Introduction UK 5 year survival from lung cancer lags behind other developed nations, and one suggested reason is late stage at diagnosis. Efforts are therefore underway to encourage patients to present earlier to primary care with chest symptoms, and to encourage earlier referral in primary care for CXR or clinic review. Stage at presentation varies between geographical areas, suggesting variability in referral even within UK regions. The reasons underlying this variability are unknown, but one possible explanation could be different patient behaviours or clinical practice in primary care.

Aim To analyse the number of CXRs done 12 months prior to a confirmed diagnosis of lung cancer according to stage at presentation.

Method We reviewed all patients with lung cancer discussed at the Lung MDT between December 2008 and May 2009. Cases of Small Cell Lung Cancer were excluded. The stage (IASLC 6th system), and number of CXRs performed within 12 month prior to diagnosis were recorded. Numbers of CXRs for each group were compared using Kruskal–Wallis test with Dunn's post-test comparison.

Results 223 NSCLC (stage 1 & 2=61, stage 3=61, stage 4=101) and 32 SCLC (limited-12 & extensive-20) were diagnosed. Median, IQR and range of number CXRs in the previous 12 months are displayed in Abstract P159 table 1. Patients presenting with stage IV lung cancer had undergone significantly fewer CXRs in the year prior to diagnosis compared to patients diagnosed with stage I/II lung cancer (p<0.05) and those diagnosed with stage III disease (p<0.05). There was no significant difference between the stage I/II and stage III disease.

Conclusion Patients with late stage disease (Stage IV) appear to have fewer CXRs in the year prior to diagnosis than patients presenting with earlier disease. This may represent a reluctance to seek medical review for persistent respiratory symptoms, or reluctance among GPs to refer for CXR. Ongoing analysis is investigating rates of GP consultation and antibiotic prescription among these patients to try to discriminate between then possible explanations.