Method of referral significantly impacts on the outcome of patients with colorectal cancer

Aims: Patients who are ultimately diagnosed with colorectal cancer are referred to secondary care from different sources and with variable degrees of urgency. Symptomatic referrals can be emergent (E), urgent (U) or in a routine(R) fashion. There are also patients who are incidentally found (I) to have malignancy during investigation of other problems or through screening (S).Those patients referred as emergencies are known to have a higher mortality rate than other patients but little is known about the impact, if any, other referral mechanisms have on mortality. We aim to address this for our unit - in a region where screening has been piloted and rolled out in 2007. Methods: All patients with a newly diagnosed cancer discussed at the colorectal multidisciplinary team meeting from 01/01/2006–31/12/2010 were included.Data on presentation, management and outcome were collected prospectively on a standardised MDT database and supplemented with data from pathology,radiology and operation reports. Analysis was performed using SPSS v 20.0 Results: Of 1,467 patients, 806 (54·2%) were male. Median follow up was 725 days. The number of patients in the I, U, R, E and S groups was 163(10·9%), 599 (40·3%), 263 (17·6%), 191 (12·8%), 251 (16·8%) respectively. Groups differed significantly with respect to mortality (p<0·001), (Kaplan-Meier) Absolute mortality rates within each group were I-36·2%, U-35·4%, R29·3%, E 55% and S 11·6%. Median age of death was different between groups(I 79·1, U 78·8, R 74·9, E 74·7 and S 68·4years)(p<0·001). Conclusions:The all cause mortality rates clearly show that screening prevents colorectal deaths. Symptomatic patients referred routinely are not harmed by the delay in their investigation compared to those referred urgently. The finding that there is a significant difference in age of death of those patients who die, has not previously been described. Whether this is as a result of the disease, the operation or other confounding factors remains unclear.