Background:: Cancer pain is highly prevalent and commonly undertreated. This study was designed to determine whether dissemination of a clinical protocol for pain management would improve outcomes in community oncology practices.

Patients and methods:: A pain management protocol was developed based on accepted guidelines. After baseline assessment, oncology practices were randomly assigned to ‘analgesic protocol’ (AP) sites, where oncologists implemented the guidelines in a group of lung or prostate cancer patients, or to ‘physician discretion’ (PD) sites, where customary treatment was continued. Patients treated on protocol and a comparison group of patients with pain due to breast cancer or myeloma were monitored for change in pain using the Brief Pain Inventory, and for change in other symptoms or mood.

Results:: The protocol terminated early because of poor accrual. We compared groups using proportions of patients who had no or mild pain at follow-up. Although measures of protocol adherence did not suggest the occurrence of major practice change, the proportion of lung or prostate cancer patients with no or mild pain increased significantly from baseline for those treated at AP sites compared with those treated at PD sites. There was no significant difference between the breast and myeloma patients treated at AP sites versus those treated at PD sites.

Conclusion:: A protocol for cancer pain management can improve pain control. Diffusion of these benefits to other patients was not confirmed. Given the small sample size, these findings require confirmation in a larger trial.