Symptom attributions in patients with colorectal cancer

BACKGROUND Symptoms of cancer may be interpreted differently by different patients before the diagnosis. This study investigated symptom attributions in Danish patients with colorectal cancer and the potential associations with symptom type, socio-demographic characteristics and patient interval. METHODS Data were collected among incident colorectal cancer patients (n = 577, response rate 64.2 %), who were asked to think back on the time before their diagnosis when completing the questionnaire. The questionnaire comprised a Danish version of the revised Illness Perception Questionnaire (IPQ-R) with questions on 19 symptom attributions. These 19 attribitutions were categorised into five causal groups for statistical analyses. The patient interval (i.e. the time from the patient's first symptom experience to presentation to the healthcare system) was assessed in the same questionnaire. Data on socio-demographic characteristics were obtained by using nationwide registers from Statistics Denmark. RESULTS Patients who experienced 'blood in stool' as the most important symptom were more likely to attribute this to cancer (PR(ad) 1.94, 95 % CI 1.46-2.58) and benign somatic causes (PR(ad) 1.36, 95 % CI 1.05-1.76), such as haemorrhoids, compared to patients who did not perceive this symptom as the most important. Socio-demographic characteristics were also associated with symptom attribution. Patients with higher educational levels were less likely to attribute their most important symptom to psychological causes (PR(ad) 0.57, 95 % CI 0.34-0.96) than patients with lower educational levels. Patients with rectal cancer attributed their most important symptom to a benign somatic cause more often than patients with colon cancer (PR(ad) 1.34, 95 % CI 1.02-1.77). [corrected]. CONCLUSIONS Symptom attribution in patients was associated with aspects of socio-demography and with the symptom type perceived by the patient as the most important. No significant associations were found between symptom attributions and patient interval. These results have implications for general practice as symptom attributions may prompt patients to present symptoms in a certain way and thereby influence the general practitioner's assessment of presented symptoms.