Tuberculosis still much more common in people with HIV in high-income settings


Tuberculosis is still one of the world’s deadliest infectious diseases, with 10.6 million people developing active (symptomatic) TB infection in 2022 and 1.3 million dying from it. That’s second only to COVID, which killed 6.95 million people last year, and far more than the 630,000 who died due to HIV.

But it still receives less publicity than HIV – even though it is the greatest cause of death on people with HIV, with more than a quarter of HIV-related deaths (167,000) directly due to TB, and not entirely in immunocompromised people.

It is not often realised that far more people are infected with inactive or ‘latent’ TB than ever develop symptomatic illness (‘active TB’). Also, the development of active TB can be effectively prevented by taking a few months’ course of one or two TB drugs (‘TB preventive therapy).


latent TB

A form of TB that is not active. Persons with latent TB are infected with M. tuberculosis but do not have any symptoms and they cannot spread TB infection to others. Only specific tests will tell if anyone has latent TB. Treatment for latent TB is recommended in people living with HIV. 

active TB

Active disease caused by Mycobacterium tuberculosis, as evidenced by a confirmatory culture, or, in the absence of culture, suggestive clinical symptoms.

middle income countries

The World Bank classifies countries according to their income: low, lower-middle, upper-middle and high. There are around 50 lower-middle income countries (mostly in Africa and Asia) and around 60 upper-middle income countries (in Africa, Eastern Europe, Asia, Latin America and the Caribbean).


When the statistical data from all studies which relate to a particular research question and conform to a pre-determined selection criteria are pooled and analysed together.

low income countries

The World Bank classifies countries according to their income: low, lower-middle, upper-middle and high. While the majority of the approximately 30 countries that are ranked as low income are in sub-Saharan Africa, many African countries including Kenya, Nigeria, South Africa and Zambia are in the middle-income brackets. 

TB prevalence, illness and deaths are very much more common in the low- and middle-income countries than high-income countries. However the higher rate of TB in low- and middle-income settings, especially in people with HIV, has meant that large TB studies have tended to be conducted there, and in some ways we know less about TB in high-income settings, especially in people with HIV,.

People diagnosed with HIV in high-income countries, especially if they do not come from areas with high HIV and/or high TB prevalence, are less systematically tested for TB and so won’t get the opportunity to take TB prevention drugs if they have latent TB. In addition, we also don’t know the extent to which antiretroviral treatment for people with HIV and latent TB lowers the risk of developing active TB so as to make TB prevention therapy unnecessary.

A team of Dutch researchers therefore conducted a meta-analysis that drew together the data they could find in studies of TB in people with HIV in high-income settings, meaning North America, western Europe and Australasia. They tried to answer three questions:

  • Prevalence: How common is latent TB infection among people with HIV in high-income countries and which people are more likely to have it?
  • Incidence: How common is progression to active TB in people with HIV?
  • Prevention: To what extent does TB preventive treatment prevent active TB developing in people with HIV in high income countries?

Prevalence – latent TB

Nearly a quarter of the entire global population – about 23% – has latent TB infection. This can be acquired at any point in a lifetime, and the annual infection rate globally is about 0.8% a year, with latent TB about twice as common in people over 50 than people under 20.

Most infections remain latent, with the immune system controlling, but not eliminating, the Mycobacterium tuberculosis organism. By definition, people with latent TB are not infectious (though about 20% of people with active TB, as confirmed by test, may not experience symptoms but may shed the bacterium.)

Latent TB is not that much rarer in the high-income world than in other countries, but it is highly skewed towards people who have come from low- and middle-income countries.

For instance, general population studies have found that 1% of people born in the US have latent TB, compared to 13% of US residents not born there. In Australia, the figures are even more skewed, with 0.4% of lifelong Australians having latent TB compared with 17% of people born outside the country. In Europe the picture is more mixed, but in western Europe the prevalence of latent TB is generally below 10% (exceptions are Portugal and Germany).

In the meta-analysis, the researchers found 51 studies covering 65,930 people with HIV. Twenty-eight were conducted in North America, 19 in western Europe and four in Australasia.

Among the 51 studies, the pooled prevalence of latent TB was 12%; ranging from about 6% in Australasia to about 13.5% in North America. The prevalence of latent TB in people with HIV declined considerably in recent years; in studies conducted before 2010 it was about 14% and after 2010, about 8%. The studies differed widely in size, design, tests used, and population selection (for instance, more people with low CD4 counts and not on antiretroviral therapy (ART) tended to be screened). Because some studies were in people newly diagnosed with HIV, the rate of viral suppression ranged from zero to 77%.

In terms of people who tested positive for latent TB, 81% of them were men; among those that recorded country of origin, 34% were born outside the relevant country and 20% came from a country with high TB prevalence. In the European studies, 14% of subjects with latent TB were from sub-Saharan Africa and in the US studies, 47% were of Black ethnicity. Two-thirds had acquired HIV heterosexually. A quarter had a CD4 count below 200 and 34% were on ART, but only four studies recorded viral load.

In multivariate analysis, some factors were not associated with having latent TB. For instance, men were 35% less likely to test positive for latent TB but this was not statistically significant (p=0.08).

