1.
No one will be permitted to publish or publicly present multi-site outcome
data or site outcome data or to discuss such data with anyone outside of
their site or outside the CC until the multi-site data collection is
complete at all sites. Comparisons
across sites (that is, comparing data from one site to another), comparisons
of the experimental to the control conditions, or examination of any multi-site
outcome data by anyone outside the CC will not be permitted until the study
is concluded.
2. Sites will not have direct access to the complete multi-site database
until one year following completion of all multi-site data collection and
cleaning, and after the core papers have been written. Multi-site baseline
data will be made available to sites by the CC when data collection of that
data is closed (that is, when no new entrants to the study will occur) and
the data have been cleaned. Analysis of baseline data can proceed under the
guidelines put forth in item number 3.
3. Presentations of preliminary findings of multi-site baseline aggregate
data (that is, with E and C combined) is permitted with the approval of the
Publications Workgroup. In
addition, examinations of these data over time, in the aggregate, may be
permitted as long as, in the opinion of the Coordinating Center, the basic
integrity of the E vs. C comparisons is not compromised and no comparisons
are revealed by site. Decisions about presentation of preliminary findings
are made by the Publications Workgroup.
4. Any analysis of multi-site data by an individual site must be turned over
to the Coordinating Center for verification and storage to maintain the
consistency of the data (for example, so that similar re-coding, collapsing,
and subscales are used). Once the funding of the Coordinating Center ceases,
this submission will no longer be required.
5. Research staff should be cautioned that preliminary examinations of the
data could be misleading and could result in demoralization among staff at a
site or changes in the way that services are delivered. Although the Study Blind policies cannot prohibit the PI or
research staff at a site from examining and interpreting their own data,
such interpretations should be made cautiously and circumspectly.
Policy amended by Steering Committee on October 16, 2001:
Future baseline data sets released by the CC to the Steering Committee
will include site identification. Site as a variable may be exclusively
included in any analysis papers and presentations that do not use follow-up
data.
5400
Arsenal Street