TB diagnosis: Improving the yield with fluorescence microscopy

This article is more than 17 years old. Click here for more recent articles on this topic

One key strategy for improving the effectiveness of smear microscopy is concerned with how to get the best sputum sample, and how to treat it to get it to reveal all its secrets. Some of these approaches could even increase the likelihood of getting a diagnosis of smear-positive TB in a person with HIV.

Getting a good morning sputum specimen is one of the most basic rules of the AFB test. Early morning specimens are more likely to contain bacilli because secretions tend to build up in the airways overnight. Other researchers have stressed inducing a productive cough with a string in the throat in one study, or with hypertonic saline using an ultrasonic nebuliser.

Researchers from MSF have investigated how to make certain that the sputum is of the highest quality and how to differentiate sputum from mucous or saliva. They stain smears and can rapidly grade them with x 10 microscopy into sputum and not-sputum, then categorise the smear sizes by how much sputum is in them using x 100 microscopy. The method is simple and fast and sputum quality and smear size were highly correlated with AFB detection (Klarkowski). Going back and looking at the quality of specimens that had been processed in their own labs without this technique they found that 60-70% were of poor quality.



A specimen of tissue or other material taken from part of the body and smeared onto a microscope slide for examination. A Pap smear is a specimen of material scraped from the cervix (neck of the uterus) examined for precancerous changes.


Material coughed up from the lungs, which can be examined to help with diagnosis and management of respiratory diseases.


When using a diagnostic test, the probability that a person who does have a medical condition will receive the correct test result (i.e. positive). 


In a bacteria culture test, a sample of urine, blood, sputum or another substance is taken from the patient. The cells are put in a specific environment in a laboratory to encourage cell growth and to allow the specific type of bacteria to be identified. Culture can be used to identify the TB bacteria, but is a more complex, slow and expensive method than others.

CAT scan

A computerised axial tomography (CAT) or computed tomography (CT) scan is a type of specialised X-ray that gives a view of a 'slice' through the body, and is used to help detect tumours, infections and other changes in anatomy.

Thus, “sputum quality needs to be assessed at the point of collection,” according to one audience member at the FIND meeting: “I think this is the cheapest and the fastest way to improve the quality of both microscopy culture and whatever the new method for analysing sputum is. As the great Chinese philosopher Confucius said, ‘It’s hard to catch a black cat in a dark room if the cat is not there’.”

Other efforts are looking at how to make the most of the sputum given. Some of these suggest that it might be possible to significantly improve the sensitivity of smear microscopy. One, the bleach (sodium hypochlorite) sedimentation method, which involves the digestion and concentration of sputum by bleach, is another of the strategies currently being evaluated in the FIND portfolio of “incremental improvements.”

In an MSF-sponsored prospective controlled study including 644 patients attending Mathare clinic (Nairobi, Kenya), the addition of an equal volume of 3.5% bleach to smear specimens (overnight) significantly improved the sensitivity of Ziehl-Neelsen microscopy. “Regardless of definition,” the investigator wrote: “the bleach method resulted in a 15% increase of new detected patients. Using a one smear-positive definition the approach was more sensitive than the standard three smear approach (Bonnet).

In another study with 232 smear specimens in Kenya, after adding a step where 3.5% (sodium hydrochlorite) was used on the smears overnight, Ziehl-Neelsen staining achieved a sensitivity approaching that of culture (27.1% were positive versus 29.3% that were culture positive). However, adding too high a concentration (5% NaOCl) reduced the sensitivity (Matu).

Getting the solution just right could explain discrepancies in the different reported results — suggesting that the practice of the technique in the field may not yet be standardised enough to give consistent, reliable results. As one researcher from Cameroon pointed out, the method may increase sensitivity of smear microscopy in people with HIV by ~ 10% — but operational research is needed in order to develop consensus on the best procedure (Eyangoh).

In another study he demonstrated that centrifugation with 3.5% NaOCl greatly enhanced the diagnosis of smear negative TB when using a fluorescent microscope (see below) (Githui).

Finally, Dr. Kevin Fennelly working on a TDR funded project in Brazil, showed that several of these approaches could be combined. His experiment involved the addition of small membrane filters to the specimen processing (which included bleach). When these small membrane filters were then put on a slide and examined with a fluorescent microscope, it resulted in an increase in the sensitivity of sputum microscopy to 88%.

