INTRODUCTION
Task-shifting tuberculosis (TB) rapid molecular testing from laboratory technicians to nurses could help decentralise TB diagnosis at primary health centres (PHCs) and allow in-ward testing to shorten treatment decision for very sick patients. We assessed the feasibility of Xpert MTB/RIF Ultra (Ultra) testing on nasopharyngeal aspirate (NPA) by nurses in children with presumptive TB at PHCs and in hospitalised children with severe pneumonia within the TB-Speed project in seven countries.
METHODS
Of 23 PHCs and 15 paediatric wards, 9 and 4 respectively had nurses trained to perform Ultra using the battery-operated GeneXpert Edge. Laboratory technicians performed testing in the other sites. We compared the proportion of samples tested, invalid or error results, TB detection yield, turnaround time (TAT) between sample reception and result at PHC and time to clinician between sample collection and result delivery to clinicians in paediatric wards between nurses and laboratory technicians. External quality assessment (EQA) and site support supervision assessed performance. Self-administered questionnaire and semistructured individual interviews assessed nurses’ perceptions.
RESULTS
Ultra was done in 253/254 (99.6%) and 258/258 (100%) samples for PHC and hospital nurses versus 895/897 (99.8%) and 874/874 (100%) for laboratory technicians, respectively. At PHC, TAT was below 1 hour 30 min for 158/252 (62.7%) samples tested by nurses versus 677/893 (75.8%) by laboratory technicians, p<0.001. Ultra results were available to clinicians within 3 hours in 201/258 (77.9%) samples for nurses versus 464/874 (53.1%) for laboratory technicians in hospitals, p<0.001. EQA results <87.5% was more common for PHC nurses than PHC laboratory technicians or hospital nurses. Technical difficulties, lack of practice and workload were the main challenges, and training and supervision the main facilitators reported by nurses.
CONCLUSION
Task shifting of Ultra testing from laboratory technicians to nurses under close supervision could support decentralisation of TB diagnosis and shorten time to treatment decision for very sick patients.