Abstract Body

Background: South Africa has phased in Xpert MTB/RIF, replacing sputum smear microscopy in centralised laboratories, as the first-line test for tuberculosis (TB). In a pragmatic cluster-randomized trial, we evaluated the effect of Xpert implementation on yield, initial loss to follow-up and proportion treated.

Methodology: 20 laboratories in 4 provinces were randomized to intervention (immediate Xpert implementation) or control (microscopy; Xpert deferred) arms. At 2 primary health clinics per laboratory, a systematic sample of adults with suspected TB, identified by clinic staff, was enrolled. Laboratory results were obtained from the National Health Laboratory Service database, and TB treatment start dates from clinic registers or participant interview. Yield was defined as the proportion with a positive index Xpert or smear result. Initial loss to follow-up among participants with a positive index result was defined as not starting TB treatment within 28 days. The intervention effect was estimated using methods appropriate for cluster-randomized trials. An adjusted analysis controlled for baseline differences in individual-level factors.

Results: From June-November 2012, 4972 persons were screened, and 4656 (93.6%) enrolled (median age 36yrs, 62% female). More participants in the intervention vs. control arm were asymptomatic (7.4% vs. 4.3%) and fewer had BMI<18.5 (8.7% vs. 12.3%). Among 4412 participants (94.8%) with a laboratory test result available, the yield was greater in the intervention vs. control arm (9.2% (200/2176) and 7.8% (174/2236) respectively, (adjusted prevalence ratio 1.49 [1.00-2.23), p=0.05). Among 374 participants with a positive index test result, initial loss to follow-up was similar by study arm (intervention: 17% (34/200), control: 14.9% (26/174); adjusted risk ratio [aRR] 0.96, 0.48-1.93). Among 4656 participants, the proportion starting TB treatment by 6 months was similar by study arm (intervention vs. control: 10.8% (250/2324) vs 12.5% (291/2332) respectively, aRR 1.04 (0.76-1.43)).

Conclusions: Xpert increased the yield of TB by 49% but did not increase the proportion treated overall or reduce initial loss to follow-up. The increased yield has potential to be cost saving, and economic analyses are in progress. This analysis does not assess time to treatment rifampicin-resistant TB, which is a limitation. Maximising the benefit of Xpert scale-up requires strengthened health systems, particularly ensuring that people with positive test results start treatment promptly.