Suspected cancer (part 2—adults): reference tables from updated NICE guidanceBMJ 2015; 350 doi: https://doi.org/10.1136/bmj.h3044 (Published 23 June 2015) Cite this as: BMJ 2015;350:h3044
- William Hamilton, professor of primary care diagnostics1,
- Steve Hajioff, director of public health2,
- John Graham, director and consultant in clinical oncology3,
- Mia Schmidt-Hansen, researcher3
- 1University of Exeter, Exeter EX1 2LU, UK
- 2London Borough of Hillingdon, Uxbridge UB8 1UW, UK
- 3National Collaborating Centre for Cancer, Cardiff CF10 3AF, UK
- Correspondence to: W Hamilton
The bottom line
In possible lung cancer, someone aged 40 or over with haemoptysis is recommended for urgent referral within two weeks for suspected cancer
In possible breast cancer, women aged 30 years or over with an unexplained breast lump or aged 50 years or over with nipple changes are recommended for urgent referral within two weeks for suspected cancer
In possible colorectal cancer, patients who do not meet criteria for suspected cancer referral should be offered testing for occult blood in faeces
Clinicians should trust their clinical experience where there are particular reasons that this guidance does not pertain to the specific presentation of the patient
It is generally believed that early diagnosis of cancer reduces mortality and morbidity. The National Institute for Health and Care Excellence (NICE) has updated its 2005 guidance on the recognition and referral from primary care of people with suspected cancer.1 The full guidance will be available on bmj.com. This summary of the full guidance is in two parts: part 1 on recommendations for children (up to 15 years old) and young adults (16-24 years), and part 2 on those for adults. Separate recommendations have been made for adults and for children and young people to reflect that there are different referral pathways. However, in practice, young people (aged 16-24) may be referred using either an adult or children’s pathway depending on their age and local arrangements.
Key changes in the updated guidance are:
Reliance on new evidence derived from primary, rather than secondary, care
Explicit use of a threshold risk of cancer to underpin recommendations for urgent investigation—the first cancer guidance to do so.2
Assumes that patients will have had a full history, clinical examination, and appropriate initial blood tests
Recommends urgent investigation in adults with a 3% or higher cancer risk, but uses a lower threshold for children and young people and when primary care testing is available
Relies on evidence mainly from moderate quality observational studies
Will increase the number of investigations or referrals in some cancer sites; the use of direct access testing should reduce the costs of this, but will require organisational change
Asks clinicians to continue to trust their clinical experience where there are particular reasons that the guidance isn’t relevant to a patient’s specific presentation.
“urgent referral”—within two weeks
“very urgent referral”—within 48 hours
“direct access”—tests that are requested and followed up in primary care
“suspected cancer pathway referral”—offers consultation within two weeks.
Key to tables
*—Refer, supported by evidence
†—Investigate in primary care, supported by evidence
Unmarked—Consider referral or primary care investigation, evidence thin or absent.
Gastrointestinal or abdominal features (table 1⇓)
Abdominal symptoms are common, and usually have a benign cause. Some abdominal symptoms are linked with specific cancers. Others, such as abdominal pain, may arise from several different cancers.
Bleeding (table 2⇓)
Abnormal bleeding is one of the highest risk presentations for cancer, with the possible cancer site usually obvious.
Gynaecological features (table 3⇓)
Recommendations for ovarian cancer were published in 20113 and are unchanged, so have been omitted for brevity. Although abnormal bleeding is a prominent feature of gynaecological cancer, other presentations are possible.
Lumps or masses (table 4⇓)
Little evidence was available, other than for unexplained breast lumps in women aged 30 or over, where referral is recommended.
Neurological features (table 5⇓)
Brain cancer is relatively rare, presenting mostly with neurological deficit or symptoms of raised intracranial pressure.
Oral lesions (table 6⇓)
Intraoral malignancies may present to doctors or dentists, the latter generally having more experience in this area.
Respiratory features (table 7⇓)
A suspected cancer referral for haemoptysis in someone aged 40 or over is now recommended (see “Bleeding⇑”). The indications for a chest x ray have been liberalised from previous guidance.
Musculoskeletal features (table 8⇓)
Back pain is a common primary care complaint, though it can rarely indicate pancreatic cancer, myeloma, or cancer metastasis.
Skin or surface features (table 9⇓)
This group of cancers differ by being visible externally. In some breast cancers, the presentation is not with a breast lump, but with nipple retraction or discharge, or Paget’s disease of the nipple (which resembles a unilateral eczema).
Urological features (table 10⇓)
Urological cancers often present with haematuria, though other features can occur, and patients with such presentations often experience diagnostic delay.
