1. Important definitions.

Epidemiological typing: this is a term in common usage. However, it is not to be encouraged. Just because two isolates are phenotypically or genotypically indistinguishable does not mean that they are related epidemiologically. This has to be proven by shoe leather epidemiological investigations performing descriptive data analysis. A preferred term would be “typing to explore epidemiological hypotheses”. There is little point using typing as a fishing tool, there has to be an epidemiological focus and an appropriate collection of organisms submitted for typing (see Criteria for referring isolates to a laboratory) to enable proper interpretation of the typing results and the likelihood that typing can provide a helpful answer to the problems posed. It is apparent in much of the literature, that the reasons for typing have not been thought through and much time, effort and resources have been spent with little to show that can stand up to scientific scrutiny. There is an extensive literature (see references) which can support the Infection Prevention and Control/Hospital Hygiene (IC/HH) community but there also needs to be training and competency assessments.

Definitions:

  • Microbial epidemiology: the study of the distribution and dissemination of microbial pathogens including their transmission patterns, risk factors for infection, and preventive and control measures of infectious disease.

  • Microbial Strain: a strain is an isolate or group of isolates that can be distinguished from the other isolates of the same genus or species by phenotypic and/or genotypic characteristics. When sampling for typing it is wise to pick five distinct colonies to reduce the chance of missing mixed strains.

  • Clone: bacterial cultures isolated independently from different sources in different locations and perhaps at different times, but still showing so many identical phenotypic and genotypic traits that the most likely explanation for this identity is a common origin.

  • Clonal variants: the recent descendants of a successful clone which have undergone minor evolutionary changes

  • Clonal complex/group: A group of bacterial isolates showing a high degree of similarity, ideally based on near-identity of multi-locus enzyme profiles and multi-locus sequence types. Identical to a clonal group.

  • Clade: This is an evolutionary term that is appearing increasingly in infection prevention and control/hospital hygiene (IC/HH) and antimicrobial stewardship/resistance (AS/AMR) literature where DNA sequencing has been performed. It refers to a grouping of microbial isolates (even those that are extinct) descended from a common ancestor.

2. Roles for microbial typing in HAIs prevention and control and antimicrobial stewardship.

Microbial typing is an essential part of the clinical microbiology laboratory’s functions in supporting healthcare associated infection (HAIs) IC/HH activities, and also can play an important role in informing AS, including the investigation of AMR. Clinical laboratories will vary in what they can offer, but they may also submit isolates to a larger central or reference laboratory, depending on ease of access, resources and ability of the service to supply useful typing information in a timely manner. There are many reasons why microbial typing might be considered. These can be considered under the following broad titles.

  1. Incident/outbreak investigation (local, national, European Union, global).

    This will include the examination of microbial isolates from epidemiologically related clusters of patients and staff colonised/infected over varying periods of time. It might also be employed to investigate what are thought to be endemic or sporadic cases. It is essential that clinicians and IC/HH team members consider the hypotheses that need to be tested. These should be informed by the cycle of infection.

    • Which part of the microbial reservoir is serving as a source of the initial outbreak or its continuation? This needs a good knowledge of the potential reservoir for different HCAI microbes. Sometimes there can be more than one source over time, and sometimes the source itself can be a mixture of strains e.g. Pseudomonas aeruginosa in sinks, Stenotrophomonas maltophilia outbreaks can involve several patients each with their own strain ab initio. This may be because they have been selected by carbapenem antimicrobials to which they are innately resistant.

    • What is the route of transmission of strains and what are the patterns of their spread?
      Although the index case may be transmitted in a particular way e.g. in the operating theatre during surgery, subsequent routes might be, for example, via the unwashed healthcare workers' hands (see hand hygiene) on the wards.

    • Examination of possible invaders: see below.

    • What are the portals of entry into patients? This might be, for example, via wounds, injection/canula sites, airborne or oral portals.

    • As the outbreak spreads, so one may need to identify the enlarged reservoir: e.g. patient carriers; in many HAIs outbreaks ~70% or more of affected individuals are colonised rather than infected either of which can be the source of further spread.

    • Informing IC/HH and as prevention and control measures of suspected outbreaks (local, national, European Union, global). A common issue during such outbreaks is to examine possible “invaders.” These are organisms isolated from patients that epidemiologically are thought to be distinct from the outbreak, but this requires confirmation. These patients might be newly admitted, but without previous involvement in the outbreak, or have already been screened in the outbreak and thought to have been unaffected. It can be further complicated, in that a hospital strain might ALSO be prevalent in the community e.g. MRSA US300, which started as a community strain but is now prevalent in many USA healthcare establishments. This is where finer sub-typing techniques can be useful e.g. next generation sequencing (see microbial epidemiology), although the clock speed of changes in the chromosome can vary between strains (this needs more research) and would also need to be considered for the specific strains being investigated. 

