Impact of information sources on expectation of cancer treatment side effects: A URCC CCOP study

Abstract

8094 Background: Cancer patients can expect many side effects from their chemotherapy or radiation treatment. This study investigated possible causes of these expectations. Methods: 1015 Cancer patients from 17 Community Clinical Oncology Program (CCOP) institutions affiliated with the University of Rochester CCOP research base were surveyed for symptoms they expected from treatment (given 12 common symptoms); the symptoms they were already experiencing; their age, education, disease site and gender; Karnofsky score; self-rated general health; and which of 25 sources of cancer information grouped into 3 broad areas of medical (10), community (7) and media (8) they have used. Results: 3 significant predictors (p<.05) of the number of side effects expected by the patients emerged from a stepwise multiple regression: the number of side effects the patient was already experiencing (Δr2=0.14), patient age (Δr2=0.05) and the number of information sources used (Δr2=0.03). Only the cancer care physician, the PCP (primary care physician) and newspapers had a negative correlation with individual side effects (PCP and newspapers related to a lower expectancy of hair loss, p<0.05) all other information sources had none or a positive correlation with the expectancy of a side effect. The total number of expected side effects was most highly related (all p<0.01) to using the Internet (r=0.20) the American Cancer Society (r=0.19) and the NCI (National Cancer Institute, r=0.18) as sources of treatment information. Use of information sources from the NCI was positively associated with expecting every side effect (9 with p<0.01, 3 with p<0.05). Conclusions: Information sources can have an impact on the expectancy of side effects by people undergoing cancer treatment. Since this expectancy may influence the actual experience of the patient during the treatment, knowing how many and which sources of information patients have used may promote optimal management of treatment related symptoms. Supported, in part, by a supplement from the Division of Cancer Control and Population Sciences, NCI, to Public Health Service grant U10 CA37420 and by grant RSG-01-071-PBP from the ACS. No significant financial relationships to disclose.