Abstract
Introduction: Surgical consent forms often exceed recommended reading levels, limiting patient understanding and hindering truly informed decision-making. Artificial intelligence (AI)"“based tools like ChatGPT offer a novel strategy to improve readability without compromising content accuracy.
Methods: Four standardized spine surgery consent forms from a tertiary academic center were simplified using ChatGPT. Readability was assessed before and after simplification using six validated metrics: Flesch-Kincaid Grade Level, Flesch Reading Ease, Coleman-Liau Index, Automated Readability Index, Gunning Fog Index, and SMOG Index. Changes were analyzed using paired t-tests. Linguistic complexity"”including total word count, character count, sentence count, words per sentence, and sentences per paragraph"”was evaluated using one-sample t-tests. Statistical significance was set at p<0.05.
Results: ChatGPT simplification significantly improved readability across all measures. Flesch-Kincaid Grade Level decreased from 11.37 to 6.67 (p<0.001), and Flesch Reading Ease improved from 44.51 to 68.88 (p<0.005). Similar improvements were noted in Coleman-Liau (12.00 to 8.09), Automated Readability Index (10.79 to 5.79), Gunning Fog Index (18.07 to 11.65), and SMOG Index (15.70 to 11.32), all with p<0.005. Linguistic complexity was also reduced: total words (−43%, p<0.001), characters (−43%, p<0.001), sentences (−23%, p<0.01), words per sentence (−26%, p<0.001), and sentences per paragraph (−18%, p<0.001).
Conclusions: ChatGPT significantly enhances the readability of spine surgery consent forms while preserving essential medical content. This AI-driven approach may improve patient comprehension and support more equitable, informed decision-making. These findings highlight the potential of AI tools to streamline clinical communication and enhance the informed consent process.