Are hazard-based approaches to chemical regulation unscientific and anti-business?February 14, 2013 at 12:34 pm | Posted in H&E Features | 3 Comments
Tags: chemicals, environment, ragnar löfstedt, risk assessment, risk vs hazard, science
Should chemical regulation be based on the intrinsic potential for a substance to cause harm, or should further account be taken of the probability that they will actually cause harm? This is a question of increasing importance in European chemical regulation, as disputes about the safety of chemicals heat up and regulators seek to reassure the general public that what they are exposed to is safe, or if not safe, that chemical hazards are at least an issue which is under control.
However, as Dr Ragnar Löfstedt, of the Kings Centre for Risk Management at King’s College London, writes: “In Europe there has been a rather long and at times acrimonious discussion, as to the merits of risk assessments for regulatory purposes especially with regard to chemical substances. There is no European consensus of when to use risk assessments and when to use hazard classifications.” (Löfstedt 2011)
This seems particularly pertinent in, for example, the case of BPA, which has been risk-assessed by the European Food Safety Authority (EFSA) and found safe at current levels of exposure, recommended for restriction by French food safety agency ANSES, banned in baby bottles by the EU, banned even in receipts by Sweden, and generally held in suspicion by environmental groups and the public at large. Approvals accord with the current risk assessment advice; bans tend to be based on a preference for regulating on the hazard presented by BPA. At least, that is the story as Löfstedt presents it.
Before we get into why BPA is variously banned or accepted, it is worth reminding ourselves of the difference between risk- and hazard-based approaches to regulating chemicals. This is important, because “risk” and “hazard” have specific, technical meanings in the context of chemical safety assessment which tend to be conflated in everyday use (indeed, the Oxford English Dictionary defines risk as “a possibility of harm” and hazard as “a danger or risk”).
David Ropeik, a consultant in risk perception and management, describes hazard as concerning “whether or not the chemical in question presents an environmental or health hazard (is it potentially harmful?); the level of hazard it presents (such as the dose at which it is toxic, and the effects of the toxic dose); and the people or environmental elements to which it poses a hazard.”
Of risk, he says that: “Risk assessment takes this information and goes one step further, to evaluate the likelihood that the hazard is likely to result in actual harm, asking questions around: in what ways are we exposed to the chemical? How often? To what degree and at what age? (Ropeik 2013)
The contention is the extra information used in risk assessment allows the risk assessor to quantify the exposure level at which a chemical is likely to be harmful, make a judgment on who is going to be exposed to that level, and propose risk management measures on that basis.
If a risk assessment is reliable, then there can be a minimum number of restrictions on chemical use, allowing the maximum advantage to be gained from the entire pool of available compounds, rather than unnecessarily restricting society to using just those which cannot pose a risk because they pose no hazard.
As Löfstedt observes, different EU Member States place different degrees of weight on the hazard and risk elements of evaluating chemical safety in making a regulatory decision. Sweden, with its early action on BPA in infant proposals and its ban on BPA in receipt papers, tends to operate at the hazard end of the scale.
The UK takes quite the opposite approach: its double-checking of EFSA’s risk assessment of neonicotinoids to see if there really is adequate degree of risk to warrant a ban (Independent 6.2.13), and its promotion of a risk-based approach to the regulatory definition of “endocrine disruptor” (DE/UK 2011, see this H&E article for a full discussion) are both testament to this.
Löfstedt characterises risk assessment as a “science-based tool” and warns that “a hazard-based approach takes away the scientific rigour of risk assessments.” (Löfstedt 2011) The implication being that hazard-based approaches to chemical regulation are less scientific.
Hazard assessments are not pulled out of a hat. They are based on the same toxicological and epidemiological research as risk assessment (obviously, since risk assessment is the next step after hazard assessment). Presumably Löfstedt does not mean they are less scientific in that sense; he must mean that the extra data used in risk assessment means it is more scientific.
Risk assessment uses an extra set of data, but whether or not that makes it “more science-based” must surely depend on the quality of the science behind the extra data. More science is not the same thing as more scientific, because adding data only makes a decision more scientific if the data is any good: an extra volume of data does not necessarily improve the management of risk from a compound, if risk is underestimated because of an absence of data or because the exposure is simply too complex to model.
Löfstedt characterises the perceptions which drive hazard- and risk-based approaches as follows: “Those favouring hazard-based regulations argue that risk assessments are complex, non-scientific, costly and based on different methodologies that can produce different outcomes,” while advocates of risk-based approaches see hazard-based decisions as “based on assumptions rather than testing”.
It is not clear if the observation does not work both ways: a sceptic about the effectiveness of risk assessment could argue that risk-based decisions are based on at least as many assumptions as hazard-based decisions. As new evidence emerges about the possibility of low-dose effects, cumulative toxicity of mixtures, endocrine disruption and so forth – all new developments unanticipated by risk assessment practices in the past – the accuracy of risk assessment in the face of knowledge gaps could start to appear questionable.
These knowledge gaps make even the hazard identification stage of risk assessment difficult enough; to then add more information about routes of exposure, environmental impacts etc. introduces so many additional assumptions and reliance on further uncertain data, that it does not in fact improve our ability to protect ourselves from harm.
As is implicit in the latest edition of Late Lessons from Early Warnings (European Environment Agency 2013), if anything, risk-based approaches make us over-confident in our ability to anticipate harm and lowers our ability to respond sufficiently quickly: it is inherently risk to rely on risk assessment. It could be argued that a hazard-based approach is a pragmatic choice to nullify the risk of getting the risk assessment wrong, by eliminating hazard as much as possible.
Risk-based approaches are obviously consistent with the economic imperatives of governments and industry, as it maximizes opportunity for commercial enterprise. But are hazard-based approaches necessarily uneconomical?
Opposition to hazard-based approaches often comes from the argument that unnecessary restrictions on chemicals pose a large economic threat. Companies such as Hewlett Packard see this very differently: scientific uncertainty about the safety of individual compounds, in which new research can rapidly change understanding of the risk posed by an individual compound leading to its being banned, produces a very risky business environment into which to introduce a compound.
See this video for HP’s presentation of its hazard-based approach to chemicals management.
Do companies like HP need risk assessment to deal with these challenges? They argue that they do not. Retailers and product managers sell and design products with electrical flexes, circuit boards, heated surfaces, chewable parts, etc. Since each of these elements is used in the same way, exposure must be more-or-less constant across each product class.
If exposure is constant, to reduce risk the only thing the retailer needs to control for is the intrinsic hazard of the chemicals used in each product class; any decision about substitution can therefore be based on hazard profile. To put it another way: exposure is constant across product class, therefore manufacturers can dispense with exposure and risk assessment, and make substitution decisions based purely on hazard.
This is obviously not an argument which will suit primary manufacturers who are tooled-up for producing any of the chemicals which are restricted on the basis of their hazard profile. The notion, however, that a hazard-based approach might in some way be intrinsically anti-business or unscientific does not seem to stand up to scrutiny when one considers the entire range of companies involved in producing and selling consumer products.