318 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS values for the shampoo decreased with time. By way of contrast, determination of D-values and Ea' values for the test organisms in a proprietary lotion preserved with MP and Quaternium-15 revealed no changes in these values for test samples stored at 3 ø to 49øC for the duration of the stability study. This indicated that the preservative system of the lotion was not changing detectably, as evaluated by the linear regression method. These examples illustrate a useful caveat: the Ea' values should deviate little from the initial, acceptable values if a product is to have satisfactory shelf life. These studies demonstrate that the linear regression method is useful for quantitating the rates of inactivation of bacteria inoculated into stability test samples and illustrate the value of this method for monitoring preservative efficacy of stability test samples. To our knowledge, this is the first report on the use of a quantitative method, which provides via the kinetics of bacterial death, a method to monitor preservative potency at various times during stability studies of cosmetic products. In most instances, preservative efficacy testing demonstrates that the cosmetic preserva- tive system inactivates test organisms rapidly and that the D-values are positive. The preservative system is judged to be inadequate when D-values are greater than accep- tance criteria (4) or when they are negative. For example, the preservative system fails when test organisms grow in the sample. The slope of the survivor curve is positive when growth occurs, which means that the D-value is negative. Negative D-values are rarely, if ever, reported in published literature because they indicate preservative system failure and the need for reformulation. Although negative Ea' values appear to be somewhat anomalous, one must keep in mind that they represent a decrease in potency of the preservative system that occurs on storage and that the rate of preservative deterioration is accelerated by increases in temperature. As noted above, negative D-values are obtained when test organisms grow in the test samples. Negative D-values cannot be used in calculating Ea' values because one cannot take the log of a negative number. This report shows that the linear regression method may be used for predicting the stability and the apparent molar concentration of a preservative system in addition to its already-documented utility in determining the cosmetic preservative efficacy. Thus, examination of the preservative efficacy of a formula after storage at 49øC for 1 to 3 mo. may indicate the system is unstable and that reformulation is necessary. It is recom- mended that formulation chemists quantirate preservative system potency during accel- erated aging studies to determine the likelihood of preservative system failure. ACKNOWLEDGEMENT The authors express their appreciation to Mr. W. E. Dickman for his assistance with stability testing of the samples. REFERENCES (1) U.S. Food & Drug Administration, Human and Veterinary Drugs, Current Good Manufacturing Practice in Manufacture, Processing, Packing, or Holding, Fed. Register, 43, 45014-45087 (1978). (2) D. S. Orth, Principles of preservative efficacy testing, Cosmet. Toilet., 96(3), 43-52 (1981).

SHAMPOO PRESERVATIVE TESTING 319 (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18). H. N. Bhargava and A. Anaebonam, Essentials of cosmetic preservation, Soap Cosmet. Chem. Special- ties, 59(10), 39-41, 125 (1983). D. S. Orth, Linear regression method for rapid determination of cosmetic preservative efficacy, J. Soc. Cosmet. Chem., 30, 321-332 (1979). Anon, Microbiological tests, antimicrobial preservatives--effectiveness, United States Pharmacopeia XIX, United States Pharmacopeial Convention, Rockford, MD, pp. 587-592 (1975). Preservation subcommittee of the CTFA microbiological committee, A guideline for the determina- tion of adequacy of preservation of cosmetics and toiletry formulations, TGA Cosmet. J., 2, 20-23 (1970). C. R. Stumbo, Thermobacteriology in food processing, Academic Press, Inc., New York, p. 59. (1965). I. Matsuura, Methods for the prediction of shelf life, Cosmet. Toilet., 96(1), 39-44 (1981). G. Jacobs, S. M. Henry, and V. F. Cotty, The influence of pH, emulsifier, and accelerated aging upon preservative requirements of O/W emulsions, J. Soc. Cosmet. Chem., 26, 105-117 (1975). M. M. Rieger, Current aspects of cosmetic science. I. The inactivation of phenolic preservatives in emulsions, Cosmet. Toilet., 96(5), 39-43 (1981). S. Tanenbaum, Pseudomonads in cosmetics,J. Soc. Cosmet. Chem., 18, 797-807 (1967). R. Bhadauria and D. G. Ahearn, Loss of effectiveness of preservative systems of mascaras with age, Appl. Environ. Microbiol., 39, 665-667 (1980). T. J. McCarthy, Microbiological control of cosmetic products, Cosmet. Toilet., 95(8), 23-27 (1980). P. A. Berke, D.C. Steinberg, and W. E. Rosen, Germaben II, a complete preservative system in clear liquid form, Cosmet. Toilet., 97(11), 89-93 (1982). S. M. Henry and G. Jacobs, Cosmetic preservatives--1982 update, Cosmet. Toilet., 97(11), 31-52 (1982). J. Cowls, American Water Works Assoc., 30, 979 (1938), Cited in: Difco Manual. Dehydrated Culture Media andReagents for Microbiology, 10th ed. Difco Laboratories, Detroit, p. 499. (1984) F. Daniels and R. A. Alberty, Physical Chemistry, 3rd ed. John Wiley & Sons, Inc., New York, p. 342. (1967). D. G. Pope, Accelerated stability testing for prediction of drug product stability, Drug Cosmet. Ind., 127(6), 48ff (1980).