Keberhasilan pengobatan Tuberkulosis Resistan Obat (TBC RO) di DKI Jakarta masih sebesar 58% dari target 80%, beberapa faktor risiko sering dihubungkan dengan tingkat keberhasilan pengobatan. Penetitian ini bertujuan untuk mengetahui hubungan kepatuhan dan riwayat pengobatan terhadap keberhasilan pengobatan TBC RO dengan paduan jangka pendek 9-11 bulan. Metode penelitian yang dilakukan adalah kohort retrospektif untuk membandingkan kelompok terpapar dan tidak terpapar (rasio risiko/RR) terhadap tiingkat keberhasilan pengobatan. Sebanyak 364 pasien dengan keberhasilan pengobatan 62,36%, 59,07% patuh pengobatan dan 51,65% ada riwayat pengobatan. Kepatuhan pengobatan memiliki RR 9,14 (95%CI: 5,50-15,20,  p=<0,001) sedangkan riwayat pengobatan memiliki RR 1,06 (95%CI: 0,9-1,24, p=0,48) terhadap keberhasilan pengobatan. Analisis stratifikasi dan multivariat model kausal dengan metode backward menghasilkan umur, jenis kelamin, status bekerja, status gizi, dan koinfeksi HIV tidak terbukti sebagai variabel perancu maupun interaksi. Diperlukan penambahan jumlah sampel untuk bisa menunjukkan hubungan variabel kovariat dengan lebih jelas. Kepatuhan pengobatan menjadi salah satu kunci dalam peningkatan keberhasilan pengobatan TBC RO di DKI Jakarta.

---

The success rate of drug-resistant tuberculosis (DR-TB) treatment in DKI Jakarta is currently 58% of the 80% target, and various risk factors are often associated with treatment outcomes. The aim of this study was to determine the relationship between adherence and treatment history to the success of treatment of DR-TB with a short-term regimen of 9-11 months. The research method used was a retrospective cohort to compare exposed and unexposed groups (risk ratio/RR) on treatment success rate. A total of 364 patients with 62.36% treatment success, 59.07% were treatment adherent and 51.65% had a history of treatment. Treatment adherence had an RR of 9.14 (95%CI: 5.50-15.20, p=<0.001) while treatment history had an RR of 1.06 (95%CI: 0.9-1.24, p=0.48) on treatment success. Stratification and multivariate analysis of causal models using the backward method resulted in age, gender, working status, nutritional status, and HIV co-infection not being proven as confounding or interaction variables. An increase in sample size is needed to show the relationship between covariates more clearly. Treatment adherence is one of the keys to improving the success of DR-TB treatment in DKI Jakarta.