Closer look at household contacts finds more TB cases

Investigating household contacts of patients who contracted tuberculosis (TB) combined with passive measurement was able to detect more cases of TB than passive measurement alone, a randomized trial in Vietnam found.

This active screening intervention, which included inviting household contacts for clinical exam and imaging, was able to uncover significantly more TB cases among household contacts than passive methods of screening, reported Greg J. Fox, MB BS, PhD, of the University of Sydney, and colleagues.

They described the passive method of investigating cases of TB as a person presenting with symptoms of cough and sputum production to a health facility. This assumes most people who have the infection will seek care because of their symptoms, but prevalence surveys indicate "this long-standing assumption is not justified," they wrote in the New England Journal of Medicine.

The reason is that some people who test positive for TB on sputum smear microscopy, and are infectious, do not have "typical symptoms of the disease," the authors noted.

Investigating household contacts has had success in high-income, low-prevalence countries, but Fox's group pointed out "limited" implementation in high-prevalence areas because there is little evidence that shows it is effective. In fact, one trial in southern Africa found that investigating household contacts did not significantly reduce prevalence of TB within the population.

"Data from randomized trials are lacking with respect to the effectiveness of adding active screening to traditional passive case finding for contacts of infected persons," they wrote.

The authors conducted a cluster-randomized trial at clinics in 70 districts in eight provinces of Vietnam. Districts randomized to receive the intervention had health workers testing patients for TB, then inviting household contacts of patients who tested positive for a clinical assessment and chest radiography, with follow-up at 6, 12, and 24 months. Passive assessment or the control group tested patients for TB only and were asked to return after 24 months for an interview.

Index case patients were eligible if they were ages ≥15 years, tested positive for TB, and visited the TB clinic in their home district. Household contacts were eligible if they lived in the house with the index patient during the 2 months prior to the TB diagnosis.

Patients were about three-quarters men, though household contacts in both districts were around 60% female. Average household size was 3.3 people in intervention districts and 3.9 in control districts.

Overall, there were 25,707 contacts of 10,964 patients with smear-positive TB. Two-thirds of the 10,069 contacts in the intervention districts came for the 6-month screening visit, while 56% attended the 12 month and a little under three-quarters attended the 24 month.

In the intervention districts, 180 contacts (1,788 per 100,000 population) were registered as having TB, while 110 contacts (703 per 100,000) were registered in control districts (RR 2.6, 95% CI 2.0-3.3, P<0.001). The number of people needing to be screened for one additional registered case of tuberculosis was 74.6 (95% CI 64.2-89.2).

Secondary analyses also found more registered cases of smear-positive TB in the intervention group versus controls (160 versus 39, RR 6.4, 95% CI 4.5-9.0, P<0.001).

Study limitations included a larger household size in the control districts versus the intervention districts, and a lower proportion of contacts that reported a prior history of TB. The authors also said they were unable to confirm diagnosis of TB for cases not listed by the National Tuberculosis Program.

An accompanying editorial by Barry R. Bloom, PhD, of the Harvard T.H. Chan School of Public Health in Boston, characterized these as important new findings, arguing that this "provides formal proof in a high-burden setting that active case finding can detect early and asymptomatic cases of tuberculosis more effectively than the current strategy."

Bloom added that interventions such as this, which can help strengthen the health system in diagnosing and treating TB are critical to help "bend the incidence curve" of the disease.

"When patients need to see three providers before being diagnosed and only a quarter of physicians can recognize the cardinal symptoms of tuberculosis or know the standard treatment, there is a major health system problem," he wrote.