Fluorescence microscopy

A fluorescent microscope would be a significant step up for most peripheral laboratories but could significantly improve the efficiency of the facility. The microscopes illuminate the smear, typically with a quartz-halogen lamp or a high-pressure mercury vapour lamp allowing a much larger area of the smear to be seen and resulting in more rapid examination of the specimen. Because the specimens are stained with auramine o/rhodamine, the bacilli shine fluorescent yellow against the blue background, which also makes them much easier to see and count.

“Repeatedly,” said Dr Richard Anthony, of The Royal Tropical Institute in Amsterdam, “fluorescence microscopy has shown to have significant advantages over ZN microscopy and ZN microscopy has almost always reported to be less sensitive, even when looking at a spot smear, single smear.”

A recent review article published in The Lancet Infectious Diseases showed a very strong trend towards higher efficacy of fluorescent microscopy (around a 10% increase in sensitivity).

“The question is, why don’t people use fluorescence microscopy more?” asked Dr Anthony (rhetorically).

The first problem, he said, is the mercury vapour lamp, which is expensive, and lasts a very short time. Technicians don’t like them because, Dr. Anthony said: “They tend to blow up as you’re sitting in front of the microscope, especially if you use them longer than you should which you tend to do to try and save money.” He added that they take a while to warm up, they burn a lot of electricity, and electrical supply problems can significantly shorten their life span. Although they produce a very bright white light, most of it has to be screened out because the fluorescent smear microscopy process needs blue light.

In short, fluorescence microscopy needs a regular supply of expensive lamps and a stable power supply which is frequently not available.

But a potential solution for many of these problems may have been found in light emitting diodes (LEDs). In contrast, to mercury vapour lamps, these have highly desirable characteristics:


  • They emit only a single colour.
  • They switch on extraordinarily quickly - in nano seconds.
  • If powered correctly (with little batteries/power units) they have an extremely long life expectancy. They are believed to have life expectancy between 50 000 and 100 000 hours, which is 10 to 20 years of use, and
  • They don’t explode.


The performance of LEDs has improved dramatically recently — and the Royal Blue colour which works best for fluorescence microscopy was only very recently developed.

A one-watt LED is very different from a 150 watt vapour lamp — it seems less bright but it is more energy efficient since LEDs only emit one wavelength compared to the bright light from the mercury vapour lamp that must be filtered out. The usable light from the one-watt LED, is approximately equivalent to a standard vapour lamp.

So Dr Anthony investigated whether he could substitute a Philips LED for the old mercury vapour lamp in their Nikon fluorescent microscopy. The process was actually simple. And, “lo and behold it worked. It worked pretty well actually,” said Dr Anthony. It was simple to install and smears showed up just as well as with the more expensive lamp.

To see how they would work in the field some LEDs and power supplies were sent to the National Hospital of Tuberculosis and Respiratory Diseases in Hanoi, Vietnam, which analysed all their sputum samples for one week, using both a mercury vapour lamp and an LED sequentially.

They analysed 492 sputum samples and found no significant difference between the performance of the LED and the mercury vapour lamp.

Dr Anthony noted: “We didn’t give them any help in converting their microscope. We found it very easy and apparently they found it very easy. So, converting an existing fluorescent microscope appears to be quite simple.”

The LED is much more cost-effective (with a longer life) and can even be powered with a battery. So it seems to be a way to deliver the benefits of fluorescence microscopy deeper out into the periphery. “The current situation is it’s relatively easy to adapt a fluorescent microscope,” said Dr Anthony. “We’re trying to establish a suitable supply of LEDs which is critical - if you use the wrong LED, it doesn’t work. My information is that there are a hell of a lot of very high quality fluorescence light microscopes out there which are not used because people can’t get lamps.”

He also believes that lower costs lamps could be made using LEDs. Indeed, FIND is working with companies to develop a low cost LED based version microscope that could be a standard issue and that is hoped to be available soon.

Dr Anthony stressed that not only is the lamp more sensitive, it improves working conditions as well.

“People much prefer to spend the day looking at fluorescence slides than ZN slides. And I think there should be a roll-out of fluorescence microscopy where possible, and it should be possible today. I think if somebody is working all day, doing lots of screens, they’re going to want a first world microscope. You can’t expect them to work with a toy.”

Other items in this series

TB diagnosis special report: why smear microscopy needs to be improved

Background on smear microscopy in TB diagnosis

TB diagnosis: Reducing the number of smears and clinic visits needed

TB diagnosis: For now, still desperately seeking better diagnostics (all references listed in this article).