Methods

The protocol for this study was published in advance athttps://entintegrate.co.uk. This manuscript has been prepared with reference to the STROBE checklist for cohort studies.8

Ethical considerations

The Health Research Authority decision tool determined the study design to fall under the remit of audit, and so no ethical approval was required (available at:http://www.hra-decisiontools.org.uk/research/).

Study design and setting

A national prospective audit of the hospital management of tonsillitis and PTA by UK secondary care ENT departments was conducted, in parallel to an audit of acute epistaxis emergency care.9 The audit was developed and run by INTEGRATE, the UK ENT Trainee Research Network. All UK ENT departments were invited to participate via advertisement, and registration with local audit and Clinical Governance Departments was required. Sites could open at any point during the prospective data collection period.

Participants

Consecutive patients with tonsillitis and PTA, aged 18 years and older, and referred to ENT secondary care, were eligible for inclusion, whether managed by telephone advice or face-to-face review.

Data collection

Eligible cases were identified over a 12 week period, between 6th April and 29th June 2020. Each case was followed-up for 10 days. A standardised electronic case report form was created using Excel software (Microsoft Corporation, Washington, USA) and made available online (Supplementary material ***), incorporating data validation to encourage valid data entry and completeness. Data were initially held offline at each centre, and patient identifiable data were removed prior to submission to the project management team. Data were collected on: patient demographics; COVID-19 status; reviewing clinicians; IV medications and their effect; discharge medications; examinations performed; and PTA drainage.

Interim reports

The 12 week audit was divided into three 4-week periods. Two interim reports were produced (after periods 1 and 2) allowing for rapid feedback of management and preliminary outcomes to the UK ENT community. Both interim reports were disseminated electronically via ENTUK mailouts within 10 days of data submission and hosted online athttps://entuk.org andhttps://entintegrate.co.uk.

Data analysis

The primary outcome was unscheduled re-presentation to hospital within 10 days for ED discharges. This intended to assess the safety of the lower rates of admission anticipated.
Univariable binary logistic regression analysis was used to identify significant determinants of the primary outcome measure. The level of significance was set at <0.05 with Bonferroni corrections applied, where applicable. Analysis was performed using R statistical software (R Foundation, Vienna, Austria).