Introduction
In 2000, the Department of Health developed the UK National Guidelines for referring suspected head and neck cancer cases through a fast-track, two-week-wait pathway. In England, this means that suspected cancer patients should be seen by a specialist within 14 days of the primary care referral, to facilitate early detection. The target in England is 93 per cent of 2-week-wait referrals seen within 14 days. In Northern Ireland, the two-week-wait pathway only applies to suspected breast cancer. This scheme does not exist in Scotland or Wales.1,Reference McKie, Ahmad, Fellows, Meikle, Stafford and Thomson2 The National Institute for Health and Care Excellence (NICE) generated guidelines for two-week-wait head and neck cancer referrals in 2005.Reference McKie, Ahmad, Fellows, Meikle, Stafford and Thomson2 This was refined in 2015, with the latest revision in 2021.1 Despite this, the cancer detection rate is not significantly different from non-urgent referrals.Reference Hobson, Malla, Sinha, Kay and Ramamurthy3,Reference Tikka, Pracy and Paleri4 A recent retrospective study of head and neck cancer showed that 11.8 per cent of two-week-wait referral patients were diagnosed with cancer and 88.2 per cent with benign disease.Reference Douglas, Carswell and Montgomery5
There are currently insufficient data in the literature regarding the patients’ journey after referral. This includes the pathway for diagnosed benign conditions before discharge back to primary healthcare.Reference Douglas, Carswell and Montgomery5 This study explored the outcomes and pathways of all two-week-wait head and neck cancer referral patients for whom no pathology was identified or who were diagnosed with benign pathology.
Materials and methods
A retrospective review audit was conducted in the first six months of 2018, from January to June, in a tertiary head and neck surgical centre. All patients referred to the ENT out-patient clinic through the two-week-wait pathway were included. All patients with suspected cancer are managed on a two-week-wait pathway, which allows data collection for comparison with national guidelines, as well as expedition of investigations and treatment in order to meet these targets. The outcome from the first specialist appointment was either a decision to continue on the two-week-wait pathway (for suspicious referrals) or to be taken off the pathway (for those with benign conditions). The durations between referral and the time of the first appointment and the time of secondary care completion were calculated. The analysis also included demographic details, social history as documented in the specialist clinic letter, and the presenting history. Compliance with 2005 NICE referral criteria and the 2015 update was analysed.
Clinical findings and outcomes were recorded anonymously. The data were collected from: clinic electronic letters stored on a shared-access network drive, the Integrated Clinical Environment platform, and Infoflex healthcare data management software (to determine multidisciplinary team meeting outcomes). This study was an audit of clinical practice and therefore formal ethical approval was not required.
The data were analysed in comparison to similar publications and UK cancer research statistics. A systematic review of all published two-week-wait head and neck centres was conducted. Criteria for study inclusion were: a single centre where the national two-week-wait head and neck cancer clinic target was met; a minimum of six months’ duration of data collection; use of the Department of Health (DoH) and NICE guidelines as standard; and reporting of results that include compliance of referral symptoms with NICE guidelines and cancer detection rates from these referrals.
Results
There were 1107 referral patients who proceeded through the two-week-wait pathway during the study period. The average number of referrals per month was 184 (standard deviation = 14). Of the patients, 98.8 per cent were reviewed by a specialist within 2 weeks, with an average waiting time of 6 days. Social history details of smoking and alcohol use were documented in 74 per cent of specialist clinic letters. Performance status was rarely documented. Sixty per cent of the referrals were female. There were no urgent referrals for children and adolescents younger than 15 years old, with only 0.8 per cent from those aged 15–20 years. Most referrals fell in the age groups 40–60 years and 60–80 years, with 37.4 per cent for each group.
Fifty-two per cent of the referrals were compliant with the NICE 2015 guideline criteria. The commonest head and neck regions of concern in the primary care letters were the larynx (33 per cent), pharynx (27 per cent) and neck (18 per cent).
