In the 1870’s and 1880’s, the German researcher, Robert Koch developed the technique known as smear microscopy, which he used to identify the rod-shaped bacterium, Mycobacterium tuberculosis (as well as a number of bacteria that cause other diseases — he was quite a prolific researcher).
The technique has changed little since his time.
Smear microscopy involves collecting a biological sample (usually sputum or some other clinical material), fixing it thinly on a glass slide and then staining it with a dye that binds specifically to mycobacteria (making them easier to identify under a microscope). Smear microscopy for TB is sometimes called AFB testing, because mycobacteria are ‘acid-fast bacilli’ (AFB). This means that certain dyes adhere to the waxy coat of mycobacteria and remain visible even after rinsing with water and being briefly treated with a solution of acid-alcohol that strips the dye from the rest of the smear (WHO 2004).
For example, when performing the most widely used staining method, Ziehl-Neelsen staining, the smear is covered with carbol fuchsin dye for a few minutes. After heating, rinsing with water and the acid-alcohol treatment, the smear is counterstained with another dye, methylene blue (to colour the background of the smear for contrast) and then rinsed again (WHO 2004). Once dried, if there was a substantial concentration of TB bacilli in the sputum sample, some of the bacilli on the slide should show up under the microscope as red characteristically rod-shaped organisms against the blue background.
The lab technician must meticulously examine each slide and then record the number of organisms observed (grading the burden — which can predict severity of disease in HIV-negative people). From sample collection until microscopic examination, the process takes at least a couple of hours (TDR 2006). The microscopic examination itself is a labour intensive and time consuming procedure.
In addition, according to the current WHO guidelines, the technician has to examine at least two or three slides per patient on successive days — two of which need to show evidence for AFBs before a diagnosis of ‘smear-positive’ TB can be made (although only one out of two smears is now required for a smear-positive diagnosis for people with HIV).
However, even though the technique is simple enough to be performed even in settings with rudimentary facilities, it isn’t very sensitive. There generally has to be a very high concentration of bacilli in the specimen in order for the lab technician to detect the ten or more organisms needed for a clear, positive result. Even a skilled eye working in a well-maintained laboratory is unlikely to detect TB when there are fewer than 5000–10,000 bacilli per ml of sputum. Such a high of a burden of bacilli is typically only found in adults with advanced pulmonary disease — and rarely in earlier disease, or when the disease is active in other parts of the body (extrapulmonary TB), in people with HIV or in children. In fact, rates of active TB detection by smear microscopy in people with HIV can be as low as 20%, and as low as 5% in children.
Increasing access to smear microscopy
Despite these shortcomings, the test is not worthless. A diagnosis made on the basis of smear microscopy is strongly predictive of a person’s need to go immediately onto TB treatment, and without access to the test, many people go without being diagnosed or treated. Even if a greater proportion of those diagnoses are in HIV-negative people, everyone in the community benefits from earlier diagnosis and treatment of people with TB because they will be at less risk of exposure to TB.
Yet even though the test may be relatively inexpensive and has long been used throughout the world, there is still inadequate access to quality smear microscopy in some high burden settings — and the distribution of microscopy facilities does not match the distribution of disease. According to a recent FIND/TDR report, there are around 45,000 microscopy centres in the 22 countries with the highest burden of TB, which initially may sound like a lot, but these facilities must service nearly 4 billion people — so that’s roughly one lab for every 73,000 people.
In addition the labs are not always well-maintained or adequately staffed; they often lack basic diagnostic equipment and sustainable consumable supplies. This in turn affects the morale of the staff and the quality of results where facilities are available, according to the recent FIND/TDR publication: “The inherent low sensitivity of the test is compounded by the conditions under which it is commonly performed: poor equipment, heavy workload, and inexpert or unmotivated staff.”
