Our study is placed within the design and development cycle of the Obesity-Related Behavioral Intervention Trials (ORBIT) model [22], specifically in phase 2a, “refine.” It involves providing material based on SH+ intervention in text, audio, and video format by a virtual coach (ie, a digital chatbot assistant, identified as ALBA [A Well-Being Assistant]) implemented within the TreC (an acronym for cartella clinica del cittadino, meaning “citizen’s electronic health record”) research app (a platform or app purposively designed to enable the delivery of several types of mobile health [mHealth] features). It could be stated that this study is innovative for 2 main reasons. First, it involves the digitalization of an intervention that was originally delivered in person and in group settings [14,15]. The ALBA chatbot version is designed to be more accessible and tailored to the individual, functioning without the presence of a helper, unlike in the RESPOND project [17,18], and includes chatbot-guided activity completion between sessions, which differs from the group session delivery format. Therefore, it is essential to evaluate how effectively this new format achieves these goals. Second, this study introduces a novel focus on a population consisting exclusively of female individuals and those considered medically vulnerable, which represents a significant new factor in evaluating the feasibility, acceptability, and efficacy outcomes. The proof-of-concept (POC) study, described in this research protocol, aims to test both the transferability and acceptability of the digitalized intervention from the original, context-specific in-person format (refer to the Primary Objective section), as well as to assess its initial efficacy on the targeted psychological variables in this newly defined population (refer to the Secondary Objective section).

ALBA is an intervention based on ACT techniques; it aims to promote the quality of life and enhance stress-coping skills in women with BC through a psychoeducational approach. The module is self-administered and available for both Android (Google LLC) and iOS (Apple Inc) devices. All the contents of the dialogues were developed by a group of researchers belonging to the Digital Health Research of the Bruno Kessler Foundation of Trento and psychologists belonging to the Pavoniano Artigianelli Institute with specific skills in the communication field starting from the SH+ manual [23] and subsequently reviewed by a psychologist manager of the psychology operational unit of the Provincial Health Services Authority of Trento (Azienda Provinciale per i Servizi Sanitari di Trento). The dialogues, videos, and audio tracks were developed according to educational logic. Furthermore, the structuring of the material and the delivery methods were supervised by professionals from the WHO Center for Research in Mental Health of the University of Verona, who have already carried out implementation studies and validation of SH+ at national and international levels.

To ensure the best possible adaptation of the intervention to our target sample, we previously conducted a co-design study based on guidelines of a previous step of the ORBIT model [22]. This process involved domain experts, such as psychologists, SH+ experts, and communication and usability experts, and a small sample of women with BC and oncological clinicians. For further details, please refer to our previous study [24].

This intervention is part of a behavior change framework [25] aimed at supporting women’s psychological well-being. The term “intervention” refers to the virtual coach, the chatbot ALBA, which delivers structured, rule-based informational content reviewed by psychological professionals. ALBA functions as a tool for information delivery, not as a conversational agent, as it does not engage in real-time, dynamic interactions. While it collects and transmits data, it is primarily a chatbot for eHealth interventions. ALBA is developed for research purposes to validate the digital health version of SH+ focused on ACT practices for women with BC.

ALBA, the technological component of this study, is built upon the general TreC platform, a public mHealth platform enabling residents of the Autonomous Province of Trento (North Italy) to access, manage, and share their health and well-being information. TreC is a robust and reliable platform designed as a “system of systems” rather than merely a data repository [26]. Indeed, its flexible architecture allows for collecting and managing diverse data types and supports developing and integrating additional subsystems to offer specific functions. A key feature of TreC is empowering users to manage their health-related data, functioning as a personal health record.

Within this general platform, a specific mHealth ecosystem has been designed to support research and digital health interventions, called TreC research (“Trec Ricerca” in Italian; Figure 1). This module is available for free download as a mobile app to all women diagnosed with BC. Authentication is required via the secure, General Data Protection Regulation–compliant national identification system (Sistema Pubblico di Identità Digitale).

