ACI began this project by compiling a comprehensive Ingredient Inventory used among our members in the manufacture of consumer cleaning products sold in the United States. Ingredient lists were compiled from over 900 products and different naming conventions standardized to produce the inventory of 588 ingredients that is searchable below. As part of this effort, information on function, chemical category, and product use were collected so that these data could be used to inform hazard and exposure assessment.
Once the Ingredient Inventory was completed, ACI began a compilation of human health and environmental hazard data for each ingredient from publicly available sources. The goal of this effort was to develop a complete hazard data set that could be used for screening level risk assessment. Over 7,800 individual records were collected from a wide range of sources. Collection sources ranged from general findings regarding safety to highly specific results of lab studies, to values adopted for use as benchmarks by governmental organizations based on research from multiple studies. Data collection focused on chronic and repeated dose toxicity testing, but included collection of any data that were available. In some cases, this included results from studies examining carcinogenicity or sensitization, although data for these endpoints was not always available. The hazard data were examined to establish a dose-response threshold against which to compare exposure;
however, the carcinogenicity and sensitization endpoints were not used to establish the threshold. Hazard data were filtered based on data quality and prioritized based on transparency and defensibility of the source of information. Where data were subject to uncertainty, assessment factors were applied to make values more conservative. In some cases, read across was applied by using data for structurally similar ingredients to fill data gaps. The end result was the identification of either quantitative threshold values or a defensible safety finding for 542 of the 588 ingredients. Of the remaining ingredients, approximately half (46) are proprietary and hazard information is maintained by manufacturer.
While hazard data were being collected, exposure models were run to quantify human health exposure to each ingredient from products in which it occurs. Assumptions regarding product types, cleaning habits and practices, and concentration of ingredients in different product forms was developed from scientific and trade literature as well as member company listings. This information was used to develop quantitative exposure models for consumer exposure to both individual product types and multiple product types over the course of the day. The end result was a database of over 2903 unique combinations of ingredient, product type, and exposure route.
Exposure estimates, hazard data, and past findings were combined to produce a human health screening level risk assessments for the 494 ingredients for which publicly available hazard data were available. A screening-level risk assessment is a highly conservative initial assessment of the chance that an ingredient in a cleaning product may pose a risk of effect during use. Each screening level risk assessment for a specific ingredient was based on the most defensible and authoritative data compiled, and was designed to err on the side of caution, making conservative assumptions wherever uncertainty or a range of circumstances might be expected. Because of this, findings of "no" or "low risk" are relatively certain, while findings that there is a potential for risk are uncertain, and will undergo further assessment. This is planned as a refinement step to be completed in early 2017 in which conservative assumptions will be re-evaluated and updated with more realistic ones where appropriate.
Two approaches were used for screening level risk assessment. The first was to rely on risk assessments already conducted using defensible and comparable methodologies. The second was to perform quantitative modeling of exposure and risk. A screening-level risk assessment presents a general indication of the potential for risk (or lack of risk) from a particular substance. The SLRAs performed for the CPISI ingredients can provide an estimate of the chance that an ingredient in a cleaning product may pose a risk during use. A total of 101 ingredients were identified as unlikely to produce risks based on the findings of past risk assessments and studies. Another 441 ingredients were assessed via a quantitative screening level risk assessment in which exposure estimates and hazard data were combined to calculate a risk characterization ratio (RCR). An RCR below 1 indicates minimal potential for risk while an RCR above 1 indicates a possible risk. A total of 436 ingredients demonstrated were found unlikely to pose risks because their RCR values were below 1 even under conservative maximum case exposure scenarios. A small number of ingredients (5) had an RCR above 1 under maximum case scenarios; these were found to warrant additional study via refining risk assessment methods. The models used to develop the RCR were examined for any areas of potential uncertainties associated with the collected hazard or exposure data. These data may have included ingredient concentration ranges, application of uncertainty factors, or the ingredient-product exposure combinations that were evaluated. Additionally, more input was solicited as needed from ACI members regarding specific ingredients in specific products to secure more updated hazard or exposure data. For most of these 5 ingredients with an RCR greater than 1 under maxiumum case scenarios. The RCR for these ingredients did not exceed 1 for the minimum case scenario or over 95% of the expected range of exposure. These risks were associated with conservative ingredient concentrations or conservative assessment factors that warrant additional research and review. When more realistic assumptions were used, the RCR was less than 1 for all 5 ingredients.
In summary, the CPISI project produced screening level risk assessment results in the form of risk models or review of past findings for over 90% of the 588 ingredients studied, and found that 99% of those ingredients were unlikely to produce risks even when highly conservative exposure models and hazard values were used.
More detail about the development of each component of this CPISI project can be found on the Frequently Asked Questions page. You may also contact ACI through the contact us form.