Diagnostic and treatment delays and associated factors among RR-TB patients under molecular drug susceptibility testing: a retrospective study in Guizhou Province, China

Rifampicin-resistant tuberculosis (RR-TB) constitutes a persistent and formidable challenge for global TB control programs, as delayed diagnosis and treatment initiation contribute to transmission and poor outcomes. Molecular drug susceptibility testing (mDST) enables rapid detection of resistance and may shorten these delays. This study evaluated the impact of mDST on diagnostic and treatment delays among RR-TB patients in Guizhou Province, China, and explored factors associated with delays under mDST. All RR-TB cases registered in Guizhou between January 2019 and December 2023 were included. Patients were classified into the mDST or phenotypic DST (pDST) groups. Diagnostic delay was defined as the primary outcome, and treatment initiation delay as the secondary outcome. Diagnostic and treatment delays were compared using chi-square and Mann–Whitney U tests. Multivariable logistic regression identified predictors of diagnostic delay, including time-varying effects, and treatment delays in the mDST group. Among 2,661 RR-TB patients, 93.3% were diagnosed using mDST. The overall median diagnostic delay was 15 (2–62) days, significantly shorter in the mDST group than in the pDST group [14 (2–49) vs. 57 (1–164) days; Z = − 4.262, P < 0.001]. The diagnostic delay rate was lower with mDST group than in the pDST group (48.3% vs. 56.8%; $${\chi}^{2}$$= 6.701, P = 0.010). Time-interaction analysis showed that the diagnostic delay associated with pDST was pronounced during 2019–2021 but attenuated over time (Pinteraction = 0.003). In the mDST group, having comorbidities (OR = 1.56, P = 0.013) were associated with diagnostic delay, while patients detected through seeking consultations (OR = 1.47, P < 0.001) or follow-up (OR = 1.35, P = 0.004) were more likely to experience delay than those referred from other facilities. Among 2,508 patients who initiated treatment, the median time to treatment was 5 (0–22) days, significantly shorter in the mDST group than in the pDST group [4 (0–20) vs. 14 (0–91) days; Z = − 6.045, P < 0.001]. A higher proportion of mDST patients began treatment within 5 days (52.9% vs. 41.3%; $${\chi}^{2}$$= 10.917, P < 0.01). In the mDST group, intra-city mobility (OR = 1.64, P = 0.038) and diagnosis at county-level hospitals (OR = 1.97, P = 0.008) predicted treatment delay. Large-scale implementation of mDST substantially shortened diagnostic and treatment initiation intervals for RR-TB patients in Guizhou Province. However, delays persisted among ethnic minorities, intra-city migrants, and patients diagnosed at county-level facilities. Strengthening diagnostic-to-treatment referral pathways and expanding mDST access in under-resourced areas are essential to further improve the RR-TB care continuum in high-burden settings. Not applicable.