Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 17
1
Background and Methodologic
Considerations
INTENT AND GOALS OF THE STUDY
The Centers for Disease Control and Prevention (CDC) charged the
Institute of Medicine (IOM) committee responsible for this report to con-
duct a comprehensive review of the scientific literature regarding the rela-
tionship between damp or moldy indoor environments and the manifesta-
tion of adverse health effects, particularly respiratory and allergic symptoms.
The request came against the backdrop of escalating public and scientific
community interest in the question of whether indoor exposure to mold
and other agents might have a role in adverse health outcomes experienced
by occupants of damp buildings. Prominent among these health outcomes
is acute idiopathic pulmonary hemorrhage in infants, cases of which were
reported in Cleveland, Ohio in the 1990s. Residence in homes with recent
water damage and in homes with visible mold (including Stachybotrys
chartarum) were among the risk factors identified in the case infants.
The CDC requested that the review focus on fungi and their second-
ary metabolites, including mycotoxins. Several issues were identified for
consideration:
· The effect of damp indoor spaces on health.
· The relationship between damp indoor spaces and fungi.
· The characterization of fungal growth in homes, including the defi-
nition of the specific ecologic niches that fungi exploit in water-damaged
areas.
17
OCR for page 18
18 DAMP INDOOR SPACES AND HEALTH
· The conditions needed for toxin and allergen production.
· Methods of detecting fungi and secondary metabolites in indoor
environments.
· Mechanisms of exposure to fungi and secondary metabolites.
· Respiratory health effects of fungal exposure, including allergic
effects.
· The role of secondary metabolites--in particular, mycotoxins--in
adverse health outcomes.
· Pathologies associated with mycotoxins in pulmonary tissue.
· The synergistic interaction of molds and their toxins with biologic
and chemical agents in the environment.
· The evidence for and effectiveness of prevention, control, and man-
agement of exposures related to fungi.
However, CDC indicated that the committee should exercise its own judg-
ment concerning the topics to address in its report.
The committee operationalized this charge by establishing 7 broad
areas of inquiry:
· How and where buildings become wet, the signs of dampness, how
dampness is measured, the risk factors for moisture problems, and what is
known about their prevalence, severity, location, and duration.
· How dampness influences indoor microbial growth and chemical
emissions, the various agents that may be present in damp environments,
and the influence of building materials on microbial growth and emissions.
· The means available for assessing exposure to microorganisms and
microbial agents that occur in damp indoor environments.
· The experimental data on the nonallergic biologic effects of molds
and bacteria, including the bioavailability of mycotoxins and toxic effects
seen in in vitro and animal toxicity studies of mycotoxin and bacterial toxin
exposure.
· The state of the scientific literature regarding health outcomes and
indoor exposure to dampness and dampness-related agents.
· Dampness prevention strategies, published guidelines for the removal
of fungal growth (remediation), remediation protocols, and research on the
effectiveness of various cleaning strategies.
· The public health implications of damp indoor environments and
the elements of a public health response to the issue.
The ensuing chapters of the report address these topics to the extent
permitted by currently available science. Because there are great differences
in the amount and type of information available on specific topics, the
discussions vary in their depth and focus.
OCR for page 19
BACKGROUND AND METHODOLOGIC CONSIDERATIONS 19
The remainder of this chapter describes how the committee evaluated
the evidence it reviewed. It discusses the committee's research approach,
and--for the epidemiologic evidence used to assess questions about health
outcomes--the methodologic considerations underlying the evaluation of
information, considerations in assessing the strength of the evidence, and
the categories used to summarize the committee's conclusions. Some text is
derived from Chapter 2 of the report Clearing the Air (IOM, 2000). Insti-
tute of Medicine (IOM) reports characterizing scientific evidence regarding
vaccine safety (IOM, 1991, 1993) and the health effects of herbicides used
in Vietnam (IOM, 1994, 1996, 1999, 2001, 2003) have used similar ap-
proaches to summarize epidemiologic evidence.
RESEARCH APPROACH
Information Gathering
To answer the questions posed by CDC, the committee undertook a
wide-ranging evaluation of the research on the determinants, characteriza-
tion, and remediation of damp indoor spaces and the possible association
of dampness or dampness-related agents with occupant health. While it did
not review all such literature--an undertaking beyond the scope of this
report--the committee attempted to cover the work it believed to be influ-
ential in shaping scientific understanding at the time it completed its task in
late 2003.
