Avaliação da usabilidade de um agente conversacional para a promoção do autocuidado em gestantes, e em pessoas com hipertensão e diabetes: gissa intelligent bot

Abstract

Digital health (eHealth) is defined as the use of Information and Communication Technology (ICT) in the work context of healthcare professionals. Among the tools used for this purpose are chatbots, conversational agents that use natural language programming, movements simulating a human conversation via text or voice. Considering the high percentage of the Brazilian population with access to smartphones and the internet, and the need to intensify actions to promote self-care in health, especially for people in need of continuous monitoring, such as pregnant women, people with hypertension and diabetes, it is important to develop digital solutions for this purpose. This study aims to evaluate the usability of a chatbot capable of interacting with pregnant women, people with hypertension and diabetes, through textual dialogues about self-care and the Primary Health Care Network/Family Health Strategy. This is a study with mixed methods, developed in two stages: Stage I – Creation of dialogue plans about self-care and the journey in primary health care for pregnant women or people with hypertension and/or diabetes, to build the computational prototype of the chatbot , and its implementation on the Facebook Messenger platform. Stage II: Conducting a cross-sectional, quantitative study, with a non-probabilistic convenience sample, to evaluate the usability of a chatbot called GISSA Smart Bot for pregnant women and people with hypertension and/or diabetes. The places where the study was carried out are: Crateús and Sobral in Ceará and Cruz Alta in Rio Grande do Sul. Data were collected from 119 hypertensive and/or diabetic patients and 61 pregnant women. In the evaluation of usability by pregnant women, the chatbot was classified as good, obtaining 79% in the parameters of the SUS scale, in the public of people with hypertension and/or diabetes, there was the same classification, and in this public the usability classification was 77 %. As a result, the degree of user satisfaction was considered good, configuring the experience of using GISSA Intelligent Bot as well evaluated by the selected audiences.