As well as working with clients within the biopharma and medical device space, Cytel supports customers in consumer industries such as cosmetics. A key challenge in this sector is to provide a constant stream of products that comply with changing market demands. In this case study we describe how we helped one client with expert statistical support.
Our client is one of the world's largest consumer goods companies. The company is committed to supporting sustainability and providing consumers around the world with products that help them look good, feel good, and maintain health. Recent environmental changes have prompted Fast Moving Consumer Goods (FMCG) companies to explore the air pollution impact on human skin and skincare and haircare products are now being developed to prevent its adverse effects. Our client planned a study to explore the historical biological, biophysical and skin appearance differences in human skin by comparing two groups in high and low pollution exposure levels. The client sought Cytel’s guidance and recommendations to help clarify their specific data collection, endpoints, objectives, and analysis requirements. They additionally sought Cytel’s support for study conduct.
The data to be collected in the high and low pollution group was not balanced in terms of the number of subjects and the covariates. Cytel needed to help ascertain which subjects should be included for analysis and to make two groups comparable.
Cytel needed to help finalize the most appropriate response variable for the statistical analysis.
This large study ultimately required the exploration and analysis of more than 40 endpoints.
Cytel carried out stratified randomization and propensity score matching to obtain two balanced pollution groups.
The sponsor was undertaking a complex study design with two phases and their respective baseline measurements. Usually, the change from baseline (CFB) measurements is considered for comparison purpose between any two groups. In this particular trial, other than CFB, many other response variables were derived depending on the extraction of the endpoints from different skin layers.
Cytel provided input to derive the new variables in line with the sponsor’s exploratory objectives.
Cytel’s statisticians presented an innovative idea to address one of the key questions raised and it was readily accepted by their statisticians.
Question: If the data is not meeting the normality assumption then ANCOVA (Analysis of Covariance: a general linear model which blends ANOVA and regression) is not appropriate. In this case, what can be done further for covariate adjustment?
Answer: For covariate adjustment exploration for non-parametric tests, one possibility is to fit a linear model and obtain residuals. Non-parametric methods can then be applied to the residuals since they will be free of covariate effects.
Cytel met all required timelines for the project and the client was delighted with the quality of work. The results obtained in this study for both biophysical and biochemical endpoints were promising and the results are set to demonstrate that air pollution is associated with skin aging.
Cytel continues to work with the client on new exploratory studies.
Cytel is extremely experienced in statistical methodologies and was able to quickly learn the fundamentals of the sponsor’s field. Having a consistent team throughout the project ensured knowledge retention and efficiency.
Cytel’s statisticians were able to quickly grasp and apply complex technical concepts, alongside carefully examining, absorbing, and questioning what was presented to them.
As a result of good communication, accurate submissions and meeting tight delivery timelines, Cytel has developed a strong relationship with the sponsor.
To learn more about how we help our clients tackle complex statistical problems, click the button below.
With thanks to Meenakshi Mahanta of Cytel.