A Machine Learning Model to Predict Patients’ Adherence Behavior and a Decision Support System for Patients With Metastatic Breast Cancer: Protocol for a Randomized Controlled Trial
Bibliographic record
Abstract
BACKGROUND: Adherence to oral anticancer treatments is critical in the disease trajectory of patients with breast cancer. Given the impact of nonadherence on clinical outcomes and the associated economic burden for the health care system, finding ways to increase treatment adherence is particularly relevant. OBJECTIVE: The primary end point is to evaluate the effectiveness of a decision support system (DSS) and a machine learning web application in promoting adherence to oral anticancer treatments among patients with metastatic breast cancer. The secondary end point is to collect a set of new physical, psychological, social, behavioral, and quality of life predictive variables that could be used to refine the preliminary version of the machine learning model to predict patients' adherence behavior. METHODS: This prospective, randomized controlled study is nested in a large-scale international project named "Enhancing therapy adherence among metastatic breast cancer patients" (Pfizer 65080791), aimed to develop a predictive model of nonadherence and associated DSS and guidelines to foster patients' engagement and therapy adherence. A web-based DSS named TREAT (treatment adherence support) was developed using a patient-driven approach, with 4 sections, that is, Section A: Metastatic Breast Cancer; Section B: Adherence to Cancer Therapies; Section C: Promoting Adherence; and Section D: My Adherence Diary. Moreover, a machine learning-based web application was developed to predict patients' risk factors of adherence to anticancer treatment, specifically pertaining to physical status and comorbid conditions, as well as short and long-term side effects. Overall, 100 patients consecutively admitted at the European Institute of Oncology (IEO) at the Division of Medical Senology will be enrolled; 50 patients with metastatic breast cancer will be exposed to the DSS and machine learning web application for 3 months (experimental group), and 50 patients will not be exposed to the intervention (control group). Each participant will fill a weekly medication diary and a set of standardized self-reports evaluating psychological and quality of life variables (Adherence Attitude Inventory, Beck Depression Inventory-II, Brief Pain Inventory, 13-item Sense of Coherence scale, Brief Italian version of Cancer Behavior Inventory, European Organization for Research and Treatment of Cancer Quality of Life 23-item Breast Cancer-specific Questionnaire, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, 8-item Morisky Medication Adherence Scale, State-Trait Anxiety Inventory forms I and II, Big Five Inventory, and visual analogue scales evaluating risk perception). The 3 assessment time points are T0 (baseline), T1 (1 month), T2 (2 months), and T3 (3 months). This study was approved by the IEO ethics committee (R1786/22-IEO 1907). RESULTS: The recruitment process started in May 2023 and is expected to conclude on December 2023. CONCLUSIONS: The contribution of machine learning techniques through risk-predictive models integrated into DSS will enable medication adherence by patients with cancer. TRIAL REGISTRATION: ClinicalTrials.gov NCT06161181; https://clinicaltrials.gov/study/NCT06161181. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/48852.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Direct model labels (unvalidated)
Per-model category and study-design labels from the labeling rounds. They are machine output, unvalidated, and the disagreement between models ships as data. No study design here is MEDLINE-validated yet.
| Model arm | Categories | Study design | Confidence |
|---|---|---|---|
| gemma | no category Domain: not available · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Randomized trial | low |
| gpt | no category Domain: not available · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Randomized trial | high |
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedLabeled directly by 2 models reading the full record.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".