required)

Shubeck SP, Diaz A. JAMA Oncol. Epub 2026 Jan 8.

Gaps in cancer screening can contribute to delayed diagnosis and increased patient mortality. This research letter calculates the substantial avoidable deaths incurred by of the loss of US federal funding for Medicaid coverage on patient access to screening programs and its projected impact on patient health at the state level.

11) Improving diagnostic safety through better measurement, standards, and clinical culture.

Slawomirski L, Kelly D, de Bienassis K, et al. J Patient Saf. Epub 2025 Dec 15.

Diagnostic safety is a global concern that requires systemic change to support excellence. This commentary summarizes an international analysis of the economic impact of problem of diagnostic error and submits strategies for improvement centered on measurement, standardization, and collective organizational change.

12) Linking patient-reported and clinician-assessed wound status via Chatbot-based digital surveillance for wound infection: retrospective observational study.

Su Y-C, Lin Y-H, Huang M-Y. JMIR Formativ Res. 2026;10:e77685.

Consumer digital devices can help patients rapidly report symptoms, which can accelerate diagnosis. This Taiwanese study finds that chatbot apps are effective for reporting potential wound infections but notes less accuracy with acute wounds than chronic wounds.