The following ISPE members helped draft these guidelines: Elizabeth
B. Andrews, Ph.D., M.P.H.; Jerry Avorn, M.D.; Edward A. Bortnichak, M.P.H.,
Ph.D.; Robert Chen, M.D., M.A.; Wanju S. Dai, M.D., Dr.P.H.; Gretchen S.
Dieck, M.P.H.,Ph.D.; Stanley Edlavitch, Ph.D., M.A.; Joel Freiman, M.D.,
M.P.H.; Allen A. Mitchell, M.D.; Robert C. Nelson, Ph.D.; C. Ineke Neutel,
Ph.D., FACE; Andrew Stergachis, Ph.D.; Brian L. Strom, M.D., M.P.H.; Alexander
M. Walker, M.D., Dr.P.H.
©International Society for Pharmacoepidemiology, 1996
These Guidelines for Good Epidemiology Practices (GEPs) have been adapted from a document prepared by the Chemical Manufacturer's Association Epidemiology Task Group (1). Wherever appropriate, we have (with permission) retained the exact text of that document. The Guidelines for Good Epidemiology Practices propose practices and procedures in the following areas:
The Guidelines have the following goals:
Clearly, large complex studies will benefit from the careful planning and thorough documentation implicit in these guidelines. Adherence to the spirit of the guidelines will be beneficial for those activities preceding protocol development as well as more informal investigations such as health hazard assessments/evaluations or small cluster investigations. Particularly in circumstances of immediate public health concern, the guidelines will provide a useful framework to ensure that all research issues are adequately addressed.
These guidelines should evolve based on the experiences gained through
their application to studies.
The protocol should include the following:
A. A descriptive title, including original date of approval.
B. The names, titles, degrees, addresses, and affiliations of all responsible parties, including at least the principal investigator and all co-investigators, and a list of all collaborating primary and other relevant institutions and study sites.
C. The name(s) and address(es) of the sponsor(s).
D. An abstract of the protocol.
E. The proposed study tasks, milestones, and timeline.
F. A statement of research objectives, specific aims, and rationale.
Research objectives describe the kinds of knowledge or information to be gained from the study. Specific aims list the measurements to be made, and any hypotheses to be tested. The protocol must distinguish between a priori research hypotheses and ones that have been based on knowledge of the data. The rationale explains how achievement of the specific aims will further the research objectives.
G. A critical review of the literature to evaluate relevant variables and gaps in knowledge.
For example, the literature review should be of sufficient depth to identify potential confounders and effect modifiers and to determine areas where new knowledge is needed. The literature review might encompass relevant animal and human experiments, clinical studies, vital statistics, and previous epidemiologic studies.
H. A description of the research methods, including:
Research designs include, for example, case-control, cohort, cross-sectional, nested case-control, or other hybrid designs. The rationale for the selection of the proposed design over others should be presented.
The population and sample to be studied.
The population is defined in terms of persons, place, time period, and exclusions. The protocol should identify any changes in population or sample that were implemented after the beginning of the study.
The data sources for exposure, health status, and risk factors.
For example, questionnaires, hospital discharge files, abstracts of primary clinical records, automated records such as prescription drug files, biological measurements, exposure/work history record reviews, or exposure/disease registries.
Clear operational definitions of health outcomes, exposure, and other measured risk factors as well as selection criteria, as appropriate, for exposed and nonexposed persons, morbidity or mortality cases, and referent groups.
An operational definition is one that can be implemented using the data available in the proposed study. For example "PCP episode" is not an operational definition, whereas "hospitalization with a primary diagnosis of ICD-9-CM code 136.3" is.
Projected study size, statistical precision, and the basis for their determination.
Present the relation between the study size and the specific aims of the study.
The methods to be used in assembling the study data.
This should include a description of, or reference to, methods used to control, measure, or reduce various forms of error -- e.g., bias due to selection, misclassification, interviewer, or confounding -- and their impact on the study. Pretesting procedures for research instruments and any manuals and formal training to be provided to interviewers, abstractors, coders, or data entry personnel should be described or referenced.
Procedures for data management.
This should include data management programs and hardware to be used in the study.
Methods for data analysis.
Data analysis includes all the major steps that lead from raw data to a final result. It comprises comparisons and methods for analyzing and presenting results, categorizations, as well as procedures to control, if possible, sources of bias and their influence on results. The statistical tests to be applied to the data and procedures for obtaining point estimates and confidence intervals of measures of occurrence or association should also be presented.
A description of quality assurance and quality control procedures for all phases of the study.
Mechanisms to ensure data quality and integrity should be described, including, for example, reabstraction of original documents. As appropriate, include certification and/or qualifications of any supporting laboratory or research groups.
Major limitations of the study design, data sources, and analytic methods.
