Head and neck cancer risk calculator (HaNC‐RC)—V.2. Adjustments and addition of symptoms and social history factors

• DOI: 10.1111/coa.13511
• Publication Date: 2020-01-27T11:54:57Z
• Authors: Theofano Tikka; Kimberley Kavanagh; Anja Lowit; Pan Jiafeng; Harry Burns; Iain J. Nixon; Vinidh Paleri; Kenneth MacKenzie
• Keywords: Triage; Univariate analysis

Abstract Objectives Head and neck cancer (HNC) diagnosis through the 2‐week wait, urgent suspicion of (USOC) pathway has failed to increase early detection rates in UK. A head risk calculator (HaNC‐RC) previously been designed aid referral high‐risk patients USOC clinics (predictive power: 77%). Our aim was refine HaNC‐RC its prediction potential. Design Following sample size calculation, prospective data collection statistical analysis criteria outcomes. Setting Large tertiary care centre Scotland. Participants 3531 new seen routine, (HaN) clinics. Main outcome measures Data collected were as follows: demographics, social history, presenting symptoms signs HNC diagnosis. Univariate multivariate regression performed identify significant predictors HNC. Internal validation using 1000 bootstrapping estimate model diagnostics included area under receiver operator curve (AUC), sensitivity specificity. Results The updated version (HaNC‐RC v.2) includes age, gender, unintentional weight loss, smoking, alcohol, positive negative It achieved an AUC 88.6% with two recommended triage cut‐offs (cut‐off: 7.1%; sensitivity: 85%, specificity: 78.3%) or 2.2%; 97.1%; specificity 52.9%). This could redistribute from current 60.9%–85.2%, without affecting total numbers each clinical setting. Conclusions use v.2 a potential both identifying at high thought but also improving health service delivery practices by reducing number inappropriately referrals.