Non-specific features of cancer (table 11⇓)
Some symptoms suggest possible cancer, though the particular cancer site may not be obvious. Patients with these features are at risk of multiple sequential referrals—the so called ping pong patient.
Primary care investigation results (table 12⇓)
Primary care investigations for cancer comprise (a) non-specific tests, particularly full blood counts, often performed without cancer being considered initially, and revealing anaemia or thrombocytosis; or (b) tests used when specific malignancy is suspected, such as a prostate specific antigen test for possible prostate cancer.
How patients were involved
Committee members involved in this guideline update included lay members who contributed to the formulation of the recommendations summarised here.
Further information on the guidance
The general view that early diagnosis of cancer is beneficial underpins many initiatives in the UK and other countries, such as awareness campaigns, cancer screening, and better diagnosis of symptomatic cancer. There is also unwarranted variation in referral rates for suspected cancer,5 investigation rates for cancer,6 and clinical outcomes both internationally and within the UK.7
As initial consideration of possible cancer typically occurs in primary care, evidence from primary care must inform the identification process. Previous approaches, including the predecessor guidance,1 have relied mostly on evidence from secondary care, partly because evidence from primary care was lacking. More primary care evidence is now available.
This update is the first cancer guidance explicitly to use a threshold risk of cancer to underpin its recommendations for urgent investigation.2 A 3% cancer risk warranting urgent investigation was agreed: this took account of several considerations, including perceived benefits and costs, patients’ preferences, and resource implications. However, the threshold is lower for children and young people, and when primary care testing is available.
This guidance includes unchanged recommendations from the ovarian cancer guidance in 2011.3 The methodology, and use of health economic evidence, followed the standard methods of the guideline body.8 The systematic reviews incorporated a primary care filter. Each study had a quality assessment using the QUADAS tool,9 and where three or more studies reported on a cancer site/symptom dyad, meta-analysis was performed to provide a summary risk estimate. For each clinical question, recommendations were developed based on summaries of the clinical evidence and, where appropriate, economic evidence.
The guideline development group included a chair, a clinical lead, patient representatives, general practitioners (some with additional roles, such as commissioning of healthcare services), an oncologist, a radiologist, and a secondary care cancer manager, supplemented by expert advisors where necessary. Stakeholder consultation was sought on an initial version of the recommendations, which were amended afterwards.
NICE also produces three other versions of the guideline: the NICE guideline (a shorter version of the guideline), NICE pathways (online tools for health and social care professionals), and “Information for the Public (IFP),” a summary of the recommendations in everyday language for patients, their families and carers, and the wider public (www.nice.org.uk/guidance/ng12/informationforpublic). All these versions are available from the NICE website (www.nice.org.uk/guidance/ng12).
The guideline group identified four priority areas for future research:
The use of age thresholds to underpin recommendations for investigation
The diagnostic accuracy of commonly performed primary care cancer tests
Specific cancers with insufficient research on symptoms in primary care
Studies on patient experience and information needs, especially in the interval between first primary care presentation with a symptom of possible cancer and attendance in secondary care.
Cite this as: BMJ 2015;350:h3044
This is one of a series of BMJ summaries of new guidelines based on the best available evidence; they highlight important recommendations for clinical practice, especially where uncertainty or controversy exists.
Members of the guideline development group were: Susan Ballard, patient and carer member (June 2012–March 2013); Nicki Doherty, general manager, Barnsley NHS Foundation Trust; Jeanne Fay, general practitioner, Oxford; Steve Hajioff, guideline development group (GDG) chair from September 2013, director of public health, London Borough of Hillingdon; Willie Hamilton, GDG clinical lead, professor of primary care diagnostics, University of Exeter; Susan Hay, patient and carer member; Georgios (Yoryos) Lyratzopoulos, senior clinical research associate/honorary consultant in public health, Department of Public Health and Primary Care, University of Cambridge; David Martin, patient and carer member; Joan Meakins, general practitioner, York; Orest Mulka, GDG chair (January 2012–June 2013), retired general practitioner, Derbyshire; Richard Osborne, consultant medical oncologist, Dorset Cancer Centre; Euan Paterson, general practitioner, Glasgow; Liliana Risi, general practitioner, London; Karen Sennett, general practitioner, London; Lindsay Smith, general practitioner, Somerset; Stuart Williams, consultant radiologist, Norfolk & Norwich University Hospital.
Competing interests: Based on NICE’s policy on conflicts of interests (available at www.nice.org.uk/Media/Default/About/Who-we-are/Policies-and-procedures/code-of-practice-for-declaring-and-managing-conflicts-of-interest.pdf), the authors’ statements can be viewed at: www.bmj.com/content/bmj/350/bmj.h3036/related#datasupp.