Six errors were identified in the 48 per cent of non-compliant referrals. The first (29 per cent of all referrals) was ‘no consistent indication’ from the guidelines. The ‘intermittent course’ was the second error. For example, 105 (9 per cent) of the referrals had intermittent hoarseness of voice. The third (6 per cent) was ‘wrong’ referrals, such as lump in the neck where there was nothing obvious or palpable to the specialist, or when there was a mismatch between the patient's complaint in the specialist appointment and the original indication in the urgent referral letter from primary care. The fourth and fifth errors were either very long history, of more than a year, or a very short history, of less than three weeks. The last error was the concern resulting from accidental findings on imaging where the patient was asymptomatic. Figure 1 illustrates the percentages of all compliant and non-compliant referrals.
Percentages of all referrals compliant and non-compliant with National Institute for Health and Care Excellence guidelines.
The commonest general symptoms resulting in referral were: voice change (27.2 per cent), neck lump (25.9 per cent) and globus sensation (18.7 per cent). All patients with intermittent hoarseness and recent neck lump, and 99.4 per cent of those with globus sensation, had a benign outcome, whereas 93 per cent and 81.8 per cent of all referred patients with persistent hoarseness and persistent neck lump, respectively, had a benign outcome.
In 489 patients, there were ‘red flag’ symptoms appropriate for referral, but the diagnosis was benign. In general, there was low sensitivity for cancer, with a high rate of a benign outcome for each symptom, of between 82.6 and 96.3 per cent (Table 1).
Relations between ‘red flag’ symptoms and benign outcomes
2ww = two-week-wait
During the study period, 1035 patients (93 per cent of all referrals) were diagnosed with benign conditions, and 47 per cent of them had been appropriately referred on the two-week-wait pathway. Table 2 illustrates the most commonly diagnosed benign conditions. Laryngopharyngeal reflux was the commonest diagnosis (21 per cent). Eighteen per cent of patients were reassured as being ‘ENT clear’ after their first specialist visit. Although 18.7 per cent of patients had globus sensation on presentation, only 4 per cent were diagnosed with globus pharyngeus.
Common benign conditions in urgent referrals
LN = lymph node
All patients with globus sensation and reflux-related findings were treated with a proton pump inhibitor and followed up before being discharged. Of all two-week-wait referrals, 243 patients (22 per cent) were discharged following their first visit. The average number of specialist visits was 1.7, after which 58 per cent of patients were discharged back to primary care. Referral to speech and language therapy was undertaken for 7 per cent of all two-week-wait referral patients.
Sixty-three patients (5.7 per cent of all referrals, 6.1 per cent of patients with benign outcomes) were referred for barium swallow study. No barium swallow study resulted in a diagnosis of cancer. Furthermore, 16.1 per cent of all patients on the two-week-wait pathway were referred for a neck ultrasound; for these patients, 96.1 per cent of the outcomes were benign.
A systematic review of relevant literature yielded nine studies.Reference McKie, Ahmad, Fellows, Meikle, Stafford and Thomson2–Reference Lyons, Philpott, Hore and Watters10 Three studies were excluded because the authors did not report the compliance with NICE criteria in their results.Reference Gardner and Wijesekara7,Reference Allam and Nijim9,Reference Lyons, Philpott, Hore and Watters10 Another cohort study between two centres was excluded.Reference Tikka, Pracy and Paleri4 Table 3 shows the five centres included and their results. The sample size in the pooled analysis was 4218 patients. The rate of benign outcomes in two-week-wait referral patients, across all five centres, was 89.8 per cent. Of all referrals, 68 per cent were compliant with DoH and NICE guidance.
Systematic review of studies investigating compliance with guidelines
Studies by Allam and Nijim,Reference Allam and Nijim9 Lyons et al.,Reference Lyons, Philpott, Hore and Watters10 and Gardner and Wijesekara,Reference Gardner and Wijesekara7 were excluded because of the lack of statistical data on compliance with National Institute for Health and Care Excellence guidelines. Study by Tikka et al.Reference Tikka, Pracy and Paleri4 was excluded because it was a cohort study of two centres.
