d. Soc. Cosmet. Chem. 29 399-412 (1978) A rapid hot-room procedure for testing the performance of antiperspirants D. C. CULLUM Unilever Research Laboratory at Isleworth, 455 London Road, Isleworth, Middlesex, U.K. Presented at the Symposium on 'Product Evaluation', 18 April 1978, at Eastbourne Synopsis Previously described hot-room methods for the evaluation of antiperspirant efficacy either take an inconveniently long time to complete or permit only a low rate of throughput of subjects. A protocol is described by which a test can be completed in 3 days. It is a shortened form of an established method, and it is shown that the shorter version satisfies the basic requirements of a valid evaluation procedure. The nature of these requirements is also discussed. INTRODUCTION During the last 20 years or so, hot-room methods for testing the performance of antiperspirants have found increasing use and are now generally accepted indeed, the OTC Antiperspirant Panel of the FDA demands a specified level of performance in either a hot-room test or an ambient temperature test for a product to be recognised as an antiperspirant (1). The various published procedures all suffer from one disadvantage, namely that the rate of throughput of test subjects is inconveniently low, particularly for a product development programme in which the decision on what to test next day may well depend on the result of the last test. For example, the method proposed by Fredell and Read (2) required 4 weeks of daily measurements for one test. All methods involving a cross-over of treatments require at least 4 weeks, of which the middle 2 are normally a rest period during which the effects of the products administered in the first week gradually wear off. The method of Fredell and Longfellow (3) could be completed in 4 days, but hot-room sittings had to be repeated daily, so that the number of subjects was limited. The same applies to the method used at Hilltop Research Inc. (4). The method of Wooding and Finkelstein (5) involved only one hot-room sitting but required daily application of product for 5 days, so that only one test per week could be done, and the maximum number of subjects was limited to that number which the hot-room facilities could deal with in 1 day. We required a method which would enable us to complete a test within a week, using a sufficiently large number of subjects to yield acceptably narrow confidence intervals as well as to give reasonable assurance that they resembled the population at large. Fairly arbitrarily this number was set at fifty, although we often work with fewer. Experience had shown that with some products no appreciable antiperspirant activity was shown in a hot-room test immediately following application, while maximum 0037-9832/78/0700-0399 $02.00 ¸ 1978 Society of Cosmetic Chemists of Great Britain 399

400 D. C. Cullum effectiveness was not reached until after at least five consecutive daily treatments. It was decided that three consecutive daily treatments, followed by a hot-room sitting on the third day, would be the best compromise between the need for repeated application of products and the need to do as many hot-room sittings as possible in a week, since it would permit sittings on Wednesday, Thursday and Friday in a normal working week. The procedure was thus essentially a modification of the 'SSEM' method of Wooding and Finkelstein. It is described in detail later. The chief purpose of this paper is first to define the basic requirements of a valid performance evaluation method and then to show that, in its modified form, our 3-day test satisfies these requirements. PRINCIPLES THE PURPOSE OF THE TEST It ought to be self-evident that the first thing to do in devising any kind of test is to define precisely the information the test is required to produce. In product development, performance tests normally serve one of three purposes: to provide support for advertis- ing claims to assess the extent to which users will be able to perceive that the product works and to tell product development scientists whether they are moving in the right direction. In practice, 'the right direction' will usually be defined in terms of either advertising support or user perception, so there are really only two purposes. It is unlikely, though not impossible, that one fixed procedure will serve both of these. The hot-room stimulus and the methods used for assessing sweat output are artificial, and the only purpose such a test can serve is to rank products in order of efficacy under an arbitrary set of conditions. For reasons to be discussed later this will not necessarily be the order of efficacy perceived by users. MODE OF APPLICATION OF PRODUCTS In recommending a test procedure we may prefer on the one hand complete control over all the variables, which does not resemble what happens in real life, or on the other hand complete realism, which will entail considerable variation in modes of application of the product, applied stimuli, and parameters and criteria of efficacy. Usually we shall strike some kind of compromise between these extremes. The question of how best to apply the products is a particularly difficult one, yet it is surprising how little attention it has received in the published literature. Its supreme importance lies in the fact that when two different products are compared, either In separate tests against a common control or directly in one test, the result in the sense of a decision as to which is better may be dictated by the amounts applied. Although the hot- room test is artificial and can achieve no more than a ranking of products, it is surely implicit that we are interested in the ranking order under conditions of actual use. The problem is most easily seen when a comparison is made between products of different types, for example a roll-on and an aerosol. It we apply eight strokes of the roll-on and a three-second spray with the aerosol, perhaps the aerosol will appear better. If we give twelve strokes and a two-second spray, perhaps the roll-on will appear better. Indeed, such a comparison has no discoverable meaning unless the products are applied in a manner which closely resembles the mode of application in real use.