Pengembangan Sistem Rekomendasi Antibiotik untuk ISPA dengan Pendekatan Data-Driven Berbasis Referensi Klinis dan Riwayat Penggunaan Obat

Tantri, Meidina (2025) Pengembangan Sistem Rekomendasi Antibiotik untuk ISPA dengan Pendekatan Data-Driven Berbasis Referensi Klinis dan Riwayat Penggunaan Obat. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5023211024-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Perkembangan resistensi antibotik kian menjadi ancaman kesehatan global, pada kasus Infeksi Saluran Pernapasan Akut (ISPA) yang dapat disebabkan oleh virus ataupun bakteri, hal berkaitan dengan penggunaan antibiotik dalam pengobatannya. Metode deteksi resistensi konvesional seringkali membutuhkan biaya yang tidak murah, waktu yang tidak sedikit, dan sumberdaya yang memadai. Penelitian ini bertujuan untuk mengembangkan sistem rekomendasi antibiotik yang mengintegrasikan berbagai sumber data klinis, seperti pedoman WHO, Comprehensive Antibiotic Resistance Database (CARD), Anatomical Therapeutic Chemical Classification System (ATC), Kyoto Encyclopedia of Genes and Genomes (KEGG), serta data Simplified Molecular Input Line Entry System (SMILES). Pendekatan rule-based diterapkan untuk mengevaluasi apakah antibiotik diperlukan pasien berdasarkan gejala yang dialami dan pedoman WHO. Rekomendasi antibiotik alternatif diperoleh melalui tahap penyaringan data dengan menggunakan kode ATC(J01 untuk penggunaan sistemik manusia) untuk memastikan relevansi klinis. Kemiripan struktur molekul dihitung dengan menggunakan koefisien Tanimoto dan validasi tambahan melalui klasifikasi KEGG (memastikan perbedaan kelas farmakologis alternatif). Sistem dirancang untuk merekomendasikan antibiotik lini pertama sesuai dengan paedoman WHO dan menyediakan alternatif melalui analisis berdasarkan kemiripan struktur molekulnya. Pengujian fungsionalitas menunjukkan sistem berhasil menyaring 57,21% antibiotik yang tidak relevan untuk penggunaan sistemik manusia dari CARD. Analisis kemiripan struktur dan sebagian besar antibiotik alternatif yang memiliki kemiripan struktural rendah (koefisien Tanimoto 0.1-0.4) berasal dari kelas farmakologis yang berbeda, hal ini mengindikasikan potensi alternatif lebih rendah untuk cross-resistance pada antibiotik terdampak dalam dokumentasi ilmiah. Implementasi Backend (FastAPI) dan frontend (HTML, CSS, JavaScript, Bootstrap) berfungsi optimal, dilengkapi dengan sistem otentikasi JWT yang aman, kontrol akses berdasarkan role, enkripsi data pasien (Fernet), pencatatan log aktivitas, dan notifikasi email. Hasil menunjukkan sistem mampu merekomendasikan antibiotik secara rasional dan berbasis data, serta mendukung pengambilan keputusan klinis dalam pengobatan ISPA yang lebih aman dan efisien.
=======================================================================================================================================
The development of antibiotic resistance is increasingly becoming a global health threat, in the case of Acute Respiratory Tract Infection (ARTI) which can be caused by viruses or bacteria, related to the use of antibiotics in its treatment. Conventional resistance detection methods often require a lot of money, time, and resources. This study aims to develop an antibiotic recommendation system that integrates various clinical data sources, such as WHO guidelines, Comprehensive Antibiotic Resistance Database (CARD), Anatomical Therapeutic Chemical Classification System (ATC), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Simplified Molecular Input Line Entry System (SMILES) data. A rule-based approach is applied to evaluate whether antibiotics are needed by patients based on their symptoms and WHO guidelines. Alternative antibiotic recommendations were obtained through a data filtering stage using ATC codes (J01 for human systemic use) to ensure clinical relevance. Molecular structure similarity was calculated using Tanimoto coefficient and additional validation through KEGG classification (ascertaining the different pharmacological classes of alternatives). The system was designed to recommend first-line antibiotics according to WHO guidelines and provide alternatives through analysis based on the similarity of their molecular structures. Functionality testing showed the system successfully filtered out 57.21% of antibiotics irrelevant for human systemic use from CARD. Structural similarity analysis and most of the alternative antibiotics that have low structural similarity (Tanimoto coefficient 0.1-0.4) are from different pharmacological classes, indicating lower potential of alternatives for cross-resistance to affected antibiotics in scientific documentation. The Backend (FastAPI) and frontend (HTML, CSS, JavaScript, Bootstrap) implementations function optimally, equipped with a secure JWT authentication system, , role-based access control, patient data encryption (using Fernet), activity logging, and email notifications. The results show that the system is able to recommend antibiotics in a rational and data-driven manner, and support clinical decision-making for the treatment of ARI that is safer and more efficient.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Resistensi Antibiotik, ISPA, Keputusan Berbasis Data, Sistem Berbasis Aturan, Precision Medicine, Antibiotic Resistance, ARTI, Data-Driven, Rule-Based System, Precision Medicine
Subjects:	Q Science > QA Mathematics > QA76.76.E95 Expert systems Q Science > QA Mathematics > QA76.9.I58 Recommender systems (Information filtering) Q Science > QA Mathematics > QA76.9D338 Data integration T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.62 Decision support systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User:	Meidina Tantri
Date Deposited:	05 Aug 2025 04:22
Last Modified:	05 Aug 2025 04:22
URI:	http://repository.its.ac.id/id/eprint/127347

Actions (login required)

View Item