Related Experiment Video
Updated: Oct 12, 2025

Digital Home-Monitoring of Patients after Kidney Transplantation: The MACCS Platform
Published on: April 12, 2021
A Data-Driven Intervention Framework for Improving Adherence to Growth Hormone Therapy Based on Clustering Analysis
Matheus Araújo1, Paula van Dommelen2, Jaideep Srivastava1
1Computer Science Department, University of Minnesota, Minneapolis, MN, USA.
A new traffic light system effectively identifies children with poor adherence to recombinant human growth hormone (r-hGH) therapy. This data-driven approach enables early intervention to improve treatment outcomes for growth hormone deficiency (GHD).
Area of Science:
- Endocrinology and Metabolism
- Health Informatics
- Pediatric Endocrinology
Background:
- Recombinant human growth hormone (r-hGH) therapy is crucial for treating growth hormone deficiency (GHD).
- Suboptimal patient adherence and persistence significantly hinder achieving full height potential in children undergoing r-hGH treatment.
- Real-world data and data-sharing agreements are essential for developing effective adherence monitoring tools.
Purpose of the Study:
- To develop and validate a data-driven clinical decision support system using "traffic light" visualizations for managing adherence risks in pediatric r-hGH therapy.
- To categorize patient adherence patterns into distinct risk levels (high, medium, low) for proactive intervention.
- To identify the most effective metrics and thresholds for an alert system to monitor r-hGH adherence.
Main Methods:
- Utilized real-world adherence data from 11,015 children receiving r-hGH therapy for ≥180 days, sourced from easypod™ connect.
- Employed cluster analysis with a Gaussian mixture model to categorize adherence patterns based on daily adherence and hours to next injection.
- Developed a traffic light system (green, yellow, red) to visualize adherence risk, optimizing thresholds using Area Under the Receiver Operating Characteristic Curve (AUC-ROC).
Main Results:
- The most effective adherence monitoring utilized the standard deviation of hours to the next injection, achieving an AUC-ROC of 0.85.
- Established optimal thresholds for daily adherence: >0.82 (high), 0.53-0.82 (medium), <0.53 (low).
- Defined optimal thresholds for hours to next injection: <27.18 (high), 27.18-34.01 (medium), >34.01 (low).
Conclusions:
- A practical, data-driven alert system using traffic light coding can effectively monitor r-hGH adherence in pediatric patients.
- Early identification of sub-optimally adherent patients allows for timely intervention, potentially improving treatment outcomes.
- This system empowers healthcare practitioners to proactively manage adherence and enhance the efficacy of growth hormone therapy.
Related Concept Videos
Therapeutic Drug Monitoring: Affecting Factors
Drug Therapy
Antianxiety Medications
SBAR II: Application of SBAR
SBAR Report from a Nurse to a Health Care Provider
S: "Hello, Dr. Smith. This is Jane, RN, from the Med Surg unit. I am calling to tell you about Ms. White in Room 210, who is experiencing increased pain and redness at her incision site. Her recent...
Therapeutic Drug Monitoring: Overview and Classification
Drug Dosing: Geriatric Patients

