Decision criteria for the selection of analytical instruments used in clinical chemistry: III Non–monetary criteria

THE selection of analytical instruments is a very complex problem for a clinical chemical laboratory. The reason for this is that medical, methodological and practical organisa-tional aspects must all be considered simultaneously. It is necessary to develop criteria for these aspects which relate to characteristic and relevant features of the instruments concerned, which can be verified, and which may be quantified , wherever possible. At present, there is no consensus agreement on criteria of this kind, but there are, however, some recommendations from international scientific associations relating to particular problem areas 1,21. After the objectives have been defined as already outlined, the examination of the data for selection should be carried out in the following order of priority: (1) Applicability to the intended purpose. (2) Performance of the instrument. (3) Practicability. Applicability to the intended purpose Clinical chemical analyses are not only required for diagnostic but also for prognostic purposes, for follow-up and above all, for decisions relating to therapy. There is therefore a wide range of requirements which the analytical instruments must satisfy. These are listed in Table 1. Table 1" Decision criteria Single or multiple testing fixed program possibility of selecting tests according to request possibility of selecting clinically orientated groups of tests Analytical mode-analytical principle (photometric, others) endpoint or reaction rate measurement Sample size (volume) use for paediatric departments Degree of mechanisation-expected number of patient specimens fully mechanised (or automated) system needed electronic data processing needed Possibility of performing statistical emergency tests The requirements will also vary with the type of hospital to be served, i.e. whether general, small, paediatric, intensive care (1) unit, etc. (2) Frequently, instrument manufacturers claim 'universal' (3) applicability as a significant advantage of their analytical (4) instruments. However, practical experience shows that this (5) is rarely the case a simple instrument designed for a specific (6) purpose is often to be preferred to a more sophisticated so-called 'universal' instrument. Once the type of analytical instrument has been determined using the clinical requirements to provide the basic data, suitable criteria must be applied to determinethe performance of the individual instrument. Examples of these criteria are shown in Table 2. Analytical reliability of the measuring part of the instrument precision accuracy specificity detection limit Speed criteria specimens/samples per time unit (frequency) analysis time throughput time Correction of interferences automated detection and correction of interferences such as two wave lengths' turbidity Contamination carry-over between samples …