Commentary

Cancer diagnosis — the role of urgent referral guidelines

This excellent review brings together evidence from primary care that will undoubtedly help clinicians in their quest for earlier cancer diagnoses for their patients.1 It is particularly welcome within the current context of the Department of Health's National Awareness and Early Diagnosis Initiative.2 Recent policy announcements have focused on easier and quicker access to investigations for GPs for suspected cancer symptoms, and the development and implementation of software for GPs to identify early cancer symptoms within 5 years. Such policies are likely to help save the estimated 5000–10 000 lives per year currently lost in the UK because of probable later diagnosis compared with some European countries.3^–5 Data from primary care, such as presented in Hamilton's review, will directly inform the development of these policies.

One key policy question is whether urgent referral guidelines are the answer, whatever the quality and quantity of the evidence about the meaning of symptoms. There is a continuing lack of evidence that cancer diagnoses overall are made quicker, and with clinical benefits, through a fast-track system that prioritises some patients over others, based upon their symptoms. Many, quite possibly a majority, of cancers are diagnosed through a more atypical symptomatic presentation, often congruent with symptoms of benign pathology. This is a major challenge for secondary care and for commissioners. Paradoxically, guidelines may prioritise those with least to gain in terms of improved outcomes because they hasten diagnosis of very early tumours that may remain curable and treatable if diagnosed later. They also prioritise those with already established aggressive disease who are beyond curative treatment at the time of diagnosis, although good palliative care for these people may be hastened.6 All patients deserve the earliest diagnosis, whatever their symptoms. Guidelines do focus GPs'attention on potential alarm symptoms for cancer.7 However, there are also still significant concerns about the implementation and use of the urgent cancer referral guidelines in primary care, which is hugely different to their target driven implementation in secondary care.

In policy terms, there are two main conclusions to be drawn. Firstly, Hamilton is correct in stating that GPs may have too high a threshold for investigation. Gatekeeping has been the cornerstone of the UK NHS for decades. However, for suspected cancer, gatekeeping may be inhibiting early diagnosis and it may be that the balance between referral for opinion or investigation, and minimising costs and the risk of iatrogenic illness needs to be shifted. Watching and waiting is no longer acceptable for certain symptom complexes. Secondly, we need faster routes to diagnostic tests and/or specialist opinion (dependent upon the available diagnostic tests for different cancers) for all patients with a suspicious symptom, above a certain, and as yet undetermined, threshold. A stronger evidence base to enhance our understanding of the predictive value of symptoms will facilitate this.