[Development of a chatbot (conversational app) for cataract surgery patients]

The goal of this study was to create and promote a type of chatbot or conversational app, for patients who need cataract surgery and to evaluate its acceptability.

Multicentric prospective clinical study in two phases. Phase I : distribution of a questionnaire of 20 questions (evaluating patients' knowledge about cataracts and cataract surgery and their expectations in terms of patient education). Statistical analysis was performed through factorial analysis with factor rotation and Cronbach's alpha calculation. Phase II : creation of a chatbot with a repertoire of question-answer sets. An acceptability analysis was performed using a second questionnaire inspired by the « SUS Score ».

One hundred and six initial questionnaires were collected. The patients were mostly women (56.6 %), aged 60 to 79 years (81 %), retired (77.4 %), with no high school diploma (33.0 %), had never used a chatbot before (95.3 %) and were accustomed to using a smartphone (66.0 %). Patients evaluated their knowledge about cataracts as insufficient (51.8 %) and felt the need to receive additional information (81.1 %). The comprehensibility score of the first questionnaire was 91 (middle school level). The baseline data of the chatbot was composed of 316 questions with a median comprehensibility score of 101 (middle school level). The first test of the chatbot included 18 patients. The median connection time was 4min and 40seconds (standard deviation 6.6). The median of number of questions asked for each connection was 6.5 (standard deviation 6.7). Acceptability was good, with a mean Sus Score of 78.6/100 (standard deviation 11.9).

This study shows the importance of information for cataract surgery patients. The creation of a chatbot for patients undergoing cataract surgery appears to be relevant in achieving this goal.