Deception, perception, and P values.

OBJECTIVE: There has been a recent proliferation of medical reference texts intended to guide practitioners whose patients use herbal therapies. We systematically assessed six herbal reference texts to evaluate the information they contain on herbal toxicity. METHODS: We selected six major herbal references published from 1996 to 2000 to evaluate the adequacy of their toxicological information in light of published adverse events. To identify herbs most relevant to toxicology, we reviewed herbal-related calls to our regional California Poison Control System, San Francisco division (CPCS-SF) in 1998 and identified the 12 herbs (defined as botanical dietary supplements) most frequently involved in these CPCS-SF referrals. We searched Medline (1966 to 2000) to identify published reports of adverse effects potentially related to these same 12 herbs. We scored each herbal reference text on the basis of information inclusiveness for the target 12 herbs, with a maximal overall score of 3. RESULTS: The herbs, identified on the basis of CPCS-SF call frequency were: St John's wort, ma huang, echinacea, guarana, ginkgo, ginseng, valerian, tea tree oil, goldenseal, arnica, yohimbe and kava kava. The overall herbal reference scores ranged from 2.2 to 0.4 (median 1.1). The Natural Medicines Comprehensive Database received the highest overall score and was the most complete and useful reference source. All of the references, however, lacked sufficient information on management of herbal medicine overdose, and several had incorrect overdose management guidelines that could negatively impact patient care. CONCLUSION: Current herbal reference texts do not contain sufficient information for the assessment and management of adverse health effects of botanical therapies.

Drug Evaluation↗

PubMed2002

Testing high SPF sunscreens: a demonstration of the accuracy and reproducibility of the results of testing high SPF formulations by two methods and at different testing sites.

BACKGROUND/PURPOSE: The goals of this study were (i) to demonstrate that existing and widely used sun protection factor (SPF) test methodologies can produce accurate and reproducible results for high SPF formulations and (ii) to provide data on the number of test-subjects needed, the variability of the data, and the appropriate exposure increments needed for testing high SPF formulations. METHODS: Three high SPF formulations were tested, according to the Food and Drug Administration's (FDA) 1993 tentative final monograph (TFM) 'very water resistant' test method and/or the 1978 proposed monograph 'waterproof' test method, within one laboratory. A fourth high SPF formulation was tested at four independent SPF testing laboratories, using the 1978 waterproof SPF test method. All laboratories utilized xenon arc solar simulators. RESULTS: The data illustrate that the testing conducted within one laboratory, following either the 1978 proposed or the 1993 TFM SPF test method, was able to reproducibly determine the SPFs of the formulations tested, using either the statistical analysis method in the proposed monograph or the statistical method described in the TFM. When one formulation was tested at four different laboratories, the anticipated variation in the data owing to the equipment and other operational differences was minimized through the use of the statistical method described in the 1993 monograph. CONCLUSIONS: The data illustrate that either the 1978 proposed monograph SPF test method or the 1993 TFM SPF test method can provide accurate and reproducible results for high SPF formulations. Further, these results can be achieved with panels of 20-25 subjects with an acceptable level of variability. Utilization of the statistical controls from the 1993 sunscreen monograph can help to minimize lab-to-lab variability for well-formulated products.

Drug Evaluation↗