  2. Informing HAIs/AMR surveillance activities: (local, national, European Union, global).

    Interacting HAIs/AMR surveillance data with microbial typing information is performed increasingly, and is of more use if there are globally agreed methods, standards and a database of strains to refer to. In this way linkage between microbial type and spreadability, severity/chroniciy of the infection caused by certain clones and association with specific diseases e.g. toxic shock, can be described and prevention and control measures harmonised and monitored. 

  3. Informing the interpretation, diagnosis and treatment/follow-up of infection/colonisation for the individual patient

    Typing can help interpret whether there are mixed type infections/colonisations, relapses with a previous strain or acquisition of a new strain; these issues can also be important in outbreak investigations and surveillance activities. 

  4. Further characterisation of strains associated with AMR issues (e.g. interpreting the success or otherwise of antimicrobial stewardship interventions).

  5. Informing vaccine development and monitoring for emergence of vaccine escape strains.

    Quite apart from helping with the original choice of strains that are to be covered by a vaccine, typing plays a vital role in monitoring the emergence of new strains e.g. influenza, Streptococcus pneumoniae, that emerge over time and so threaten the success of vaccine programmes.

  6. Informing interpretation of events in Medico/Legal cases.

    Microbial typing can help expert witnesses and the courts to interpret the possible cause of infections. There are often constraints, however, as to the availability of isolates that can be tested.

  7. Analysis of evolution of strains; genetics, lineages, antimicrobial resistant or severe infections.

 

3. Criteria for assessing microbial typing systems.

Such criteria can vary depending on the level at which typing is being used:

  • locally at the healthcare institution e.g. hospital or long term care facility.

  • regionally and nationally, in reference laboratories and research centres

  • globally: increasingly there are now dedicated networks and internet-accessible databases.

The choice of methods may depend on this level. Whatever level this is, typing methods should be tested rigorously as part of laboratory quality improvement systems. An issue which faces many laboratories is that there are no systems accessible to them to help ensure that there are regular external quality assessments. This can be a challenge to the laboratory and the accreditation authorities.

  • Performance criteria of typing methods.

  1. Type ability: the ability to assign isolates to a type. Some strains may not type at all. For example, that phages might have lysogenised Methicillin-resistant Staphylococcus aureus (MRSA) making them more resistant to typing phages or they may have a large capsule that makes them non typable by serotyping systems. Sometimes a strain can initially be typable but can then become non typable e.g. S. aureus serotyping (also relevant to vaccines) and phage typing systems.

  2. Discriminatory power: the ability to assign a different type to two unrelated strains sampled randomly from the population of a given species. There are various indices to measure this e.g. Simpson’s Diversity Index.

  3. Reproducibility: the ability to assign the same type to an isolate tested on different occasions.

  4. Stability: the stability of marker(s) assessed by the typing method. Sometimes the strain is sub-cultured several times to explore this, it can also indicate that the strain per se has a short “clock speed” for change. This can be a feature of a strain or sub strain and may be either useful of confusing depending on the typing hypotheses that are being explored.

  5. Epidemiological concordance: the results should reflect, agree with, and further illuminate available epidemiological information. Sometimes the term “Epidemiological Typing” is used but really one should use the term “Typing to explore epidemiological hypotheses”. There is little point using typing as a fishing tool, there has to be an epidemiological focus and an appropriate collection of organisms submitted for typing to enable proper interpretation of the typing results and the likelihood that typing can provide a helpful answer to the problems posed.

  6. Convenience criteria: (fit for purpose: includes quality assurance/accreditation criteria); for example can the typing method cope with workloads in different laboratory settings e.g. reference laboratories will have larger workloads than a typical clinical laboratory; flexibility (can it address small and large batches adequately on the different days? can a typing run be easily interrupted if an urgent specimen arrives?); satisfactory turnaround times (these need to be audited regularly, there are many different ways of measuring them and this needs to be considered in quality assurance systems e.g. time from the ward, time to the clinical laboratory, time to the reference laboratory and time the result is returned to the clinical laboratory and the ward. Another interesting area to explore is how long it then takes for appropriate IC/HH or AS measures to be implemented); technically easy (training, can it be performed to a high standard even when not done regularly, can the same system be used to type many different microbes); easy to analyse and communicate results (e.g. via the internet); economically viable (e.g. costs of equipment, servicing, reagents and staff: note that in some countries reagents are expensive and staff cheaper whilst in others the opposite is the case); electronic databases available with required utility e.g. interact with other national or global databases?

4. Potential issues in clinical laboratories.

  • Specimens not being taken.

