Journal of Advanced Health Informatics Research

Development of Smartotc Mobile Application with Clinical Decision Support System (CDSS) to Support Pharmacists in Managing Common Cold

Ikhwan Yuda Kusuma — 2026-07-31

Common cold is a highly prevalent self-limiting respiratory illness and a frequent reason for self-medication and community pharmacy consultations. Variability in pharmacists’ clinical decision-making may affect the quality of care, highlighting the need for evidence-based decision-support tools. This study developed SmartOTC, a mobile application incorporating a Clinical Decision Support System (CDSS) to support pharmacists in managing common cold cases through symptom assessment, over-the-counter medication recommendations, patient education, and referral support. A pilot quasi-experimental before-and-after study was conducted among 10 community pharmacists in Indonesia as part of a preliminary evaluation of the application. The application was developed using Android Studio and a Python-based CDSS algorithm and evaluated through functional feasibility, task performance, usability, user experience, and expert validation assessments. SmartOTC demonstrated high technical performance, achieving a 92.73% success rate, a 0% crash rate, and a mean response time of 2.89 seconds. Significant improvements in task performance and response times were observed following implementation. However, these findings should be interpreted within the context of a pilot study with a limited sample size.

The application demonstrated positive user experience and strong content validity, while usability outcomes suggested opportunities for further improvement. The findings are further limited by the single-site study setting and the absence of patient-level outcome evaluation. SmartOTC shows promise as a digital decision-support tool to support pharmacists in the management of common cold cases and warrants further evaluation in real-world pharmacy practice.

Development of a Pilot Chatbot to Detect Interactions Between Monotherapy and Combination Therapy with Captopril Using a Local (LLM) in Cardiovascular Therapy

Ikhwan Yuda Kusuma — 2026-07-31

Cardiovascular diseases (CVDs), including hypertension and heart failure, remain major contributors to global morbidity and mortality and frequently require long-term polypharmacy. The concurrent use of multiple medications increases the risk of drug–drug interactions (DDIs), which may reduce therapeutic effectiveness, increase toxicity, and contribute to adverse drug reactions. This study aimed to develop and validate a locally deployed large language model (LLM)-based chatbot for detecting drug–drug interactions involving captopril in both monotherapy and combination therapy settings. A development and validation study was conducted using DDI data obtained from DrugBank and Drugs.com. The chatbot was developed using the LLaMA-3 model integrated with LangChain, Ollama, and FastAPI and was evaluated through iterative testing and 5-fold cross-validation. System performance was assessed using accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), while usability was evaluated using the System Usability Scale (SUS) and Single Ease Question (SEQ) questionnaires completed by pharmacists. The chatbot demonstrated progressive performance improvement throughout development and achieved excellent performance during final validation involving 300 drug pairs, with an accuracy of 99.0%, sensitivity of 100.0%, specificity of 98.0%, PPV of 100.0%, and NPV of 92.0%, exceeding all predefined acceptance thresholds. Usability testing indicated only fair to moderate usability, with a mean SUS score of 65.0 and a mean SEQ score of 5.0, suggesting that further refinement of the user interface and workflow may be required. The locally deployed LLM-based chatbot demonstrated satisfactory diagnostic performance and preliminary feasibility for captopril-related DDI screening. Although the system showed promising classification performance, additional usability optimization and evaluation in real-world clinical workflows are needed before broader implementation can be considered as a pharmacist-supportive screening system.

Knowledge Distillation in Lightweight U-Net Transformer Architectures for Brain Tumor Segmentation

