Water Journal : Water Journal December 2011
refereed paper microbiology water DECEMBER 2011 85 or, iii) use an indicator or surrogate for pathogenic viruses (e.g. bacteriophages). On occasions where there are a number of options regarding the micro-organism of choice to meet the objectives of the monitoring program, the selection of test micro-organism will depend not only on the cost of analysis but also on the supporting scientific evidence (e.g. documented relationship between levels of indicator(s) and target virus) and the operating characteristics of the methods (e.g. recovery efficiency, false positives/ negatives, reproducibility, whether viable viruses only are detected etc). In circumstances where there are budgetary and/or other reasons for the choice of a surrogate micro-organism or analyte rather than the 'true' target of the monitoring program, all assumptions used in decision-making should be documented and supporting evidence (e.g. references to scientific papers and/or to results of prior monitoring programs) should be retained to assist future users of data. This is important to contextualise the monitoring program and results in terms of knowledge existing at the time for both future users of data and designers of monitoring programs. Where the budget amount is insufficient for the monitoring of the preferred target micro-organism and there is no evidence to support the use of an alternative, the budget and/or scope of monitoring should be reviewed. Step 6: What Methods are Available and 'Fit for Purpose'? Having selected the microorganism for testing, the next step is to determine which of the available methods (where more than one is offered) to employ. Where multiple methods are available, a standard method should be preferentially employed. In using a standard method, the operating characteristics of the method are more likely to be detailed; its access (documentation) to a wide number of laboratories is assured and data sharing via a data repository or other mechanism is promoted. From a laboratory accreditation perspective, standard methods for microbiology are accepted from recognised standard writing bodies, from compendia and pharmacopoeia texts, and method validation bodies such as the Association of Analytical Chemists. When used without modification, minimum verification to assure the proficiency of laboratory staff in the execution of the method is required for these methods to be accredited. Non-standard methods (e.g. methods developed for research purposes), and standard methods that have been modified or are used for testing outside of their validated scope (e.g. used for biosolid instead of water samples), must be fully validated, including statistical analysis of results in order for these methods to be accredited (NATA, 2009). In those instances where an accredited laboratory offers testing using methods that are outside the scope of their accreditation, the laboratory is required nonetheless to implement practices which ensure the integrity of reported results. Accordingly, where the objective of a monitoring program requires different method sensitivity to that of the standard method and/or where changes to the standard method are made, the laboratory should be able to inform the customer about the contribution that additional steps and/or modifications have on the final reported result (e.g. change in detection limits). Similarly for research methods, the laboratory should have undertaken relevant seeding experiments, performed relevant staff training and, in the performance of the method itself, incorporated relevant quality controls to allow the customer to be provided with information about the detection limit of the method and the recovery efficiency of the target micro- organism in different matrix types (e.g. turbid versus non- turbid waters). On those occasions (e.g. waterborne disease outbreak or special investigations) where it is imperative to monitor for a specific micro-organism, despite there being no standard method available for its detection/ enumeration, a newly developed and experimental method may need to be employed. In such circumstances, the suite of controls must cover the key elements of the methodology known or suspected to be problematic. For example, PCR virus (RNA) methods should incorporate process controls; a negative RNA extraction control; positive RT-PCR and RT-PCR inhibition controls and negative and positive PCR controls. In the design of a monitoring program, adequate provision must be made for the collection of method recovery efficiency datasets. The recovery dataset should be both adequate (enough samples) and relevant (recovery data should be available for the water type of interest). Questioning about the availability of recovery efficiency data for a proposed method is important. Where datasets are unavailable, discussion should follow about the the way in which a recovery efficiency dataset, relevant to the water location/ type in question, will be collected. Step 7: Determining the Number of Samples Where statistical analysis of output data is to be performed, awareness of the minimum required sample numbers prior to commencing a water quality monitoring program will ensure that sufficient samples are analysed or, if this is not possible due to budget constraints, will allow for re- consideration of the budget amount and/or the scope of the monitoring program. In some instances determination of the minimum number of samples may require preliminary (pilot) monitoring to scope the levels of target micro-organisms and/or their prevalence prior to a 'full' monitoring program being conducted. For example, where the purpose of monitoring is to construct a probability density function (pdf) of the level of a specific pathogen in a particular water type, pilot monitoring can ascertain the proportion of samples where test results are expected to be at, or below, the method detection limit. With this knowledge the minimum number of samples to construct a pdf can be calculated and budgeted for. In addition, consideration can be given to modifying the method so that the proportion of samples with pathogens at, or below, the detection limit is reduced. Where the purpose of monitoring is to assess compliance with a guideline value, the minimum number of samples required to assess compliance will depend upon the prevalence of the microorganism in the water type being tested. The lower the prevalence of the target organisms, the greater is the number of samples required. Step 6: Summary • Is a standard method available? • If no standard method is available are there available research methods that may be employed? • Does the laboratory have validation/verification data to support the use of the proposed method? • Can the laboratory provide information about the operating characteristics of the method (recovery efficiency, method sensitivity, MU etc)? • Are there any method recovery datasets for my matrix? • What is the consistency of performance of the method?
Water Journal April 2012
Water Journal November 2011