Statistics from Altmetric.com
I read the two letters from McCaughey and Curry and their respective colleagues1,2 with astonishment. Their debate on a policy for electron microscopy (EM) use reminded me of those mediaeval ones about the number of angels that could dance on the head of a pin. Interesting debate, pity it missed the point. They failed, I think, to address three important aspects:J Clin Pathol 2000;53:722–725
The nature of virus diagnosis. No one can pretend that virus laboratories can investigate every individual “viral” illness in the community, but they have an obligation to monitor what is prevalent in it. Although attractive to a cash strapped service, selecting specimens on the basis that some are more likely to yield positives than others misses the point of diagnosis—more than one virus can cause many “virus like” syndromes and virus excretion does not parallel exactly the presence of symptoms. To discard specimens taken from those who are recovering because the yield may be low, or where the cause is apparently obvious, strikes me as arrogant. If someone has taken the trouble to send a specimen, it seems reasonable to look at it, if only in the hope that the sender might be encouraged to send others in the future, especially if it turns out to be positive. It is a constant battle to get worthwhile virological specimens sent to the laboratory—choking them off is daft. Moreover, I would further confirm the Irish view that solid stools may often yield positives and also that making the diagnosis by holding the specimen up to the light (metaphorically speaking) is very unrewarding. Trends in infection can only be given some credence if the specimen base remains more or less constant. Arbitrary and variable selection of what will be examined, and what will not, destroys this base. Elaborating “a rational policy” is always a recipe for cutting down what is done (why else have one?), usually in pursuit of saving money, and is rarely based on sound science.
The question of money. EM is unique among virological laboratory techniques in that the major costs of using it diagnostically (equipment, staff) are incurred in setting it up in the first place. Running costs, in comparison, are trivial, but savings are thought to be possible by allowing the operator to work only part time. Used full time, the technique becomes less expensive for each specimen the more it is used, but this use must be sensible, and humane to the operator. Microscopy, and EM in particular, can never really be a part time occupation—the operator needs to keep in constant practice and to be committed to it. The cream of his/her work should not be skimmed off by using other techniques (enzyme immunoassays, etc)—to detect rotavirus or adenovirus—for example. Like everyone else, electron microscopists thrive on getting positive results and these reward immensely what can be a lonely working existence. Moreover, the laboratory gets several simultaneous tests for its money, results are produced more quickly than with other tests and, in most cases, with greater certainty.
The need to retain EM as a non-centralised resource. With changes in travel and climate, “new” viruses can appear anywhere at any time. Part of characterising a putative new virus, whether truly novel or merely transferred to a new habitat, is to know what it looks like—other properties correlate surprisingly well with structure—and having only a national, or even a regional, facility for this purpose is usually a recipe for serious delay. Specimens from the periphery do not carry the same urgency as the facility's own work, especially if the latter has to meet internal performance targets. Passing perhaps the most interesting part to another laboratory is hardly likely to encourage the local virus hunters either.
Clamping EM into a “rational policy”, instead of encouraging its use wherever possible, will ultimately cause it to die a slow death from a downward spiral of discouragement. Put in rugby terms, use it or lose it. EM is the only truly “catch all” technique in diagnostic virology. If lost, it will be replaced (partially, even at best) with a battery of tests which, collectively, will cost more, take longer to do, and will frequently miss the dual or triple gut infections that are common in children.
The authors reply
Madeley's perception of the continuing usefulness of diagnostic electron microscopy (EM) in virus laboratories is very different from ours. We believe that diagnostic EM is becoming largely redundant and will continue to be replaced by other methods. Increasingly, these alternatives are molecular. Such assays are preferable to EM on grounds of cost, sensitivity, and the inherent versatility of generic methodology. In our laboratory, we have replaced EM as our routine first line test for adult cases of gastroenteritis with nested reverse transcription polymerase chain reaction (RT–PCR) for small round structured virus (SRSV). We have replaced EM as the first line test for skin material with a multiplex nested PCR for herpes simplex virus (HSV) types 1 and 2 and varicella zosta virus (VZV).1 This approach has yielded a sixfold increase in sensitivity over EM.1
Madeley appears to confuse the cost of expensive commercial molecular assays with that of much more economical in house assays. The change from the use of EM and virus isolation to nested PCR in our laboratory was on the basis that the latter approach was more economical, in addition to the improved assay performance.2 A typical reagent cost for our in house nested PCR assays would be less than £3, including extraction, and the manpower costs are considerably less than for traditional methods.2 The cost of EM is not “trivial”; the man hours spent are a valuable resource that can be used more productively within the diagnostic laboratory.