Discussion
Comparison with existing literature
The numbers of two-week-wait referrals have increased dramatically across all specialties in the National Health Service. Over 2.5 years, from January 2004 to December 2006, 1079 patients were referred to a tertiary centre.Reference McKie, Ahmad, Fellows, Meikle, Stafford and Thomson2 The same tertiary centre showed 4460 referrals in the following four years (January 2007 to December 2010).Reference Tikka, Pracy and Paleri4 Another busy tertiary centre in Glasgow had 2116 annual referrals from June 2015 to May 2016.Reference Douglas, Carswell and Montgomery5 A fourth tertiary centre recorded 622 over one year, from July 2009 to 2010.Reference Prichard, Whitehead, Moorthy and Pracy6 In a district general hospital, the number dropped to 177 patients over a one-year study period.Reference Hobson, Malla, Sinha, Kay and Ramamurthy3,Reference Gardner and Wijesekara7 The national two-week-wait targets should be met in 95–100 per cent of cases.Reference Hobson, Malla, Sinha, Kay and Ramamurthy3,Reference Gardner and Wijesekara7 In our centre, two-week-wait clinics achieved the national target, despite the extremely high volume of referrals in this six-month period because of one or more of the six errors identified in the referrals.
Reasons for non-compliant referral
The appropriateness of referrals to head and neck two-week-wait clinics remains low despite NICE guidelines. The cancer detection rate in ENT is lower than in other specialties; for instance, rates of 23 per cent and 12.8 per cent have been reported for urology and breast two-week-wait referrals.Reference Gardner and Wijesekara7 Possible contributing factors include the poor predictive value of head and neck symptoms for cancer; the lack of ENT experience among general practitioners; the poor availability of diagnostic aids in primary care; and the high rate of general practitioner consultations for ENT problems (25–50 per cent).
Tikka et al. implemented an online calculator in 2016, which was refined recently for primary care, with a predictive value cut-off of 0.08 before referral to secondary care.Reference Tikka, Pracy and Paleri4,Reference Tikka, Kavanagh, Lowit, Jiafeng, Burns and Nixon11 They showed that common red flag symptoms included in NICE guidelines have low sensitivity and low positive predictive values: 45.4 per cent and 17 per cent for neck lump, 17.38 per cent and 7.76 per cent for hoarseness, and 7.3 per cent and 13 per cent for dysphagia, respectively.Reference Tikka, Pracy and Paleri4 Our study showed a high percentage of benign outcomes associated with all red flag symptoms in NICE guidelines.
Globus sensation and intermittent hoarseness were the commonest symptoms resulting in non-compliant referrals in the cohort study by Tikka et al. and others.Reference McKie, Ahmad, Fellows, Meikle, Stafford and Thomson2,Reference Tikka, Pracy and Paleri4,Reference Douglas, Carswell and Montgomery5 These authors suggested that two-week-wait referrals were unnecessary.Reference Tikka, Pracy and Paleri4,Reference Tikka, Kavanagh, Lowit, Jiafeng, Burns and Nixon11 As in our study, 9.2 per cent of the referral patients in Douglas and colleagues’ study had intermittent hoarseness.Reference Douglas, Carswell and Montgomery5 The lack of experience in history-taking that can differentiate globus from odynophagia, explains the persistently high referral rate. Globus sensation documented in the referral letter is frequently found to be dysphagia in the patient's own words during consultation, and vice versa.Reference Douglas, Carswell and Montgomery5 Furthermore, patients frequently confuse intermittent sore throat with persistent odynophagia. The latter has a five-times increased risk of cancer.Reference Tikka, Pracy and Paleri4 The three-week duration might not be sufficient for patients to establish an intermittent course. Our study showed positive predictive values of 1 per cent for globus sensation and 0 per cent for intermittent hoarseness. All patients referred with odynophagia and/or sore throat were pooled together in this study, showing a predictive value of 4 per cent. Regular e-learning modules for healthcare on head and neck cancer presentations, and communication between ENT consultants and general practitioners, would improve history-taking and reduce inappropriate referrals.Reference Prichard, Whitehead, Moorthy and Pracy6
