Food, water and environment

The examination of nonclinical specimens is beyond the scope of this book and readers are referred to appropriate reading. However, an outbreak of gastrointestinal disease inevitably leads to the question of identifying the source and comparing strains isolated. When disease is due to preformed toxins, as with staphylococcal food poisoning, faecal examination is unhelpful, and the diagnosis can only be made by testing the food. Frequently, the offending item has been discarded and the examination of food and water related to specific patients is often unrewarding. Routine sampling of water sources and potentially contaminated food such as poultry at various critical points of production is essential in maintaining good public health. As part of the investigation of an outbreak it may be necessary to collect samples under controlled conditions which can be used as evidence in any prosecution. Laboratories need to be able to receive and process such specimens in an approved way.

Microscopy

Microscopy is an important part of the examination of many specimens. For bacteriology, the Gram and acid-fast (Ziehl–Neelsen or Auramine) stains are usually sufficient, but for the demonstration of fungi or parasites special stains or concentration techniques may be required. ‘Wet’ mounts, i.e. unstained preparations of fluid material, are widely used in looking at cells in urine, cerebrospinal fluid (CSF), faeces and vaginal secretions. None of these procedures takes more than 5 or 10 min and all are inexpensive in reagent costs and capital equipment. They are therefore ideal rapid methods and new diagnostic techniques have to be judged against microscopy. An initial report, such as ‘Gram-negative diplococci and pus cells seen’ from meningitic CSF, can be issued within minutes of receiving the specimen and will aid the clinician in confirming the diagnosis and starting appropriate antibiotics.

Similarly, a rapid diagnosis of falciparum malaria can be lifesaving. Indeed, suspected pyogenic meningitis and falciparum malaria are among the few conditions for which it is clearly justifiable to call upon emergency laboratory services outside normal working hours.

Electron microscopy (EM) requires more elaborate preparation than does light microscopy and it is therefore much slower. It was valuable in the diagnosis of certain viral infections, including viral diarrhoea but the availability of antigen (ELISA) and molecular probes has largely replaced EM. The nonspecific nature of electron microscopy gives it an advantage in that any type of viral agent, if present in sufficient numbers, may be recognized.

Nonculture methods

There are many situations where isolation of microbes in vitro or in tissue culture is impossible or is insufficiently sensitive to make a microbiological diagnosis. Some microbes are so fastidious or slow-growing that useful information cannot be given to clinicians. The isolation of many viruses requires laborious tissue culture methods that are too slow to influence patient management. Microscopy is usually of low sensitivity, e.g. the threshold of detection of acid-fast bacilli in sputum is about 10⁵ microbes per millilitre. Microscopy also has low specificity; for example, Gram staining of faeces would yield millions of Gram-negative rods, but it is not possible to recognize those that are pathogenic by this means.

To overcome these deficiencies probes have been developed which combine a part that reacts with a specific microbial structure and a part that will produce a signal (colour, fluorescence or radioactivity) after the reaction. Many microbes are detectable in this way. Antibodies labelled with fluorescent molecules are widely used in diagnostic virology, e.g. for respiratory secretions to find respiratory syncytial virus. Enzyme-linked antibodies are available as enzyme-linked immunosorbent assay (ELISA) kits for chlamydia detection, and many other immunoprobes are commercially produced or under investigation. The explosion in molecular biology has led to the widespread availability of DNA and RNA probes, which have changed from complicated research techniques requiring radioisotopes to relatively simple methods that can be carried out with minimal expertise. Use of amplified nucleic acid tests (e.g. polymerase chain reaction, PCR) has greatly increased the sensitivity of probes. Provided a unique sequence of nucleotides is used, the method is highly specific. None the less, as the new technology becomes more commonplace traditional methods will not be discarded. It is more likely that it will complement them.

One of the chief advantages of probes – their specificity – is also one of their major disadvantages. For example, in the investigation of diarrhoea, it would be impossibly laborious to have to use a separate probe for each of the possible microbial causes. Furthermore, use of probes leaves little scope for detection of the unexpected and hampers the discovery of previously unsuspected aetiological agents of disease. The advent of ‘chip’ technology (microarrays) may overcome this deficiency (see Ch. 6).