Factors that were significant were birth in a high-TB prevalence country (odds ratio 4.67); being Black African (OR 3.26); being non-native born (OR 3.28); having been previously exposed to TB (OR 2.88); and having caught HIV heterosexually (OR 1.85). People on ART were 45% less likely to have latent TB and this just failed to be statistically significant (p=0.06).

Incidence: active TB

Only 8-10% of people with latent TB will go on to develop symptoms and have active TB disease: 5% in the first two years after infection, the others more slowly. Factors such as smoking make it more likely. People with HIV and latent TB, however, are more likely to develop active TB.

The World Health Organization estimates that the incidence of active TB in the general population is under 10 per 100,000 people a year in most countries in western Europe. In central and eastern Europe incidence is considerably higher, at 29 per 100,000 in Lithuania, 46 per 100,000 in Russia, and 64 per 100,000 in Romania.

But in a large cohort study in The Lancet HIV from 2015, the incidence of active TB in all people with HIV in the UK was 6.7 cases per one thousand people a year, not per 100,000 – in other words, 670 per 100,000, or nearly 100 times that in the general population. It was 13.6 per 1000 in sub-Saharan Africans and 1.7 per 1000 in people of White British ethnicity. However it fell to 1.9 per 1000 in African people who were taking HIV treatment and 0.5 per 1000 in White British people taking treatment – though this is still more than seven times the incidence in the general White UK population.

A second UK cohort study in 2020 found a similar incidence of active TB in people with HIV of 0.6 per 1000, but double that (1.2 per 1000) in people of Black ethnicity. In this study, 283 out of 704 people with active TB were on ART with viral loads below 50.

In the present meta-analysis, only seven studies looked at active TB incidence, among 10,629 people with HIV. Among all study participants, regardless of latent TB status, 1.25% developed active TB at some point.

In terms of annual incidence, in people who had tested positive for latent TB, the annual incidence of active TB ranged from 12.7 per 1000 a year to 48.4 per 1000 a year. This variation most likely represents heterogeneity in study design as much as real differences in incidence.

The tests for latent TB are not perfectly reliable, and some people who test negative for latent TB can still develop active TB; in this group, incidence varied from zero to 10.4 per thousand a year. Taking all the studies together and allowing for variations in design, the pooled incidence on active TB in people testing positive for latent TB was 28 per 1000 a year and in those testing negative, 4 per 1000 a year.

Five studies found that progression to active TB was considerably more common in people with HIV who were not on ART, but the varied design of studies did not allow the authors to come up with a figure for ART’s contribution to preventing active TB.

Prevention: TB preventive therapy

Only five studies reported on people with latent TB who had been on TB preventive therapy (TPT). Among those, two found no cases of active TB in people on TPT, making it difficult calculate relative reduction in risk, a small study found no cases of active TB at all, and one study had no control arm. The only study reporting cases in both TPT and control arms found TPT (in this case, with 63% adherence) prevented 70% of cases of active TB that might otherwise have happened.

Pooling together all the studies, however, the researchers were able to determine that the incidence rate of active TB was 65 per 1000 a year in people not taking TPT and six per 1000 in people taking it, or more than 90% efficacy.

Numbers needed to treat and conclusions

Finally, the researchers calculated three figures.

  • Firstly, the number of HIV-positive people needed to screen to detect one case of latent TB. Overall, this was eight. However, only six people born outside the relevant country would need to be screened to detect one case, compared to 14 people born within the country. Four people with previous exposure to TB would need to be screened to detect one case.
  • Secondly, the number of people with latent TB needed to treat with TPT in order to prevent one case of active TB. As TPT effectiveness only depended on adherence, this came out as 20, for all populations.
  • Finally, the number of HIV-positive people to screen for latent TB, assuming that all who tested positive were given TPT, in order to prevent one case of active TB: this ranged from 111 in people who were born abroad to 285 in people born in the country.

Although WHO guidelines recommend screening all people with HIV for TB, in practice this is not always done. For instance, in 2017 a UK survey found that only 57% of HIV clinics were offering TB screening, and this rate was no higher in higher-prevalence areas of the UK. In a 2015 audit at a major London HIV clinic, 12% of people with HIV had regular TB screening, despite this being recommended by the National Institute for Health and Clinical Excellence (NICE).

The authors of the meta-analysis endorse a selective screening strategy for latent TB. This kind of approach is currently recommended by bodies such as the European AIDS Clinical Society and the British HIV Association, with screening only recommended for people with HIV from countries with high and medium rates of TB, especially the newly diagnosed. Other people are only screened if they have risk factors such as recent TB exposure.

However when it comes to the reverse case – screening people diagnosed with TB for HIV – public health bodies recommend universal HIV testing and these guidelines are followed in many settings. In 2021, the UK Health Security Agency (UKHSA) found that 98.4% of people diagnosed with TB were offered an HIV test.

It could be argued that it might be better to offer a TB test to all people with HIV. After all, as the authors of the current paper note, the incidence of active TB in people with HIV was approximately 100-fold higher than the general TB incidence in low-incidence countries, so “screening appears favourable in this population.”


Van Geuns D et al. Screening for HIV tuberculosis infection and effectiveness of preventative treatment among people living with HIV in low-incidence settings: a systematic review and meta-analysis. AIDS, first online publication, 12 October 2023.

A non-technical summary of the British HIV Association’s TB guidelines can be found here.