The Food and Drug Administration (FDA) has long regulated the production, distribution and advertising of tobacco products such as cigarettes, roll-your-own tobacco, and smokeless tobacco under authority of the Tobacco Control Act. However, a new proposed rule would deem additional products, including electronic cigarettes and possibly cigars, to be subject to requirements such as ingredient listing, pre-market clearance, free sampling prohibition, minimum age requirement, and limit on sales by vending machines, as well as required health and addiction warning statements.
In general, regulations rely on indirect causal mechanisms to achieve their goals. From an evaluative standpoint, this makes them complex interventions, as processes of change are often non-linear; they rely on behavior change through causal mechanisms such as: deterrence, prohibition, information, etc. In the particular context of this FDA regulation another set of complex factors comes into play as there are limited data on the actual impact of the novel tobacco products on health outcomes and addiction patterns. Naturally, there are many unknowns in the proposal, in particular, a lack of evidence specifically about differentiated public health impacts of various tobacco products and of baseline usage patterns and risks. Given the uncertainty and inherent complexity of a regulation such as this one, it is paramount to ensure that the rule is written in a way that allows ex post feedback on whether intended goals have been reached.
In a public comment filed with FDA, I assess whether the proposed rule, as presently written, meets the minimum preconditions for retrospective review, as required by executive order: (i) clear assessment of the problems at stake; (ii) transparent logic of intervention; (iii) measurable metrics to track progress; (iv) commitment to collect performance information; (v) reasonable timeframe; and (vi) distinction between outputs and outcomes. Regarding the first point, while the proposed rule clearly identifies the multifaceted problems it is intended to address, FDA is unable to provide the public with a clear estimate of the magnitude of the public health risks that these unregulated tobacco products represent, partly due to a lack of baseline data. More importantly, FDA does not fully and coherently explain how the proposed rule intends to solve the complex issues at stake, nor does it transparently lay out the assumptions underlying its regulatory logic. Third, the rule does not contain any measurable metrics that FDA intends to monitor over the medium to long term, nor does it set up a timeline for assessing whether the rule's objectives have been met. Finally, while the agency lists a number of expected benefits, it does not systematically distinguish between output and outcome levels.
We thus invite the agency to take three key principles into account in the final version of the proposed rule.
First, before adjudicating between the two proposed options, FDA should lay out clearly the overarching goals sought out by the regulation as well as the assumptions underlying the intervention logic. The two co-proposals presented in the proposed rule are motivated by two different logics. On the one hand, option 1 (include all tobacco products in its regulatory framework) applies a precautionary approach to rule-making and, in the absence of full scientific certainty about the risks incurred by novel tobacco products, would treat as potentially dangerous all products that contain nicotine. Option 2, on the other hand, by exempting ‘premium cigars’ from coverage and additional access provisions, relies on the logic of a ‘continuum of nicotine delivery.’ Under this latter option, some products are considered more dangerous than others and therefore exempt from regulation. The FDA decision of exempting premium cigars from the regulatory framework of the FD&C’s chapter IX does not seem fully evidence-based given that recent research seems to indicate that e-cigarettes are also relatively more innocuous than other tobacco products. The agency should also seek to strengthen its baseline data, notably with regards to the magnitude of the public health challenge posed by the various novel tobacco products.
Second, FDA should add language to its final rule committing to measure efficacy at two-year intervals following implementation of the rule, measured notably as percent reductions in use of the various tobacco products subject to coverage. This information will tell both the agency and the public how valid its theory of change was, and will provide information for future rulemakings on how to tailor regulatory provisions to achieve desired outcomes. In addition, retrospective review efforts may be able to provide information on whether the exemption of some products (potentially premium cigars and e-cigarettes) was appropriate for maximizing net benefits.
Third, while the agency rightfully acknowledges the relatively weak evidence about the health risks linked to many of the currently unregulated products, it shows only a half-hearted commitment to bridging this knowledge gap. In order to better address the uncertainty surrounding this rule-making, the agency should take three commitments: (i) Transparently expose the theory of change underlying the various provisions as well as the causal mechanisms which need to be activated for the rule to be successful; (ii) Clearly delineate the various scenarios and their expected impact; and (iii) Commit to systematically collect information on the implementation of the various provisions, their outputs and medium-term outcomes. Moreover, in regulations attempting to trigger behavior change amongst addicted people, causal processes are bound to be complex, with non-linear and non-proportional effects. Consequently, the agency should commit to assessing potential unintended consequences of its rule making as an inherent part of the retrospective review’s scope.
Estelle Raimondo is a PhD Candidate in Public Policy and Evaluation Research.