Many people who report with TB symptoms at South Africa’s public sector clinics do not receive TB tests. But according to a major new modelling study, testing more of these people is one of the most effective things we can do to reduce infection rates and death, writes Marcus Low for Spotlight.
While South Africa has made substantial progress against TB, we are not on track to meet key targets set for 2030. This is according to the recent paper, published in the journal Global Health Action.
TB incidence in South Africa was projected to decline by 46% by 2030, relative to a 2015 baseline. The World Health Organisation (WHO) End TB Target adopted by South Africa aimed at an 80% reduction in incidence. TB mortality was projected to decline by 54%, against the End TB target of 90%.
The researchers also published uncertainty intervals around these estimates. Even though these intervals are relatively wide, their upper limits are below the End TB targets, suggesting that South Africa is unlikely to reach the targets by 2030.
The modelling, however, also identified interventions that could help South Africa get closer to the targets.
The new research is an extension of Thembisa, an existing mathematical model of HIV and TB in South Africa. University of Cape Town epidemiologist Dr Leigh Johnson, the lead author of the new study, is also the key driving force behind Thembisa.
Identifying what works
The researchers estimated which of a long list of interventions would have the greatest impact against TB. In technical terms, they looked at how varying 27 different parameters linked to TB interventions impacted projected TB incidence and mortality from 2025 to 2040 by calculating correlation coefficients (see the paper for a more nuanced explanation).
They found that the greatest reductions would be achieved if a high percentage (89%) of people reporting with TB symptoms were tested using new point-of-care TB tests.
This would result in a 49% reduction in average TB incidence and a 67% reduction in average TB mortality, relative to a scenario in which there are no changes to current programmes. TB symptoms include persistent cough, chest pain, night sweats, fever, and unexplained weight loss.
There is good evidence that many people who report with these symptoms at clinics are not offered molecular TB tests.
Earlier this year the WHO recommended the use of such new point-of-care TB tests. These tests aren’t yet in general use in South Africa, although they are being evaluated in pilot projects.
The new tests can be run on both sputum and tongue swabs. For now, most molecular TB tests in South Africa still involve someone producing a sputum sample – with which some people struggle – and the sample being transported to a lab for testing.
The researchers also found that high rates of door-to-door testing with the new point-of-care tests could reduce TB incidence and mortality by 38% and 49% respectively, while combining high rates of door-to-door testing and digital chest X-ray screening in asymptomatic individuals could reduce TB incidence and mortality by 17% and 21%.
Even without the new tests, there are substantial gains to be had if testing were stepped up in people with symptoms. TB incidence could be reduced by as much as 29% if the level of sputum testing in individuals seeking care for TB symptoms were increased to the upper bound considered by the researchers.
Once people have been diagnosed with TB and start taking TB treatment, they typically become non-infectious within around two weeks. Diagnosing them more quickly thus results in people remaining infectious for shorter periods, slowing transmission.
Other influential parameters identified by the researchers include the extent to which Covid-19-related behaviour changes are sustained, the rate at which people with HIV start taking ARV treatment (untreated HIV is a major driver of TB), and the combined uptake and efficacy of 3HP, a relatively new form of TB preventive therapy.
Testing is the key
“We found that the rate of testing in people seeking treatment for TB symptoms is the most important driver of future TB incidence and mortality,” wrote the authors. “This finding is not surprising given the historically low rates of testing.”
In studies of patients seeking treatment for TB symptoms in South African health facilities, the proportion who received a TB test has been highly variable, ranging from 3% to 84%, with a median of only 30%. Similarly, low rates have been reported in other countries with high TB burdens.”
The authors write that these low rates of testing reflect the non-specificity of TB symptoms and healthcare workers’ concerns about the cost and time required to collect sputum specimens and to send them to a central lab for molecular testing.
In one South African study, they point out that 36% of all people attending clinics had symptoms suggestive of TB.
“Conducting testing in such a large fraction of primary care attendees may be infeasible,” they write.
“Achieving high testing levels in people with TB symptoms could require levels … beyond current capacity. However, new point-of-care tests performed on tongue swabs or sputum present an important opportunity to increase levels of testing in symptomatic individuals at reduced cost, without placing additional strain on labs.”
In addition to reduced cost, the new point-of-care tests also offer quicker results and have lower operational requirements. The main potential benefit is that they’re likely to be used more frequently.
Regardless of the testing platform, the study authors argue there is a need to strengthen adherence to existing guidelines for systematic testing of all patients with symptoms suggestive of TB and that we need a larger healthcare workforce dedicated to diagnosis, better supply chains, and better monitoring systems to identify where and why symptomatic patients are not being tested.
See more from MedicalBrief archives:
New TB testing strategy showing promising results in SA
More than half of South Africans do not seek TB treatment – HSRC survey