Table 6.

Cluster-randomised trials evaluating effect of ACF on tuberculosis infection incidence or prevalence in children

	Country, population	ACF delivery	Diagnostic method	Tuberculosis infection measurement	Intervention population	Control population	Adjusted analysis
Ayles et al (2010)36	Zambia and South Africa, general population (high tuberculosis prevalence districts)	Community mobilisation and mobile clinics	Sputum smear if symptoms for ACF; culture for all for prevalence survey	Schoolchildren evaluated had TST in 2005 (before ACF) and same children had TST in 2009 (after ACF)	391 (7·9% of 4934 children who were TST-negative at baseline had >15 mm TST induration at endline; geometric mean per cluster incidence of TST conversion was 1·41 per 100 000 person-years	342 (6·6%) of 5169 children who were TST-negative at baseline had >15 mm TST induration at endline; geometric mean per cluster incidence of TST conversion was 1·05 per 100 000 person-years	Adjusted rate ratio for incidence of tuberculosis infection: 1·36 (95% CI 0·59–3·14)
Marks et al (2019)37	Vietnam, general population	Door to door	Sputum Xpert regardless of symptoms (ACF and prevalence survey)	Prevalence of positive IGRA among children born in 2012 (who would have been 1–2 years old when intervention started)*	23 (3·3%) of 701 children were IGRA-positive	18 (2·6%) of 705 children were IGRA-positive	Prevalence ratio 1·29 (95% CI 0·70–2·36)*

None of the studies had any co-interventions. ACF=active case=finding. TST=tuberculin skin test. IGRA=interferon γ release assay.

^*The study also included a post-hoc infection outcome of IGRA positivity among children born between 2004 and 2011 (who would have been 3–10 years old when intervention started); the IGRA positive prevalence ratio for intervention vs control clusters for these older children was 0·50 (95% CI 0·32–0·78).