International urogynecology journal · 2026
ChatGPT Health and ChatGPT Plus in Urogynecology: A Blinded Comparison of Response Quality, Guideline Concordance, and Safety.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
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.
Tayarer Aybüke
Аннотация
INTRODUCTION AND HYPOTHESIS: ChatGPT Health is a health-focused conversational artificial intelligence (AI) environment, but its performance in patient-oriented urogynecology has not been directly compared with standard ChatGPT using both validated quality assessment and objective guideline-based criteria. We hypothesized that ChatGPT Health would provide higher patient-facing response quality while maintaining comparable guideline concordance and safety.
METHODS: Twenty-four guideline-informed patient queries covering common urogynecological and lower urinary tract concerns were submitted independently to ChatGPT Plus and ChatGPT Health on 14 August 2026. Forty-eight responses were anonymized and independently evaluated by five blinded clinicians using the validated Quality Analysis of Medical Artificial Intelligence (QAMAI) instrument, a prespecified five-element guideline-concordance checklist, and a safety scale. Blinded head-to-head preference was assessed secondarily. Prompt-level paired comparisons used Wilcoxon signed-rank tests.
RESULTS: Median guideline concordance was 100% for ChatGPT Health and 98% for ChatGPT Plus (p = 0.298). ChatGPT Health had a higher blinded five-domain QAMAI subtotal (25.0 vs 24.2; p = 0.026) and greater clarity (p = 0.020), completeness (p = 0.016), and usefulness (p = 0.033), with all three remaining significant after correction for false discovery rate. Accuracy did not differ (p = 0.453). No clinically meaningful safety concerns were identified in 240 ratings. ChatGPT Health was preferred in 47.5% of 120 assessments, ChatGPT Plus in 18.3%, and no meaningful difference was reported in 34.2%.
CONCLUSIONS: Both platforms generated highly accurate, guideline-concordant, and safe responses. ChatGPT Health showed advantages primarily in clarity, completeness, usefulness, and overall quality of patient-facing communication rather than objective concordance of clinical content.
METHODS: Twenty-four guideline-informed patient queries covering common urogynecological and lower urinary tract concerns were submitted independently to ChatGPT Plus and ChatGPT Health on 14 August 2026. Forty-eight responses were anonymized and independently evaluated by five blinded clinicians using the validated Quality Analysis of Medical Artificial Intelligence (QAMAI) instrument, a prespecified five-element guideline-concordance checklist, and a safety scale. Blinded head-to-head preference was assessed secondarily. Prompt-level paired comparisons used Wilcoxon signed-rank tests.
RESULTS: Median guideline concordance was 100% for ChatGPT Health and 98% for ChatGPT Plus (p = 0.298). ChatGPT Health had a higher blinded five-domain QAMAI subtotal (25.0 vs 24.2; p = 0.026) and greater clarity (p = 0.020), completeness (p = 0.016), and usefulness (p = 0.033), with all three remaining significant after correction for false discovery rate. Accuracy did not differ (p = 0.453). No clinically meaningful safety concerns were identified in 240 ratings. ChatGPT Health was preferred in 47.5% of 120 assessments, ChatGPT Plus in 18.3%, and no meaningful difference was reported in 34.2%.
CONCLUSIONS: Both platforms generated highly accurate, guideline-concordant, and safe responses. ChatGPT Health showed advantages primarily in clarity, completeness, usefulness, and overall quality of patient-facing communication rather than objective concordance of clinical content.