A pilot of artificial intelligence telephone triage of patients with
suspected head and neck cancer
Abstract
1. The overwhelming majority of patients referred into
secondary care with suspected head and neck cancer (HNC) do not have
cancer (~95%). 2. During the COVID-19 pandemic
telephone triage of patients with suspected HNC was necessary. During
this time, a validated HNC risk-calculator, HaNC-RC-v2 (a set of
symptomatology-based questions) was recommended by ENT UK to stratify
patients into high or low risk of having HNC via telephone triage
[(1)](#ref-0001) 3. Ufonia, a digital health company which uses an
Artificial Intelligence (AI) voice assistant to automate clinical
conversations via telephone, and *INSTITUTION*, were awarded an SBRI
Healthcare grant to help develop an AI-delivered HNC triage telephone
call. This was based on the HaNC-RC-v2 and co-created with HNC patients
from the Heads2Gether charity via round-table discussions and one-to-one
sessions. 4. Twenty-nine patients underwent a clinician-supervised
AI-delivered HNC triage conversation as part of their standard telephone
consultation. 100% of calls were completed with an average agreement of
89% between the clinician and the AI system for all symptoms asked. The
technology was highly acceptable to patients with a median net promoter
score (NPS) score of 8 out of 10. 5. Novel technologies involving AI
automated telephone calls can be generated to remotely triage suspected
HNC patients. This technology may offer an exciting opportunity to help
departments triage suspected HNC referrals in an ever increasingly
resource pressurised NHS.