ZAMSTAR was a 2×2 factorial design cluster randomised trial of 2 interventions; community enhanced case finding of tuberculosis (ECF) and a household level integrated TB/HIV intervention.
The household intervention reduced both the prevalence of culture confirmed tuberculosis in adults and the incidence of tuberculous infection to school-children. The ECF intervention had no effect on either outcome. The study found that the ECF intervention, while identifying 25% of new smear positive cases of tuberculosis during the intervention period, had no effect on either prevalence of tuberculosis in adults or incidence of tuberculous infection in children, when measured at the end of the study.
Conversely the HH intervention reduced the prevalence of tuberculosis by 18% and the incidence of tuberculous infection by 55%. This intervention identified less cases of tuberculosis during the intervention but had the added benefit of identifying many cases of HIV and enabling individuals to start ART.
Community TB case-finding alone, as conducted in the ZAMSTAR study, was not sufficient to reduce the burden of TB whereas a more intense intervention, targeted at those with highest risk, demonstrated the power of involving households and ultimately communities in TB control. While community TB case finding, such as ECF, may have benefits for individuals we have no evidence that it reduces the burden of tuberculosis at community level. It may be that a more intense intervention, such as the HH intervention, that engages households and ultimately communities in a more meaningful way is needed to reduce the burden of tuberculosis at community level.