|
Experience, Expertise, Excellence |
|
Repeated Measures Analysis |
|
Many toxicology studies are designed to measure the same response in a given animal at multiple time points during the study. Neurobehavioral studies evaluate, among other responses, motor activity, startle response, and the functional observational battery multiple times during the course of a study. Cardiovascular and ECG measures are monitored continuously during safety pharmacology studies. Nearly all studies include multiple measurements of body weight. A common trait among these different studies is that animals are randomly assigned to a dose group and a response is repeatedly measured on the same animal.
Repeated Measures in Statistics Traditional statistical analyses that are conducted individually for each time point (e.g., one-factor ANOVA for each time point) evaluate the effect of different dose levels on a response at a given time point (i.e., a snapshot assessment of dose effect). The appeal of repeated measures data analysis is that it allows for the evaluation of two factors, dose level and time, on a response. Extending the ANOVA to a repeated measures analysis allows for the assessment of changes in the dose effect over time (i.e., a profile of dose-related effects over time).
From a statistical model standpoint, responses from a repeated measures study are influenced by 3 controlled and 2 uncontrolled sources of variation. Controlled sources of variation (fixed effects) include the dose level that an animal received, the time point at which the response was measured, and the interaction of dose level and time. It is this interaction of dose and time that defines the profile of dose-related effects over time. Uncontrolled sources of variation (random effects) include the variability that exists between different animals and that which exists within the same animal measured at different times. It is this latter source of variation (within animal) that adds a level of complexity, in terms of assumptions and requirements, to the repeated measures analysis when compared to that of the simple ANOVA.
The BioSTAT Solution Governing regulatory agencies recognize the advantages and complications associated with repeated measures data analysis. Enhanced methodology and computer software have improved the approach to repeated measures analysis compared with that which was available as recently as 5-6 years ago. BioSTAT has the experience of utilizing the most current methodology for repeated measures data analysis. Call for more detail. |
|
To contact us: |
|
BioSTAT Consultants, Inc |
