Can ParC Ser83Ile status predict fluoroquinolone efficacy in Mycoplasma genitalium infection?

We commend Sizulu Moyo and colleagues for their insightful analysis of South Africa’s national tuberculosis prevalence survey 2017–19. Key findings included the high estimate of tuberculosis prevalence that led to an upward revision of WHO’s incidence estimate for South Africa; that tuberculosis prevalence was higher in men than in women; and that the majority of tuberculosis was found in individuals with abnormal chest radiographs who did not report symptoms and in HIV-negative individuals. An important finding not addressed by Moyo and colleagues was that people who reported a history of previous tuberculosis treatment con tributed disproportionately to prevalent tuberculosis. Of the 35 191 indi viduals participating in the survey, 2964 (8·4%) reported previous tuberculosis treatment. However, of 234 participants with bacteriologically confirmed pulmonary tuberculosis, 62 (26·5%) reported previous treatment, all of whom were culturepositive for Mycobacterium tuberculosis complex as per the study’s case definition. The crude prevalence of bacteriologically confirmed pulmonary tuberculosis was thus 3·2-times higher in participants who had previously had treatment compared with those who had not. Previous tuberculosis treatment was approximately twice as common in participants with tuberculosis who reported symptoms compared with participants with tuberculosis who did not report symptoms (37·8% for symptoms vs 18·4% for no symptoms), which is consistent with earlier findings, suggesting that there is a higher potential of onward transmission for participants with tuberculosis who had previously received treatment compared with those who had not received treatment. High rates of prevalent tuberculosis among people who were previously treated have repeatedly been documented at the sub-country level in South Africa. Of the 728 individuals with tuberculosis detected during prevalence surveys in ten communities, 155 (21·3%) reported previous treat ment. The findings reported by Moyo and colleagues confirm these earlier findings at the country level. We emphasise that previously treated people are a key high-risk group for tuberculosis in South Africa, representing less than 10% of the adult population but contributing more than a quarter (26·5%) of undetected tuberculosis, and they should therefore be considered for targeted interventions. Using mathematical modelling, we showed that posttreatment follow-up with secondary preventive therapy among people who had previously completed tuberculosis treatment could reduce transmission, tuberculosis incidence, and mortality in high-incidence settings, and potentially save costs. Based on these findings, adults with a history of previous tuberculosis treatment were included as one of three highrisk groups eligible for facilitybased, targeted universal testing for tuberculosis in a large clusterrandomised trial. Considering Moyo and colleague’s findings, we note that such targeted approaches could add substantial value to South Africa’s fight against tuberculosis. Integrating prevention and case finding among previously treated people into national efforts to end tuberculosis should be a priority.