In many high-income countries, reactivation of latent tuberculosis infection is estimated to account for more than 80% of all incident cases of tuberculosis, and prevalence of latent tuberculosis might exceed 50% in certain populations.1 There is no gold-standard test to diagnose latent tuberculosis, but based on currently available immune-based tests, a third of the world's population has presumptive latent tuberculosis.2, 3 Although an updated estimate of the global burden of latent tuberculosis would be very helpful,4 it is accepted that there is a vast reservoir of latent infection, from which it is estimated that 100 million people will develop active, contagious tuberculosis over their lifetimes.5
Management of latent tuberculosis is considered to be one of the core interventions for tuberculosis elimination. In the 1960s and 1970s, several studies showed that isoniazid for 6–12 months could significantly reduce the risk of reactivation of active tuberculosis in people with a positive tuberculin skin test (TST).6, 7 However, the length of therapy, need for close follow-up, and risk of potentially fatal hepatotoxicity8 reduced uptake, acceptance, and completion of therapy.9 These problems substantially reduce the cost-effectiveness10 and the population-level epidemiological impact of this approach.11, 12 As a result, in the past two decades, randomised trials have aimed to identify shorter regimens that are as effective as, yet safer and more acceptable, than isoniazid.13 These trials have identified several alternative rifamycin-based regimens that have recently been recommended for treatment of latent tuberculosis infection.14, 15, 16
Factors associated with non-completion of treatment of latent tuberculosis have received considerable attention: a recent systematic review identified 68 studies investigating non-completion from North American centres alone. The authors noted the importance of strategies to improve treatment adherence that are specific to the context and populations being served, and that a one-size-fits-all approach was unlikely to be successful.12 However, there has been very little recognition for the impact of losses during the many steps in patients' trajectories before therapy is begun. People with latent tuberculosis might not be identified for screening, and even if they are, might not be tested, or a TST might be done but not read, or an interferon-gamma release assay (IGRA) result might not be received by providers. Individuals with a positive TST or IGRA might not complete medical evaluation (eg, symptom check, physical exam, and chest radiography),17, 18, 19, 20, 21 and providers might not recommend therapy or treatment might not be started or completed. Given the lack of recognition of this problem, it is not surprising there have been very few studies of interventions to prevent the losses and drop-outs at these steps. With the new End TB Strategy to eliminate tuberculosis by 2035, there has been increased recognition of the importance of addressing latent tuberculosis infection. This systematic review offers evidence for the importance of addressing the losses along the cascade of care, if efforts to eliminate tuberculosis are going to be successful.
Research in context
Evidence before this study
Latent tuberculosis infection is estimated to affect more than a third of the world's population. Although methods exist to diagnose and effectively treat latent tuberculosis infection, they are slow and imprecise. This scarcity leads to difficulties in the identification and treatment of this vast pool of infected people, which has been identified as a key barrier to global tuberculosis control. There are multiple steps in the care process from initial identification of people with latent tuberculosis infection who could potentially benefit from therapy, until treatment completion. Patients can, and do, drop-out or are lost at each of these steps. We searched three electronic databases, for studies that were published between 1948 and June 20, 2016, describing the full procedures of diagnosis, evaluation and treatment of latent tuberculosis. There have been multiple studies of the problems leading to non-completion of therapy once it has been started: one systematic review identified 68 studies from North America alone. This systematic review showed inconsistent associations between adherence and patient factors or treatment characteristics. Far fewer studies have estimated the losses and drop-outs at earlier steps, largely because these patients are not seen by health-care personnel, and so it is unknown how often and why these problems occur. We undertook this systematic review to understand the extent, and reasons for patient losses, during the entire latent tuberculosis cascade of care.
Added value of this study
To our knowledge, our study is the first to conceptualise the latent tuberculosis cascade of care, and develop an explicit framework of analysis to account for the losses during each individual step in this cascade, from initial identification of risk of infection, through to completion of treatment. We show estimated losses at each step, and identify the patient and health system factors associated with those losses. Notably, losses before starting therapy accounted for greater net reduction of the public health benefit of latent tuberculosis infection management than did patient non-adherence with therapy once started.
Implications of all the available evidence
Interventions that aim to reduce losses at the early steps of the latent tuberculosis cascade of care should enhance the public health impact of diagnosis and treatment of infection more than will interventions that focus on improving patients' completion of treatment. To achieve the goals of the new WHO End TB Strategy of reducing the tuberculosis incidence rate to less than ten infections per 100 000 by 2035, latent tuberculosis management needs to be substantially scaled up. Our findings suggest that every step in the entire cascade of latent tuberculosis care will need improvement to achieve that goal. And, although we clearly demonstrated the extent of the problem and some health-system factors associated, very little published evidence was found of successful interventions to reduce the losses at every step.
We aimed to systematically review the published research about the so-called cascade of care in latent tuberculosis diagnosis and treatment. Specific outcomes of interest included: the number of people eligible for testing for latent tuberculosis infection; the number who initiated and completed screening with IGRA or TST; and the number with positive tests who had chest radiographic and medical evaluation; and who were prescribed, started, and, completed treatment for latent tuberculosis infection.