The intervention spans a total duration of 6 weeks. For each week, a session of approximately 40 minutes is delivered, while this can be divided into two 20-minute subsessions according to users’ preferences. Phase 3, represented in Figure 2, provides a graphical depiction of the conversational protocol delivered to the participants and its chronological structure. The intervention’s contents are presented through a gradual process designed to progressively increase the participant’s self-awareness and mindfulness-oriented skills, including accepting and normalizing internal states, managing stress, and enhancing psychological well-being. Participants are also required to complete tasks and fill in a dedicated e-diary. The program is structured into 5 main coaching sessions, following the operational chapters of the SH+ manual, each addressing a specific strategy for managing stress and promoting psychological well-being, which are the main aims of the intervention. These 5 modules, translated into weekly e-sessions, are grounding (mindfulness), unhooking (defusion), acting on your values (values-based behavioral activation), being kind (gratitude), and making room (acceptance; phase 3 in Figure 2).

Therefore, this protocol paper aims to describe the POC developed to deploy the ALBA app to women with a BC diagnosis. The POC aims to evaluate the acceptability and feasibility of delivering SH+ through ALBA, a platform designed to provide ACT-based psychoeducational sessions. ALBA supports users in completing exercises and tasks assigned between coaching sessions and periodically delivers structured information—curated by professional psychologists—in audio, text, and video formats. This material is independently accessed and managed by users. The virtual coach serves as an information transmission tool, while the app also administers questionnaires and manages data collection.

This behavioral change intervention [25] aims to empower women with adaptive strategies for managing stress. Studying its feasibility can represent a first step to evaluate the possibility of including a digital service to support and promote psychological well-being in the context of stepped care for women in the BC pathway.

The main objective of the POC study is to evaluate how women experience the chatbot-based version of the WHO-validated intervention SH+. It aims to assess the following areas: (1) the user experience (UX) and the usability of the virtual coach, ALBA, installed in the TreC research app environment and (2) how users felt during the intervention, using ALBA, as reported through semistructured interviews. The IT evaluation of the TreC research app itself will be presented in another study.

The secondary goal is to assess the efficacy of the intervention on women’s psychological well-being through the administration of self-report pretest-posttest and follow-up questionnaires.

This phase will take place at the hospital’s Breast Unit, and it will be carried out by the health professionals of the Breast Unit (physicians and case manager nurses).

The screening phase is carried out by psychologists, and subsequent included participants are given access to the TreC research app.

At the beginning of the path, ALBA will administer 5 self-report questionnaires to establish the baseline level of the examined variables (preassessment questionnaires). The participant can choose the best day and time to interact with ALBA (first interaction with ALBA).

Specifically, there is 1 psychoeducational session per week, and the contents will be presented in different formats (text, images, audio, and video tracks). Moreover, ALBA provides users with instructions on how to carry out simple exercises and fill in a personal diary. The questionnaires will be further administered by ALBA at the end of the course (at week 6) to verify whether participants have perceived a change in terms of psychological well-being and quality of life. In total, 4 questionnaires will be administered to evaluate usability and UX (weeks 3 and 6).

Four months after the conclusion of exposure to SH+ intervention, women who have given consent will be invited to participate in a semistructured interview to investigate their experience of using ALBA and the TreC research app to understand how the women felt during the intervention. In addition, the long-term effects of stress will be collected by administering a follow-up questionnaire.

The study will take place in the Province of Trento (North Italy), where approximately 5000 women have or have had a diagnosis of BC, with almost 500 new cases every year [27]. In January 2018, the Breast Unit was established in the Province of Trento. It represents an innovative organizational model that merges various professionals, structures, and services transversally to make it more effective and homogeneous and coordinates the clinical path of patients with BC.

Recruitment will target women diagnosed with BC who are enrolled in the Breast Unit program, with additional support from local voluntary associations where possible. A convenience sampling approach will be used. Textbox 1 [14,17,28-30] outlines the study’s inclusion and exclusion criteria.

Health professionals at the Breast Unit will verify the inclusion and exclusion criteria related to the sociodemographic data (phase 0), while psychologists will assess the psychological screening data (phase 1).

Participants will be required to sign an informed consent form during recruitment (phase 0), following a comprehensive explanation of the project, its objectives, data collection, management, processing procedures, the required level of involvement, the study duration, and confidentiality matters. They will be informed that they can withdraw from the study at any time without providing a reason and that doing so will not affect the quality of their care or interfere with their treatment. Informed consent must be given freely and in writing, accompanied by a data processing authorization form, before joining the study. Participants must provide written informed consent. They will also receive a privacy policy statement explaining how their personal data will be processed during the study and must provide explicit consent for its use. Copies of the informed consent and privacy forms will be provided to participants and available in a dedicated section of the app.

