Accuracy of AI App, Breast Cancer, Breast Neoplasm, Chemotherapy-Related Toxicities, Medication Adherence, Quality of Life in Cancer Patients, Symptom Management
Conditions
Keywords
Artificial Intelligence, Mobile Application, digital health, supportive care, chemotherapy, patient education, Symptom Monitoring, Human-in-the-Loop AI, Breast Cancer Chemotherapy, Oncology, Arabic Language App, Pharmacy Iraq, Randomized Controlled Trial
Brief summary
The goal of this clinical trial is to learn if an Arabic-language mobile application that uses artificial intelligence (AI) can help women with breast cancer during chemotherapy. The app is designed to give personalized support by reminding participants about their medications, teaching them how to manage treatment side effects, and alerting their healthcare team about serious symptoms. The main questions this study aims to answer are: 1. Does the AI-based mobile app provide accurate and safe recommendations for the patients? 2. Does using the AI-based mobile app help lower treatment-related symptoms and side effects compared to usual care? 3. Does the app help participants take their medications more regularly? 4. Does it increase participants' understanding and satisfaction with the information they receive about their treatment? Researchers will compare two groups: Group 1: Participants who use the AI-based mobile app plus usual oncology care. Group 2: Participants who receive usual care only. Participants will: 1. Use the mobile app daily for 12 weeks while receiving chemotherapy. 2. Complete short questionnaires about symptoms, medication use, and quality of life at the start and end of the study. 3. Report any problems or feedback about using the app. The AI app is for support and education only. It does not make treatment decisions. All information from the app will be reviewed by oncologists and pharmacists to ensure participant safety.
Detailed description
Despite advances in oncology care, breast cancer patients in Iraq face significant challenges regarding medication adherence and symptom management during the inter-cycle chemotherapy periods. This randomized controlled trial aims to bridge this gap by evaluating the efficacy, safety, and feasibility of a specialized, Arabic-language Artificial Intelligence (AI) mobile application. Current standard care in the local setting often relies on episodic clinic visits, leaving patients without real-time support for side effects experienced at home. This study hypothesizes that a continuous, AI-driven digital intervention can reduce symptom burden and improve adherence to chemotherapy and supportive care medications (e.g., antiemetics) compared to standard care alone. The application utilizes Natural Language Processing (NLP) to provide conversational support tailored specifically to the cultural and linguistic context of Iraqi patients. The intervention integrates a "Human-in-the-loop" safety model to ensure clinical accuracy. The AI algorithms are trained on clinical practice guidelines adapted for the local formulary. Symptom Triage Logic: The app utilizes an algorithm based on the CTCAE grading system. Low-grade symptoms trigger self-care advice (e.g., hydration, dietary changes), while high-grade symptoms trigger immediate alerts to the patient to seek care and a notification to the study investigators. Adherence Algorithms: Unlike static alarms, the notification system adapts to the specific chemotherapy cycle (e.g., AC or Taxane-based regimens) to remind patients of specific supportive medications required on specific days. Control Group Specification (Standard of Care) Participants randomized to the control arm will receive the institutional standard of care. This includes routine oncologist consultations, standard written or verbal discharge instructions regarding chemotherapy side effects, and pharmacy dispensing counseling. They will not have access to the interactive AI features but will undergo the same schedule of outcome assessments to ensure rigorous comparison. This study represents the first empirical effort to integrate AI-driven digital health tools into the public oncology sector in Iraq. It aims to validate whether automated, algorithmic triage is a feasible addition to the healthcare infrastructure in low-resource settings.
Interventions
BEHAVIORALAI-Based Mobile Application for Personalized Chemotherapy Support
The intervention is an Arabic-language mobile application powered by artificial intelligence (AI) designed to provide personalized chemotherapy support for women with breast cancer. The app assists participants by monitoring symptoms, sending medication adherence reminders, and offering educational content on managing side effects and improving treatment understanding. It uses a conversational interface based on natural language processing (NLP) to communicate with users. Participants are asked to use the app daily for 12 weeks while receiving chemotherapy. A human-in-the-loop system ensures oncologists and pharmacists review AI-generated advice for accuracy and safety.
OTHERUsual Care
Standard oncology care provided by the hospital team, including chemotherapy administration, routine follow-up, and patient education according to local protocols.
Sponsors
Dena h. Al-Tameemi
Baghdad Medical City
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
SUPPORTIVE_CARE
Masking
NONE
Eligibility
Sex/Gender
FEMALE
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Confirmed diagnosis of breast cancer stages I, II, or III. * Patients must be currently scheduled to initiate their first-ever cycle of chemotherapy. * Age 18 years or older. * Ability to understand and provide informed consent. * Possession of a smartphone (Android or iOS) and functional digital literacy, defined as the ability to independently navigate mobile applications, read on-screen text in Arabic, and input daily health data. (for the intervention group). * Willingness to comply with study procedures and follow-up schedules. * Ability to communicate in Arabic, as the mobile application and chatbot will be developed in Arabic.
