Ward, S.E., Serlin, R.C., Donovan, H.S., Ameringer, S.W., Hughes, S., Pe-Romashko, K., & Wang, K.K. (2009). A randomized trial of a representational intervention for cancer pain: Does targeting the dyad make a difference? Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association, 28(5), 588–597.

To test the effectiveness of the RIDcancerPain program at overcoming attitudinal barriers to pain management among patients; to determine if the program is more effective when delivered to a patient with a significant other present than to a patient without a significant other present

Patients were blocked by setting (teaching vs. nonteaching facility) and then randomized to one of three study conditions: the dyad, in which both the patient and a significant other received the educational intervention; the solo condition, in which the patient received the intervention without a significant other present; and the care-as-usual condition. The RIDcancerPain intervention consisted of seven elements:

1. The patient was asked to describe beliefs regarding the cause, timeline, consequences, cure, and control of cancer pain.
2. The intervener addressed gaps, confusion, and misconceptions about using analgesics and reporting pain. Those present discussed the issues.
3. The patient discussed limitations and losses that result from the misconceptions.
4. The intervener provided information to fill gaps and eliminate confusion, using scripted messages about common attitudinal barriers.
5. The intervener summarized the benefits of adopting recommendations derived from the information.
6. The patient created a plan for changing the way he or she managed pain.
7. During a follow-up telephone consult 2–4 weeks after the first session, the intervener evaluated and revised the coping plan.

In the solo condition, the significant other was asked to leave the room during the intervention. In the dyad condition, the intervention was provided to both individuals. In the care-as-usual condition, project staff answered the questions that participant pairs had about the project. Measurement was done at baseline (T1), at five weeks (T2), and at nine weeks (T3).

• The sample was composed of 161 patients.
• Mean patient age was 58.54 years; age range was 20–85 years. Mean age of significant others was 55.58 years; age range was 18–85 years.
• Of all patients, 60.2% were female and 39.8% were male.
• In the dyad group, lung and breast cancers were the most frequent diagnoses; in the solo group, lung, breast, and gastrointestinal cancers were the most frequent; in the control group, gastrointestinal cancer was the most common. Of all patients, approximately 90% were Caucasian. More than 50% of patients had an annual income below $50,000. In the majority of cases, the significant other was the patient's spouse.

• Multisite
• Outpatient
• Midwest, United States

Randomized single-blind controlled trial

• Barriers Questionnaire, to measure attitudes about analgesic use
• Composite score of five items, on a five-point Likert scale, to measure pain severity (Items included amount of time in pain and usual, worse, least, and current pain intensity.)
• Brief Pain Inventory (Short Form), to measure pain-related interference with activities
• Functional Assessment of Cancer Therapy-General (FACT-G)
• European Organization for Research and Treatment quality of life questionnaire QLQ-C30, sections to measure pain relief (0 = no relief, 4 = complete relief)

• Before T2, 37 pairs dropped out of the study. Only 124 pairs completed the study through T2. Comparison of dropouts to others showed a significant difference in pain relief ratings (p = 0.02): Those who dropped out had higher pain relief scores.
• At T2 there were no differences in attitudinal barriers between the two intervention groups.
• Pain relief change was higher in the dyad condition than in the solo condition (p = 0.008). However, neither group differed significantly from the control group in regard to this outcome.
• There were no significant differences across study groups in barriers at five weeks. Analysis of variance demonstrated no main effects of the intervention over time. However, authors did note a significant intervention-mediator effect, between baseline and nine weeks, in regard to changes in barrier scores and pain relief (p = 0.026), patient interference (p = 0.003), and negative mood (p = 0.033).

This intervention did not result in a significant effect regarding pain management or barrier reduction. Whether the patient received the intervention with a significant other present did not seem to affect the result of the intervention.

• The study had a limitation due to no appropriate control group and no attentional control for the care-as-usual group.
• The study had a large number of dropouts before completion of measures at nine weeks, so impacts over a long term could not be evaluated.
• Authors did not examine changes in analgesic medications. Changes in regimen may have affected results.
• Authors did not provide details about the care-as-usual situation. The report did not cite the amount of education and counseling this group received.

Findings suggest that the authors' hypothesis—that RIDcancerPain intervention changes attitude, which in turn affects pain outcomes—oversimplifies pain management.