Ethiopia has the ninth highest tuberculosis (TB) burden in the world. Given the country’s limited resources, decision-makers need to optimise case detection without overloading the health system. Patients currently have to produce three sputum specimens for testing. Are the second and third sputum tests really necessary?

The diagnosis of pulmonary TB in low income countries often relies on studying smears of sputum samples under the microscope. This is less sensitive than tests that involve bacterial culture, but it is relatively simple and will identify most infectious cases of pulmonary TB. Laboratory equipment and supplies are allocated according to guidelines that recommend screening an average of 10 suspects (three samples from each) to identify one positive case. But is this number of tests correct? Researchers from Ethiopia’s TB/Leprosy and Blindness Control Programme and the Liverpool School of Tropical Medicine reviewed the laboratory records of 42 TB diagnostic centres in the southern region of Ethiopia.

They found that:

• 15 821 TB suspects submitted three smears each.
• Prevalence rates by zone range from 11.4 to 110.8 per 100 000, with a mean rate of 38 for the whole region.
• On average, only four suspects have to be screened to diagnose one positive case.
• 26 % of suspects have at least one positive smear. 91.6 % of these are positive from the first specimen. A further 7.4 % are positive on the second but not the first and 1 % are positive on the third specimen only.

In Ethiopia, technicians are responsible for all routine laboratory activities and are often short of time. National guidelines suggest that each TB suspect should submit three sputum smears and technicians should examine at least 100 microscopic fields before deciding a smear is negative. In practice, they rarely have enough time to do this. Drop-out of the diagnostic process is another common problem. Most patients come from far away and need to stay overnight or walk for hours to submit specimens and collect results.

A reduction in the number of smears required could reduce technicians’ workload and the number of times patients have to visit the laboratory to produce samples or collect results. As 99 % of cases were identified by the first two specimens, the third seems to provide a negligible gain.

To decide whether to recommend the widespread use of fewer sputum samples, policy-makers should consider the:

• local prevalence of pulmonary TB
• duration of the programme, as new programmes will have more sputum positive suspects
• quality of the diagnostic procedures
• likelihood of defaulting patients with only one or two smears to contact health services in the future
• regional prevalence of HIV.

Source(s):
‘How many sputum smears are necessary for case finding in pulmonary tuberculosis’, Tropical Medicine and International Health 8 (10): 927-932, by M. Yassin and L. Cuevas, 2003

id21 Research Highlight: 12 January 2004

Further Information:
Luis E Cuevas
Liverpool School of Tropical Medicine
Pembroke Place
Liverpool L3 5QA
UK

Tel: +44 (0) 151 708 3219
Contact the contributor: lcuevas@liv.ac.uk

Liverpool School of Tropical Medicine (LSTM), UK

Other related links:
'What's the diagnosis? Testing for tuberculosis in Kenya' >

'Coughing up for TB tests - cost-effective diagnosis in Zambia' >

'Test match: the search for a better test for TB infection' >

'Coughing up for TB tests - cost-effective diagnosis in Zambia' >

See id21's collection of links relevant to infectious diseases.