NOTE: Not every database record is associated with a corresponding PDF file. Per Data Providers, there are 3397 records in the Oak Ridge Dose Reconstruction Database and only 1803 PDF files.
NOTE ABOUT DOWNLOADING LARGE PDF FILES: Some of these database
records include hyperlinks to PDF files which may be very large. Some computers
or networks may not be configured to handle large files well.
FOR BROWSING THE ENTIRE COLLECTION:
Use this link for
a complete listing of ORDR documents, by title. The download of the complete
listing of ORDR documents is a large file. Please give it a few moments to download
to your browser.
FOR SEARCHING BY KEYWORD:
Use this link for searching the Oak Ridge Dose Reconstruction Database.
CONTROLLED CODES USED IN RECORDS IN THIS DATABASE:
This link is a download (image below) of a text searchable
PDF document which provides abbreviations and explanations of information contained
within the ORDR database.
A critical element of the Oak Ridge Dose Reconstruction was the systematic searching of repositories containing records of historical operations at the ORR. The purpose of this methodical review of all records was to ensure that there were no significant, previously unidentified releases that should be included in a comprehensive dose reconstruction. This exhaustive approach to document review was a critical step towards achieving credibility with members of the public by increasing their confidence that all activities that might have affected their health have been revealed for the scientists to include in their study. The level of effort in the document search component of the dose reconstruction was unprecedented, representing the most extensive records review ever undertaken at the ORR.
The first step undertaken by the document team was to develop a comprehensive list of all repositories maintained at the three major complexes comprising the ORR: X-10, K-25 and Y-12, and any off-site repositories that housed records dealing with ORR. Table 6-1 gives a list of the 44 repositories identified. The project team then ranked the repositories by the apparent likelihood that they contained previously unidentified relevant information and placed them into three categories: High Priority Classified, High Priority Unclassified, and Lower Priority. Two of the key factors affecting rank were types of records (classified vs. unclassified) and age of records. Repositories with older (pre-1970) classified records were ranked highest.
The basic strategy for review of repository holdings involved the following steps:
Repository-specific search plans were developed based on each repository's characteristics. Two key factors affecting search strategies were the type and quality of cataloging systems available at a repository (electronic or non-electronic) and the size of its holdings.
Where possible, the project team obtained copies of electronic cataloging systems from repositories to aid in systematic searching. Any listings available for the contents of the repositories were used to make the initial selection of the documents for review. For example, if a list of titles was available for a facility, an analyst initially reviewed that list to determine which documents should be examined. If the repository was not well documented or contained voluminous records, an appropriate method for sampling boxes was developed. Typically, these repositories had standard, 1 cubic foot boxes containing records. When this occurred, samples of boxes were selected and closely inspected to improve understanding of the repository holdings.
For the purposes of Task 5 activities, three domains or classes of records at ORR were identified: a priori relevant, non-relevant, and relevance undetermined. Using a given repository's finding aids (keyword search of databases, card files, bibliographies, et.), a subset of potentially relevant records was identified. These were called a priori relevant. This project team made every effort to physically review all of the potentially relevant records.
The finding aids also allowed identification of classes or records that were clearly not relevant for a dose reconstruction and that could be excluded from the review. However, before these records were excluded, they were sampled and inspected to confirm that there was indeed no likelihood of relevant information in that class of records. Fr example, at the DOE Records Holding Center, the unclassified records database showed that there were classes or records that could be grouped, such as records from the accounting branch and the acquisition branch. There were 139 boxes that contained accounting/acquisition records. Twenty-two of the 139 boxes were chosen for review - approximately 16%. In most cases, at least 10% of the boxes were randomly chosen to represent a group of boxes. All of the 22 boxes turned out to be not relevant to the Oak Ridge Health Study; therefore, all 139 boxes were considered not relevant and were not physically reviewed.