746 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS irritancy of a test substance would be based, not on a score, but on the average time taken for a visual skin reaction to be produced. In this way, a severe irritant would be detected after 1 or 2 days, whereas a mild irritant might require up to 10 or 20 days before being detected. The authors believed that the technique considerably reduced errors due to subjective assessment which are bound to occur when skin reactions are scored after single exposures of 24 h. The technique might prove misleading, however, in distinguishing between primary irritants and substances likely to bring about skin fatigue. To distinguish primary irritants from agents which cause skin fatigue, two groups of volunteers could be used (54). One group might be given the standard Draize test with the cutaneous responses being scored after 24 or 48 h exposure, whereas the other group would be treated repeatedly for 1 or 2 weeks by Lanman's test (24). In this type of investigation two controls are required, namely a known primary irritant and a skin fatigue inducing substance. An agent which fails to give a response by the Draize test but gives a reaction by Lanman's test after 10-20 days would be regarded as a mild irritant or skin fatigue inducing agent. As a definitive test in man, substances are tested under conditions which are more closely related to the product in its intended conditions of use (33, 55). One such test designed to incorporate features of the patch testing procedure and the recommended use of the final product, was suggested by Rieger and Battista (5). This test takes the form of daily applications of the test compound to the skin for 30 consecutive days, followed by a rest period of 7 days. Then the substance is applied under its exact conditions of use for an unspecified period. During the initial period, the skin reactions are examined daily whereas in the latter stage, reactions are noted as and when they occur. The results given by such a test might be misleading since the test is likely to also show up agents producing contact sensitization. Fisher (55) suggested a test in which substances are rubbed into areas of the arm or back three times daily for 3 consecutive days. A test frequently employed in the safety testing of detergents takes the form of repeated daily immersion of the arms and hands of volunteers in suitable solution of the test substances, for periods which may vary from a few days to several weeks (48, 56-58). In this test the arms and hands would be ex- amined daily and scores allocated for any reactions present. Satisfactory safety data obtained from these confirmatory tests might be accepted as being indicative of the suitability of a product for marketing. However, in order to be satisfied that a new product is without harmful

APPRAISAL OF METHODS FOR DETECTING PRIMARY SKIN IRRITANTS 747 effects, Schwartz and Peck (59) suggested that yet a further precaution might be taken, namely, that the product should be distributed to between 5 000 and 10 000 people. If no adverse reactions are reported, the product could be deemed safe for public use. The outcome of human patch testing may be influenced by variations in the sensitivity of the individual subjects and environmental factors. Bjornberg (60) reviewed the various human factors which could influence the response of human skin to primary irritants. He found that females were more sensitive to irritants than males, and white people more than negroes. In another study, Caucasians were found to give very consistent results in patch tests (11). It is likely that the variations in response to sex or race are due to inherent differences in the morphology of the skin, its biochemical constitution (61, 62), the rigidity of the skin at the test site and the degree of muscle action beneath it (23). This explanation may also account for the greater sensitivity of the skin in one area of the body than in another (24, 63). For example, the skin of the eyelids, which is devoid of hair and has only a thin layer of keratin (64), is considerably more sensitive than the skin on the soles of the feet, where it is thick and has a substantial layer of keratin. The sensitivity of the skin is dependent also upon the health of the individual at the time of testing. Skin which is subject to psoriasis is considerably more sensitive to irritants than normal skin (55, 65). The environmental influence on the outcome of human patch testing is such that the cutaneous response to irritants is greater in winter than in summer (4, 66, 67). In summer, when the skin is well hydrated as a result of sweat and sebaceous gland activity, a subject is less likely to respond to a mild irritant than in winter, when the sweat glands are less active and the skin is subject to chapping and increased tenderness (68-70). Thus, it would seem preferable to conduct patch tests in winter rather than in the summer, particularly when wishing to detect mild irritants (4, 24). Aids for detecting tissue damage in patch testing Vital dyes Frequently, the skin reactions in animals are difficult to detect due to the erythema being masked by the colour of the skin (71). In an effort to overcome this difficulty, vital dyes have been used. These bind to serum proteins which then diffuse through the walls of the blood vessels where, due to inflammatory changes, there is increased permeability (72).