MedElevator Войти в исследования
← Поиск публикаций

DOI 10.3390/jcm15103878

Медицинская библиотека · Journal of clinical medicine · 2026

The Use of Chat-GPT 5.2 by Patients Affected by Rotator Cuff Tears Leads to Inaccurate Diagnosis and Treatment Suggestions: A Study by SICSeG (Italian Society of Shoulder and Elbow Surgery).

de Giovanni Roberto, Gaj Edoardo, Mottola Luciano, Cecere Antonio Benedetto, Coppola Martina, Garofalo Raffaele, Cozzolino Andrea

СтатьяПеревод пока не подготовлен

Аннотация

Background

Patients increasingly rely on freely available artificial intelligence tools, such as Chat-GPT, to obtain diagnostic and therapeutic information for different orthopedic conditions. While preliminary studies have evaluated its educational potential, evidence regarding its diagnostic accuracy and treatment recommendations in rotator cuff pathology remains limited. This study aimed to assess the ability of Chat-GPT 5.2 to correctly diagnose rotator cuff tears and propose treatment strategies, comparing its performance with that of expert shoulder surgeons. Materials and

Methods

Five clinical cases representing common rotator cuff tear patterns were retrospectively selected in an exploratory pilot analysis. For each case, MRI images, radiologic reports, and clinical information were sequentially provided to Chat-GPT 5.2 using standardized prompts simulating a patient inquiry. Responses were compared with those of three experienced shoulder surgeons. Diagnostic accuracy, treatment recommendations, postoperative management suggestions, and complication descriptions were analyzed. Krippendorff's alpha was used to assess interobserver agreement.

Results

Chat-GPT 5.2 did not correctly diagnose any case using MRI images alone, whereas human examiners reached accurate diagnosis in most cases. Treatment recommendations provided by Chat-GPT were non-specific when based solely on imaging or radiologic reports, but became more defined after a detailed lesion description. Interobserver agreement between Chat-GPT and surgeons regarding treatment was inconsistent across cases. Postoperative rehabilitation advice and complication descriptions were accurate and comprehensive, demonstrating high consistency with published data.

Conclusions

The results of this pilot analysis suggest that the use of Chat-GPT 5.2 by patients affected by rotator cuff tears could potentially lead to misdiagnosis and wrong treatment advice, while on the contrary it could be used by physicians to better illustrate postoperative protocols, complications and expected outcomes.

Для цитирования: de Giovanni Roberto, Gaj Edoardo, Mottola Luciano, Cecere Antonio Benedetto, Coppola Martina, Garofalo Raffaele, Cozzolino Andrea. The Use of Chat-GPT 5.2 by Patients Affected by Rotator Cuff Tears Leads to Inaccurate Diagnosis and Treatment Suggestions: A Study by SICSeG (Italian Society of Shoulder and Elbow Surgery).. Journal of clinical medicine. 2026. DOI: 10.3390/jcm15103878.
Подготовка и перевод публикации
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…

История · было → стало

Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.

Тип публикации указан по данным источника. Уровень доказательности автоматически не присваивается.