We will answer Dr Nader Al-Hassan's points in turn:
1. In this series of cases, no patient had an iron deficiency anaemia (IDA).
2. The new NICE guidelines for referral of patients with symptoms suspicious of bowel cancer specifically says that a family history in a symptomatic patient is of no proven value in identifying a higher risk patient.
3. In Portsmouth patients, who have had significant polyps will already be on a colonoscopic follow-up programme. No such patient was in our series.
4. Colonoscopy is used in Portsmouth if the patient is found to have a significant adenomatous polyp.
Barium enema is usually reserved for patients with an IDA or an easily palpable abdominal mass.
In patients presenting with rectal bleeding having a normal flexible sigmoidoscopy to 60 cm the residual risk of cancer is 1:500. In patients with this pretest probability of cancer the difference in the value of barium enema and colonoscopy is small. Where resources for colonoscopy are restricted to those patients most likely to benefit, as is mainly in the UK, then a barium enema may be acceptable particularly in patients in whom the bleeding has stopped.
5. The sensitivity of the symptom combination for cancer was 100% as all the patients in this small series presenting with rectal bleeding had an associated change in bowel habit.
6. The total number of patients developing cancer in this cohort was determined by follow-up at 18 months. This is essential because even colonoscopy misses cancers. This is an acceptable technique to determine the prevalence of cancer in patients not having investigation. We had 100% follow-up of these patients at 18 months.
7. Our gold standard was 18 month follow-up to catch all cancers from all groups including those having colonoscopy.
8. We can give the data about the duration of the symptoms, but again in this small group of patients this did not affect cancer risk, although this might become apparent in a larger group of patients.
9. The number of patients that flexible sigmoidoscopy will miss depends on the epidemiology of patients referred to the clinic. In our clinic in Portsmouth only 12% of all cancers referred are proximal to the sigmoid colon, probably because proportionately more right-sided lesions compared with cancers in the sigmoid and rectum present as emergencies not to Outpatients or with a significant IDA to other clinics. Of those right-sided lesions in our clinics the majority either have an associated IDA or a palpable abdominal mass, which means that any patients who present to our clinic with symptoms, but no IDA or an abdominal mass with a normal flexible sigmoidoscopy to 60 cms only very small numbers of patients have cancer. But this was not the subject for discussion in this paper. This is a stage after GP referral, and was not addressed by the data in this paper, but it does explain why our experience over the last 15 years in Portsmouth that flexible sigmoidoscopy alone in patients without an IDA or an abdominal mass is a powerful way of identifying most of the cancers which attend our clinic.
10. The method of recruitment to the trial was not on the basis as is stated, but was offered to all patients presenting with rectal bleeding. The group was divided up into those that accepted a flexible sigmoidoscopy, those that simply filled out a questionnaire and the third group that refused both. It was interesting that 40% of patients refused a flexible sigmoidoscopy.
11. Patients refusing referral to hospital and examination were indeed followed up after a reasonable time to check that they came to no harm.
12. It is interesting that our conclusions fall completely in line with the National Institute of Clinical Excellence Guidelines for referral of patients suspected of cancer recently published in the UK.
- © British Journal of General Practice, 2006.
