Effects of Respiratory Isolation for Tuberculosis to Reduce Community-based Transmission: A Systematic Review

Background: Respiratory isolation of people with pulmonary tuberculosis (TB), including after treatment initiation, is used to prevent community-based transmission; yet guidelines on duration are limited and implementation is heterogeneous. This systematic review synthesized evidence on respiratory isolation for TB to inform National TB Coalition of America guidelines.

Methods: After searching 6 databases, 8 reviewers screened and extracted data in duplicate on effects of respiratory isolation compared to no isolation or masking. Studies were stratified by outcomes: TB infection or disease in contacts, mortality, hospitalization duration, patient and health system costs, and impact on mental health or stigma. We used a convergent integrated approach to synthesize quantitative and qualitative findings and assess limitations.

Results: Seventeen studies were included. There were limited data directly comparing isolation to non-isolation interventions, including effects after treatment initiation. One randomized controlled trial suggested treatment in a sanatorium versus at home did not affect TB incidence in contacts. Modeling studies suggest isolation may reduce transmission but relied on various assumptions, and isolation was implemented alongside other interventions. Descriptive, mixed-methods, and qualitative studies described adverse impacts of isolation on employment, education, food/housing security, and mental health due to transmission fears, stigma, and social isolation. Impacts were compounded in marginalized groups including Indigenous and incarcerated persons.

Conclusions: Data to support current isolation practices, particularly after effective treatment initiation, to reduce TB transmission in communities are limited. Public health guidance should weigh the negative impacts on people with TB against decreased community transmission to make evidence-based decisions about respiratory isolation.