Memorandum by Dr Chris Gliddon, School
of Biological Sciences, University of Wales
1.1 The following evidence relates to the problems
associated with the lack of consistency of implementation and
interpretation of Directive 90/220/EEC. Issues of harmonisation
with the EU are addressed and suggestions for amendment of the
Directive are addressed.
2.1 A widely accepted definition of Risk
is a measure of the effects (economic, injury, environment) of
an occurrence in terms of both its probability and the
magnitude of its consequences. This simple definition immediately
requires that risk be decomposable into two main components: the
hazard and its likelihood.
2.2 Hazard can be defined as the property
of a substance or a process that can cause harm. Hazard can never
be zero; it can be direct or indirect, immediate or delayed, natural
or technological, intentional or unintentional. Harm is
the manifestation of that hazard. Given this, risk can be defined
risk as the product of (1) the magnitude of the harm (generated
by a hazard) and (2) the frequency with which the hazard occurs.
3. RISK ASSESSMENT
3.1 Risk assessment is the scientifically based
process consisting of the identification and characterisation
of hazards, the assessment of exposure and the characterisation
of the resulting risk. Risk management is the process of weighing
policy alternatives in the light of the results of the risk assessment
and if required, selecting and implementing appropriate control
options including regulatory measures. That is, risk assessment
seeks to identify the hazards and determines the likelihood of
their occurrence, i.e., the probability of something going wrong.
Risk management then can, as a result of this process, allow necessary
controls to be put into place to eliminate the hazard or, if this
is not possible, to reduce the risk of the hazard causing harm.
3.2 Environment risk assessment is the systematic
evaluation of the risks associated with hazards to human health
and safety and the environment, arising from human activities
capable of impacting on the environment on a continuous or accidental
4. The process of environmental risk assessment
4.1 The following sequence of procedures for
risk assessment has been proposed:
(i) identification of hazards;
(ii) estimation of their magnitude;
(iii) estimation of the frequency of their
(iv) evaluation of risks.
4.2 Risk assessment is feasible even in situations
where current information is limited, providing its purposes and
limitations are realised. Risk assessment should be an iterative
process. The earliest analysis might determine where more information
is needed to support credible risk assessments in future iterations
and provide limited guidance in reducing risks to health and the
environment through risk management and in the continuing decision-making
process associated with remediation.
4.3 The way to present the results of a risk
assessment in which variability and uncertainty are acknowledged
is controversial. This is often as a result of confusion regarding
the difference between variability and uncertainty. Variability
comprises a population's natural heterogeneity or diversity, which
does not change through further measurement or study, although
better sampling can improve an estimate of its magnitude. Uncertainty,
in contrast, reflects gaps in information about scientifically
observable phenomena. Uncertainty sometimes can be reduced through
further measurement or study and several quantitative methods
to describe risk-assessment uncertainties are currently being
explored. Although there is general agreement as to the value
of qualitative statements describing critical uncertainties in
health risk assessment, formal quantitative approaches to uncertainty
analysis are complex, difficult to perform, difficult to understand
and often unnecessary. Variability, in contrast, can be described
much more readily and can be based on actual measurements.
4.4 Appropriate risk assessment consists of
the application, in a systematic manner, of a wide range of scientific
methods in order to master:
(i) how information is gathered systematically;
(ii) how its uncertainty is determined;
(iii) how potential future outcomes and their
impacts are explored in an objective and reproducible manner;
(iv) how the likelihood of these outcomes
is displayed clearly and comprehensively.
5. THE ACTUALITY
5.1 A review of the current situation regarding
applications for consent to market GMO's within the EU, with almost
every application eventually being referred to the Article 21
Committee, shows clearly that environmental risk assessments are
not performed and/or interpreted consistently among the various
5.2 This lack of consistency seems to derive
from (1) several weaknesses both in the present regulatory framework
and the way it has been implemented nationally as well as from
(2) significant gaps in the underlying science base.
6. THE REGULATORY
6.1 Hazards cannot be zero and, as such, there
is a risk associated with every release of a GMO. In this context,
the use of the products of modern biotechnology is often perceived
as a possible threat. It would be useful, therefore, if both regulators
and consumers were able to balance potential risk against possible
benefit. The scope of Directive 90/220 is to allow the release
of a GMO based on an assessment of environmental safety and it
would be in the interests of both consumers and regulators to
broaden this scope to permit environmental risk to be balanced
against environmental benefit.
6.2 Environmental benefits can be direct or
indirect and they should be assessed against direct as well as
indirect environmental risks. As such, the scope of Directive
90/220 should be broadened in order to accommodate direct as well
as indirect effects.
6.3 When balancing risks against benefit we
must also take into account that the "currencies" applied
must be identical. Furthermore, a pragmatic approach is required
when taking distant indirect effects into consideration. Their
consequences become weak and therefore these should not get the
same level of attention as close indirect effects or as direct
6.4 The regulatory framework should promote
the exchange of views among experts on pertinent issues such as
hazard definition in relation to crops with relevance for European
biotechnology or for genes that are currently being released in
experimental field trials.
