Discussion

Compared to pre-pandemic data, there was a decline in 2-week-wait referrals to our regional specialist unit during the study period, of 7.3 per cent. This trend conforms with other studies, which also showed a decrease in urgent head and neck cancer referrals during the pandemic.^{Reference Bhamra, Gorman, Arnold, Rajah, Jolly and Nieto3} The trend likely results from limited access to primary care and altered health-seeking behaviours during the Covid-19 pandemic.^{Reference Greenwood and Swanton4}

Nine per cent (9.2 per cent) of patients referred on the two-week-wait pathway were diagnosed with cancer. This conversion rate is comparable to that in current literature, which is reported to be between 6 per cent and 11.8 per cent.^{Reference Langton, Siau and Bankhead5} Recent studies have observed a national downwards trend of head and neck cancer two-week-wait conversion rates, which is suggested to be mainly due to an increase in the number of two-week-wait referrals from the community.^{Reference Langton, Siau and Bankhead5} This reflects the trend seen in our regional unit, with a rise in the number of two-week-wait referrals over the years, outside of pandemic times. There are concerns that this trend could potentially overwhelm secondary specialist services.^{Reference Pullyblank, Silavant and Cook6}

In our study, despite pandemic-related disruption, 100 per cent of the urgent referrals were able to be accommodated within 14 days, without significant impact on the waiting times of routine referrals. It is possible that telephone clinics are more efficient than traditional face-to-face appointments and result in an improved patient flow.^{Reference Weerasooriya, Stewart and Thomas7}

Historically, the most common types of malignancy diagnosed from the head and neck cancer urgent pathway have been oral cavity cancer, lymphoma, laryngeal cancer and thyroid cancer.^{Reference Mettias, Charlton and Ashokkumar8} The same pattern was observed in this study. This may suggest that the Covid-19 pandemic has not caused any major deviations from the pre-pandemic head and neck cancer referral patterns.

Under the current 2-week-wait pathway, patients with cancer should receive their initial diagnosis within 28 days of the initial referral.^{Reference Mettias, Charlton and Ashokkumar8} Three cases in the study cohort were diagnosed outside of this window and therefore are considered to be late diagnoses. There are limited data on the late or missed diagnosis rate of the head and neck cancer two-week-wait pathway. Recent communication from ENT-UK suggests that telephone consultations are 1.4 per cent more likely to result in a missed cancer diagnosis compared to face-to-face consultations.¹ A recent 16-week prospective study suggested that the late diagnosis rate of the head and neck cancer 2-week-wait pathway was 0.6 per cent overall.^{Reference Hardman, Tikka, Paleri and ;9} This is comparable to the 0.28 per cent rate found in our study.

The inability to perform physical examinations and diagnostic FNE during telephone consultations is unequivocally a major reason why these cancer cases were missed. It could be speculated that the missed cancer cases would otherwise have been diagnosed by direct visualisation in a face-to-face consultation.

Interestingly, two out of the three missed cancer cases had presented with a hoarse voice. A recent study showed there is limited value in assessing voice disorders over the telephone.^{Reference Watters, Miller, Kelly, Burnay, Karagama and Chevretton10} Clinicians can easily misjudge the severity of hoarseness during a telephone consultation and offer these patients deferred appointments. Future studies could examine associations between presenting symptoms and late diagnoses in urgent head and neck cancer telephone clinics on a larger scale.

A late diagnosis does not inevitably lead to a worse clinical outcome. Among the delayed diagnoses in the study, only one case, with a longer time to follow up, required escalated intervention because of the late diagnosis; overall prognosis was, however, deemed to be unaffected. The other two cases, although managed outside of the 31- and 62-days pathway, did not receive altered or escalated treatment. This is likely accounted for by the closer follow-up reviews to their first appointments, at four and six weeks.

The limitation of the study lies within the relatively small sample size. Only three late diagnoses were identified, which makes any further analysis into factors associated with late diagnoses non-meaningful. A larger sample size could be obtained by extending the study period. However, this would defeat the purpose of the study, which aimed to evaluate the impact of the Covid-19 pandemic on the head and neck cancer two-week-wait pathway. Whilst it was not the objective of this study to determine the predictive accuracy of the Head and Neck Cancer Risk Calculator, as this has already been covered by multiple studies,^{Reference Paleri, Hardman, Tikka, Bradley, Pracy and Kerawala2,Reference Hardman, Tikka, Paleri and ;9} it is interesting to observe that 20 per cent of the study cohort with a positive cancer diagnosis were stratified as low risk, including all three of the late diagnosis cases.

Although cases of Covid-19 are on the decline, telephone clinics will likely remain in some capacity after the pandemic. This study reported our initial experience of adapting telemedicine for the head and neck cancer urgent referral pathway, and found that a small proportion of cancer cases were missed. Clinicians will be mindful that telephone clinics, whilst a pragmatic means to maintain patient flow during the pandemic, could result in late diagnosis and harm compared to traditional face-to-face appointments.