Assessing ChatGPT's responses to office ergonomics and spine health questions

The objective of this study was to assess ChatGPT's responses to common office ergonomics and spine health questions. ChatGPT was asked the 50 most frequent questions about 'office ergonomics and spine health' and evaluated by 10 experts on a scale of 1-4 points. The consistency between the experts' ratings was calculated with the intraclass correlation coefficient (ICC). The Flesch-Kincaid reading level was used to evaluate the readability of answers. The results indicated that ChatGPT demonstrated the capacity to provide generally satisfactory information concerning ergonomics and spinal health in office workers, with good inter-rater reliability (ICC = 0.838) as well as a readability level equivalent to a middle school level. This study indicated that although the responses were considered adequate, that they cannot replace experts' clinical judgment, particularly in the personalization of exercise prescriptions and the provision of individualized and comprehensive approaches.