At a minimum, issues relating to confounding, bias, generalizability, and random error should be considered. The likely success of efforts to reduce errors, presented in section H6, should be discussed.
This should include information about whether study subjects will be placed at risk as a result of the study, under what circumstances informed consent will be required, provisions for maintaining confidentiality of information on study subjects, and circumstances, if any, under which personally identifiable information may be provided to entities outside the study.
J. A description of plans for disseminating and communicating study results, including the presence or absence of any restrictions on the extent and timing of publication.
There is an ethical obligation to disseminate findings of public health importance.
K. Resources required to conduct the study.
Describe, for example, time, personnel, and equipment required to conduct the study, including a brief description of the role of each of the personnel assigned to the research project.
L. The bibliographic references.
M. Dated amendments to the protocol.
Significant deviations from the protocol should be documented in writing.
For example, this should include delineating the roles and responsibilities to be assumed by the study sponsor and the contractor(s) in communicating various aspects of the study as well as data ownership, archiving, etc.
The unusual decision to terminate a study prematurely should be taken with great caution, and should be based on good scientific and ethical reasons and documented in writing. There may be rare instances in which administrative reasons require study termination. Such decisions must be made independent of any study results. Investigators and sponsors should specify and agree in advance about the circumstances under which the study could be terminated early. Included should be a mechanism for resolution of any disagreement.
Investigators will ensure that personal identifiers will be removed from any study files that are accessible to non-study personnel. All personnel with access to data containing personal identifiers will sign a pledge to maintain the confidentiality of study subjects. They will maintain an ability to verify the origin and integrity of data sets from which personal identifiers will have been removed.
All procedures used to verify and promote the quality and integrity of the data shall be outlined in writing. A historical file of these procedures shall be maintained, including all revisions and the dates of such revisions. Any changes in data entries shall be documented.
The final report shall include at a minimum:
An abstract.
Purpose (objectives) of the research as stated in the protocol.
The names, titles, degrees, addresses and affiliations of the principal investigator and all co-investigators.
Name(s) and address(es) of sponsor(s).
Dates on which the study was initiated and completed.
Introduction with background, purpose, and specific aims of the study.
A description of the research methods, including:
a. the selection of study subjects and controls,A description of circumstances that may have affected the quality or integrity of the data.b. the data collection methods used,
c. the transformations, calculations, or operations on the data, and
d. statistical methods used in data analyses.
Describe also the initially identified limitations of study approach and the methods used to address them (e.g., response rates, or missing data)(See Protocol, Sections H6 and H9)
A summary and analyses of the data.
Include sufficient tables, graphs, and illustrations to present the pertinent data and to reflect the analyses performed.
A statement of the conclusions drawn from the analyses of the data.
A discussion of the implication of study results.
Cite prior research in support of and in contrast to present findings. Discuss possible biases and limitations in present research.
References.
Government agencies and all sponsors shall be informed of study results in a manner that complies with applicable regulatory requirements.
There is an ethical obligation to disseminate findings of public health importance. Scientific peers shall be informed of study results by publication in the scientific literature or presentations at scientific conferences, workshops, or symposia.
Potential conflicts of interest should be disclosed.
Access to the archives shall be controlled and limited to authorized personnel only. Special procedures may be necessary to ensure that access to confidential information is limited and that the confidentiality of information about study subjects is protected (see Protocol, Section I).
The archive should be maintained for at least five years after final report or first publication of study results, whichever comes later. At a minimum, the study archive should contain, or refer to, the following:
A. Study protocol and copies of all approved modifications.
B. A final report of the study.
C. All source data and, where feasible, specimens. A printed sample of the master computer data file(s) with reference to the location of the machine readable master. All "source data" should comprise the raw data that provided the basis for the final analysis of the study. The archival material should be sufficiently detailed to permit re-editing and re-analysis.
D. Documentation adequate to identify and locate all computer programs and statistical procedures used, including version numbers where appropriate (see section IV(C): Study Conduct).
E. Copies of computer printouts, including relevant execution code, that form the basis of any tables, graphs, discussions, or interpretations in the final report. Any manually developed calculations shall be documented on a work sheet and similarly retained.
F. Correspondence pertaining to the study, standard operating procedures, informed consent releases, copies of all relevant representative material, copies of signed institutional review board and other external reviewer reports, and copies of all quality assurance reports and audits.
Include, for example, questionnaires, name, make and model numbers of relevant measurement instruments, calibration information and procedures.
G. Documentation relating to the collection and processing of study
data, including laboratory/research notebooks, training and reference documents
for abstracts, interviews, and coders.
Would you like to return to the DDIL Homepage?
This document has been provided through the Drugs and Devices Information
Line, http://www.hsph.harvard.edu/Organizations/DDIL/ddil.html