How and for whom? Asking questions about the utility of psychosocial interventions for individuals diagnosed with cancer.

A recent review revealed 155 randomized controlled trials of psychological interventions for cancer patients published through 1998, and many additional trials have been published since that time. Even a quick examination of the literature on the effects of psychosocial interventions on outcomes related to quality of life, such as psychological adjustment and fatigue in cancer patients, reveals that some randomized controlled trials are impressive in their positive outcomes, some produce mixed results, and some yield null findings. In recent qualitative and quantitative meta-analytic reviews of the relevant body of intervention research, the reviewers draw disparate conclusions, suggesting that the existing evidence demonstrates the utility of psychological interventions, yields a mixed picture, or supports no strong recommendations for their effectiveness. Just as the intervention studies included within the reviews differ dramatically on a number of dimensions, so do the reviews themselves on such factors as criteria for inclusion, outcomes examined, and approaches to the data. In this issue of the Journal of Clinical Oncology, Chan et al provide an example of a randomized controlled trial that yielded negative effects. Rather than continuing to ask the omnibus question of whether psychological interventions are effective for cancer patients and chalking up the trial by Chan et al as one for the “loss” column, posing questions that address the issues of how and for whom specific interventions do or do not carry positive effects might better illuminate directions for this important domain of psychosocial oncology. Asking how an intervention produces its effects involves an exploration of mechanisms or explanations for a trial’s findings. Characterizing the pathways through which an intervention produces positive effects can shed light on how the intervention might be refined further, and how seemingly distinct treatments might produce similar effects through common mechanisms. Asking how an intervention might have yielded disappointing findings also can be revealing, both in providing information on how much weight should be accorded to the findings and in helping researchers learn which paths not to pursue in developing interventions in the future. For both positive and negative trials, explanations for results are likely to lie in such factors as the nature of the sample studied, the intervention implemented, and the outcomes examined. For example, Chan et al acknowledge the high baseline functioning of the patient sample as a potential explanation for the lack of significant effects. If a sample of cancer patients reports low distress and otherwise high functioning at the outset of a trial, then it is statistically impossible for a significant intervention effect to emerge on those outcomes, at least without a very large sample. A related issue, particularly for trials that target patients during or shortly after medical treatment, is that many individuals tend to recover over time without psychological intervention, as also occurred in the trial by Chan et al, again working against the likelihood of detecting significant intervention effects over a standard care control. With regard to the nature of the intervention, it is important to note that the intervention by Chan et al was multimodal, encompassing such multiple core components as psychoeducation, supportive care, stress management, relaxation, and pain management. Furthermore, the intervention was individually tailored to each participant, with the specific content left to the discretion of the treating psychologist. Although many published psychological interventions also are multimodal and some include individual tailoring, the reader is left with a number of questions JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 23 NUMBER 22 AUGUST 1 2005