6.5 These experts could form a scientific committee
that proposes to the European Commission projects that are worthy
of being carried out. A demanding mandate could lead to a more
efficient use of the committee that is composed of the representatives
of the Member States and that has been set up under Article 21
of Directive 90/220/EEC to assist the Commission.
6.6 The information requirement for Annexes
II and III of Directive 90/220/EEC should be structured so that
they can be directly related to the definition of the actual hazards
related to the application.
6.7 Directive 90/220/EEC should take account
of the difference in data requirements between an application
for a deliberate field trial and an application for a marketing
dossier due to differences in both scale and location. A new annex
III that is specific for multi-location, large scale marketing
applications should be elaborated.
6.8 Annex II and Annex III should distinguish
between requirements for different groups of organisms.
6.9 Administrative provisions should be made
to allow for relevant data collection. An example is to allow
for a single approval system for multi-state field trials that
are considered as very important prerequisites for any commercialisation
7. SCIENTIFIC LACUNAE
7.1 While well over 8,000 releases of GMOs into
the environment have been made world-wide, of which about 1,000
have taken place in Europe, many of these releases have been carried
out with monitoring purely designed to ensure efficacy of any
confinement requirements attached to the consent. As such, much
scientific information of potential value for future risk assessment
remains uncollected and there is a tendency to over-value the
so-called familiarity with particular classes of GMOs.
7.2 Hazard identification must depend largely
on the availability of high quality data. Actions resulting in
increasing the availability of data or improving their quality
lead ultimately to an improvement of the processes of hazard identification
and of risk assessment and should be encouraged.
7.3 Expert advice should be organised at the
Community level in relation to applications for consent to market.
7.4 Apart from the need for the organisation
of the expert advice, hazard identification becomes a more scientific
process as more data becomes available. Therefore, both the continuing
generation of high quality data and its exchange on an international
basis will lead ultimately to a streamlining of the hazard identification
process and should be encouraged.
7.5 Data collection, data sharing and the compilation
of databases that allow for the evaluation of the interaction
of a GMO with a given ecosystem are generally recognised as an
essential step forward. There is a need for a systematic collection
and storage of a thoroughly investigated set of information so
that hazard and the concomitant risk analysis can be performed
on an internationally accepted basis.
7.6 Data should be obtained from the field testing
of genetically modified organisms in which some of the environmental
and safety concerns match those of the marketing stage. Broad
geographical field testing therefore enhances European wide experience
with a specific GMO. This in turn should have a positive effect
on the consideration of possible hazards involved at the marketing
stage and at the approval process in all Member States.
7.7 It is not possible to compile a positive
list of all hazards that need to be taken into consideration in
any risk assessment dossier. Some types of hazards, such as persistence
of the GMO in nature might always appear in a risk assessment
dossier whereas other hazards, such as the creation of new plant
pests, are dossier-specific.
7.8 Hazards might be specific for a geographical
location or for a given ecosystem. A product dossier, although
originating from one Member State, should ideally take into consideration
all possible hazards for all possible European environments.
7.9 Hazards differ when formulated by different
individuals. It is extremely important to assemble a list of perceived
hazards that is complete on the one hand but sufficiently restricted
on the other hand.
7.10 Directive 90/220/EEC should be modified
with respect to hazard definition for the commercialisation of
transgenic crops. At present, only those hazards that are directly
related to the protection of human health and the environment
are taken into consideration. All hazards and hazard categories
pertinent to commercialisation should fall within the scope of
any future modifications to the Directive.
7.11 Definition and identification of hazards
are based on knowledge. An appropriate framework should be constructed
to permit the definition of hazards with an intellectual input
that is as high as possible.
8. LIMITATIONS OF
8.1 The traditional framework for risk assessment
and management, drawn from expertise with chemical products, involves
a methodological progression through a rigorous sequence of analytical
steps, including hazard identification, exposure assessment and,
ultimately, risk-cost benefit assessment. The bio-ecological phenomena
related to environmental releases of GMOs, however, do not submit
neatly to this quantitative approach, due to the complexity of
the phenomena and the scarcity of relevant data.
8.2 When organisms are released into an ecosystem,
it must be realised that the various pathways of outcomes and
consequences are far too numerous for detailed empirical investigations.
Scientists must rely on a judgmental analysis and on reasoning
8.3 Environmental risk assessment is still far
from providing a standardised methodology that is based on data
on occurrence probabilities and data from tests on effects of
8.4 High quality review and test data provide
the basis for decision making. Data should be collected/generated
such that they can be interpreted and in a hazard-oriented model.
8.5 Experience should be fully appreciated in
the risk assessment process and, for those areas where experience
is lacking, experts should clearly define the need for generating
8.6 The responsibility for evaluating an environmental
release must thus invoke a considerable degree of scientific knowledge
and qualitative judgment in order to anticipate potential harmful
consequences and to balance it against the available alternatives.