Toat Tuloh — 2026-07-31

Brain tumor segmentation from Magnetic Resonance Imaging was an essential step for therapy planning, prognosis evaluation, and treatment monitoring in glioma patients. Manual delineation required substantial time and was prone to inter-observer variability. Although deep learning models achieved high segmentation accuracy, performance improvements were often accompanied by increased computational complexity, limiting their applicability in resource-constrained clinical environments. To address this issue, an efficiency-oriented segmentation framework was developed based on a lightweight three-dimensional U-Net enhanced with a shallow Transformer module and guided by knowledge distillation. The main contribution of this study was the integration of logit-level and feature-level distillation to improve segmentation capability while maintaining low inference complexity. The framework emphasized a balanced trade-off between segmentation accuracy and computational efficiency rather than benchmark maximization. Experiments were conducted using the BraTS 2020 and BraTS 2021 datasets. The official training sets were internally split into eighty percent for training and twenty percent for validation. The student network was trained using a hybrid segmentation loss combined with temperature-scaled logit distillation and bottleneck feature alignment from a higher-capacity teacher model. Model performance was evaluated using Dice score, Intersection over Union, and computational complexity measured in floating-point operations. On the BraTS 2021 dataset, the proposed model achieved Dice scores of 0.6538 for Whole Tumor, 0.5382 for Tumor Core, and 0.5304 for Enhancing Tumor. Per-class Dice values were 0.9706 for background, 0.3028 for necrotic or non-enhancing tumor core, 0.4938 for edema, and 0.4936 for enhancing tumor. The corresponding Intersection over Union values followed similar trends. The model maintained an inference complexity of approximately 54.38 gigafloating-point operations for input patches of size 64 × 64 × 64. These findings indicated that the proposed framework achieved a stable balance between segmentation performance and computational efficiency, supporting practical deployment under limited computational resources

Usability and Acceptability of a Co-Designed Menstruation Tracker App for Indonesian Adolescent Girls

Arlyana Hikmanti — 2026-07-31

Background: Many adolescent girls experience limited access to accurate and age-appropriate menstrual health information, which may contribute to anxiety, stigma, and suboptimal menstrual hygiene practices. Existing menstrual tracking applications are often designed for adult users and may not address adolescent needs in local cultural contexts. Objective: This study evaluated the perceived usability and acceptability of Srikandi Sehat, a co-designed menstruation tracker application for Indonesian adolescent girls. Method: A prospective usability evaluation was conducted among 151 adolescent girls aged 13-18 years who used the application during three menstrual cycles. The application was developed through a co-design process involving adolescent users and health workers. Data were collected using electronic questionnaires on menstrual knowledge, technology-related attitudes, and application usability. Descriptive statistics and Spearman correlation analyses were used because knowledge and usability scores were not normally distributed. Results: The modified usability scale showed a mean total score of 133.35 out of 160 (SD 19.72), indicating favorable perceived usability within the scoring framework used in this study. Participants reported that the application was easy to navigate, visually understandable, age-appropriate, and useful for recording menstrual cycles and symptoms. Attitude toward menstrual health showed a stronger association with usability (rho approximately 0.53, p < 0.001) than knowledge (rho approximately 0.21, p = 0.010). These findings indicate association rather than causation. Conclusion: Srikandi Sehat was perceived as usable and acceptable by participating adolescent girls. Further studies should validate the adapted instruments, include objective usage analytics, and evaluate educational or behavioral outcomes using controlled and longer-term designs.

Predicting Childbirth Complications in High-Risk Pregnant Women Using Machine Learning

Surtiningsih Surtiningsih — 2026-07-31

Childbirth complications remain one of the leading causes of high maternal mortality rates in Indonesia, primarily due to delayed detection at referral healthcare facilities, which hinders timely intervention. Most previous machine learning studies have focused on predicting a single outcome, even though high-risk pregnancies are prone to experiencing multiple complications simultaneously. This study aims to develop an explainable machine learning-based multi-class classification model utilizing Electronic Medical Record (EMR) data to support the early identification of six major childbirth complications: hypertension, preeclampsia, prolonged labor, premature rupture of membranes, hemorrhage, and infection. The study employed a retrospective cohort design involving 10,644 maternal records from the 2020–2024 period at Margono Soekarjo General Hospital. A total of 16 maternal, obstetric, clinical, and laboratory variables were analyzed using the Extreme Gradient Boosting (XGBoost) algorithm. The dataset was divided into training (80%) and testing (20%) sets, with class imbalance addressed through oversampling. Model evaluation was performed using accuracy, F1-score, and AUC, while interpretability was analyzed using Shapley Additive Explanations (SHAP). The model achieved an accuracy of 70%, a macro F1-score of 0.65, and a macro-AUC of 0.92, demonstrating good discriminatory ability. SHAP analysis identified metabolic, hemodynamic, and hematological factors as the primary predictors. These findings suggest the model’s potential as a tool for early risk stratification and clinical decision support, although external validation is still required