ABSTRACT
BACKGROUND: Clinical practice shows that certain patients may underestimate and others overestimate some initial symptoms of their disease. In studies on the interval between first symptoms and treatment onset, estimating the date in which symptoms first appeared is crucial. The study analyzed patient-physician agreement in assessing first symptom attributable to cancer. METHODS: During two years, two physicians personally interviewed, through a structured questionnaire, all symptomatic patients with a neoplasm of the digestive tract admitted to Hospital del Mar (Barcelona, Spain). Patients had a mid-low sociocultural profile and most had been admitted through the Emergency Department. RESULTS: Absolute agreement (symptom and date) occurred in 85% of the 183 subjects. In most discordant cases, patients had overlooked some component of the "toxic syndrome", and the date of symptom onset was, based on physician's assessment, chronologically prior to the date elicited from the patient. Disagreement was directly related to the patient's health status (p < 0.05) and to the number of reported symptoms until hospital admittance (p = 0.002), but not to tumour stage. Agreement increased with the importance attributed by the patient to the first symptom (p < 0.05). CONCLUSIONS: In spite of difficulties inherent to measuring symptomatic onset of diseases, structured patient interviews appear to be a reasonably valid method and deserve further development in this and other areas of research.
