J. Soc. Cosmet. Chem. 25 307-323 (1974) ¸ 1974 Society of Cosmetic Chemists of Great Britain Relative merits of 'in use' and laboratory methods for the evaluation of antimicrobial products R. A. COWEN* Presented on 13th November 1973 in Nottingham at the Symposium on 'Evaluation of Product Performance', organized by the Society of Cosmetic Chemists of Great Britain. Synopsis--The relationship between the resistance of cultures grown in the laboratory and on the skin is examined in the light of factors known to influence RESISTANCE OF BACTERIA under laboratory culture. Examples are taken from tests designed to measure the efficacy of HAIR SHAMPOOS, DEODORANTS, BATH ADDITIVES and ANTISEPTICS to show that little correlation exists between activity measured in the laboratory and in practice. Both LABORATORY TESTS and 'IN-USE' TESTS should be carried out side by side as a means of identifying factors which influence the results. With this information design of more realistic tests should be possible. INTRODUCTION The methods used for the evaluation of the antimicrobial activity of a product can be divided into three groups. Simple laboratory tests. Simulated 'in-use' tests. In-use tests. The latter, by definition, refer solely to the evaluation of the product under conditions of usage and not for the description of other types of * Bacteriological Laboratory, Pharmaceutical Division, Reckitt and Colman, Dansore Lane, Hull. 307

308 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS test (e.g. simulated 'in-use' tests). The distinction between the simple labora- tory and simulated 'in-use' test is more difficult. It may be argued that any modification of a simple laboratory test, for instance by adjustment of pH, or by addition of body fluids, which bring it closer to the conditions of usage convert the test from a simple laboratory test to a simulated 'in-use' test. Equally, whilst a particular test may be classified as an 'in-use' test for one group of products the same test can only be regarded as a simulated 'in-use' test when applied to a totally different group of products designed for a different purpose. Thus the 'Glove Juice Test' as used by Lowbury, Lilly and Bull (1) is an 'in-use' test when applied to the evaluation of a surgical scrub. However, if the same test is applied to the evaluation of some other type of antimicrobial product, designed for instance as a de- odorant, the test must be regarded as a simulated 'in-use' test since factors are present in the test which will not be present in the 'in-use' situation. For the purposes of the present paper simulated 'in-use' tests will be considered along with laboratory tests, and these will be compared with true 'in-use' tests. The trend, in evaluating finished products making antibacterial claims, is to move away from the use of laboratory and simulated tests and to employ whenever possible 'in-use' tests. This follows the belief that only an 'in-use' test can positively evaluate the true antimicrobial activity of a product. Unfortunately, 'in-use' tests are more difficult to perform than laboratory tests or simulated 'in-use' tests. Moreover, they involve con- siderably more time and usually a panel of volunteers is required. Conse- quently, laboratory tests are still widely used during the development stages of a product. It may be argued, however, that if these tests fail to evaluate the final product satisfactorily, they may equally well fail to select the most suitable product during the development period. The reason for the failure of laboratory tests in the evaluation of an antimicrobial product rests largely with a lack of knowledge of the factors governing performance in actual practice. Once these factors are established there is little doubt they can be applied to design of more realistic in vitro methods. At the same time more information is required of the factors influencing performance in in vitro methods. By this interplay it should be possible to design much more realistic laboratory tests, but in order to achieve this there is a need for both laboratory tests and 'in-use' methods to be carried out side by side. Illustrations from laboratory tests and 'in-use' tests being carried out by the author's laboratory are used to indicate the kind of factors which influence performance.