Exclusion criteria
* Patients with Stage IV (Metastatic) breast cancer. * Patients receiving concurrent hormonal therapy during the chemotherapy phase, to isolate chemotherapy-induced adverse events. Patients with cognitive impairment or severe psychiatric disorders that would preclude effective interaction with the mobile application or questionnaire completion. * Patients receiving palliative care where symptom management is the sole focus and active chemotherapy is not being administered with curative or life prolonging intent. * Patients participating in other interventional clinical trials that might confound the outcomes of this study. * Patients with severe comorbidities that could significantly impact their ability to participate or bias outcome measures.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Change in Symptom Burden and Chemotherapy-Related Toxicities (Arabic PRO-CTCAE) | Up to 12 weeks | Symptom burden and treatment-related toxicities will be assessed using the validated Arabic version of the Patient-Reported Outcomes-Common Terminology Criteria for Adverse Events (PRO-CTCAE). Participants will report frequency, severity, and interference of common chemotherapy-related symptoms such as fatigue, nausea, vomiting, pain, and neuropathy. Mean score changes between baseline and 12 weeks will be compared between the AI app group and the usual care group to evaluate whether the intervention lowers symptom burden and improves self-management.Data will be collected at Baseline and weekly thereafter. |
| Medication Adherence Score (Arabic MMAS-8 | Baseline and at weeks 6,12. | Medication adherence will be measured using the 8-item Morisky Medication Adherence Scale (MMAS-8) in its validated Arabic version. Participants' self-reported responses generate a score from 0-8, with higher scores indicating better adherence. The mean change in adherence scores from baseline to 12 weeks will be compared between groups to determine the app's effectiveness in promoting medication adherence during chemotherapy. |
| Change in Patient Knowledge and Information Satisfaction (EORTC QLQ-INFO25) | Baseline and at weeks 6,12. | The Arabic version of EORTC QLQ-INFO25 questionnaire will assess participants' perception of the adequacy, clarity, and usefulness of information provided about their disease, treatment, and care. The total score ranges from 0 to 100, with higher scores reflecting greater satisfaction with information. The AI-based mobile app is expected to increase knowledge and satisfaction through personalized, accessible education. |
| Change in General Quality of Life (EORTC QLQ-C30) | Baseline and at weeks 6,12. | Quality of life (QoL) will be measured using the validated Arabic version of the EORTC QLQ-C30. The EORTC QLQ-C30 is a 30-item questionnaire used to assess the quality of life of cancer patients. It incorporates five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain, and nausea/vomiting), and a global health status/QoL scale. Raw scores are transformed to a linear scale ranging from 0 to 100. Functional Scales \& Global Health Status: Higher scores represent a higher/better level of functioning and quality of life. Symptom Scales: Higher scores represent a higher/worse level of symptomatology. |
| Change in Breast Cancer-Specific Quality of Life (EORTC QLQ-BR23) | Baseline and at weeks 6,12. | The Arabic version of EORTC QLQ-BR23 will be used to assess the change in Breast Cancer-Specific Quality of Life.The EORTC QLQ-BR23 is a supplementary module specifically for breast cancer patients. It consists of 23 items covering functional scales (body image, sexual functioning, future perspective) and symptom scales (side effects of systemic therapy, breast symptoms, arm symptoms, hair loss). Raw scores are transformed to a linear scale ranging from 0 to 100. Functional Scales: Higher scores represent better functioning. Symptom Scales: Higher scores represent worse symptoms. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Accuracy of AI-Generated Advice Compared with Oncologist Assessment | Through study completion, an average of 12 weeks. | The accuracy and clinical appropriateness of the AI-generated recommendations will be evaluated by comparing the app's advice logs to oncologist judgments on the same patient-reported scenarios. Agreement will be assessed using Cohen's kappa coefficient to determine the reliability of the AI-based symptom triage. |
| User Satisfaction and App Usability Score | At 12 weeks (end of intervention) | Participants will evaluate the application using a structured questionnaire assessing multiple domains including ease of use, perceived usefulness, information clarity, and trust in AI recommendations. Responses are measured on a 5-point Likert scale: 1 = Strongly Disagree 2 = Disagree 3 = Neutral 4 = Agree 5 = Strongly Agree The final outcome is reported as an aggregate average score across all items. Scale Range: 1 to 5. Interpretation: Higher scores indicate higher user satisfaction |
| App Usage Frequency | Continuously during the 12-week intervention. | Engagement is assessed by monitoring the backend usage logs to determine the average number of distinct application sessions (log-ins) per participant. Unit of Measure: Sessions per week |
| Identification of Technical, Ethical, and Implementation Barriers | After 12 weeks of intervention use and upon study completion. | Qualitative data will be collected through semi-structured interviews to identify challenges related to app usability, privacy concerns, data reliability, and clinical integration. Thematic analysis will be conducted to inform strategies for safe and effective implementation of AI-based supportive care tools in Iraqi oncology settings. |
Countries
Iraq
Contacts
CONTACTDena A. Al-Tameemi, MSc.
CONTACTSamer Imad Mohammed, PhD
PRINCIPAL_INVESTIGATORSamer Imad Mohammed, Assistant Prof
University of Baghdad-College of Pharmacy
Outcome results
None listed