The use of next-of-kin to estimate pain in cancer patients.

The clinical management of pain could be improved if more were known about the intensity and duration of the pain. The cancer patient, however, is often unable to communicate this information to caregivers. Relying on next-of-kin to provide information about the patient's pain could help, but little has been done to verify next-of-kin responses with respect to subjective experiences. For the present study 42 pairs of cancer patients and their next-of-kin were independently surveyed in 1982 in Washington state regarding their cancer pain and various aspects of medical treatment to determine whether proxies can give reliable responses. Close agreement between subject and next-of-kin was observed for the items which were salient and had a limited choice of responses, such as the presence of pain. Agreement in the aggregate was achieved for the items having several possible responses, such as the intensity and frequency of pain. Correspondence was virtually random for those items which had a variety of listed responses. The use of proxies in obtaining information, future research directions, and difficulties with measuring agreement are discussed.