Near point-of-care TB test: 85% sensitive, but weaker in HIV and smear-negative disease

Design

Cochrane systematic review and bivariate meta-analysis of diagnostic accuracy

Population

3,291 people with presumptive pulmonary TB in five studies, nine countries

Primary outcome

Sensitivity and specificity for pulmonary TB against culture

Effect

Sensitivity 85.4% (95% CI 82.2 to 88.1); specificity 97.6% (96.9 to 98.1)

This Cochrane diagnostic accuracy review, supported by the World Health Organization, pooled five studies with 3,291 participants from nine countries with moderate to high tuberculosis burden. All five assessed one marketed, battery-powered near point-of-care nucleic acid amplification test on sputum. Reference standards were culture and a composite standard.

Summary sensitivity against culture was 85.4% (95% CI 82.2 to 88.1) and specificity 97.6% (96.9 to 98.1), with moderate-certainty evidence. In a hypothetical 1,000 symptomatic people with 100 true cases, the review estimates 15 cases would be missed and 22 non-cases wrongly labelled positive. Sensitivity was 79.2% (68.7 to 86.9) in people with HIV and 66.7% (58.8 to 73.7) in smear-negative, culture-positive disease.

Compared with a laboratory-based automated molecular test, it was less sensitive (85.6% vs 93.1%) with similar specificity; compared with smear microscopy it was far more sensitive (86.7% vs 64.7%) but less specific. Certainty was downgraded because most studies ran the test in laboratories rather than in peripheral clinics. Data for children were sparse and sputum was the only specimen.

• A positive result is reliable (specificity 97.6%); a negative result does not exclude TB, especially with HIV or smear-negative disease.
• Where a negative test conflicts with a persuasive picture, pursue culture or a more sensitive molecular test.
• Compared with smear microscopy, sensitivity was 22 percentage points higher.
• Do not rely on the test in children; data were very limited.
• Check which tests your local TB programme supports before changing your diagnostic pathway.

Why it matters

It shows what a battery-powered test can and cannot rule out in peripheral clinics.

Don't overread it

Only one test met the WHO class definition, most studies ran it in laboratories, and this review does not change national TB programme guidance.

The statistics, in plain English

Sensitivity of 85.4% means about 15 in 100 people with TB test negative. Specificity of 97.6% means about 2 or 3 in 100 without TB test positive, which is why, among 1,000 people with 10% prevalence, 22 false positives appear. Both figures shift with how common TB is among those tested.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.