The committee consulted several sources of information in the course
of its work. For conclusions regarding health outcomes, the primary source
was epidemiologic studies. Most of those studies examined general popula-
tion exposures to dampness or indoor agents at home, reflecting the focus
of researchers working in this field. Some clinical research was considered
where appropriate. Animal (in vivo) and cellular (in vitro) studies were
examined in the review of toxicologic literature. The literature of engineer-
ing, architecture, and the physical sciences informed the committee's dis-
cussions of building characteristics, exposure assessment and characteriza-
tion, indoor dampness, pollutant transport, and related topics; and public
health and behavioral sciences research was consulted for the discussion of
public health implications. These disciplines have different practices regard-
ing the publication of research results. There are, for example, relatively
few papers in the peer-reviewed literature that address building construc-
tion or maintenance issues. The committee endeavored in all cases to iden-
tify, review, and consider fairly the literature most relevant to the topics it
was charged to address.
Studies and reports were identified for review through extensive searches
of relevant databases. The majority of these were bibliographic, providing
OCR for page 20
20 DAMP INDOOR SPACES AND HEALTH
citations to peer-reviewed scientific literature. Factual databases were also
searched to provide toxicological, demographic, and other information. Com-
mittee staff examined the reference lists of major review articles, books, and
reports for relevant citations. Reference lists of individual articles were also
scanned for additional relevant references. Committee members indepen-
dently compiled lists of potential citations, based on their areas of expertise.
The input received both in written and oral form from participants at the
public meetings served as a valuable source of additional information. If an
initial examination revealed that the study addressed agents or means of
exposure that were not relevant to the indoor environments being evaluated;
details regarding the subjects, research methodology, or some other aspect of
the study that lead the committee to conclude that it would not inform the
review; or that the study replicated information in papers that were already
being reviewed, it was not further evaluated.
Publication Bias
An important aspect of the quality of a review is the extent to which all
appropriate information is considered and serious omissions or inappropri-
ate exclusions of evidence are avoided. A primary concern in this regard
is the phenomenon known as publication bias. It is well documented
(Begg and Berlin, 1989; Berlin et al., 1989; Callaham et al., 1998; Dickersin,
1990; Dickersin et al., 1992; Easterbrook et al., 1991) that studies with
statistically significant findings are more likely to be published than studies
with nonsignificant results. Where such bias is present, evaluations of
exposuredisease associations based solely on published literature could be
biased in favor of showing a positive association. Other forms of bias
related to reporting and publication of results have also been suggested.
These include multiple publications of positive results, slower publication
of nonsignificant and negative results, and publication of nonsignificant
and negative results in non-English-language and low-circulation journals
(Sutton et al., 1998). For example, several researchers have addressed the
specific topic of possible bias in the publication of studies regarding the
health effects of exposure to environmental tobacco smoke (Bero et al.,
1994; Kawachi and Colditz, 1996; Lee, 1998; Misakian and Bero, 1998).
The committee did not in general consider the risk of publication bias
to be high among studies of the health of people exposed to indoor damp-
ness or dampness-related agents, because
· Numerous published studies reported no association.
· The committee was aware of the results of some unpublished research.
· The committee felt that the interest of the research community, pub-
lic health professionals, government, and the general public in the issue of
OCR for page 21
BACKGROUND AND METHODOLOGIC CONSIDERATIONS 21
mold exposure and health is so intense that any studies showing no associa-
tion would be likely to be viewed as important by investigators. In short,
there would also be pressure to publish negative findings.
Nonetheless, the committee was mindful of the possibility that studies show-
ing a positive association might be over-represented in the literature.
The Role of Judgment
The examination of evidence went beyond quantitative considerations
at several stages: assessing the relevance and validity of individual reports;
deciding on the possible influence of such factors as error, bias, confound-
ing, or chance on the reported results of empirical studies; integrating the
overall evidence within and between diverse types of studies; and formu-
lating the conclusions themselves. Those aspects of the committee's re-
view required thoughtful consideration of alternative approaches at several
points and could not be accomplished by adherence to a narrowly pre-
scribed formula.
