23Jul2012

From PHUSE Wiki
Jump to: navigation, search


When: 3:00 PM , 23 July 2012
Place: Teleconference
Facilitator: Jen Feldmann
Scribe: Lauren Mihalcik
Agenda:

  1. Comments/modifications to the HC definition
  2. Update on relevant publications

Introduction

Who's here?

Lauren Mihalcik, Jen Feldmann, Judy Henck, Joyce Zandee, Nancy Everds

General discussion

Definition Several changes were made to the draft definition for historical control data. Notably, "animals" was changed to "subjects" (since the unit of investigation can be ,e.g., litters), the definition was reworded to remove "metadata" (which was considered too jargony), and a sentence about how the data is usually used was added to make the point that HC data, as typically consumed, is a summary.

Final definition: Test results and observations from individual subjects not treated with test article with known health status (such as control subjects) with descriptive data that adequately characterizes the subject and the conditions under which the data was collected. Data are usually evaluated, curated, and summarized so that the data can be more easily compared with results from other populations.

Publications Lauren Mihalcik: Keenan paper on HCD in rodent carcinogenicity studies did a survey that is similar in concept to what we are envisioning. They asked how respondants (~100) used HC data and came up with a list of best practices. The most frequent free text comment they recieved was that respondants wished there was an accessible HC database with controlled terminology and standardized data. This led to a discussion of why such a database does not exist (not profitable, no champion in non-profit world). There would be significant challenging in setting up such a collection, given all the parameters that can affect various outcomes.

Joyce Zandee: The Hall paper discusses the potential pitfalls of "lumping" and "splitting" HC data in the context of clinical pathology data (and why you should be wary of reference intervals). The second paper is an example of histopathology data over time at a single laboratory.

Others: We ran out of time

Follow up

For next call: In preparation for surveying the wider nonclinical community, consider how you (and your organization) use HC data and how you wish it could be used. If possible, record your list on the wiki here: Working group HC use and ideas


Last revision by Lauren.mihalcik,07/23/2012