Using Artificial Intelligence-based ChatBot to Improve Women's Participation to Cervical Cancer Screening Programme (AppDate-You)

Using Artificial Intelligence-based ChatBot to Improve Women's Participation to Cervical Cancer Screening Programme

Artificial Intelligence-assisted Decision-making to Improve Women's Participation to Cervical Cancer Screening in Occitanie Region-France

Centre Régional de Coordination des Dépistages des Cancers, centre-Carcassonne, Université de Bretagne Occidentale, Brest-France

This project looks to improve the return rate of HPV self-sampling (HPVss) as well as the management of women HPVss positive.

France has already implemented a cervical screening programme based on HPV test for the whole female population aged 30-65 years. Non-participant women can perform HPV test at their home. However, less than 20% of French women performed vaginal self-sampling when a kit was sent to their home. Women with lower income and educational levels participate less to cervical screening. A variety of personal, practitioner, test-related and logistical barriers negatively impact the screening participation of French women. Key barriers to participation could be addressed by overcoming disparities in HPV-related knowledge and perceptions about cervical cancer screening. This study aims to evaluate the impact of sending HPV self-sampling kit at women's home, associated with providing, through multiple mobile channels, a multi-language decision aid designed for women with lower education accessible via artificial intelligence-based ChatBot. This decision aid will help women review high quality evidence on cervical cancer disease and screening modalities.

cervical neoplasia, cervical cancer screening, chatbot, artificial intelligence, cluster randomized controlled trial, HPV vaginal self-sampling

A cluster randomized controlled trial with two arms (one experimental arm and one active control arm) will be conducted in the Occitanie region, France to assess the effectiveness of a decision aid accessible via artificial intelligence-based Chatbot platforms to improve the "return" rate of HPV self-samples and the proportion of invited women "well managed" among disadvantaged women. A cluster is defined by aggregated units for statistical information (Ilots Regroupés pour l'Information Statistique, IRIS) and it refers to a target size of 2000 inhabitants per unit. Only IRIS classified as 4 and 5 according to the French version of the European Deprivation Index (EDI) will be included. The study will be conducted as a two-armed nested into the French cervical cancer screening programme in the Occitanie Region. Clusters will be randomly assigned by a computer-generated randomized schedule following simple randomization procedures in a 1:1 ratio.

Women randomized to this group will be sent screening reminder letters to perform HPV self-sampling test, with access to a decision aid tool tailored to those with low education levels. This tool will be available via Chatbot platforms.

Women randomized to this group will be sent screening reminder letters to perform HPV self-sampling test (Standard care)

Decision aid tailored to women with lower educational attainment,accessible via artificial intelligence-based Chatbot platforms

Non-compliant women those that have not been screened for more than 4 years and living in the disadvantaged areas will be sent screening reminder letters to perform HPV self-sampling (HPVss) test, with access to decision aid tailored to women with lower education attainment and available via Chatbot platforms. ChatBot will deliver information in a responsive, conversational way via text and spoken language in multiple languages through multiple mobile channels.

Percent of women recalled who "return" HPVss over 12 months "Return" means: 1) the receipt of the vaginal self-sample at the laboratory; 2) the performance of clinician-collected HPV testing for those who prefer to be screened by a health professional; and 3) Information about a recent screening or a cervical screening exclusion reason.

Percent of recruited women well managed (Percent of women screened and completing full pathway of HPV screening and treatment over 12 months). "Well managed" is defined as: completed a valid screening test (sample processed successfully to positive or negative result, including recollection of invalid samples) AND for HPV negative women, woman advised of negative result for HPV positive women, completed assessment.

Costs will be estimated using an ingredients approach whereby resources used for the respective interventions are identified and valued. The incremental cost-effectiveness ratios (ICERs) will be calculated as the mean difference in total costs between the intervention groups with the mean difference in effects, and expressed as both euros and US dollars per percentage change in screening participation.

Inclusion criteria: eligible women will be: aged 30-65 did not perform a pap smear in the last 4 years are living in deprived clusters in the Occitanie region. did not answer to a first "invitation" letter to perform a clinician-collected HPV testing Exclusion criteria: ineligible women will be those: outside the target age group had a Pap smear in the past 3 years had hysterectomy including cervix are pregnant beyond 6 months already scheduled a screening appointment or had just attended a HPV screening had a cervical abnormality that was under exploration and/or treatment. participated to the study pilot

Selmouni F, Guy M, Muwonge R, Nassiri A, Lucas E, Basu P, Sauvaget C. Effectiveness of Artificial Intelligence-Assisted Decision-making to Improve Vulnerable Women's Participation in Cervical Cancer Screening in France: Protocol for a Cluster Randomized Controlled Trial (AppDate-You). JMIR Res Protoc. 2022 Aug 2;11(8):e39288. doi: 10.2196/39288.