No one with HIV should die from tuberculosis.

Tuberculosis is the leading cause of mortality among individuals infected with HIV, killing more than 1000 people every day. Even if they receive treatment for tuberculosis, people with HIV are more likely to die from tuberculosis than people without HIV, especially if they are not receiving antiretroviral therapy or if they have multidrug-resistant tuberculosis. They do not die because we cannot treat HIV or cure tuberculosis. They die because of substantial gaps in the delivery of care and innovation, despite decades of knowledge about the synergy between tuberculosis and HIV, about how to stop the spread of tuberculosis, and how to optimise HIV treatment. In 2008, WHO endorsed the Three I’s strategy— intensifi ed case-fi nding, isoniazid prophylaxis therapy, and infection control—to address the crisis of tuberculosis deaths among people with HIV. Intensifi ed case-fi nding and isoniazid prophylaxis therapy save both lives and resources, given the number of tuberculosis cases prevented. A “fourth I”, representing integrated care at the facility level for individuals co-infected with HIV and tuberculosis, has also been shown to improve treatment outcomes for both tuberculosis and HIV. Despite endorsement of these almost decadeold strategies, people with HIV continue to die from tuberculosis at an alarming rate because not enough is being done to ensure optimum prevention, detection, and treatment. How do we change this dynamic? The Stop TB Partnership’s Global TB Plan 2016–2020, now under development, calls for mass scale-up of tuberculosis screening, diagnosis, and treatment for people living with HIV—with coverage targets of at least 90%. But much more is needed than just ambitious targets. First, and at a minimum, known strategies for stopping the spread of tuberculosis have to be prioritised, implemented, and scaled up in lowincome and middle-income settings for both adults and children. As the papers in the Lancet Series on tuberculosis show, these strategies include active case-fi nding, rapid diagnosis, post-exposure treatment (both isoniazid prophylaxis therapy and treatment for drug-resistant strains), and early initiation of optimum treatment for all strains of tuberculosis. These strategies have driven rates of tuberculosis down substantially among vulnerable patients in settings such as New York City and Baltimore in the USA, and Rio de Janeiro in Brazil. Second, innovative approaches must be adopted to halt the deadly toll of tuberculosis in people with HIV. For example, initiation of antiretroviral therapy needs to become an urgent priority among all people living with HIV, including in areas of high tuberculosis and HIV coinfection, to prevent tuberculosis incidence, progression, and mortality. An overwhelming evidence base now clarifi es the need for any patient with HIV, irrespective of CD4 count, to be started on antiretroviral therapy as soon as possible after diagnosis. Some in the medical and public health communities have added immediate initiation of antiretroviral therapy as the “fi fth I” in the strategy to halt tuberculosis mortality. There is also a need for better ways to treat tuberculosis in patients with HIV, which will involve evaluating innovative treatment strategies and improving the science of tuberculosis clinical trials. Furthermore, there should be greater inclusion of people with HIV in tuberculosis drug trials, since this population is often excluded from such research. These exclusions mean there are limited indications for the use of new tuberculosis drugs in HIVinfected individuals—those with the highest rates of mortality—and there is little information about drug– drug interactions with antiretroviral therapy. Third, better screening and diagnostic strategies are needed to detect tuberculosis in people with HIV