Acknowledgements

Provenance of the ‘symptom checker’ for the three primary symptoms of bowel cancer
Provenance of the ‘symptom checker’ for the three primary symptoms of bowel cancer ,
The need for a ‘symptom checker’

The Internet based ‘symptom checker’ - for people with one of the three key symptoms of bowel cancer - is designed for patients with mild to moderate symptoms who are unsure whether they need to see their general practitioner

Patients with severe symptoms that prevent them working, eating or sleeping require treatment regardless of the nature of their symptoms and these patients clearly need to see their GP. 
However as over 99% of the large number of people with mild to moderate symptoms will not have any serious condition such as bowel cancer, it would be a great burden on primary care services if they all decided to see their GP

The data suggests that the majority of people currently manage their symptoms quite appropriately in the community without medical advice although increasingly, because of greater awareness of the occasional significance of these symptoms,  increasing numbers of people will need access to reliable information so that they can still appropriately avoid a consultation in primary care

It is also well known that 10% of patients with bowel cancer have a delay in treatment of over a year and 30% of over 6 months and,  although some of this delay is unavoidable, in a significant number of patients the long delays are because patients are unaware of the significance of their symptoms [9]. 
These delays have not been reduced by previous public awareness campaigns over the last 60 years [9] and new methods such as this bowel cancer symptom checker need to be tried if any reduction in the delay to diagnosis and treatment is to be achieved in the future

As 95 % of patients with mild to moderate symptoms have self-limiting benign disease it is best for the health care system as well as the majority of patients with these symptoms that they remain self-caring in the community. The question is - can the current ‘natural’ patient’s decision process be improved so that more patients with self limiting benign disease avoid investigation in hospital while not putting those with serious disease including cancer at risk? 

The “checker” grades the speed at which patients see their GP so those at very low risk of having cancer have longer periods of Treat Watch and Wait so that those with transient symptoms from self limiting disease have time to get better whereas those at higher risk see their GPs without delay

The symptom checker should be able to:

1. Identify the majority of patients at higher risk of having severe disease including cancer and encourage these patients to see their GP. The checker should also make clear that although the majority of these patients do not have serious disease it may be worth them having simple investigations in hospital

2. Identify those patients at very low risk of serious disease, reassure them their symptoms are likely to resolve in time either without treatment or with simple self-administered treatments, perhaps with the help and advice of a local pharmacist

3. Provide all patients seeing their GP with a printed record of their symptoms, which includes advice to the GP on whether the patient is eligible for the 2-week clinic or that it would be worthwhile, if the patient is anxious or has persistent low risk symptoms, for them to be referred to a routine clinic. This could be with a request for an urgent appointment, the ‘third way’ of referral if the GP feels the patient has other reasons to be seen quickly - for example: they have a strong family history. In very low risk patients the GP will be advised that with the patient’s consent a further period of  ‘treat watch and wait’ in primary care is safe and may be appropriate

4. Advise on simple dietary and pharmacological treatments of rectal bleeding and pile symptoms, changes in bowel habit and abdominal pain

5. Not excessively delay diagnosis of cancer patients presenting with ‘low risk’ symptoms including those with incidental asymptomatic cancers. This will be achieved by advising all patients, particularly those over 50 years of age with persistent or recurrent rectal bleeding to see their GP. It is likely that significant adenomatous polyps will also be diagnosed by a policy focused on this sub-set of patients

6. It is also an opportunity to raise awareness of the high prevalence [1,2] of these three symptoms in the community and to point out how difficult it is for GPs to identify those patients with serious disease and cancer [2]. It is impossible to reduce the risk of delay to zero but by following simple rules it should be possible to reduce the risks of delayed diagnosis to very low levels while protecting the majority of patients with benign disease from unnecessary investigation
 
 
 
 

Essential requirements to develop a symptoms checker
Essential requirements to develop a symptoms checker ,
It is necessary to stratify a patient’s risk of having a cancer or serious disease on the basis of the patient’s symptoms, age and length of history. Patients at higher risk should be advised to see their GPs whereas those at lower risks should be advised to treat watch and wait for reasonable and varying lengths of time for no more than up to 4 weeks according to risk. It is necessary to define what is higher and low or very low risk

Risk stratification. 

In the checker this is based on estimation of risk based on studies of the nature of rectal bleeding in the community [1] and patients with colorectal cancer [3,4] . These estimates are in keeping with numerous peer reviewed studies [2-7] and published abstracts over 18 years [10-20] on subsets of 31,000 patients seen in Portsmouth out patient clinics over 25 years and studies from other centres. In the Portsmouth studies records of the patient’s symptoms were collected prospectively before the diagnosis was known and the final diagnosis determined either by flexible sigmoidoscopy or whole colonic imaging. Missed cancers, regardless of whether or not they had whole colonic imaging were identified by comparison with the Wessex Regional Cancer Registry’s Database 2 years after the patients were seen

Establish that short delays in diagnosis and treatment of bowel cancer do no harm

A large review suggested that the theoretical basis for the benefit of earlier diagnosis after the onset of symptoms was weak and that this was supported by the majority of observational studies which however cannot not avoid a number of potential biases. However many bowel cancer patients die in spite of prompt diagnosis and many survive in spite of delays of over a year which suggests that at least for the majority of patients reduction in delay will not improve overall survival. These findings are supported by the fact that many bowel cancers have long natural histories, which means many may have been present for at least 2 years before they develop symptoms to indicate the need for investigation. This is long enough for most of the aggressive cancers, which cause the majority of deaths from bowel cancer to have disseminated and become incurable

