Clinician versus nurse symptom reporting using the National Cancer Institute--Common Terminology Criteria for Adverse Events during chemotherapy: results of a comparison based on patient's self-reported questionnaire.

Background: Monitoring adverse events during chemotherapy by clinicians is a standard practice but clinicians mayreport fewer side-effects or lower symptom severity than patients. Our aim was to compare symptoms self-reported bypatients with symptoms registered by clinicians and nurses, to assess validity of a nurse reporting.Methods: From April to August 2007, a double-blind questionnaire with 13 common items graduated according tothe National Cancer Institute’s Common Terminology Criteria for Adverse Events was completed by clinicians andnurses for outpatients undergoing chemotherapy at our Medical Oncology Day Hospital Unit. Patients completeda modified questionnaire with simplified terms. They were requested to specify seriousness of symptoms witha subjective scale varying from 1 to 4. Every patient–nurse–clinician questionnaire was registered for the statisticalanalysis. Agreement was evaluated by Cohen’s kappa coefficient.Results: Eighty-five paired questionnaires were completed. Patients, nurses and clinicians agreed on mostsymptoms and toxicity grade. Agreements between patients and nurses were stronger than those between patientsand physicians for the six most common symptoms: asthenia (kappa 75% versus 37%), constipation (83% versus45%), neuropathy (82% versus 55%), mucositis (78% versus 46%) and diarrhoea (90% versus 77%). Thesedifferences mainly reflected differences in the proportion of positive agreement: nurses were more able to detectsymptoms self-reported by patients than physicians. The only exception was nausea, as kappa coefficient was verygood for both health professionals (91% versus 89%). When considering the different grade of toxicity by the weightedkappa coefficient, we scored again the highest agreement between patient and nurse, with weighted kappa rangingfrom 55% (asthenia) to 86% (diarrhoea), and the lowest agreement between patient and physician, with weightedkappa ranging from 32% (asthenia) to 74% (nausea). The agreement between physician and nurse slightly improved,with weighted kappa ranging from 41% (constipation) to 77% (nausea).Conclusion: Our results support the validity of nurse toxicity reporting and that the nurse staff could be successfullyemployed in collecting toxicity data because of a greater ability to elicit information from patients than the medical staff