Write an SOP for Validation of Examination Methods and Equipment. Below follows some information about validation.

Validation of an Examination Method

To find out if an examination method provides the information the laboratory expects from that method, the laboratory first has to formulate the performance specifications for that examination procedure. With these performance specifications the laboratory indicates which results the laboratory wants to obtain by performing that examination and what information these results must give. This can be supported by quantitative data. For example: a laboratory uses the smear microscopy method to determine severity of infection with acid-fast bacilli. The requirements of the laboratory regarding this method are: the method must enable the laboratory to identify acid-fast bacilli in a sputum smear and it must enable the laboratory to count the number of bacilli per microscopy field/per 100 microscopy fields.

The next step is validation of the examination method to see if it is indeed able to comply with the performance specifications formulated by the laboratory.

Most routinely used diagnostic examination methods have been validated elsewhere. This applies to methods that have been published in established textbooks, peer-reviewed journals, or in international, national or regional guidelines. In this case a report of this validation needs to be obtained and archived by the laboratory.

The laboratory has to perform a validation of the examination method itself when the…

• Examination method has been modified from the original, validated method
• Examination method has been developed by the laboratory itself
• Standard, validated methods are used outside their intended scope

In this case the laboratory should determine if another, already validated method, exists that can be used as a gold standard to compare results with those from the method that must be validated.

In a method validation the following criteria may need to be recorded by the laboratory (not all criteria are applicable to all types of examination methods):

• Robustness of the test:
  • Tested directly by varying the method-critical parameters, for example: testing in different environmental temperature or with reagents of different shelf-lives.
  • Tested indirectly by evaluating stability of results of control samples over a longer period of time.
• Selectivity: determining if the method only measures the variable the laboratory wants to measure and the results is not influenced by interfering substances.
• Correlation: determining a relation between the measurement signal and the amount of the component to be detected.
• Detection limits and quantitation limits: identifying the limits of linearity.
• Reproducibility: compare the method by repeated testing a specific number of samples spread over the detection interval and statistically determining that results of repeated tests per samples are not significantly different.
• Repeatability: repeating a measurement of the same set of samples at least 3 times to determine that results remain the same with every repeat measurement. It is important that examinations are repeated by different persons to rule-out inter-reader variation.
• Uncertainty: determine for examinations where this has practical utility the uncertainty of measurement (standard deviation).
• Correctness: to determine correctness of the result produced by the laboratory and execution of the method the laboratory has to participate in proficiency testing (PT) schemes for the tests performed (this activity is included in phase 2). If no PT scheme is available, the laboratory has to compare the method with another standard/validated method on a periodical basis. At least the laboratory should perform internal quality controls (IQA) where possible to determine if the test results remains stable (see IQA activity).

For a more elaborate explanation of laboratory method validation it is recommended to read the technical report of the International Union of Pure and Applied Chemistry: Thompson M., Ellison S.L.R., Wood R. Harmonized Guidelines of Single-Laboratory Validation of Methods of Analysis. Pure Appl Chem 2002 74(5):835-855. A copy of the report is provided in right-hand column. Permission to reproduce the report was kindly provided by the Pure and Applied Chemistry Editorial Office.

As validation is a complicated procedure, it is also recommended to have the persons who will design and perform the validations in your laboratory follow a course or other training in validation.

Equipment validation

The laboratory needs to know whether the equipment used provides valid and desired data. To find out if a piece of equipment provides the desired data the laboratory first has to formulate the performance specifications for that piece of equipment. With the performance specifications the laboratory explains which data or which specifications the laboratory wants to obtain by using that piece of equipment. This can be supported by quantitative data. For example: the laboratory wants to know that if a centrifuge is programmed to spin samples at 8000xg, the centrifuge indeed spins the samples at 8000xg in reality.

The next step is validation of the piece of equipment.

New equipment is generally installed at the laboratory by the supplier. Before release of equipment for use the supplier must provide a validation report in which the supplier shows evidence of compliance of the equipment with the required specifications.

The laboratory has to determine validity of data produced by the piece of equipment. Among other things, it has to define the possible systematic and random error size (i.e. measurement uncertainty). For this the laboratory must try to use a reference. This can be an already existing piece of equipment of which validity has already been determined and recorded.

Note that validation of laboratory information systems (LIS) must also be counted under equipment validation. In writing of the SOP take into account specific actions that are necessary for validation of LIS. To obtain an impression of the procedure of validation of LIS, it is recommended to read the article of Biljak et al., reporting about LIS and hospital information system (HIS) validation at the Institute of Clinical Chemistry and Laboratory Medicine of the Merkur University Hospital, Zagreb, Croatia. A copy of the paper is provided in the right-hand column. Permission to reproduce the publication was kindly provided by Biochemia Media.