PURPOSE: Identification of supportive care needs in patients with cancer is essential for planning appropriate interventions. We aimed to determine patient-physician concordance in perceived supportive care needs in cancer care and to explore the predictors and potential consequences of patient-physician concordance. PATIENTS AND METHODS: A national, multicenter, cross-sectional survey of patient-physician dyads was performed, and 97 oncologists (participation rate, 86.5%) and 495 patients (participation rate, 87.4%) were included. A short form of the Comprehensive Needs Assessment Tool for Cancer Patients was independently administered to patients and their oncologists. Concordance and agreement rates between physicians and patients were calculated. Mixed logistic regression was used to identify predictors of concordance and to explore the association of concordance with patient satisfaction and trust in physicians. RESULTS: Physicians systematically underestimated patient needs and patient-physician concordance was generally poor, with weighted κ statistics ranging from 0.04 to 0.15 for individual items and Spearman's ρ coefficients ranging from 0.11 to 0.21 for questionnaire domains. Length of experience as oncologist was the only significant predictor of concordance (adjusted odds ratio for overall concordance [aOR] = 2.09; 95% CI, 1.02 to 4.31). Concordance was not significantly associated with overall patient satisfaction (aOR = 1.24; 95% CI, 0.74 to 2.07) or trust in physician (aOR = 1.17; 95% CI, 0.76 to 1.81). CONCLUSION: Our findings revealed significant underestimation of patient needs and poor concordance between patients and physicians in assessing perceived needs of supportive care. The clinical implications of this discordance warrant further investigation.
PURPOSE: Identification of supportive care needs in patients with cancer is essential for planning appropriate interventions. We aimed to determine patient-physician concordance in perceived supportive care needs in cancer care and to explore the predictors and potential consequences of patient-physician concordance. PATIENTS AND METHODS: A national, multicenter, cross-sectional survey of patient-physician dyads was performed, and 97 oncologists (participation rate, 86.5%) and 495 patients (participation rate, 87.4%) were included. A short form of the Comprehensive Needs Assessment Tool for CancerPatients was independently administered to patients and their oncologists. Concordance and agreement rates between physicians and patients were calculated. Mixed logistic regression was used to identify predictors of concordance and to explore the association of concordance with patient satisfaction and trust in physicians. RESULTS: Physicians systematically underestimated patient needs and patient-physician concordance was generally poor, with weighted κ statistics ranging from 0.04 to 0.15 for individual items and Spearman's ρ coefficients ranging from 0.11 to 0.21 for questionnaire domains. Length of experience as oncologist was the only significant predictor of concordance (adjusted odds ratio for overall concordance [aOR] = 2.09; 95% CI, 1.02 to 4.31). Concordance was not significantly associated with overall patient satisfaction (aOR = 1.24; 95% CI, 0.74 to 2.07) or trust in physician (aOR = 1.17; 95% CI, 0.76 to 1.81). CONCLUSION: Our findings revealed significant underestimation of patient needs and poor concordance between patients and physicians in assessing perceived needs of supportive care. The clinical implications of this discordance warrant further investigation.