Serology

In situations in which microbial isolation is impossible and probes are unavailable, evidence of infection may be obtained by finding a rise in antibody titre or the presence of specific IgM. Serology is still used for diagnosing the causes of ‘atypical’ pneumonia (mycoplasmal pneumonia, psittacosis, Legionnaires’ disease), syphilis, brucellosis and many viral infections, including HIV. It is preferable to take a blood sample early in the illness (the acute serum) and another 10–14 days after the onset (the convalescent serum); a four-fold or greater rise in antibody titre in the second specimen is diagnostic of acute infection. With some infections, such as HIV and hepatitis B and C, much longer times must elapse.

Isolation of micro-organisms

The basis of the study of medical microbiology was laid over a century ago by the isolation of microbes in pure culture outside the host animal. The methods used by the fathers of bacteriology have been adapted, simplified and, in some cases, automated for the modern diagnostic laboratory. The principles remain the same: use sterile equipment and media (with cell lines if necessary) and add clinical material. After incubation at 37°C, for a variable time, from a few hours for enterobacteria to weeks for mycobacteria and some viruses and fungi, a visible effect will be produced. This might be colonies growing on agar or a cytopathic effect (CPE) in tissue culture. The skill of microbiology is in identifying the microbes responsible for the effect.

There are limits to the methods that can be used in a routine hospital laboratory to isolate microbes. The choice of media used is dictated by the specimen and by the clinical condition of the patient. For example, in some areas of the world there is little point in looking for Corynebacterium diphtheriae in every throat swab, so that the specific media needed are not routinely used. It is therefore incumbent on the doctor to make sure that the laboratory is alerted to look for diphtheria bacilli in any suspicious case. Similarly, clinicians need to know which microbes are routinely sought from particular specimens so that they can make a special request if they suspect the unusual.

‘New’ causes of illness are often found and new methods of investigating old diseases regularly appear on the market. It is a difficult decision for the laboratory manager to assess at what point a ‘new’ pathogen becomes sufficiently important to be routinely sought, or to balance the advantages of new (and expensive) technology against cheap and well tried techniques especially in resource poor countries.

Identification

The full identification of each microbe isolated in a clinical laboratory is both uneconomic and unnecessary. Shortcuts must be made to satisfy the clinical demand of a final report which the doctor can understand and which is available in time to influence management of the patient. In practice, most laboratories use simple and incomplete methods of identification, depending on the level of useful information required. Typing of isolates is for epidemiological or other special reasons, and this is usually done in national reference centres using standardized methods.

Examples of the extent of identification are shown in Table 66.3. This shows that the same organism may not be identified even to the genus level, or it may have extensive genetic investigation depending on the reason for the request. In a cost-conscious climate you get what you pay for and what the service thinks you need!

Table 66.3 Examples of bacterial identification in medical laboratories

Antimicrobial sensitivity testing

One advantage of good bacterial identification is that it helps greatly in choosing antibiotics and in some situations in-vitro tests of susceptibility may be unnecessary. However, the widespread occurrence of bacterial resistance, even in genera such as Neisseria and Haemophilus in which sensitivity to β-lactam antibiotics was previously assumed, has meant that tests are usually performed on all significant clinical isolates of bacteria. The relevance of information generated in the extremely artificial conditions of the laboratory is discussed in Chapter 67.

As there are technical problems in carrying out sensitivity tests at the same time as the primary culture, the report will be delayed for at least 24 h after isolation of the pathogen. Tests of slow-growing bacteria such as mycobacteria take much longer and tests involving fungi, viruses and protozoa are not ordinarily available. In general, all patients with acute symptoms will receive treatment before the report returns to the doctor so that the result will merely confirm that correct treatment had been chosen. Sometimes, the laboratory report will allow speculative antimicrobial therapy to be modified, e.g. by allowing one or more drugs in a precautionary mixture to be discontinued.

A frequent difficulty facing microbiologists is which, out of a rapidly increasing number of agents, should be tested and, of those tested, which should be reported. Most laboratories test a few representative compounds from among the many penicillins, cephalosporins, aminoglycosides, tetracyclines, quinolones, etc., and restrict those reported to the clinician to two or three agents selected for their appropriateness in the particular infection. In this way, institutional antibiotic policies are reinforced and the impact of the promotional activities of pharmaceutical companies is lessened.