Development of Risk Assessment Criteria for Contamination with Hazardous Substances in Facial Cosmetics

Abstract

Objective: To develop the criteria for assessing the risk for contamination of hazardous substances (CHS) in facial cosmetics for consumers to use as a tool for self-protection by choosing safe cosmetics, and for authorities to use as a tool in surveillance of CHS. Methods: This research was a cross sectional analytical study in cosmetic samples for facial skin sold in Pakphanang municipality area, Nakhon Si Thammarat. Logistic regression analysis was conducted to identify factors relating to CHS in order to generate regression equation to predict the probability of CHS.  The analysis derived 2 criteria for risk assessment (CRA) i.e., CRA for authorities and that for consumers by eliminating some factors difficult for consumers to assess. The study tested the predictive ability of CRAs using analysis of receiver operating characteristic (ROC) curve. Results: 72.34% of cosmetics selling places sold cosmetics with legally violating labels or with no labels, expired or deteriorated cosmetics, or cosmetics with CHS according to the announcement of the Ministry of Public Health. Among 210 cosmetic samples, CHS was found in 127 samples (60.48%). CHS was identified in 31.01% of cosmetics with legally complied labels and in 71.71% of those with legally violating labels or no labels. There were 7 factors significantly related to CHS including misleading labels in the essence of cosmetics, distribution locations (fresh markets, flea markets or stalls), presence of indications for acne, absence of the registration number, absence of lot number, absence of the address of manufacturers/importers, and presence of a fake registration number or the terminated one. CRA for CHS for authorities showed area under the ROC curve (AUC) at 84.80% (95% CI 79.70-90.00). The cosmetics with low risk of CHS (scores 0.0-3.0), medium risk (scores 3.5-9.0) and high risk (9.5-11.5) showed LR at 0.08 (95% CI 0.03-0.20), 1.06 (95% CI 0.80-1.41) และ 9.31 (95% CI 3.51-24.70), respectively with the prevalence of CHS of 11.36, 61.90 and 93.44%, respectively. Three CRAs for consumers showed AUCs at 81.20-83.60% (95% CI 75.30-89.10), sensitivity of 82.70-85.80% and specificity of 65.10-67.50%. Conclusion: CRAs for CHS in facial cosmetics are reliable for predicting the likelihood of CHS with acceptable sensitivity and specificity, despite the removal of some factors difficult for consumers to evaluate. Consumers can therefore use the CRA as a tool in choosing safe cosmetics. Authorities could also use the CRA for surveillance of CHS in cosmetics.