br The number of referrals correlates with
The number of referrals correlates with socioeconomic status (SES),3 which is not surprising given that cancer of the head and neck is known to be more prevalent in patients from lower-status groups.4 However, to our knowledge, the association between SES and referrals on the two-week path-way has had little attention, and this might be important, as delays could be linked to later stages of disease at diagnosis and poorer survival.5
As the catchment area for our hospital includes some of the most deprived wards in England, the aim of this audit was to analyse the two-week referrals over a 616-91-1 of three months and to focus on the association with SES in terms of clinical characteristics, source of referral, and cancer rates.
Patients and methods
The sample comprised patients with suspected cancer of the head and neck who were referred to our hospital on the two-week pathway between 3 July and 29 September 2017. We obtained information on smoking status, intake of alcohol, date of birth, postcode (for small area deprivation), and delay since the symptom(s) were first noticed, from both electronic and paper records. The data on smoking, alcohol, and time since the patient first became aware of the symptoms varied considerably and required broad categorisation for the pur-poses of analysis rather than numerical analysis of pack years, units of alcohol consumed, and number of days since the symptoms started. Details on the latter depended on patients’ recall about when they were first seen by their general prac-titioner (GP) or dentist.
The Index of Multiple Deprivation (IMD)6 is the official measure of relative deprivation for small areas in England. The IMD 2015 is based on seven domains that were all based on indicators that related mainly to the tax year 2012/13. Every neighbourhood in England is ranked from 1 (most deprived) to 32 844 (least deprived), but there is no defini-
tive cut-off below which an area is considered “deprived”. We analysed the deprivation quintiles and, since most of our patients lived in the most deprived quintile, compared them with patients who lived in less deprived neighbour-hoods.
Fisher’s exact test was used to compare groups with regard to delay (delay by patients of less than 30 days, and profes-sional delay of more than 14 days), a diagnosis of cancer, the presence of specific symptoms, and referral by a den-tist. Missing data are reflected by the varying denominators in the tables. Probabilities of less than 0.05 were considered significant.
The University Hospital Clinical Audit Department approved the study.
Patients’ and professional delay. Data are number (%).
Total cases Delay by patients from first
p value* Patients’ delay
Professional delay from p value
awareness of symptoms to decision
referral to first
Asymptomatic Unknown >14 days
Source of referral:
Specialty referred to:
Alcohol intake (units/week):
No. of symptoms:
IMD2015: quintile for neighbourhood deprivation; GDP: general dental practitioner; GMP: general medical practitioner.
p values from Fisher’s exact test throughout the table.
∗ p values compare patients’ groups with regard to the percentage with a patient delay < 30 days. ∗∗ p values are for symptom present compared with symptom absent.
Outcome of referral. Data are number (%).
Delay by patients between first being aware of their symp-toms and presenting to primary care was known for 297 while three were asymptomatic. One quarter (77/297, 26%) of these delays were under 30 days (Table 1), 34% were between one and three months, 23% between three and 12 months, and 18% were over one year. Professional delay from referral from primary care to first being seen by a hospital specialist (Table 1) was more than 14 days for 42/388 (11%), with a median (IQR) overall delay of 9 (7-13) days, and a median (range) of 20 (15-43) days for those delayed more than 14 days. In 20 of the 42 delayed cases aneuploidy was because of cancellation or non-attendance by the patient. The remain-der reflected a lack of clinic capacity. A swelling or lump indicated a greater likelihood of presentation within 30 days (p = 0.001). Otherwise, there were no significant associations, but there were trends towards earlier presentation in those who had never smoked and those who lived in the most deprived neighbourhoods. Other trends suggested later pre-sentation by patients with hoarseness or a sore throat, those who consumed more than 40 units of alcohol/week, and those who were over 75 years of age. There were no notable asso-ciations between patient factors and professional delays of more than 14 days (Table 1) or between delay by patients and professional delay (results not shown).