332 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table I Comparison of D-Values and ED-Values for Cosmetic and OTC-Drug Products Test product Correlation Test organism D-value coefficient ED-value Shower gel Cleanser Cleanser Night cream Sunscreen Sunscreen Sunscreen Sunscreen Facial toner Facial toner Sunscreen Sunscreen Sunscreen Moisturizer Moisturizer Moisturizer Moisturizer Moisturizer Moisturizer Facial toner Shower gel Face product Antidandruff shampoo Antidandruff shampoo Sunscreen Sunscreen Moisturizer Moisturizer Moisturizer Moisturizer Moisturizer Moisturizer Facial toner Facial toner Facial toner Sunscreen Sunscreen Night cream Night cream Night cream Night cream Night cream Sunscreen Shower gel Shower gel Shower gel Moisturizer Moisturizer Moisturizer Moisturizer Moisturizer S. aureus 2.1 - O. 965 S. aureus 7.5 -0.997 S. aureus 6.8 - 0.989 P. aeruginosa O. 5 - 1. O0 S, aureus 2.9 -0.991 Bacillus 5.2 - O. 967 C, albicans 9.0 - 0.993 A. niger 4.5 - 0.999 S. aureus 25 - 0.994 S. aureus 30 - 0.978 S, aureus 3.0 - O. 996 C. albicans 4.0 - 0.995 A. niger 3.5 -0.934 S, aureus 3.1 - 0.993 Bacillus 7.1 - 0.943 E. coli 2.4 - 1.00 P, aeruginosa 1.0 - 1. O0 C, albicans 4.9 - 1. O0 A. niger 4.7 - 0.995 S, aureus 33 -0.974 S, aureus 3.0 -0.992 S. aureus 39 -0.978 S. aureus 8.6 - O. 999 C. albicans 11 - O. 995 S, aureus 4.0 -0.995 S. aureus 2.9 - O. 999 S. aureus 3.0 -0.99 E. coli 11 -0.96 P. aeruginosa 1.0 - 1. O0 P, cepacia 1.0 - 1. O0 C. albicans 4.2 - 1. O0 A. niger 4.5 -0.99 S. aureus 3.7 - O, 993 S, aureus 6.7 - O. 999 S, aureus 3.9 - 0.999 C, albicans 1.8 - 1. O0 Ao niger 5.1 - 0.99 S, aureus 17 - O. 998 Bacillus 10 - 1. O0 P. cepacia 0.1 - 1.00 C. albicans 4.9 - 0.994 A, niger 4.5 -0.993 S. aureus 3.0 - O. 991 S, aureus 15 - O. 999 S. aureus 3.0 - O. 991 S. aureus 7.3 - 0.969 S, aureus 5.1 - O. 99 Bacillus 8.9 - O. 99 E. coli 5.5 -0.99 C, albicans 2.6 - 1. O0 A. niger 3.0 - O. 99 3.1 7.7 7.3 5.4 3.0 4.5 8.6 4.6 27 27 3.2 4.4 4.8 3.3 6.2 3.2 3.8 4.7 4.9 31 3.2 42 9.1 11 4.2 3.3 3.2 9.7 3.5 3.9 3.9 4.7 3.9 6.8 4.3 4.7 5.1 18 10 3.4 4.8 4.6 3.2 14 3.2 7.1 5.4 8.4 5.5 3.5 4.6

PRESERVATIVE EFFICACY TESTING 333 Table I (continued) Correlation Test product Test organism D-value coefficient ED-value Sunscreen S. aureus 3.1 - 0.990 3.4 Sunscreen Bacillus 6.8 - 0.999 7.0 Sunscreen E. coli 6.6 - 0.989 6.4 Sunscreen P. aeruginosa 0.6 - 1.00 3.6 Sunscreen C. albicans 3.8 - 0.99 3.8 Sunscreen A. niger 7.3 - 0.99 7.4 Sunscreen S. aureus 5.1 - 0.997 5.3 Sunscreen Bacillus ! ! - 0.99 9.9 Sunscreen E. coli 6.3 - 0.97 5.8 D-values and ED-values are in hr. Correlation coefficient is for linear regression used in determining D-values for each sample. 3 hr when the D-values were 10 hr. The mean D-values obtained by use of both methods differed by 0.5 hr (6.6%). The greatest differences in D-values and ED-values were seen in samples in which the test organisms died so quickly that D-values were indeterminant for both the linear regression method (because no viable organisms were recovered at 2 hr) and the rapid screening method. For example, the D-value and ED-value for P. aeruginosa in a night cream were 0.5 and 5.4 hr, respectively. The ED-values were larger than D-values estimated by the linear regression method because data points at 2 and 4 hr allowed closer discrimination of the time at which the test population of organisms was killed as a result of exposure to the preservative system of the product, which enabled a more accurate determination of the D-values than estimating D-values using two APCs. When the D values were indeterminate, (e.g., 0.5 hr, 2.9 hr, 4.5 hr), ED-values were greater than D-values in all but one sample, in which case they were equal. This is desirable for use of the rapid screening method because obtaining ED-values that are larger than the D-values demonstrates that the rapid method generally errs on the conservative side. (Note: The estimated rates of death provided by indeterminant ED- values make it more difficult for the test product to meet acceptance criteria, which means that the rapid screening method provides a conservative estimate of the rates of death of the test organisms). The ED-values obtained by using the rapid screening method were plotted as a function of the D-values in a scatter diagram (Figure 1). The proximity of the data points to a line with a slope of 0.95 hr/hr illustrates the close agreement of the ED-values with the D-values. Linear regression analysis of these data gave a correlation coefficient of 0.98, which indicates excellent goodness of fit of the data to the linear regression. The ED-values tend to level off at around 4 hr. This is due to the rapid screening method being unable to give ED-values much less than 4 hr when using approximately 106 organisms/g test sample and determining APCs at time 0 and 24 hr (see below). The survival curve slope method allowed determination of antimicrobial synergism of formula components and preservatives in multicomponent systems (6). This method allowed determination of D-values and survival times (STs) when APC sampling times are known. In addition, it allowed estimation of the maximum possible D-value (MPD-