UNIVERSITY OF PITTSBURGH
GUIDELINES ON DATA RETENTION AND ACCESS


Introduction

The University of Pittsburgh has both rights and responsibilities toward
scientific data generated by research at the University.  Data produced
from federally sponsored research are increasingly subject to requirements
of the federal Freedom of Information Act (FOIA) and private sponsorships
often generate potential conflicts of interest and litigation over
ownership of research data.  In addition, increased mobility of University
researchers in recent years has changed the continuity of scientific
research necessitating the development of  institutional policies on data
access, retention, and transfer.


Definition of Data

Data means recorded information, regardless of form or the media on which
it may be recorded.  The term includes computer software (computer
programs, computer databases, and documentation thereof), and records of
scientific or technical nature.  The term does not include information
incidental to award administration, such as financial, administrative,
cost or pricing, or management information.  In practice, scientific data
include both intangible data (statistics, findings, conclusions, etc.) and
tangible data.  Tangible data include, but are not limited to notebooks,
printouts, computer disks, photographs, slides, negatives, films, scans,
images, autoradiograms, electrophysiological recordings, gels, blots,
spectra, samples, specimens, IRB consent forms, and other materials that
are relevant to the research project.


Data Retention

The retention of accurately recorded and retrievable research data is of
utmost importance for the progress of scientific integrity.  The
investigator must have clearly defined responsibility for recording,
retaining, and storing research data.  These records should include
sufficient detail to permit examination for the purpose of replicating the
research, responding to questions that may result from unintentional error
or misinterpretation, establishing authenticity of the records, and
confirming the validity of the conclusions.  

The experimental notebook is the most common medium for documentation of
experiments and its proper maintenance is of utmost importance.  In
addition to the study title, the investigators' names, and the study
hypothesis, the experimental notebook should include detailed information
on the materials used, sources of the materials, experimental methodology,
statistical treatments, results and conclusions so as to enable
replication of the experiments by others at any time.  Bound notebooks
with consecutively numbered pages are recommended for data recording and
maintenance.  Whenever possible, raw data should be stored together with
the experimental notebook.  In the event that this is not possible,
explicit instructions as to where the data can be found (e.g., location of
disks, samples, specimens, etc.) should be included in the notebook.  For
studies involving several investigators/collaborators, possibly in more
than one laboratory, it is recommended that the principal investigator
maintain a master log that catalogues the experiments of the whole study
and provides the location of other experimental notebooks, data, and
relevant materials stored in other laboratories or locations.

There are some governmental guidelines prescribing the length of time
researchers must maintain the original data.  The times required to retain
data vary from three to seven years depending on the governmental
organization.  In accordance with these guidelines, the University of
Pittsburgh requires that research records be archived for a minimum of
five years after final reporting or publication of a project.  The
archived records should be the originals whenever possible.  In addition,
the records should be kept for as long as may be required to protect any
patents resulting from this work or as required by an external funding
source.  If any questions regarding the research are raised during the
five-year retention period, the records should be kept until such
questions are fully resolved. 


Data Ownership and Access to Data

Both the principal investigator and the University have responsibilities,
and hence, rights concerning access to, use of, and maintenance of
original research data.  Research data belongs to the University of
Pittsburgh, which can be held accountable for the integrity of the data
even after the researchers have left the University.  Although the primary
data should remain in the laboratory where it originated (and hence at the
University),  consistent with the precepts of academic freedom and
intellectual integrity, the investigator may be allowed to retain the
research records and materials created by him/her.  In the event that the
investigator leaves the University, an Agreement on Disposition of
Research Data may be negotiated by the investigator and the Department
Chair or Dean to allow transfer of research records. However, consistent
with the same precepts, it should be specified in the agreement that the
University has the right of access to all research records and materials
for a reasonable cause after reasonable prior notice regardless of the
location of the responsible investigator.  The University is ultimately
responsible for producing original records in case of allegations of
misconduct or fraud against its researchers and protecting research
integrity.

Some circumstances may warrant an exception, requiring that the primary
data be retained by the University.  In the case of multi-institutional
studies, the institution of the primary study director is responsible for
retention and access to original data.  Extramural sponsors providing
support for research may also have the right to review the data and
records resulting from that extramural support.  When the data are used
for a patent application filed by the University, it may be necessary for
the original data to be kept at the University.  Patient medical records
also may not be removed from the University.  Primary research data and
unique materials (such as DNA, cell lines and genetic mapping information)
developed with funds from the Public Health Service (PHS, of which the
National Institutes of Health form a component) would be accessible to the
research community after publication.

Split of collaborative team:
When a collaborative team is dissolved, University of Pittsburgh policy
states that each member of the team should have continuing access to the
data and materials with which he/she had been working, unless some other
agreement was established at the outset.  The unique materials prepared in
the course of the research should be available/accessible under negotiated
terms of a transfer agreement.

Co-investigators, trainees, and students:
The original data must be retained by the senior investigator and
co-investigators.  Trainees and students who are an integral part of the
research project should be allowed continued access to all records and
data which pertain to their part of that project.

Note:  These guidelines have been developed based on similar policies now
being implemented at other universities.


Reviewed and approved by:



Mark A. Nordenberg
Chancellor of the University

February 3, 1997