Fast track referral for cancer

BMJ 2007; 335 doi: http://dx.doi.org/10.1136/bmj.39293.453090.AD (Published 09 August 2007) Cite this as: BMJ 2007;335:267
  1. Moyez Jiwa, professor of primary care1,
  2. Christobel Saunders, professor of surgical oncology2
  1. 1WA Centre for Cancer and Palliative Care, Health Research Campus, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia
  2. 2School of Surgery and Pathology, QEII Medical Centre, Perth, WA 6009
  1. Correspondence to: M Jiwa m.jiwa{at}curtin.edu.au

    Has not improved patient outcomes in the UK

    Most patients with cancer who are not diagnosed by screening will be diagnosed after symptoms develop, so timely referral of patients to specialists is essential. In this week's BMJ Potter and colleagues assess the long term impact of the “two week wait” rule in the United Kingdom for breast cancer on referral patterns, diagnoses of cancer, and waiting times.1 Their study found that the diagnosis of cancer in people referred within two weeks significantly decreased in the period 1999-2005 (12.8% v 7.7%, P<0.001) and diagnoses in people referred through the routine route increased (2.5% v 5.3%, P<0.001). The study suggests that the current cancer referral policy in the UK—whereby patients with a given set of symptoms are seen within two weeks—results in more patients who have cancer being seen on routine waiting lists than on the fast track list. This means diagnosis is delayed even further. Why is this so, and how can it be rectified?

    In formulating such rules, it is important to remember that waiting times for urgent appointments are only a surrogate end point. The main aim is to improve cancer survival, improve psychosocial outcomes for people with cancer and those who turn out not to have cancer, and to improve the general practitioner's ability to diagnose cancer. This rule seems not to have led to any of these outcomes, although it has perhaps increased the proportion of patients with benign pathology referred for urgent specialist appointments.

    Introduction of the two week standard clinics has not improved the outcomes for patients in some of the commonest cancers.2 The reason for this may be the complexity of the medical consultation. Many factors affect the decision to refer for an expert opinion, including a patient's help seeking behaviour, doctor-patient communication, eliciting and interpreting signs and symptoms, applying evidence to decision making, negotiation with the patient about the need for and most appropriate route of referral, and conveying the information in sufficient detail to allow the patient to be fully informed about the need for urgency or otherwise. It has been calculated that if the practitioner successfully negotiates each of the above stages on 80% of occasions then only a small percentage of decisions will be evidence based.3

    A substantial proportion of patients with common cancers present as emergencies with advanced disease; in the case of colorectal cancer this has been estimated to be as high as 20%.4 This is especially true in deprived communities or in communities where people have cultural and linguistic differences.

    A major step towards earlier diagnosis of cancer would be to raise awareness of the symptoms of cancer in the community. Furthermore, given that cancer is an uncommon diagnosis in general practice, practitioners are unlikely in most cases to opt to investigate symptomatic patients.5 Therefore patients with altered bowel habit may not be tested for iron deficiency anaemia or other signs of occult bleeding as symptoms are far more likely to be due to irritable bowel syndrome or diverticular disease than to colorectal cancer.6

    Bayes's theorem demonstrates that the probability that a patient has cancer is affected by the prevalence. The prevalence of cancer in a primary care population depends on the symptoms in question. Such considerations should influence the selection of patients for referral.7 In practice, however, doctors will act on the basis of personal experience, respected local opinion, and anecdotal evidence rather than on high quality published research.8 Research about the positive predictive value of signs and symptoms of the common cancers in primary care does exist. Implementing these findings in practice, however, will require substantial effort.9 10

    General practitioners must also be able to persuade patients with suspicious symptoms that a specialist opinion is required. In practice, however, it is the patients with worrisome symptoms of benign pathology who are likely to demand an urgent specialist appointment. Given that cancers also present with symptoms associated with benign conditions and in view of the rising tide of litigation and complaint from patients sensitised to sensational stories of misdiagnosis, it is hardly surprising that patients are being inappropriately referred through the fast track route.11 This phenomenon can also be explained by Braess's paradox, whereby “adding extra capacity to a network, when the moving entities selfishly choose their route, can in some cases reduce overall performance” (http://en.wikipedia.org/wiki/Braess'_paradox).

    So what is the best strategy for deciding which patients to refer for specialist opinion? Maybe one day we will have a reliable and valid test to help identify cancer patients in primary care. Until then, general practitioners should make a provisional diagnosis on the basis of a history and a physical examination, paying particular attention to genetic predisposition, exposure to carcinogens, and the type and duration of symptoms. Effective lines of communication between general practitioners and cancer specialists or networks will help to relay these clinical details, leading to earlier and more accurate diagnosis. Technological innovations that facilitate this process without being intrusive or cumbersome in practice may be more successful than published guidelines or schemes to ration urgent appointments only to those with a limited list of signs and symptoms.


    • Competing interests: None declared.

    • Provenance: Commissioned; not externally peer reviewed.