Abstract / Summary
Background and objectives Childhood tuberculosis remains underdiagnosed in India. Although gastric aspiration is recommended to improve microbiological confirmation, its use at lower-tier health facilities is limited by training, infrastructural, and logistical constraints. This study evaluated whether a structured, training-focused implementation strategy could integrate gastric aspiration into routine paediatric TB diagnostic workflows at primary and community health settings, and assessed its feasibility, acceptance, uptake, and diagnostic yield. Methods We conducted a mixed-methods controlled before-and-after implementation study in Rajnandgaon district, Chhattisgarh, involving training of healthcare providers across 39 primary health centres and 10 community health centres. The intervention included hands-on training, procedural standardisation, mentorship, limited logistical support and community engagement, guided by the consolidated framework for implementation research. Outcomes assessed included procedural uptake, caregiver acceptance, diagnostic yield, and contextual factors influencing implementation. Results Of 652 eligible children with presumptive TB, 567 (87%) underwent gastric aspiration, versus 18 pre-intervention, indicating a marked increase in uptake. Most samples (75%) were collected at peripheral facilities, reflecting decentralisation of the procedure. However, microbiological confirmation was low, with a cartridge based nucleic acid amplification test (CBNAAT) positivity rate of 0.9%. Quality audits identified persistent operational challenges, including suboptimal cold-chain maintenance, delayed specimen transport, inadequate sample volume, and variable staff sensitisation, which likely contributed to the low diagnostic yield. Interpretation and conclusions A structured training and implementation approach enabled integration of gastric aspiration into routine practice at peripheral health facilities with high uptake and acceptance. However, low microbiological yield highlights substantial pre-analytical and system-level constraints that limited diagnostic effectiveness.