Commentary on McCarthy et al. (2015): Ecological momentary assessment--Reactivity? Intervention?

Abstract

McCarthy et al. find modest reactivity effects from Ecological Momentary Assessment (EMA) monitoring, diminishing hopes for monitoring as an intervention, while raising theoretical issues about the mechanisms of reactivity and about the effect of conscious awareness on craving. In any case, the findings do not undermine the utility of EMA for process research. McCarthy et al. 1 randomized smokers trying to quit smoking to intensive monitoring of their experiences via EMA 2, 3 several times a day versus just once. Intensive assessment did not affect smoking cessation outcomes, but was associated with lower reports craving, anxiety, anger, positive affect and hunger. Because they have particular utility for studying drug use and addiction 4, EMA methods are used increasingly in addiction science, making it useful to assess reactivity, or how the assessments might affect the very behavior being studied. It is both reassuring and disappointing that EMA monitoring did not affect smoking cessation outcomes: reassuring, because it suggests that we can continue to use EMA to study smoking cessation processes without being concerned that the measures will somehow distort the outcomes. It is disappointing because we are always eager to identify interventions that help people to quit… and EMA monitoring is apparently not one. The contrast between the worry and the hope is striking, particularly for us clinical researchers: as research methodologists, we are concerned about anything that might unintentionally disturb natural behavior, and every hypothesized influence seems large and powerful. As clinicians we learn again and again, to our dismay, that almost nothing we do seems to change addictive behavior outcomes. Clearly, in this instance, our clinical pessimism is validated. A priori, it is not clear that intensive assessment would improve, rather than degrade, clinical outcomes. While self-monitoring may provide a degree of self-awareness that might support successful self-management 5, intensive assessment also imposes behavioral burdens on the participants—burdens that could hypothetically undermine participants’ success by draining the scarce and limited resource of attention and self-control 6. It is reassuring that this was not observed here. Nevertheless, we should be mindful of the burden of assessment, and even the burden of intervention. Technological and methodological developments are feeding an explosion of interest in Ecological Momentary Interventions (EMI) that can deliver interventions in individuals’ daily lives, ideally just when they are needed. However, because it is difficult to actually know when they are needed, because there seems little cost to delivering an intervention when it only involves moving electrons around, and because it seems intuitive that ‘more is better’, there is a risk of adding more burden than help. Especially in light of this concern about assessment response burden, it may seem surprising that the effects of monitoring seemed mostly favorable, rather than harmful. Perhaps frequently evaluating their feelings helped participants to gain some perspective on them, and allowed them to rise above them. Perhaps monitoring their craving implicitly taught smokers ‘urge surfing’ 7—a mindful approach to craving involving simply observing craving without over-reacting to it. If this were the case, however, one would expect EMA assessment to improve pain, which it does not 8. It is hard to know what processes were evoked by the frequent self-assessment, but worth exploring; there may yet be a kernel of an intervention lurking in the findings. The reduction in craving is surprising from another perspective. Tiffany 9 has argued that craving represents the experience of becoming aware of drug-seeking processes that are normally automatic and out of awareness. On this model, asking about craving, thus forcefully driving awareness, should actually provoke craving 10, yet the results here were just the opposite. The theory may need some adjusting. How troubling should it be if intensive EMA assessment has small effects on craving and mood? Not very troubling, in my view. EMA methods are seldom used to assess mean levels of mood or craving, which might be affected by reactivity. Almost always, EMA methods are used to understand relationships among variables, rather than their mean levels 11, and there is no reason to believe such relationships are changed by the effects documented in this paper. Nevertheless, investigators should be sensitive to potential reactivity effects in designing EMA protocols and in interpreting EMA data. S.S. has a small financial interest in eRT, which provides electronic diary services for research.