Diagnostic Testing and Airborne-Isolation Outcomes After Implementation of an Xpert-First Reflex- Smear Policy: A Before-and-After Study
Background: The value of reducing routine acid-fast bacilli (AFB) smear testing where molecular testing is already widely used remains uncertain. We evaluated diagnostic testing and airborne-isolation outcomes after implementation of an Xpert-first, reflex-smear policy for suspected pulmonary tuberculosis (TB). Methods: We conducted a single-centre retrospective before-and-after study of adults evaluated for suspected pulmonary TB. The primary analysis compared 1 November 2023–31 October 2024 with 1 December 2024–30 November 2025, excluding November 2024 as the transition month. Extended follow-up continued through 31 March 2026. The revised policy reserved AFB smears for Xpert-positive specimens and permitted de-isolation after two negative Xpert results without awaiting negative smears. The primary outcome was AFB smears per 100 eligible admission episodes. Analyses accounted for patient clustering but were not adjusted for baseline covariates. Results: The primary analysis included 3,408 admission episodes (1,580 pre-implementation; 1,828 post-implementation). AFB smear utilization decreased from 121.5 to 59.8 per 100 episodes (rate ratio 0.49, 95%CI 0.46–0.52; P < 0.001). Culture-confirmed TB among episodes undergoing mycobacterial culture remained similar (4.5% vs 4.4%; P = 0.881), as did Xpert and culture utilization. Median airborne-isolation duration decreased from 54.6 to 43.3 hours (median difference − 11.2 hours, P < 0.001), and isolation > 48 hours decreased from 56.8% to 41.1%. Hospital length of stay and intensive care unit admission were similar between periods. Reduced AFB smear utilization persisted through March 2026 (121.5 vs 56.5 per 100 episodes; rate ratio 0.47, 95%CI 0.44–0.49), with similar culture-confirmed TB detection rates (4.5% vs 4.2%). Among 169 culture-confirmed TB episodes, the proportion without a smear was higher post-implementation (1.5% versus 10.9%), although the estimate was imprecise. Conclusions: In this uncontrolled before-and-after study, implementation of an Xpert-first policy was associated with lower AFB smear utilization and shorter airborne isolation. The uncontrolled design and unmeasured pathway adherence limit causal interpretation. Safety equivalence was not established. Therefore, continued monitoring of smear omission and clinical outcomes is warranted.
Authors
- Dipika Agrawal
- Ronnie Voon Shiong Tan
- Woon Hean Chong (ORCID: https://orcid.org/0000-0002-2070-3585)
- Wei Jun Dan Ong (ORCID: https://orcid.org/0000-0003-4055-5222)
- Noel Stanley Wey Tut Tay
- Norris Ling
- Su Gin Douglas Chan
Institutions
- Ng Teng Fong General Hospital (SG)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-18
- DOI
- https://doi.org/10.21203/rs.3.rs-11069795/v1
- Primary Topic
- Tuberculosis Research and Epidemiology
- Type
- preprint