The vast majority of patients are reassured after their visit to the specialist. Douglas et al. showed that 42.6 per cent of patients were discharged after one clinic appointment, while another 26.8 per cent were followed up for suspected benign pathology.Reference Douglas, Carswell and Montgomery5 Only 21.8 per cent of patients were actively investigated for cancer. The commonest diagnoses in discharged patients were: reflux (7.1 per cent), globus pharyngeus (5.5 per cent), functional dysphonia (5.1 per cent) and reactive lymphadenopathy (2.6 per cent). The authors did not discuss patients investigated for benign pathology.Reference Douglas, Carswell and Montgomery5 Our study showed the same types of benign conditions, but with a different order of frequency, whereby reflux and functional dysphonia were the commonest diagnoses. There was a higher rate of ‘ENT clear’ outcomes in our study than the 5.5 per cent of asymptomatic cases reported by Douglas et al.Reference Douglas, Carswell and Montgomery5
Ultrasound and ultrasound fine needle aspiration (FNA) during the clinic have shown advantages in terms of immediate reassurance and discharge from clinic, which benefit both the clinician and patient.Reference Douglas, Carswell and Montgomery5 Investigation with ultrasound prior to clinic has been shown, within a Scottish setting, to be of limited value and a waste of resources, so it has been discouraged in primary care.Reference Douglas, Carswell and Montgomery5 The one-stop clinic, with both a cytologist and a radiologist present in the clinic, proved to be an overuse of resources, as only 15 per cent required a cytologist for FNA cytology and only 12.8 per cent required a radiologist for interpretation of the ultrasound.Reference Prichard, Whitehead, Moorthy and Pracy6 However, it can still be of good value in neck lump clinics.Reference Prichard, Whitehead, Moorthy and Pracy6 The one-stop clinic is used in our centre, but is not fully integrated within all urgent clinics. It is run by a sonographer with a special interest in the head and neck, alternating with a consultant radiologist; however, the clinic lacks a cytologist, so another patient appointment is inevitable to discuss the FNA results.
The pooled analysis of the systemic review showed that 32 per cent of referrals were not compliant with NICE guidelines, with 89.8 per cent of referred patients having a benign outcome. The chi-square test showed no significant difference (χ2= 0.67, p = 0.4) in benign outcome in our centre from the pooled analysis, whereas there was a significant difference in terms of guideline compliance (χ2= 25.9, p < 0.00001).
Implications for research and practice
The lack of knowledge shown through the six referral errors, in particular the ‘inconsistent’ and ‘wrong’ referrals, causes rapid and unnecessary saturation of the service. The large number of such errors and the abundance of non-red-flag symptoms contributed to the high rate of benign outcomes (Table 1).
• Two-week-wait referrals are still associated with a low rate in the early identification of cancer
• More than half of the referrals are not compliant with national guidelines
• The commonest reasons for non-compliant referrals were: inconsistencies in referral and intermittent disease course
Regular teaching sessions for primary healthcare on guidelines, e-learning modules on head and neck cancer presentations, and tertiary site placement of general practitioner trainees to build special interest, could improve the use of resources. The regular application of clinical governance in every general practice, together with feedback and communication from secondary or tertiary centres, might help spot an early surge in service use. Furthermore, patient education may have more effect than striving for a two-week-wait target.Reference Hobson, Malla, Sinha, Kay and Ramamurthy3,Reference Prichard, Whitehead, Moorthy and Pracy6
Strengths and limitations
The use of national guidelines such as the NICE guidance would ensure unified assessment in many general practices and would prevent a confounding bias. This study was longitudinal, retrospective and observational. Selection bias, by limiting data collection to six months, might miss any anticipated fluctuation in referrals throughout the year. The data collection did not consider the analysis of the total head and neck cancer consultations in primary practice before referral to the tertiary centre.
Conclusion
Most patients who undergo two-week-wait referral for head and neck symptoms are diagnosed with benign conditions. In this study, the reasons for referrals that did not comply with NICE guidelines were mostly related to misinterpretation of non-red-flag symptoms and the intermittent course of the disease. Errors in the referral scheme could be reduced by training primary care and improving awareness within the population regarding explanations for their presenting complaint.
Data availability statement
The data used to support the findings of this study are available from the corresponding author upon request.
Competing interests
None declared