To promote participant engagement and retention throughout the 6-week intervention, specific features were designed to sustain interest and minimize dropout, such as incorporating gamification elements, providing regular feedback or progress tracking, scheduling personalized reminders, and offering small recognition for continued participation. No monetary compensation was provided.

This research project is designed as a POC study, where a limited sample is sufficient to achieve the intended aims. On the basis of calculation assuming a nonnormal distribution and using nonparametric statistics, a sample of 24 women with BC diagnoses is sufficient when applying a Bonferroni-corrected significance level (α=.025) to control for type I error and multiple comparisons. Nevertheless, to validate this estimate, the sample calculation was run with the Power Analysis and Sample Size 2021 (NCSS, LLC) and Stata (StataCorp LLC) programs. The resulting sample size is 41 participants, based on the following assumptions: a significance level (α) of .025, statistical power of 0.80 (to ensure sufficient sensitivity to type 2 error), a mean difference between groups (eg, pre- and postintervention questionnaires) of 1.00, an expected effect size (δ) of 0.50, and an SD of 2.00.

The chosen effect size (Cohen d=0.50) reflects a moderate change [31], which is consistent with previous digital mental health interventions in populations with psychological disorders [32]. This approach aligns with recommended practices in early-stage mHealth research, where the focus is on identifying potential signals of efficacy rather than definitive effect estimates [22]. Therefore, considering a 20% possible dropout [33], 50 women were estimated to need to be recruited to carry out the study. This approach is also coherent with previous studies adopting this type of methodology in the field of mHealth.

The primary outcomes are UX and usability. They will be assessed using 4 questionnaires, namely, the User Engagement Scale-Short Form (UES-SF) [34], the System Usability Scale (SUS) [35], the Italian Chatbot Usability Scale, version B (ITA BUS B) [36], and the User Mobile Application Rating Scale (uMARS) [37]. They will be administered at different time slots throughout the study, in particular, at the end of week 3 (UES-SF, SUS, and ITA BUS B) and at the end of the study (week 6: UES-SF, SUS,; ITA BUS B, and uMARS; Textbox 2).

The end points to be measured include the scores of individual questionnaire items, the average scores of the questionnaires, and the differences between average scores of the questionnaires at various survey times. In addition, the participants’ experiences with the app and interactions with ALBA, as well as their use of the intervention, will be evaluated through semistructured interviews conducted 4 months after the study concludes. These interviews will collect qualitative data to supplement the primary outcome information.

The variables expressing the psychological well-being outcomes will be assessed by administering self-report questionnaires at the beginning (screening and week 0) and at the end (week 6) of the intervention (Textbox 2 provides a detailed overview of the questionnaire adopted). The end points to be measured include the scores of individual questionnaire items, the average scores of the questionnaires, and the differences between the average scores of the questionnaires at 2 survey times. In addition, the Perceived Stress Scale (PSS) [38] will be administered during a follow-up conducted 4 months after the study’s conclusion.

Both the clinical and usability tests will yield within-subject and between-subject comparison results.

Sociodemographic data collection will be carried out by the health professionals at the Breast Unit (physicians and case manager nurses) using a prestructured form. After informed consent is obtained, each form will be assigned a unique alphanumeric code. The parameters requested from participants will be (1) date of birth, (2) oncological disease history (ie, date of first cancer diagnosis, date of tumor removal surgery, date of any other interventions carried out or planned, and date of last neoadjuvant or adjuvant chemotherapy or radiotherapy treatment or if still ongoing), (3) maximum education grade, (4) employment, (5) marital status, and (6) current or previous experience of psychological support (exclusion criterion 7 verification).

To fulfill the verification of the inclusion criteria from 7 to 9 and the exclusion criteria 5 and 6, the researchers will administer the screening questionnaires through LimeSurvey (LimeSurvey GmbH; Textbox 2).

After processing the results of the screening questionnaires and the various sociodemographic inclusion criteria, the women will be contacted again, and the following scenarios will open.

