A battery-powered TB test picked up 85% of culture-positive cases, but fell short of Xpert Ultra

The edition in brief

A Cochrane review for the WHO assessed near point-of-care nucleic acid amplification tests for pulmonary tuberculosis: battery-operated devices that need no laboratory infrastructure. Only one device met the class definition. Across five studies and 3,291 people, sensitivity against culture was 85.4% and specificity 97.6%. That is lower sensitivity than Xpert MTB/RIF Ultra (85.6% vs 93.1%) but well above smear microscopy (86.7% vs 64.7%). Sensitivity fell to 79% in people with HIV and 67% in smear-negative disease, and most testing was done in laboratories rather than peripheral clinics. The test's place is where neither Xpert nor reliable culture is within reach. A meta-analysis of 1,078 people treated with two-drug direct-acting antivirals for acute or recent hepatitis C found 97.7% cure overall, but only 85% with four-week and 92% with six-week courses, against 99% at eight weeks. In DORAL, a small phase 2 switch trial in 114 suppressed people with HIV, one virological failure occurred by week 48 on doravirine plus raltegravir, an option when nucleoside drugs or interactions are a problem. In the AZTEC trial, ten days of prophylactic azithromycin in babies born before 30 weeks reduced gut microbial diversity and increased macrolide resistance genes that persisted to discharge. The pearl covers what a negative molecular TB test does and does not rule out.