The approach to evaluating evidence therefore evolved throughout the
committee process and was determined, to a large extent, by the nature of
that evidence. The committee informed its expectations for the literature by
the reality of the state of the science--for example, the lack of valid quan-
titative exposure assessment methods and a lack of knowledge of which
specific microbial agents might primarily account for any presumed health
effects. Although the quantitative and qualitative aspects of the process that
could be made explicit were important to the overall review, ultimately the
conclusions expressed in this report are based on the committee's collective
judgment. The committee has endeavored to express its judgments as clearly
and precisely as the data allowed.
EVALUATING THE EPIDEMIOLOGIC EVIDENCE
Methodologic Considerations
Several methodologic considerations underlie the evaluation of the epi-
demiologic studies reviewed in this report. Three of these--uncertainty and
confidence, analytical bias, and confounding--are addressed below.
Uncertainty and Confidence
All science is characterized by uncertainty; scientific conclusions con-
cerning the result of a particular analysis or set of analyses can range from
highly uncertain to highly confident. In its review, the committee evaluated
OCR for page 22
22 DAMP INDOOR SPACES AND HEALTH
the degree of uncertainty associated with the results on which it had to base
its conclusions.
In the epidemiologic studies reviewed in this report, statistical signifi-
cance is a quantitative measure of the extent to which chance--that is,
sampling variation--might be responsible for an observed association be-
tween an exposure and an adverse event. The magnitude of the probability
value or the width of the confidence interval associated with an effect
measure, such as the relative risk or risk difference, is generally used to
estimate the role of chance in producing the observed association. Confi-
dence intervals are a function, in part, of the sample size: all else equal,
increasing the number of samples increases the precision of the estimate.
This type of quantitative estimation is firmly founded in statistical theory
on the basis of repeated sampling.
Empirical measures do not, however, necessarily capture all relevant
considerations that should be applied when evaluating the uncertainty of
conclusions about an association between an exposure and a health out-
come. Therefore, to assess the appropriate level of confidence to be placed
in conclusions, it is useful to also consider qualitative aspects.
Analytic Bias
Analytic bias is a systematic error in the estimate of association; for
example, between an exposure and an adverse event. It can be categorized
as selection bias, information bias, confounding bias, and reverse causality
bias. Selection bias refers to the way that the sample of subjects for a study
has been selected (from a source population) and retained. If the subjects in
whom an exposureadverse event association has been analyzed differ from
the source population in ways linked to both exposure and development of
the adverse event, the resulting estimate of association will be biased. Infor-
mation bias is the result of a systematic error in the measurement of infor-
mation on an exposure or outcome. It can result in a bias toward the null
hypothesis (that is, that there is no association between the exposure and
the adverse event), particularly when ascertainment of either exposure or
outcome has been sloppy, or it may create a bias away from the null
hypothesis through such mechanisms as recall bias or unequal surveillance
of exposed and non-exposed subjects. Confounding bias--addressed in
greater detail below--occurs when the exposureadverse event association
is biased as a result of a third factor that is both capable of causing the
adverse event and is statistically associated with the exposure. Finally, re-
verse causality bias occurs when it is possible that the outcome in question
influences the probability of experiencing the exposure being studied. It is
not always possible to quantify the impact of such nonrandom errors in
estimating the strength of an association.
OCR for page 23
BACKGROUND AND METHODOLOGIC CONSIDERATIONS 23
Another form of bias is possible in studies that use surveys or question-
naires to obtain information about exposures or adverse events. Response
bias is "a systematic tendency to respond to a range of questionnaire items
on some basis other than . . . what the items were designed to measure"
(Paulhus, 1991). In this context, it may be considered a form of information
or reporting bias. It would be a factor in the studies evaluated here if, for
example, those who had experienced an adverse health outcome were more
likely to seek out and report instances of indoor dampness than those who
had not.
Confounding
In any epidemiologic study comparing an exposed with a non-exposed
group, it is likely that the two groups will differ in characteristics other than
exposure. When the groups differ with respect to factors that are also
associated with the risk of the outcome of interest, a simple comparison of
the groups may either exaggerate or hide the true difference in disease rates
that is due to the exposure of interest. For example, people with low socio-
economic status may be more likely to be exposed to a particular indoor
pollutant than other people. A simple comparison of the incidence of the
health outcome among the exposed and non-exposed may exaggerate an
apparent difference because socioeconomic status is also thought to influ-
ence the incidence of several health problems. If exposed people were of
higher socioeconomic status, the simple comparison would tend to mask
any true association between exposure and outcome by spuriously increas-
ing the risk of disease in the non-exposed group. This phenomenon, known
as confounding, poses a major challenge to researchers and those evaluat-
ing their work.