But several poster presentations at the TB conference demonstrated the benefits of very recent efforts scaling up smear microscopy capacity at peripheral laboratories where there had been little before. For example, in 2001 in Bangledesh, there were only five microscopy centres in the public Chest Disease Clinics in four major cities, but during 2006 the number of microscopy centres increased to 67 (Hyder). This involved training a total of 1040 laboratory technicians (Sultana). The scale up was gradual, but in 2005, 7113 new smear-positive patients were diagnosed compared to 5804 in 2004 (Hyder).
In China, the FIDELIS project helped establish smear microscopy centres in township hospitals in several prefectures leading to a 2.16-fold increase in the number of new smear positive TB cases diagnosed (Jian).
China and Bangladesh aren’t alone. Quite a few of the HBCs still have to go a long way to scale up basic laboratory capacity outside of the major medical centres. However, a new effort could assist them in the scale-up of laboratory capacity.
Namely, the Global Drug Facility, (GDF) a component of the Stop TB Partnership established in 2001 to help developing countries access quality anti-TB drugs, has now added ‘smear microscopy kits’ to its repertoire (although these kits, unlike the drugs, are not free).
According to Dr. Robert Matiru, GDF’s Operations Manager, over the last couple of years, GDF has collaborated with Management Sciences for Health to develop kits as “a way of repackaging and ensuring the quality of and availability of existing diagnostics in such a way that programmes have what they need, when they need it and where they need it so as to improve overall case finding and diagnosis.”
There are several kits:
- A microscope kit (the Olympus CX21, with all the accessories and spare parts — US $1,300.
- An equipment starter kit – containing important accessory equipment needed for staining slides —US $300.
- A consumables kit, which can be used to prepare stains for 1000 smears (with ready to use ZN stains, and 1000 sputum collection containers — US $175.
In addition, GDF developed a quantification tool: an Excel-based commodity management tool to help determine how much of the product to order - depending on how much work needs to be done in terms of diagnosis.
GDF piloted projects to validate delivery of the smear microscopy kits in Congo-Brazzaville, Nigeria and Tajikistan and performed assessments pre- and post-kit introduction. Dr. Matiru reported that the kits were most suitable for use in the peripheral labs (as opposed to central labs) in the settings studied and that they did indeed improve the quality of smear microscopy and increased case detection — provided they were used properly. He stressed that training and proper use of the kits was essential in order to maximise its use.
GDF used the experience “to get practical feedback on these kits as tools - their usefulness, their faults… in order to come up with a package and a tool that’s ultimately useful for programmes,” said Dr. Matiru. Now TB programmes can access the refined smear microscopy kits directly from GDF— and a fringe benefit of establishing a working relationship with this centralised procurement process is that GDF will likely also serve as a conduit in resource-limited settings for other new methodologies and diagnostics as they move past the demonstration phase.
The importance of establishing more microscopy laboratories was underscored by another study from Bangladesh that found that people with TB are diagnosed sooner when services are available close to their residence. In the study, patients in the Netrakona district who registered for anti-TB treatment in 2005 were all interviewed to determine how long they had been ill at the time of diagnosis. The mean duration of illness for the 1391 patients turned out to be 10.76 weeks — but the distance between a patient’s residence and the microscopy centre was directly related to delays in the time to diagnosis. When the patients resided 5 km, 5–10 km and 10 km away from the microscopy centre, the mean time from falling ill to diagnosis was 7.76 weeks, 11.24 weeks and 12.08 weeks respectively (Daru).
Such a delay could be a particular problem for people with HIV and smear-positive TB, not only because disease can progress rapidly during that time, but, according to one presentation at the conference, because there is a better chance of diagnosing active TB infection as smear-positive in people with HIV “if patients are screened at an earlier stage with cough more than 2 weeks” (Siddiqi). This is only because TB in people with HIV often follows an atypical course. In standard pulmonary TB, the reverse is usually true — the diagnostic yield/concentration of bacilli in sputum is greater as the disease advances.
Other items in this series
TB diagnosis special report: why smear microscopy needs to be improved
TB diagnosis: Reducing the number of smears and clinic visits needed
TB diagnosis: Improving the yield with fluorescence microscopy
TB diagnosis: For now, still desperately seeking better diagnostics (all references in this series are listed in this article)