This study linked national policy, hospital systems, and provider behaviour in Indonesia, Nepal, Thailand, and Viet Nam during 2022–2023. It included 486 participants across 16 hospitals. All countries had antimicrobial-stewardship policies, but financing, enforcement, training, microbiology-data use, and inter-programme coordination were uneven. Hospitals averaged 72% on G-ASET, with tracking and reporting highest at 75% and training lowest at 66%. The two Thai hospitals scored 81.3% and 80.6% overall, yet both scored only 50% for training. These results identify actionable system gaps; they do not experimentally show that any intervention reduced antibiotic use, resistance, or mortality.
Key findings
- All four countries had national AMS policies, but only Indonesia and Nepal incorporated financing mechanisms in the assessed frameworks; no overarching policy funding mechanism was identified for Thailand or Viet Nam in 2023. Shared gaps included primary-care guidance, programme links, training, prescription-only enforcement, shortage tracking, and data-informed feedback.
- Across 16 hospitals, mean G-ASET performance was 241/335 (72%). Tracking, monitoring, and reporting was highest at 56/75 (75%), while training was lowest at 17/25 (66%). A 24-hour microbiology laboratory existed in 14/16 hospitals and electronic AMU/AMR data in 15/16, but decision-support information systems in only 5/16.
- The two Thai hospitals scored 272.5/335 (81.3%) and 270/335 (80.6%), yet both scored 50% for training. Across the network, 117/292 (40%) KAP respondents perceived antibiotic overuse and only 79/161 (49%) prescribers considered non-physician views. Thailand contributed only two hospitals and no focus groups, so these are not national estimates.
Why this matters globally
The study is globally useful because it demonstrates a multilevel WHO/G-ASET approach for comparing actionable stewardship gaps in middle-income settings without reducing the problem to prescriber behaviour alone. It supports coordinated investment in guidance, workforce, laboratories, data systems, financing, and patient communication, especially for smaller hospitals and primary care. However, it maps implementation rather than effectiveness; trials or time-series evaluations are still needed to measure antibiotic use, prescribing appropriateness, resistance, costs, and patient outcomes.
Thai researcher contribution
Direk Limmathurotsakul of MORU, Faculty of Tropical Medicine, Mahidol University, contributed to conceptualisation, methodology, funding acquisition, data curation, formal analysis, investigation, supervision, and manuscript review. Anucha Apisarnthanarak of the Faculty of Medicine, Thammasat University Hospital, contributed to investigation, supervision, and manuscript review. Neither is identified as corresponding author by the publisher. The study included two Thai hospitals and 47 Thailand-based participants and received ethics approval from Mahidol University's Faculty of Tropical Medicine, MUTM 2022-017-02.
Limitations to consider
Only 16 hospitals were selected through pre-existing networks, and providers were convenience sampled, limiting national representativeness. Several measures were self-reported, allowing selection and social-desirability bias, and not all practices were independently observed or audited. Focus groups covered 11 hospitals in Indonesia, Nepal, and Viet Nam but were infeasible in Thailand, leaving no Thai provider-level qualitative data. KAP administration after some G-ASET discussions may have primed responses. Sensitive influences such as pharmaceutical incentives may be under-reported, one national framework cannot capture all antimicrobial consumption, and regression with few hospitals is hypothesis-generating. The cross-sectional design cannot establish causality or effects on resistance or patient outcomes.