Jump to main content.

Proposed Procedures for Combined Analyses of Epidemiologic Studies as Part of the Great Lakes Human Health Effects Research Program

Note: EPA no longer updates this information, but it may be useful as a reference or resource.

 Proposed Procedures for Combined Analyses of Epidemiologic 



[Federal Register: June 27, 1995 (Volume 60, Number 123)]
[Notices]               

[Page 33217-33219]

>From the Federal Register Online via GPO Access [wais.access.gpo.gov]







DEPARTMENT OF HEALTH AND HUMAN SERVICES
[ATSDR-95]

 


Proposed Procedures for Combined Analyses of Epidemiologic 
Studies as Part of the Great Lakes Human Health Effects Research 
Program



AGENCY: Agency for Toxic Substances and Disease Registry (ATSDR), 
Public Health Service (PHS), Department of Health and Human Services 
(HHS).



ACTION: This notice announces the procedures ATSDR will use for 
conducting combined research analyses as part of the ATSDR Great Lakes 
Human Health Effects Research Program.






SUMMARY: This notice describes the proposed procedures, meta- analyses 
and pooled data analyses, to be used by ATSDR to conduct combined 
analyses of epidemiologic studies supported by the ATSDR Great Lakes 
Human Health Effects Research Program. ATSDR may choose to utilize one 
or both procedures, depending on the data and the results of the future 
feasibility studies. The procedures will be used for both new and 
existing research investigations. Comments on this notice are 
requested. The procedures outlined herein will be used on an interim 
basis, subject to change based on comments received and experience 
gained during implementation of these procedures.



DATES: Public comments concerning this Federal Register notice must be 
received on or before December 26, 1995.



ADDRESSES: Comments on this notice should bear the docket control 
number ATSDR-95 and should be submitted to: Division of Toxicology, 
Research Implementation Branch, Agency for Toxic Substances and Disease 
Registry, Mailstop E-29, 1600 Clifton Road, NE, Atlanta, Georgia 30333.
    Comments on this notice will be available for public inspection at 
the Agency for Toxic Substances and Disease Registry, Building 4, Suite 
2400, Executive Park Drive, Atlanta, Georgia (not a mailing address), 
from 8 a.m. until 4:30 p.m., Monday through Friday, except for Federal 
legal holidays.



FOR FURTHER INFORMATION CONTACT: Dr. William Cibulas, Research 
Implementation Branch, Division of Toxicology, Agency for Toxic 
Substances and Disease Registry, Mailstop E-29, 1600 Clifton Road NE., 
Atlanta, Georgia 30333, telephone (404) 639-6306.



SUPPLEMENTARY INFORMATION:



Background



    The ATSDR Great Lakes Human Health Effects Research Program is 
[[Page 33218]] authorized in sections 104(I)(5)(A) and (15) of the 
Comprehensive Environmental Response, Compensation, and Liability Act 
of 1980 (CERCLA) as amended by the Superfund Amendments and 
Reauthorization Act of 1986 (SARA) [42 U.S.C. 9604(I)(5)(A) and (15)]; 
and section 106, subsection 118(e) of the Great Lakes Critical Programs 
Act of 1990 [33 U.S.C. 1268(e)]. This research program is designed to 
investigate and characterize the association between the consumption of 
contaminated Great Lakes fish and associated short- and long-term 
harmful health effects. The research objectives of the program are to 
(1) build upon and amplify the results from past and ongoing research 
in the Great Lakes basin; (2) develop information databases and 
research methodology that will provide long-term benefit to human 
health effects research in the Great Lakes basin; (3) provide direction 
for future health effects research; (4) provide health information to 
State and local health officials, and to the concerned public and their 
medical health care professionals; and (5) in concert with State and 
local officials, increase the public awareness regarding the potential 
health implications of toxic pollution in the Great Lakes basin; and 
(6) coordinate as necessary with relevant Public Health Service (PHS) 
research programs and activities, including those of the Food and Drug 
Administration (FDA), Centers for Disease Control and Prevention (CDC), 
National Institutes of Health (NIH), and the Indian Health Service 
(IHS), as well as the Environmental Protection Agency (EPA), and State 
and local health departments, to ameliorate adverse public health 
impacts of persistent toxic substances in the Great Lakes basin.
    Toward this end, ATSDR has developed a Great Lakes Health Effects 
Research Strategy. The goals of this strategy are to identify human 
populations residing in the Great Lakes basin that may be at greater 
risk of exposure to chemical contaminants present in one or more of the 
Great Lakes, and to ameliorate or prevent any adverse health effects. 
This strategy is built upon the five traditional elements of disease 
prevention: identification, evaluation, control, dissemination, and 
infrastructure. This strategy has been endorsed by the Council of Great 
Lakes Research Managers and has been adopted by the International Joint 
Commission as a framework for the study of human health and other 
ecosystem effects in the Great Lakes basin.
    In fiscal year 1992, ATSDR funded nine research grants to study the 
potential adverse human health effects of consuming contaminated fish. 
These studies include eight epidemiologic investigations in presumed 
susceptible populations, that is, Native Americans, sport anglers, the 
urban poor, pregnant women, fetuses and nursing infants of mothers who 
consume contaminated Great Lakes fish, infants and children, and the 
elderly. The ninth study focuses on developing more sensitive methods 
to detect persistent Great Lakes contaminants, such as polychlorinated 
biphenyls, dioxins, alkylated lead, mirex, and methylmercury, in human 
biologic tissues and fluids. In fiscal year 1993 ATSDR funded ten 
grants which included nine continuation awards for investigations 
funded in 1992 and one new award that established an interlaboratorybased, 
quality assurance/control program for the ATSDR research 
program. In fiscal year 1994, ATSDR funded continuation awards for all 
10 research grants.

