The Side Effect Analysis of Anti-Tuberculosis Drugs Using K-Means Clustering at RSUD Prof. Dr. H. Aloei Saboe Kota Gorontalo / Analisis Efek Samping Obat Anti Tuberkulosis Menggunakan K-Means Clustering di RSUD Prof. Dr. H. Aloei Saboe Kota Gorontalo

Background: The side effects of anti-tuberculosis drugs (OAT) experienced by patients often interfere with their daily activities and affect their compliance in completing relatively long-term treatment. This is one of the reasons patients discontinue OAT therapy unilaterally, which can lead to treatment failure for tuberculosis (TB). Objective: To analyse the side effects of OAT in patients at Prof. Dr. H. Aloei Saboe General Hospital in Gorontalo City using k-means clustering. Methods: Data including gender, age, OAT regimen, laboratory results, comorbidities, and types of OAT side effects were analyzed using the k-means method to determine patient clustering patterns. Research Results: The analysis yielded three clusters. Cluster 1 (25 patients) was predominantly male (80%), aged 45–54 years, received first-line OAT (88%), experienced elevated SGPT/SGOT levels (88%), had hypertension as the most common comorbidity (28%), and primarily experienced liver dysfunction as the main side effect (96%). Cluster 2 (348 patients) was predominantly male (58%), aged 35–44 years, receiving OAT line 1 (96.1%), no increase in SGPT and SGOT levels (0%), almost no increase in urea and creatinine levels (0.2%), the most common comorbidity being diabetes mellitus (22.1%). The main side effect being gastrointestinal disorders (58.9%). Cluster 3 (40 patients) was predominantly male (70%), aged 45–54 years, received OAT line 1 (97.5%), experienced an increase in urea and creatinine levels (97.5%), and the most common comorbidity was diabetes mellitus (47.5%). The main side effect was renal dysfunction (95%). Conclusion: The k-means algorithm is effective in generating clusters of patient characteristics. This clustering supports specific interventions such as comorbidity therapy management and monitoring of side effect risks, thereby optimising individualised tuberculosis (TB) treatment.

Latar Belakang: Efek samping obat anti tuberkulosis (OAT) yang dialami pasien sering mengganggu aktivitas sehari-hari pasien dan berdampak pada kepatuhan pasien dalam menyelesaikan pengobatan yang relatif lama. Hal ini menjadi salah satu penyebab pasien menghentikan terapi OAT secara sepihak, akibatnya akan berdampak pada kegagalan pengobatan tuberkulosis (TB). Tujuan: Menganalisis efek samping OAT pada pasien di RSUD Prof. Dr. H. Aloei Saboe Kota Gorontalo menggunakan k-means clustering. Metode: Data meliputi jenis kelamin, usia, lini OAT, hasil laboratorium, komorbiditas dan jenis efek samping OAT, di analisis menggunakan metode k-means untuk menentukan pola pengelompokan pasien. Hasil Penelitian: Analisis menghasilkan 3 (tiga) klaster. Klaster 1 (25 pasien) mayoritas laki-laki (80%), usia 45-54 tahun, menerima lini OAT 1 (88%), mengalami kenaikan nilai SGPT SGOT (88%), komorbiditas terbanyak hipertensi (28%), dan efek samping utama gangguan hati (96%). Klaster 2 (348 pasien) didominasi laki-laki (58%), usia 35-44 tahun, menerima lini OAT 1 (96,1%), tidak ada kenaikan nilai SGPT SGOT (0%), hampir tidak ada kenaikan nilai ureum kreatinin (0,2%), komorbiditas terbanyak diabetes melitus (22,1%) dan efek samping utama gangguan gastrointestinal (58,9%). Klaster 3 (40 pasien) mayoritas laki-laki (70%), usia 45-54 tahun, menerima lini OAT 1 (97,5%), mengalami kenaikan nilai ureum kreatinin (97,5%), komorbiditas terbanyak diabetes melitus (47,5%) dan efek samping utama gangguan ginjal (95%). Kesimpulan: Algoritma k-means efektif dalam menghasilkan klaster karakteristik pasien. Pengelompokan tersebut mendukung intervensi spesifik seperti manajemen terapi komorbiditas dan pemantauan risiko efek samping, sehingga dapat mengoptimalkan pengobatan tuberkulosis (TB) secara individual.