Considerations in Assessing the Strength of Epidemiologic Evidence
Evaluation Criteria
A widely used set of criteria has evolved for the assessment of epide-
miologic evidence (Hill, 1965; Hill and Hill, 1991; Susser, 1973; U.S. Pub-
lic Health Service, 1964); they are also often used to inform public health
policy recommendations and decisions (Weed, 1997). These criteria in-
formed the committee's review of the studies addressed in Chapter 5.
1. Strength of Association: Strength of association is usually expressed
in epidemiologic studies as the magnitude of the measure of effect, for
example, relative risk or odds ratio. Generally, the higher the relative risk,
the greater the likelihood that the exposuredisease association is "real";
OCR for page 24
24 DAMP INDOOR SPACES AND HEALTH
that is, the less likely it is to be due to undetected error, bias, or confound-
ing. Where the study population is small, selection bias may also play an
important role. Small increases in relative risk that are consistent across a
number of studies, however, may also provide evidence of an association.
2. Biologic Gradient (DoseResponse Relationship): In general, a po-
tential association is strengthened by evidence that the risk of occurrence of
an outcome increases with dose or frequency of exposure. In the case of
allergic diseases, this is complicated by the central roles that susceptibility
and sensitization play in the disease. The same exposure may have very
different effects in susceptible and nonsusceptible people and in sensitized
and nonsensitized people. Thus, multiple dose-response curves may be
needed to characterize a particular exposure-disease association.
3. Consistency of Association: Consistency of association requires that
an association be found regularly in a variety of studies, for example, in
more than one study population and with different study methods. Findings
that are consistent among different categories of studies are supportive of
an association. However, consistency does not necessarily mean that one
should expect to see exactly the same magnitude of association in different
populations. Rather, consistency of a positive association means that the
results of most studies are positive and that the differences in measured
effects are within the range expected on the basis of all types of error,
including sampling, selection bias, misclassification, confounding, and dif-
ferences in exposure.
4. Biologic Plausibility and Coherence: Biologic plausibility is based on
whether a possible association fits existing biologic or medical knowledge.
The existence of a possible mechanism increases the likelihood that the
exposuredisease association in a particular study reflects a true associa-
tion. In addition, in evaluating exposures such as those addressed in this
report, one might consider such factors as evidence that an outcome was
associated with documented high exposure levels outside the home.
5. Temporally Correct Association: If an observed association is real,
exposure must precede the onset or exacerbation of the disease by at least
the duration of disease induction. Temporality can be difficult to evaluate
for some indoor agents because exposure to them is recurrent and perva-
sive. If people are exposed to an agent almost every day in an environment
where they spend most of their time, it can be difficult to discern a relation-
ship between exposure and effect. The lack of an appropriate time sequence
is evidence against association, but the lack of knowledge concerning varia-
tions in exposure and exposure magnitude limits the utility of this consider-
ation. One might also consider whether the outcome being studied occurred
within a period after exposure that was consistent with current understand-
ing of its natural history.
OCR for page 25
BACKGROUND AND METHODOLOGIC CONSIDERATIONS 25
Other Considerations
As noted earlier, it is important also to consider whether alternative
explanations--error, bias, confounding, or chance--might account for the
finding of an association. If an association could be sufficiently explained
by one or more of these factors, there would be no need to invoke the
several considerations listed above. Because these alternative explanations
can rarely be excluded sufficiently, however, assessment of the applicable
considerations discussed in this chapter almost invariably remains appro-
priate. The final judgment is then a balance between the strength of support
for the association and the degree of exclusion of alternatives.
Bornehag et al.'s (2001) review of the literature regarding building
dampness and health includes an extensive discussion of the potential influ-
ences on study outcomes. These authors concluded that, while bias or
confounding may have contributed to some results, there was no reason to
believe that the reported associations were primarily driven by these factors.