    The impact of this research program will be felt most directly by 
the communities within the Great Lakes basin. Collectively, these 10 
research projects will (1) build upon and extend six existing human 
health studies in the Great Lakes basin that include high-risk 
populations; (2) establish two new subpopulations that include AfricanAmerican 
women, and men and women of reproductive age between 18 and 
34; (3) improve analytical methodology for detecting low levels of 
Great Lakes contaminants in human biologic tissues and fluids and in 
environmental media; (4) characterize exposure to all 11 critical Great 
Lakes contaminants identified by the International Joint Commission, as 
well as other pollutants; (5) determine profiles and levels (body 
burden) of Great Lakes contaminants in high-risk populations; (6) 
identify sensitive human health end points from exposure to Great Lakes 
pollutants, i.e., behavioral, developmental, reproductive, neurologic, 
endocrinologic, and immunologic effects; (7) investigate paternal and 
maternal exposure to Great Lakes pollutants and assess the potential 
for related health effects in their children (transgenerational 
effects); (8) increase collaboration, cooperation, and communication 
between the researchers in the Great Lakes basin; and (9) provide 
public health information to the study populations, health care 
providers, and State and local health departments concerning human 
health effects that may result from exposure to Great Lakes pollutants 
by fish consumption.

    Additionally, the research conducted by this program will help 
delineate the relationships among contaminant levels in the 
environment, exposure pathways, tissue levels, and potential human 
health effects; allow for evaluation and interpretation of data across 
human health studies to facilitate a basin-wide analysis of the 
pollution problem in the Great Lakes; and provide a ``model'' for other 
ecosystem-level studies intended to determine human health impacts of 
hazardous waste.



Rationale for Combined Analyses of ATSDR Research Investigations



    Combined analyses of the research studies of the ATSDR Great Lakes 
Human Health Effects Research Program will provide qualitative and 
quantitative research synthesis of the ATSDR- supported investigations. 
It is expected that combined analyses of the studies will improve the 
science base for investigations of consumption of fish contaminated 
with persistent toxic compounds from the Great Lakes, strengthen the 
scientific foundation for informed decision-making regarding public 
policy, and improve coordination and linkages between research 
activities and public health practices.



Procedures for Combined Analyses of ATSDR Research Studies



    The combined analyses (research synthesis) of epidemiologic 
investigations may be accomplished by meta-analysis of published 
results or pooled analysis of primary data. Both methods use explicit 
criteria, can be replicated, and provide a quantitative result. The 
following procedures will address key methodologic issues that are 
relevant to both methods of research synthesis, as well as their 
advantages and limitations.

    Meta-analyses attempt to analyze and combine the results of 
previous independent studies of a given scientific issue. Metaanalyses 
can be used to increase the power of statistical tests for 
important end points and subgroups, to resolve uncertainty when studies 
have conflicting conclusions, and to improve estimates of effect size. 
Meta-analyses rely on the published reports of previous studies and are 
relatively easy and inexpensive to perform. However, they are also 
susceptible to many sources of bias and are influenced by statistical 
methods. Six major areas have been identified as critical elements of 
scientifically valid meta-analyses. Proposed meta-analyses of ATSDR 
studies will be conducted according to a predetermined protocol which 
will address the six major areas as follows: (1) study design, 
including protocol and [[Page 33219]] literature search; (2) 
combinability of results of separate studies; (3) control and 
measurement of potential bias; (4) statistical analysis including 
significance tests and point and interval estimation; (5) sensitivity 
analysis to confirm final results; and (6) application of results which 
provides perspective of pooled results.
    Pooled data analyses attempt to analyze and combine the results of 
individual subject level data across studies. Pooled data analyses can 
facilitate the study of rare exposures as well as confounding and 
interactions between established and suspected risk factors. Common 
definitions, coding, cutpoints for variables, and adjustment for the 
same confounders can be accomplished in pooled data analyses. 
Consistency of findings and previously unrecognized errors, 
inconsistencies, and associations may also be examined. However, pooled 
data analyses are more difficult to conduct because they are labor- and 
time-intensive. In addition, important methodologic issues remain 
regarding the influence of study populations and methods on the results 
of the pooled data analyses, and the integration of qualitative 
assessments of research studies with quantitative estimates of the 
results. Guidelines for a systematic methodology for the pooled 
analysis of subject level data from previously conducted epidemiologic 
studies focus on eight critical areas. Proposed pooled data analyses 
for ATSDR studies will be conducted according to a predetermined 
protocol which will address the eight critical areas as follows: (1) 
location of all studies conducted on the topic of interest; (2) 
selection of the studies for the pooling project; (3) obtaining the 
primary data from original investigators and preparing the data for the 
pooled analysis; (4) estimation of study-specific effects; (5) 
examination of heterogeneity of these study-specific effects and how 
they should be pooled; (6) estimation of the pooled effects with the 
appropriate statistical model; (7) examination of heterogeneity between 
studies if this exists; and (8) conduct of a sensitivity analysis.



    Dated: June 20, 1995.

Claire V. Broome,

Deputy Administrator, Agency for Toxic Substances and Disease Registry.
[FR Doc. 95-15659 Filed 6-26-95; 8:45 am]
BILLING CODE 4163-70-P

 
 

Local Navigation

Jump to main content.