    This is not a minor issue and can be a threat to the alert organism and alert condition surveillance systems that should be in place and accessible by the IC/HH teams for all the healthcare systems for which they are responsible. Specimens may not be taken for a variety of reasons. For example: costs, lack of training, lack of effective policies and procedures. More difficult to address in many countries, are the shortening of lengths of patient stay. Infections are thus not presenting in hospitals and patients may not be sampled in the community.

  • Insufficient/incorrect organism identification e.g. some laboratories only identify “coliforms”, some organisms are more difficult to identify correctly and so are labelled for example as a species“ e.g. “Acinetobacter spp”

    5. Criteria for referring isolates to a laboratory.

    These were agreed in Harmonisation of Antibiotic Resistance measurement, Methods of typing Organisms and ways of using these and other tools to increase the effectiveness of Nosocomial Infection control (HARMONY) a Directorate General responsible for Science, Research and Development (DG R&D) funded project and c/should also be considered locally for clinical laboratory “in-house“ typing systems.

    1. Send the minimum number of isolates from an outbreak (this should rarely be more than half).

    2. Temporally related isolates should be stored. This is especially useful for potential medico legal cases e.g. blood culture isolates. Also where there are regular local HCAI issues that need investigation e.g. post caesarian wound infections, venous line infections.

    3. Give priority to isolates from invasive or serious infection

    4. Avoid sending multiple isolates from single patients or the environment. This might be necessary where a hyper-mutating strain is perhaps an issue and one wants to identify what molecular or other typing characteristics might be more reliably measured.

    5. Wherever possible, use biochemical tests or antimicrobial susceptibility profiles (antibiograms) as a markers to assist in the selection of isolates for referral


    6. Phenotypic typing.

    Quite a few techniques are basic technology but CAN be useful, especially if validated by more sophisticated typing methods. Increasingly DNA molecular techniques are used (see molecular epidemiology).

    • Advantages: cheap, can often cope with large batches, rapid turnaround time as not need access to a reference laboratory.

    • Disadvantages: considered old fashioned. Can be species-specific (e.g. phage and serotyping), can be complicated to setup, may require the use of animals, can have reproducibility and discrimination issues. The same phenotype can contain different genotypes and of course different phenotypes can comprise the same genotype. However, one can also argue this is an advantage in that such phenotypic variation can cost-effectively inform the type of genotyping required to explore such phenotypic variations i.e. it is a hybrid phenotyping and genotyping system. This is particularly important in exploring new AMR problems.

    Phenotyping methods:

    1. Antimicrobial susceptibility testing. Utilises the usual clinical antimicrobials but can be potentially made more useful if extra agents are included e.g. extra antibiotics, metals, disinfectants.

    2. Biochemical markers: can also be used or added to antimicrobial markers above.
      For both they can be assisted by other typing methods e.g. phage or DNA methods to inform the clinical laboratory which extra agents to use e.g. in ongoing outbreak(s).    

    3. Serotyping: specialised, animals are required.

    4. Phage typing: difficult to do to a high standard. Cheap and can deal with large batches.
      Many issues regarding typability and reproducibility. Changes of phage type over time can also be related to other changes within the organism e.g. toxin carriage.

    5. Bacteriocin typing: often in-house systems. Standardisation poor.

    6. Toxin assays: can be difficult to setup and perform reproducibly unless commercial reagents are available. Discrimination poor but can be valuable in e.g. toxic shock cases, C. difficile infection.

    7. Multi-locus enzyme electrophoresis (MLEE): labour intensive, not easy to setup.
      Most use DNA approaches (MLST) instead (see molecular epidemiology).

    8. Proteomics: become more available with MS, MALDI TOF and SELDI TOF (see molecular epidemiology): discrimination needs to be validated for different species and shown to be of use for local strains of these species. DNA methods are currently more popular. Attractive in that many more laboratories have this equipment and it has other uses e.g. diagnostics.


      Link to European IC/HH Core Competencies: area 4 Infection control activities: Advising appropriate laboratory testing and use of laboratory data ICA 1-3

      References

       

    • Cookson B, the HARMONY participants. HARMONY – the International Union of Microbiology Societies’ European Staphylococcal Typing Network. Euro Surveill. 2008;13(19):pii=18860.

    • Available online: http://www.eurosurveillance.org/ViewArticle.aspx?ArticleId=18860

    • Struelens MJ. Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems.ClinMicrobiol Infect. 1996; 2:2-11.

    • Van Belkum A, Tassios PT, Dijkshoorn L, Haeggman S, Cookson B, Fry NK, et al. Guidelines for the validation and application of typing methods for use in bacterial epidemiology. Clin Microbiol Infect. 2007;13Suppl 3:1-46 http://onlinelibrary.wiley.com/doi/10.1111/j.1469-0691.2007.01786.x/pdf

    Original contribution from:

    Barry Cookson, University College London

    Smilja Kalenic, University of Zagreb School of Medicine