After the permission survey was distributed, and the responses were received, we decided to reduce the time and costs of the curating procedure, and only curated those manuscripts that we had permission to use. The curating process included cleaning the individual manuscript and review files of extraneous information, ensuring that revisions and reviewers matched up, and redacting the names of all reviewers plus all authors who did not want their names included. We developed a protocol for protecting the anonymity of authors who did not wish their names to be used and all reviewers. Thus, more than 50% of curated documents contained names and affiliated institutions of reviewers whose anonymity we agreed to protect.
We have been curating scanned manuscript PDF files according to the following protocol.
We processed manuscripts only if author(s) provided us permission to participate the project. Some agreed to be used with their names, but others agreed with for inclusion if their names were not used. The Log File contains the curator’s notes, the original file name, and the page numbers for each discrete component of the original file. The status of files was logged in this file (i.e., processed, unprocessed, unmatched, cannot process) so that we would know which manuscripts were curated and which were not.
Each un-curated file is labeled with a correct manuscript number (MSN) to make sure that the PDF is recorded in the log file and, if the file is a revised manuscript, a corresponding revision number (RN) is provided. For example, an unrevised manuscript would be titled "03-095"; whereas a revised manuscript's title would be "03-095-r1" or "03-095-r2" (and so on, depending on the number of revisions).
In the PDF file, we checked a page from the manuscript review process with the title "Manuscript Review Status" to confirm that the MSN and RN used in the PDF name, the log file, and this status page match. Also, we checked the title of the manuscript and the number and name of reviewers. We ensured that the title page of the manuscript file had the same title as the one listed on the "Manuscript Review Status" page.
Next, we scrolled through each manuscript to ensure that page numbers are sequential. We removed extraneous pages from the manuscript (usually belonging to a different manuscript). We examined the last page of the manuscript, to ensure that references, appendices, tables, and figures were included with the proper manuscript.
We followed the same process for the Review Cover Sheets and each anonymous review. Typically, Review Cover Sheets will be labeled as "Confidential Report to the Editor" and followed by an anonymous review (approx. 1-3 pages of text written by the reviewer). Neither Review Cover Sheets nor Anonymous Review should contain names or any identifiable information. Use “Redact” function of Adobe to erase it. When a review cover sheet and anonymous review appear on the same page, redactions were performed on both the anonymous comments for the review cover sheet file and the editorial comments for the anonymous file.
Response to Reviewers appears only in revisions, not in the original manuscript. These are often a letter from authors to reviewers, explaining what changes they have made in the manuscript. We use them to substitute reviews from participants granting no permission to be used in this project.
Each original file contained many pages that were not Manuscripts, Review Cover Sheets, Anonymous Reviews, or Response to Reviewers. These materials were considered extraneous and were divided out into their own separate file (X-files). Anything that did not mesh with the curation protocol or seemed out of place for any reason was recorded in the “Curator Notes” column in the Log File corresponding MSN-RN.