Agreement between patient-reported and doctor-reported patient intervals and date of first symptom presentation in cancer diagnosis - a population-based questionnaire study

Abstract

Background: The concept of delay in cancer diagnosis has been a scientific issue for decades, and there is still no standardised and validated way to measure the time intervals. One of the intervals that are difficult to measure is the patient interval (i.e. the period from the patient's first symptom until the first presentation to the health care system) because dates of symptom onset and first presentation are difficult to establish precisely. Further, since patients may have another experience of the diagnostic pathway than e.g. the general practitioner (GP), a reasonable question remains whether patients and GPs agree on these important milestones. The objective of this study was to analyse the agreement between patient-reported and GP-reported patient intervals and date of first presentation of cancer-related symptom(s) to the GP.

Methods: On the basis of a cohort study, we included incident cancer patients from the former Aarhus County from 1 September, 2004 to 31 August, 2005. Both patients and GPs reported the length of the patient interval and the date of the first presentation to the GP with a cancer-related symptom measured by self-administered questionnaires. Agreement was measured using agreement-survival plots and Lin's concordance correlation coefficient (CCC).

Results: There was full agreement between GP- and patient-reported patient intervals in 21.0% of all the cancer cases. In 50.1% of cases, patients and GPs agreed about the patient interval within a margin of one month. There was full agreement between GP- and patient-reported date of first presentation in 37.5% of the cancer cases and within one week in 52.0% of all the cancer cases. Overall, the agreement on the length of the patient interval was poor (CCC=0.513), but better for patients presenting with alarm symptoms. The agreement was moderate between GP- and patient-reported dates of first presentation (CCC=0.924).

Conclusion: We found that GPs systematically reported a longer patient interval than patients did. We found moderate agreement on reported date of first presentation of symptoms to the GP, meaning that the disagreement in reported patient interval is related to date of first symptom rather than date of first presentation to the GP.

Keywords: Early detection of cancer; Epidemiology; Research design.