Considerations of biologic plausibility informed the committee's deci-
sions about how to categorize associations between relevant indoor expo-
sures and health outcomes. However, the committee recognized that research
regarding mechanisms is still in its infancy, and it did not predicate decisions
on the existence of specific evidence regarding biologic plausibility.
The committee did not feel that there was sufficient evidence to support
confident quantitative estimates of the risk associated with relevant indoor
exposures. It is not possible to make general statements about the relative
risk posed by various exposures, because this depends heavily on the char-
acteristics of a particular environment and its occupants. Fungi are ubiqui-
tous and can be the primary source of allergens in some arid climates.
Endotoxins may be found in humidifiers in urban settings or in organic
dusts that infiltrate rural homes from outdoors. Occupant choice has a role
in determining indoor humidity and temperature levels as well as the spe-
cific building materials and furnishings present.
Much of the literature regarding damp indoor spaces and health out-
comes focuses on some measure of indoor moisture or exposure to mold or
particular dampness-related agents. Indoor environments, however, are
complex. They subject occupants to multiple exposures that may interact
physically or chemically with one another and with other characteristics of
the environment such as temperature and ventilation levels. Synergistic
effects--that is, interactions among agents that result in a combined effect
greater than the sum of the individual effects--may also occur. Information
on the combined effects of multiple exposures and on synergist effects
among agents is cited in this report wherever possible; however, rather little
information is available on this topic, and it remains one of active research
interest.
OCR for page 26
26 DAMP INDOOR SPACES AND HEALTH
SUMMARIZING CONCLUSIONS REGARDING
EPIDEMIOLOGIC EVIDENCE
Categories of Association
The committee summarized its conclusions regarding health outcomes
by using a common format, described below, to categorize the strength of
the scientific evidence. The five categories were adapted by the committee
from those used by the International Agency for Research on Cancer (IARC,
1977) to summarize the scientific evidence of the carcinogenicity of various
agents. Similar sets of categories have been used in National Academies
reports characterizing scientific evidence regarding vaccine safety (IOM,
1991, 1993), the health effects of herbicides used in Vietnam (IOM, 1994,
1996, 1999, 2001, 2003), and the association between asthma and various
indoor exposures (IOM, 2000). The distinctions reflect the committee's
judgment that an association would be found in a large, well-designed
study of the outcome in question in which exposure is sufficiently high, well
characterized, and appropriately measured on an individual basis.
The categories address the association between exposure to an agent
and a health outcome, not to the likelihood that any individual person's
health problem is associated with or caused by the exposure.
Sufficient Evidence of a Causal Relationship
Evidence is sufficient to conclude that a causal relationship exists be-
tween the agent and the outcome. That is, the evidence fulfills the criteria
for "sufficient evidence of an association" and, in addition, satisfies the
evaluation criteria discussed above: strength of association, biologic gradi-
ent, consistency of association, biologic plausibility and coherence, and
temporally correct association.
The finding of sufficient evidence of a causal relationship between an
exposure and a health outcome does not mean that the exposure would
inevitably lead to that outcome. Rather, it means that the exposure can
cause the outcome, at least in some people under some circumstances.
Sufficient Evidence of an Association
Evidence is sufficient to conclude that there is an association. That is,
an association between the agent and the outcome has been observed in
studies in which chance, bias, and confounding could be ruled out with
reasonable confidence. For example, if several small studies that are free
from bias and confounding show an association that is consistent in magni-
tude and direction, there may be sufficient evidence of an association.
OCR for page 27
BACKGROUND AND METHODOLOGIC CONSIDERATIONS 27
Limited or Suggestive Evidence of an Association
Evidence is suggestive of an association between the agent and the out-
come but is limited because chance, bias, and confounding could not be ruled
out with confidence. For example, at least one high-quality study shows a
positive association, but the results of other studies are inconsistent.
Inadequate or Insufficient Evidence to Determine Whether or Not an
Association Exists
The available studies are of insufficient quality, consistency, or statisti-
cal power to permit a conclusion regarding the presence or absence of an
association. Alternatively, no studies exist that examine the relationship.
Limited or Suggestive Evidence of No Association
Several adequate studies are consistent in not showing an association
between the agent and the outcome. A conclusion of "no association" is
inevitably limited to the conditions, magnitude of exposure, and length of
observation covered by the available studies.
