PubMed1983
The objective of any experimenter should be to obtain the maximum amount of useful information in the most economical manner possible. In this context, 'useful' means study both of the principal or anticipated actions of the compound under examination, and also detection of additional responses to it. For efficient evaluation, including greater statistical rigour, a continuous or graded response and quantitative measurement is preferable, although much can be done by simply analysis of the presence or absence of responses, even if these are complex or dependent upon subjective criteria. Consistency of experimental conditions, and of diagnostic criteria are essential, whether in single, exploratory or comparative tests. Acute toxicity tests, like other experiments, should be designed pragmatically, therefore, confirming to the properties of the substance under examination, but also with general or holistic coverage, albeit less specific, to detect unanticipated actions. Greater or lesser weight may be given to these two classes of method, depending on whether the test is done as part of a general screen of a previously unknown substance, or as a specific control measure done to assay one particular property. Extrapolation of the results to man or other species is only possible with the caution proper to any experiment done under restricted and highly artificial conditions, but the value of acute toxicity data has often been shown by experience gained in man.