Minutes from the Informatics Working Group of the Texas Cancer Genetics Network

March 3/17/1999

Minutes from the Informatics Working Group of the Texas Cancer Genetics Network.

The informatics working group of the TCGN met on Thursday, March 18, 1999 at 11:00 AM via conference call. Present were Chris Amos, Jihong Zong, Jerry Kyle (U.T.MD Anderson Cancer Center), Sue Hilsenbeck, other names ? (U.T. Health Science Center San Antonio), Kyle Larson, Susan Fernbach (Baylor College of Medicine). Unable to attend were Sharon Plon and Gail Tomlinson.

Dr. Amos provided introductory discussion of the core data elements working group of the CGN and the core data elements questionnaire. Dr. Amos indicated that an initial pilot is nearly complete and a conference call is scheduled to discuss the results on March 24. Initial findings from the pilot indicate that the questionnaire is not well formatted for self-administration. In addition, some questions may need further clarification but it is anticipated that neither the questionnaire nor the data dictionary will change much as a result of the current pilot. A second pilot is planned in which data will be processed into databases and then transmitted to the central CGN site at UC Irvine. The target dates for this pilot have not been specified.

The first item discussed was availability of resources and plans for developing software to support collection of data for the core data elements questionnaire. The UTMDACC group is in the process of hiring an additional support person to assist with development of software tools. BCM has recently hired a new data manager, but database programming and management will be divided duties at their site. At UTHSCSA, existing personnel will likely be used to further develop code. At the UTMDACC, further development of an existing data resources that have been developed in ACCESS is planned. BCM stated that they expect that they will use SQL server technology to support their needs. Dr. Amos stated that the Duke group already had an SQL application to support CGN-type databases. He urged BCM to contact Duke for more information. UTHSCSA will develop software from their existing Informix based applications. BCM and UTHSCSA plan to use web-based front ends to assist in data capture. UTMDACC does not plan to create web interfaces for this application as they are not required. Further negotiations between UTHSCSA and BCM are required because Sue Hilsenbeck, PI for UTHSCSA is moving to Baylor.

Unknown at the time of the call were the specific deadlines under which UC Irvine would like computerization of the pilot. In addition, Sue Hilsenbeck pointed out the problem of flaws in the UCI database such that it is not yet programmed to handle the structure of the family history information that is currently being collected.

The development of informed consent documents was briefly discussed, but is not a major focus of work for this group.

The development of identification numbers for TCGN sites was discussed. The structure for the CGN identification numbers has not clearly been described. Once this information is provided, the TCGN will establish naming conventions for identifiers. The basic plan is that there will be a CGN number, a subcenter number, and then individuals specific numbers for each site. It was anticipated that numbers within the TCGN for each site would not deviate from those used by the CGN. (Note written in transcription) This may raise some issues of confidentiality, however, unless the UTMDACC recodes data prior to transmission to UCI, which can be easily accomplished. Dr. Amos mentioned that the TCGN investigators plan to routinely collect tissues in the future even though this is not a part of the CGN. Therefore, the database structures that are developed at each site need to include a capacity to store tissue id numbers.

Finally the need to develop a website was discussed. The website will be developed by UTMDACC (Jerry Kyle) and BCM (Kyle ). To be included in this website will be investigators names and addresses, lists of working groups, minutes from working group and steering committee meetings, commonly used questionnaires such as the core data elements questionnaire. In addition, information for the general public about the TCGN will be kept on this site. Dr. Amos indicated that for the moment, the website at www.epigenetic.com can be used for this purpose. Jerry Kyle indicated that he is the local web provider for the CGN. He has assigned passwords for TCGN members to gain access to the Massachusetts General Website. He will also assign passwords for the UCI site, once UCI contacts him to begin this process. He did not have the e-mail addresses for all participants, but will provide the necessary information to TCGN investigators as soon as he is able.

The TCGN Informatics working group will meet episodically depending upon the need for further contact among the sites. Once a more clear description of the work required by UCI is available, meetings may be held more regularly. In the meantime, approximately monthly meetings are adequate. The next meeting was scheduled for 11:00 A.M., April 8.