EVALUATION OF COSMETIC PRESERVATIVE EFFICACY 327 D-values for S. aureus were 2.5 and 4.2 hr for the lotion containing the mono- and di-esters, respectively. The predicted times for complete inactivation of the S. aureus populations in these samples were 14 and 22 hr for the lotion containing the mono- and the di-ester, respectively. No S. aureus were recovered at the 1-d sampling. Preservative efficacy testing with shampoos revealed that the test bacteria had similar D-values and that the molds had significantly larger D-values. The D-values for S. aureus, P. aeruginosa, Bacillus sp., A. niger and A. flavus were 0.67, 0.81, 0.81, 7.7 and 16.7 hr, respectively. The differences in rates of inactivation are readily apparent by examination of the survivor curves for S. aureus and A. flavus in Figure 4. The Hours Figure 4. Survivor curves showing the rates of inactivation of S. aureus and A. flavus in formaldehyde- containing shampoo. Symbols: ©--© = S. aureus and I--I = A. flavus.

328 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS predicted times for complete destruction of the test bacteria were less than 6 hr and less than 80 hr for the molds. No bacteria were recovered at 24 hr and no molds were recovered at 5 d. The addition of BHI broth to lotion decreased the rate of die-off of S. aureus. The D-value of S. aureus in lotion with 4% BHI broth was 7.7 hr, whereas the D-value for this organism in the control lotion was 4.2 hr. There was perfect agreement of the predicted times for complete inactivation of the test organisms in the samples with the observed times to find 10/ml in all samples except for protein-containing shampoos and hair conditioners. Here, low levels (i.e., 500/ml) of bacteria were recovered from the protein-containing formulations at 24 hr when the predicted times for complete destruction of the bacteria were 24 hr. DISCUSSION The linear regression method was used to evaluate preservative efficacy of creams, lotions, shampoos and hair conditioners. The salient feature of the method is that the D-value provides a quantitative expression of the rate of death of a specific test organism in a particular product when using defined test conditions. The rationale for using the D-value is that every organism has a characteristic rate of death when subjected to any lethal treatment (6). The D-value concept is used extensively in thermobacteriology, where it has been useful in computing death rates of microorgan- isms and in establishing processing conditions (7). The use of D-values in preservative efficacy testing enables one to compare the rates of inactivation of several different organisms in one or more cosmetic products. The time required for complete destruction of any size population of a particular organism may be predicted when the rate of death is known for a product. For example, the mean D-value for S. aureus in one type of lotion was 2.5 hr. By use of this D-value, the time for 106 S. aureus/ml to be totally inactivated is given by the product of the log number of organisms/ml times the D-value, or 6 x 2.5 hr = 15 hr. The recovery system used in these studies consisted of LBT diluent and TSALT agar plating medium, as recommended by the CTFA (3). Complete dispersion of the test lotions, creams and shampoos was obtained by vigorously shaking the 1:10 dilution of test material in LBT 100 times. It is believed that mechanical mixing (i.e., by use of a blender or mortar and pestle) may be required for products with higher solids contents than the test materials used here and for products with substantial quantities of materials which are not water soluble. Many of the test organisms were inactivated rapidly in the cosmetic products examined consequently, the test materials were sampled immediately after inoculation in order to minimize the contact time of the organisms with the preservative system under study. An attempt was made to handle all samples in a similar manner (i.e., to standardize the procedure) in order to reduce variation in the method. A 0.1-ml saline suspension inoculum of each test organism was used because it was a convenient amount to add to the 50-ml test samples. The United States Pharmacopeia XIX method specifies that a saline suspension of the test organisms should be introduced into the test material (2), while the CTFA method allows the use of a broth culture or a saline suspension of the test organisms (3).