Guidance on referral needs to be evidence based

EDITOR—We were surprised at the amount of media interest in the data given by Thrush et al in their letter on grading referrals to a specialist breast unit.1 We have two main concerns. Firstly, the data presented do not support their conclusions. Secondly, the issue is not simply about the inappropriateness of general practitioners to determine urgency, but more about the development of evidence based guidance and general practitioners' adherence to this guidance.

The authors say that the two week wait initiative is not ensuring that most patients with symptomatic cancer are seen within two weeks of referral. This is not supported by the data presented in that over half of patients (56%) found to have breast cancer were in fact referred as urgent cases. No data are presented concerning delays experienced by the non-urgent group. Their data show that 9.3% of urgent referrals were diagnosed with cancer compared with 1.7% of non-urgent referrals. This equates to a diagnosis of cancer being six times more likely in urgent compared with non-urgent referrals. Hence it may be valid to conclude that general practitioners' use of the guidance is working to some extent.

Thrush et al do not differentiate according to whether patients fulfilled the urgent referral criteria; this makes it impossible to distinguish between the effectiveness of the guidance and the adherence of general practitioners to the guidance. It is well established that the two week referral guidance overall lacks an evidence base, especially as there is a lack of good quality data on the predictive value of symptoms.2

Longer delays for breast cancer patients have been found to be associated with worse survival rates.3 However, the delay between referral by the general practitioner and first appointment in secondary care is only one component of the total diagnostic delay. Evidence is lacking on the association of individual components of delay (patient delay, primary care delay, and secondary care delays) with stage at diagnosis and survival.

High quality research is clearly needed to investigate diagnostic delay and clinical outcome to produce evidence based referral guidance and develop effective strategies to ensure that such guidance is adhered to.

Communication has been degraded to exchange of dataset

EDITOR—The contents of Thrush et al's letter will come as no surprise to general practitioners, who are well aware of the difficulties in selecting patients for the two week wait.1 As long ago as 2000 at the conference of local medical committees general practitioners supported my motion criticising the policy for being politically inspired and being unlikely to improve the overall care for patients with cancer.

All a general practitioner can do is sort symptomatic patients into two groups—those more likely and those less likely to have cancer. It is inevitable that, wherever the line is drawn, patients will be incorrectly sorted, and the real worry is that the two week wait requirements result in an increased time for those unfortunate to have cancer without typical signs.

In such cases the patient has to deal not only with the diagnosis of cancer but also with the knowledge that she waited a longer time than ideal for assessment. The relationship between the general practitioner and the patient can only suffer, and I am sure I am not the only general practitioner to have received a complaint because a patient referred exactly within guidelines unexpectedly had a malignancy.

The tick box referral forms favoured by breast units have degraded communication between general practitioners and consultants to the level of a transfer of a basic dataset, and so the opportunity for an alert consultant to expedite a routine referral on clinical grounds is missed. Perhaps there should be a reassessment of the role of professionals within the referral process, with consultants taking responsibility for reading letters on receipt, and assigning waiting times, and general practitioners in return providing a narrative style letter containing adequate information to allow the consultant to perform this task. Come to think of it, wasn't that what we used to do?