In contrast those cancers that are still curable after this length of time are unlikely to become incurable immediately after the onset of symptoms. This suggests that ‘treat watch and wait’ diagnostic strategies including ‘delays’ of up to 6-8 weeks, which help the majority of patients with transient symptom from benign disease avoid investigation will not reduce the chance of cure of cancer patients

Advice on the management of the patient’s symptoms

This is based on one consultant surgeon’s clinical experience (MRT; see CV) of working in specialized colorectal clinics in Portsmouth over 27 years. Advice to patients on the management of their symptoms has evolved over the last 20 years through working with Lynn Faulds Wood on public awareness campaigns

Peer Reviewed articles published in journals

1. Thompson JA, Pond CL, Ellis BG, Beach A, Thompson MR. Rectal Bleeding in General and Hospital Practice; 'The tip of the iceberg'.  Colorect Dis 2000; 2: 288-293

2. Thompson MR, Heath I, Ellis BG, Swarbrick ET, Faulds Wood L, Atkin WS.  Identifying and managing patients at low risk of bowel cancer in General Practice.  BMJ 2003; 327: 263-265
 
3. National Referral Guidelines for Bowel Cancer  www.acpgbi.org.uk  Colorectal Dis 2002; 4 (4): 287-297
 
4.  Flashman K, O’Leary DP, Senapati A, Thompson MR. The Department of Health’s ‘two-week standard’ for bowel cancer: is it working?  Gut 2004; 53(3): 387-391

5.  Ellis BG, Thompson MR. Factors identifying higher risk bleeding in general practice.  Br J Gen Pract 2005; 55(521): 949-55

6.  Thompson MR, Perera R, Senapati A, Dodds S. The diagnostic and predictive value of the common symptom combinations of bowel cancer. BJS 2007: 94: 1260-1265

7.  Thompson MR, Flashman KG, Wooldrage K, Statistician, Rogers PA, Senapati A, O’LearyDP, Atkin WS. Flexible sigmoidoscopy and whole colonic imaging in the diagnosis of cancer in patients with colorectal symptoms.  Br J Surg 2008; 95: 1140-114

8.  Thompson MR, Asiimwe A, Flashman K, Tsavellas G,  Is earlier referral and investigation of bowel cancer patients presenting with rectal bleeding associated with better survival?  Accepted 19 July 2010

9.  Thompson MR, I. Heath, E. T. Swarbrick, L. Faulds Wood and B. G. Ellis Earlier diagnosis and
 treatment of symptomatic bowel cancer: can it be  achieved and how much will it improve survival?  Colorectal Disease 2011; 13, 6–17


Published Abstracts and Posters presented to National and International Meetings

10. Dodds SR, Thompson MR. The significance of rectal bleeding in symptomatic colorectal disease Poster.  Tripartite Meeting.  Sydney, Australia. 1993 

11.Vakis SA , Dodds SR, Thompson MR.  Presentation of colorectal cancer at a Surgical Out-Patients.  How occult a sign is rectal bleeding?  Int.J.Colorect.Dis. 1995; 10:240
 
12. Thompson MR, Armstrong-James D, Moss S, Prytherch D. Prevalence of colorectal      cancer in patients presenting with anorectal bleeding.  Effect of age and dark red bleeding. Gut 1996; 39 (Supple.1), F181.

13. Dodds S, Dodds A, Vakis S, Flashman K, Senapati A, Cripps NPJ, Thompson MR. The value of various factors associated with rectal bleeding in the diagnosis of rectal cancer. Gut 1999; 44 (Supple 1): Th396.

14. Nichols PH, Dodds S, Dodds A, Flashman K, Senapati A, Cripps NPJ, Thompson MR. Painless rectal bleeding or bleeding without pile symptoms. Colorect Dis. 1999; 1(Supple1) 23

15. Chave H, Flashman K, Cripps NPJ, Senapati A, Thompson MR. The Relative Values of the Characteristics of Rectal Bleeding in the Diagnosis of Colorectal Cancer. Colorect Dis 2000; 2 (Supple 1): 1/01

16. Branagan G, Senapati A, Thompson MR, O’Leary DP.  Dark vs bright red bleeding in the diagnosis of colorectal cancer.  Colorect Dis 2004; 6 (Suppl 1): 47

17. Ellis, B, Baig MK, Cripps NPJ, Senapati A, Flashman K, Jaral M, Thompson MR. Common modes of presentation of colorectal cancer patients Colorect Dis 1999; 1 (Supple 1): 24/P26

18. Dodds S, Dodds A, Vakis S, Flashman K, Senapati A, Cripps NPJ, Thompson MR.  Effect of age and sex on value of symptoms in diagnosis of colorectal cancer.  Gut 1999; 44(Suppl 1): A15/T60.

19. Chave H, Flashman K, Senapati A, Cripps NPJ, Thompson MR. Characteristics of the change in Bowel Habit in Patients with Colorectal Cancer. Colorect Dis 2000; 2 (Supple 1): 1/0ral 2

20. Baig MK, Whatley P, Thompson MR. Delay in diagnosis and treatment of colorectal cancer; Does it affect outcome? Gut 1999; 44(Suppl.): A143. (Abstract TH572)
Tip Title