EVALUATION OF EFFICACY AND SAFETY OF COSMETICS 23 be dispensed with if we know the formulation of the product. The final patch test is done on at least fifty, and preferably one or two hundred subjects. The conditions of the patch test vary with the product being tested and its method of use. Some materials, such as lipsticks, are applied to the skin and left uncovered, although usually the patch is covered with an Elasto- patch, which is a strip of elastic adhesive tape, the centre of which is protected with a cell, ophane disc to prevent the adhesive material com- ing in contact with the product being tested. We ordinarily use the flexor surfaces of the arms, and although we prefer to use the lower arms be- cause they present a more rigid surface and the patches remain in place better, we are often obliged to accede to the demands of female subjects that they be patched on the upper arm. Occasionally we use the back instead of the arms, especially when a relatively large test area is desired. Some products are applied as is, others are mois- tened with water or artificial per- spiration. ,The patches are worn for periods of time ranging from several hours, in the case of certain solvents and plasticisers, to five days, in the case of textile materials. But for most cosmetics, the patches are worn for one or two days. The subjects are examined for any irritation, erythema or other derma- titis immediately after removal of the patch. The test area is then thoroughly cleaned with a suitable non-irritant solvent, followed by soap and water, and the subjects again examined after one or two hours. They are again examined the following day and at two-day intervals for one week, in order to catch any delayed reactions. If performed on a large enough number of subjects, this test will indicate whether or not the product is apt to produce an irritation on primary contact, that is whether it is a pri- mary irritant. Obviously, the greater the number of subjects patched, the greater is the accuracy and dependability of the test. If the results are questionable, or if they are apt to involve court testi- mony, they are checked by our staff physician or preferably. by a con- suiting dermatologist. Some prod- ucts are not primary irritants but they are sensitizers. In other words, a person may not react to a product when he first comes in con- tact with it, but he may thereby be sensitized to it so that subsequent exposure may produce a reaction. Although this is really an allergic reaction, the manufacturer should guard against the use of materials that may be sensitizers to even a small percentage of the population. The period during which a person is most apt to show maximum sensi- tivity is ten days to two •eeks after the original exposure. For this reason, if we wish to determine whether a product is apt to be a sensitizer, the same subjects are re- patched after ten days to two weeks with the same products in the same place as previously. The technique is exactly the same as that used in

24: JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS the original patch test. We owe a great debt to our next speaker, Dr. Louis Schwartz, who is the out- standing authority on this subject, and in whose presence I hesitate even to mentior• patch tests. His publications and written advice have been of inestimable value to us, and before tackling any unusual problem, we always ask ourselves, "How would Louis Schwartz do it ?" Some substances, particularly cer- tain dyes, may not be either primary irritants or sensitizers, but they may be so-called photosensitizers, that is they may cause a reaction only when activated by ultra-violet light. Thus any pigmented cosmetic to be used on the face or hands and which may be exposed to the action of sun- light during use, such as lipstick, rouge, face powder, cake make-up, nail lacquer, etc., should be tested forphotosensitization. This is done by exposing the patched area to the action of ultra-violet light for a period of time calculated to produce a mild erythema on the subject. This varies from five to fifteen min- utes, depending on the complexion of the subject. It is preferable to use natural sunlight, but because it is impossible to standardize, we always use artificial ultra-violet light. We use a 1 per cent solution of' eosin as a positive control, and we find that anywhere from 10 to 25 per cent of the subjects are photo- sensitive to eosin. We must bear in mind that the results of patch tests are much more dependable when they are positive than when they are negative. In other words it is much easier to say that a product should not be used because of positive reactions, than it is to say it is probably safe to use because of n. egative reactions. All products passing the patch test should be followed up by actual use tests before the product is placed on the market. 'So much for safety • tests. Now let us turn our attention to perform- ance tests. There are no set rules to go by, but we must adapt our tests to the •product and what is ex- pected of it. Stability tests are of prime importance. The product should not change its characteristics on aging for a reasonable period of time. If it is a cream, it should re- main as a cream and not harden or cake. If it is an emulsion, it should not break into two phases. Acceler- ated aging tests are run at 52øC. (about 125øF.) when. the product will stand that temperature. In general, we have found that if a product is stable for three weeks at 52øC., it will probably be stable for one year of normal shelf life. Sta- bility at low temperatures is also a factor. These tests are run at 4øC. (about 40øF.). It is often advis- able to determine whether a product that has been frozen will regain its original consistency on thawing. Comparative accelerated aging tests are also run on emulsions in a cen- trifuge, where different competitive products are centrifuged at a defi- nite speed and the time required for each emulsion to break is deter- mined. Stability to bacterial and fungous