To Help People Quit, Sometimes Less is More

Baker, Bolt and Smith
From left: Baker, Bolt, and Smith published a paper that makes a significant contribution to science. It includes findings from studies that are the result of a lot of hard work by many UW-CTRI staff, among other researchers.

It appears that piling on multiple treatments to help people to quit tobacco use often produces diminishing returns. The findings, published in the journal Clinical Psychological Science, were discovered by UW-CTRI Research Director Dr. Tim Baker, UW-CTRI Director of Biostatistical Operations Dr. Stevens Smith and Professor of Educational Psychology Dr. Dan Bolt. They came to this conclusion when they examined results from three factorial experiments: the BREATHE 1 Motivation Study, the BREATHE 1 Maintenance Study, and an older population-based UW-CTRI study involving SmokeFree.gov.

They found a pattern of negative interactions, such that adding an intervention component to other interventions resulted in the whole being less than the sum of the parts. That is, as more components were combined, on average, the effectiveness of the individual components declined. This effect could be so strong as to eliminate any benefit from the added components.

These studies were conducted as part of the Multiphase Optimization Strategy (MOST) treatment development strategy in partnership with Dr. Linda Collins and others. Researchers designed the screening experiments to identify intervention components that would form an especially effective integrated, multicomponent treatment that could later be evaluated in subsequent research.

“We found a consistent pattern of effects across multiple factorial experiments: multiple negative interactions and few positive or synergistic interactions,” Baker said. “That is, we found that combining intervention components tends to result in decreases in the effects of the components.”

For example, there were no main effects in the motivation study but there were numerous significant interactions, all of them negative.

For the new paper in Clinical Psychological Science, the authors sought to examine additional evidence that combinations of smoking intervention components tend to produce negative interactions. They also used more informative analytic methods that yield more accurate estimates of the cumulative effects of such interactions. Finally, they explored hypotheses regarding such interactions.

The authors tested several hypotheses about why adding intervention components together reduces their effectiveness. For instance, more intense treatment could impose burdens that affect both the patient and clinical staff, which could then affect treatment adherence, attrition, and delivery fidelity. For instance, a recent study showed that 24 weeks of nicotine patch treatment produced higher quit rates than 52 weeks of treatment. It may be that 52 weeks of medication use was so burdensome that it led to disuse of important treatment elements.

They also addressed the possibility that different treatment components were not really different in terms of how they affected change mechanisms; they were, in fact, highly redundant with one another and therefore did not contribute anything new. Finally, they examined the idea that some persons will change readily with little treatment but others are so resistant to change that adding intervention components does not help. Thus, one or two intervention components can benefit those who are easily helped but more components do not help those who are treatment resistant.

“The results were mixed, but not promising for combining most intervention components,” Baker said. “While combining a couple of components—such as nicotine replacement therapy plus counseling—can work well, combinations of greater numbers of components pretty consistently eroded the effectiveness of the combined components.”

Moreover, these negative interactions may suppress estimates of component main effects when multiple components are used together. That means that, if a component is effective when used by itself, when it is tested along with many other components in the same study, the other components can hide its effectiveness.

The researchers hope this research informs the development of more effective treatments. It may, for instance, help researchers screen intervention components to eliminate those that produce strong negative interactions when used with other components. This research also suggests that programs might focus on treating more people who smoke with less treatment rather than fewer people who smoke with more treatment.