Introduction The All Wales guidelines for management of dyspepsia (2004) advises urgent endoscopy for those patients who have ‘alarm features’, and routine investigation in patients aged over 55 years with unexplained and persistent dyspepsia of recent onset. However there is concern that strict adherence to this guideline may lead to a delay in diagnosis, with early curable cancers not being discovered until alarm features develop and progression to an advanced stage has occurred. The aim of this study is to evaluate the timescale of symptoms in upper gastrointestinal (GI) cancer – to determine if patients experience dyspepsia prior to developing their ‘alarm features’, and hence whether the current guidelines delay diagnosis.

Methods This prospective study included patients diagnosed with upper GI cancer at endoscopy between May 2004 and January 2007. They were questioned directly after the procedure about the nature and duration of their presenting symptoms, and the number of primary care reviews prior to referral. Patient outcome data has been collected until September 2010.

Results 60 patients diagnosed with upper GI cancer were included. There were 43 males and 17 females; mean age was 77 years (range 44-92); 36 had oesophageal cancers and 24 gastric cancers; median survival was 284 days (range 8-1998). At referral to endoscopy 56 patients had alarm features, 1 patient had dyspepsia alone, 1 patient presented with chest pain, and 2 patients whose endoscopy referrals were due to incidental findings on CT scans had no upper GI symptoms or alarm features. Of the 56 patients with alarm features 17 reported co-existing dyspepsia; in 3 of these the dyspepsia had been present for many years and clearly pre-dated the cancer, in 4 cases dyspepsia had shortly preceded the alarm feature – the maximum preceding duration being only 8 weeks, and in 10 patients the dyspepsia started at the same time or after development of an alarm feature. None of the 5 patients who were younger than 55 years old complained of dyspepsia. Overall, 20 patients had more than one primary care review prior to their referral for endoscopy. When this group was compared to the 38 patients who were referred at their first primary care review, there was a significant delay in referral (p<0.001) but there was no significant difference in tumour stage at diagnosis (p=0.771) or length of survival (p=0.120).

Conclusion Patients with upper GI cancer who are referred to endoscopy with alarm features rarely report a significant period of preceding dyspepsia that would have allowed earlier diagnosis. Early upper GI cancer is largely asymptomatic and population screening is the only reliable method to detect these cases. The data from this study supports the use of current guidelines to investigate dyspepsia.