First, the participant does not meet the inclusion criteria due to demonstrating a markedly positive state of psychological well-being, as indicated by a stress level <15.9 on the Kessler Psychological Distress Scale (K-10) [28], a general well-being score <3 on the General Health Questionnaire (GHQ) [39], and a general well-being score >80 on the “Mental Health” scale of the Short Form Health Survey 36 (SF-36) [29]. In such cases, the women will not be enrolled in the study, and the reasons for exclusion will be shared along with the screening questionnaire results, but upon the participant’s request.

Second, the participant does not meet the inclusion criteria due to exhibiting a markedly negative state of psychological well-being or signs of psychopathology, as indicated by a score <30 on mental health scale of the SF-36 and a K-10 score >29.9. In addition, in the second case, the participant will not be able to be enrolled in the study, and the reasons for exclusion from the study will be returned, and only upon request of the participant will the results obtained from the screening questionnaires be returned. Furthermore, indications of local services to contact for problems will be appropriately provided. This is to provide the participant with a structured and agreed system of referral to a qualified support service and taking charge where appropriate.

Third, the participant meets the inclusion criteria. In this case, the participant is enrolled in the study. The participant will also be asked if they would like to participate in a qualitative interview and in compiling a last questionnaire about stress 4 months after delivery. In case of positive feedback, the telephone number will be collected to facilitate contacts and follow-up interviews.

The follow-up interview will be carried out 4 months after the conclusion of the SH+ process; will last approximately 20 minutes; and, subject to the patient’s consent, will be audio-recorded to allow subsequent analysis. The interviews are structured with ad hoc items specifically designed to capture key characteristics of the study and participants’ experiences. They are conducted by psychologists from the design team. During the qualitative interview, the patient will also be asked if someone supported them during the intervention, how easy it was to carry out the exercises, and general personal perceptions about the usability and real-world application of this app. Interviews will be conducted 4 months after the end of the intervention and will last approximately 20 minutes.

Preliminary data analysis is scheduled for early November 2024, while the final results will be available at the end of the study. The publication of the results will take place within 1 year after the conclusion of the study. Given the potential for dropout, which is commonly observed in digital interventions, and the sensitive nature of our sample population, missing data will be carefully addressed during the analysis phase. Where appropriate, multiple imputation methods will be applied to account for missing values and reduce potential bias. In addition, participants who drop out or fail to complete the follow-up assessments will be included in the analysis using an intention-to-treat approach, wherever feasible.

Statistical data processing will be conducted using the software R (version 4.0.0; R Foundation for Statistical Computing) [47], SPSS (IBM Corp) statistics [48], Stata (version 17) [49], and Jeffreys’s Amazing Statistics Program [50].

Categorical variables will be summarized using absolute frequencies and percentage distributions, while quantitative variables will be summarized using specific measures of central tendency and variability. Descriptive analyses will cover psychological variables (stress level, anxiety level, depression level, emotional regulation, psychological flexibility, coping strategies, well-being, and health status) as well as UX and usability variables. Relationships between variables will be analyzed primarily using ad hoc statistical tests, including the chi-square test, Fisher exact test, paired t test, Wilcoxon signed rank test, and sign test, depending on the assumptions met. These tests will help identify differences between the beginning and end of the course within the same study sample for the variables under investigation. In addition, univariate logistic or multinomial regression models will be presented. To account for potential confounding factors, multiple regression models will be proposed, adjusting the effects of explanatory variables on outcome variables for possible confounders. Statistical significance will be found for each analysis with a P value ≤.05. The McNemar test [51] and Cochran Q test [52] will also be used to evaluate paired qualitative data.

A qualitative analysis will be conducted using a combination of text mining techniques and thematic analysis. With regard to the analysis of the final semistructured interviews, the text mining approach [53] will be used to extract the answers that appear repeatedly from the interviews that will be conducted with respect to the women’s experience of interacting with ALBA and with respect to how they felt during the study participation. In particular, frequencies will be analyzed by identifying recurring words or phrases and exploring the dominant lexicon. Classifying texts according to emotional tone (positive, negative, and neutral) will also allow us to analyze sentiment and polarity. This is very useful for analyzing feedback, reviews, and opinions [54]. The specific software or tools to support this analysis (eg, NVivo [55]; Lumivero) are currently being evaluated. The final selection will be based on their ability to efficiently code and interpret open-ended responses, ensuring alignment with the study’s objectives and methodological framework.

Finally, significant segments (quotes) will also be extracted, and then, all results will be interpreted in the light of the literature.