Molecular assay delivery does not require operators with more training or experience than traditional methods such as EM, virus isolation, and antigen detection. The use of generic methodology simplifies the training and skill mixes required to provide a comprehensive service compared with that required for a service delivered via diverse traditional methods.
The term “catch all” was first applied to diagnostic EM in the 1970s and is now inappropriate. EM is insensitive and non-specific. Our nested multiplex PCR assay can distinguish between HSV-1, HSV-2, and VZV in a slide from a vesicular skin lesion. EM will often miss the diagnosis, and even when positive will only indicate the presence of a virus from the Herpesviridae family.
In 30 years of diagnostic EM we have not discovered any new viruses. In the first year of routinely using nested PCR on all skin specimens we have discovered a deletion mutant of HSV type 1.3 Looking for new viruses via routine diagnosis is not a function of a diagnostic virology laboratory. Producing results that affect patient management and infection control in an efficient, timely, and economic fashion is a major function.
If Professor Madeley considers the rewarding aspect of diagnostic virology to be the positive result, then he should welcome the availability of more sensitive methodology. There is a responsibility for diagnostic laboratories continuously to improve their performance, rather than to continue with assays whose appeal is aesthetic rather than functional.
The authors reply
Professor Madeley's comments are very welcome but his letter opens up the debate about electron microscopy (EM) to a much greater extent than that originally intended by either the rationalised EM policy paper1 or the recent correspondence relating to it.2,3 We strongly agree that EM should not be allowed to wither and die, but maintaining the relative abundance of EM units that existed into the late 1980s was untenable for two main reasons. First, the development of new investigative and diagnostic technologies (initially enzyme immunoassays and subsequently the polymerase chain reaction) and secondly financial pressures that have affected, not only the PHLS, but also hospital pathology departments and universities. Both these factors have meant that the use of EM in many centres has been critically reviewed. Many universities have centralised their EM facilities into units either specialising in materials science or biomedical science. Within pathology, many laboratories have given up locally provided EM services because of cost considerations and buy only the EM services that they require from established units, which is more cost effective than maintaining a local EM facility. Within the PHLS, a strategy for EM was formulated, which resulted in the formation of a strategic network of EM units in England and Wales being retained. This rationalised EM service provided significant improvements in virus surveillance and showed—for example, the true importance of small round structured viruses (SRSVs) as a cause of outbreaks of gastroenteritis in the UK.
A fundamental limitation of EM is that every specimen needs individual examination by a skilled microscopist. This aspect of EM cannot be automated and restricts the number of specimens that can be examined by a microscopist within a working day. Only those specimens that warrant individual attention should be examined, and this is at odds with Professor Madeley's view that if a specimen is submitted it should be examined. Newer diagnostic methods are very sensitive, much cheaper for each test, can handle greater numbers, and often require less skilled staff to perform them. It is because of such developments that EM would no longer be considered as a front line test for—for example, group A rotaviruses.
Professor Madeley also makes the point that EM is relatively cheap after the initial capital investment. However, this is over simplistic because microscopes and associated ancillary equipment all have finite lives. Even if an electron microscope were in operation for 20+ years, few organisations would have the money to replace such an instrument on anything near a one to one basis. Even in the remaining microscope units, the cost of annual maintenance contracts for such instruments is high, and when taken with the limited number of specimens that can be examined and the costs of a skilled electron microscopist, this means that a virological examination costs tens of pounds to perform. We would agree that EM is a “catch all” method, but specimen numbers and turnaround times must be considered if an efficient diagnostic service is to be provided to the customer, who has to pay for such investigations.
What we have tried to achieve with our specimen testing policy is to provide a range of diagnostic techniques to our customers, which necessitates testing only the most appropriate specimens by EM; this is underpinned by epidemiological evidence of the relative benefit of discriminating between samples as opposed to the “shotgun” effect of unconsidered and unstructured sampling.
EM does have a future, but this probably lies with units that have a broad range of ultrastructural interpretative skills, carry out several preparative methods (such as negative staining and thin sectioning), and have a high throughput of specimens, thus maximising the use of these expensive facilities. Units able to undertake a variety of work, with skilled interpretation, will prosper. However, an important problem in the medium term is how to pass on established EM skills to a new generation of electron microscopists. Few organisations that have retained EM facilities have grappled with this problem of succession planning.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.