Colorectal cancer is the second most common cause of death from cancer in the UK, with 30,000 new cases and 19,000 deaths every year.
Current NHS guidelines for the referral of patients with colorectal symptoms often over-classify patients as high risk. More importantly, they fail to identify a significant number of cancers that are wrongly identified as low-risk.
S. N. Selvachandran, David Cade, and colleagues from Leighton Hospital, Crewe, England, investigated the value of using a specific questionnaire (completed by patients) and a computer-generated risk score in order to prioritize symptoms indicative of possible colorectal disease.
A total of 2268 patients with adverse colorectal symptoms (such as rectal bleeding, increased and/or looser bowel movements) were studied.
They had been referred to a colorectal specialist by their general practitioner, and completed a patient consultation questionnaire linked to a computerized record.
Referrals were prioritized with a cancer risk score according to the guidelines by a senior colorectal surgeon separately for the general practitioner's letter and for the questionnaire.
A weighted numerical score was derived from the weighting of the main disease symptoms and symptom complexes (presence of more than one major symptom). It was calculated automatically when the questionnaire data were entered into a computer program.
| The scoring system identified 99% of colorectal carcinomas.
It was found that 95 of the 2268 patients had colorectal cancer.
The average weighted numerical score was substantially higher (76.5) for patients with cancer than for patients who did not have cancer (44.5).
The malignancy risk score derived from the patient consultation questionnaire and the weighted numerical score had a good cancer pickup. However, it resulted in fewer patients being identified at high risk (around 40%) compared with a 50% high-risk grading for the current NHS guidelines.
The new system identified a further 13% of cancer patients who would have been categorized as low risk by the current NHS guidelines. Thus it detected nearly all cancers (99%) by investigating just over half (57%) of the patients referred.
David Cade comments, "The patient consultation questionnaire with the weighted numerical score produces an accurate system to prioritize patients with colorectal symptoms referred by their general practitioners.
"The questionnaire and the weighted numerical score might also have a place in general practice to help family doctors identify individuals at substantial risk of colorectal cancer."