REFERENCES
Begg CB, Berlin JA. 1989. Publication bias and dissemination of clinical research. Journal of
the National Cancer Institute 81(2):107115.
Berlin JA, Begg CB, Louis TA. 1989. An assessment of publication bias using a sample of
published clinical trials. Journal of the American Statistical Association 84:381392.
Bero LA, Glantz SA, Rennie D. 1994. Publication bias and public health policy on envi-
ronmental tobacco smoke. Journal of the American Medical Association 272(2):133
136.
Callaham ML, Wears RL, Weber EJ, Barton C, Young G. 1998. Positive-outcome bias and
other limitations in the outcome of research abstracts submitted to a scientific meeting.
Journal of the American Medical Association 280(3):254257. [Published erratum ap-
pears in JAMA 1998 280(14):1232.]
Dickersin K. 1990. The existence of publication bias and risk factors for its occurrence.
Journal of the American Medical Association 263(10):13851389.
Dickersin K, Min YI, Meinert CL. 1992. Factors influencing publication of research results:
follow-up of applications submitted to two institutional review boards. Journal of the
American Medical Association 267(3):374378.
Easterbrook PJ, Berlin JA, Gopalan R, Matthews DR. 1991. Publication bias in clinical
research. Lancet 337(8746):867872.
Hill AB. 1965. The environment and disease: association or causation. Proceedings of the
Royal Society of Medicine 58:295300.
Hill AB, Hill ID. 1991. Bradford Hill's Principles of Medical Statistics (Twelfth Edition).
London: Hodder & Stoughton.
OCR for page 28
28 DAMP INDOOR SPACES AND HEALTH
IARC (International Agency for Research on Cancer). 1977. Some Fumigants, the Herbicides
2,4-D and 2,4,5-T, Chlorinated Dibenzodioxins and Miscellaneous Industrial Chemi-
cals. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to
Man, Vol. 15. Lyon: IARC.
IOM (Institute of Medicine). 1991. Adverse Effects of Pertussis and Rubella Vaccines. Wash-
ington, DC: National Academy Press.
IOM. 1993. Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Cau-
sality. Stratton KR, Howe CJ, Johnston RB, eds. Washington, DC: National Academy
Press.
IOM. 1994. Veterans and Agent Orange Health Effects of Herbicides Used in Vietnam.
Washington, DC: National Academy Press.
IOM. 1996. Veterans and Agent Orange: Update 1996. Washington, DC: National Academy
Press.
IOM. 1999. Veterans and Agent Orange: Update 1998. Washington, DC: National Academy
Press.
IOM. 2000. Clearing the Air: Asthma and Indoor Air Exposures. Washington, DC: National
Academy Press.
IOM. 2001. Veterans and Agent Orange: Update 2000. Washington, DC: National Academy
Press.
IOM. 2003. Veterans and Agent Orange: Update 2002. Washington, DC: The National Acad-
emies Press.
Kawachi I, Colditz GA. 1996. Invited commentary: confounding, measurement error, and
publication bias in studies of passive smoking. American Journal of Epidemiology
144(10):909915.
Lee PN. 1998. Difficulties in assessing the relationship between passive smoking and lung
cancer. Statistical Methods in Medical Research 7(2):137163.
Misakian AL, Bero LA. 1998. Publication bias and research on passive smoking: comparison
of published and unpublished studies. Journal of the American Medical Association
280(3):250253.
Paulhus DL. 1991. Measurement and control of response bias. In JP Robinson, PR Shaver, LS
Wrightsman, eds. Measures of personality and social psychological attitudes. Volume 1.
pp. 1759. San Diego, CA: Academic Press.
Susser M. 1973. Causal Thinking in the Health Sciences: Concepts and Strategies in Epidemi-
ology. New York: Oxford University Press.
Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. 1998. Systematic reviews of trials and
other studies. Health Technology Assessment 2(19):1276.
U.S. Public Health Service, U.S. Department of Health, Education, and Welfare. 1964. Assess-
ing Causes of Adverse Drug Reactions with Special Reference to Standardized Methods.
Venulet J, ed. London: Academic Press.
Weed DL. 1997. On the use of causal criteria. International Journal of Epidemiology 26(6):
11371141.
