Background Delirium is a serious and distressing condition, which commonly affects terminally ill people. Delirium often goes undiagnosed and hence under-treated. Delirium detection tools have been shown to improve detection, over clinical judgement alone (Woodhouse, Siddiqi, Boland, et al. BMJ Support Palliat Care. 2022;12(2):187). Delirium guidelines exist (Scottish Intercollegiate Guidelines Network. SIGN 157. Risk Reduction and Management of Delirium. 2019; Scottish Palliative Care Guidelines. Delirium [internet]), yet there are challenges with implementing these into routine clinical practice. We sought to improve delirium care in two hospice inpatient units in Scotland. As a starting point, we conducted an audit to describe current practice, and provide a baseline against which future audits can compare.

Aim To audit delirium assessment practices within two hospice inpatient units. Delirium prevalence was retrospectively assessed using an approximation of the DSM-5 delirium criteria (American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Publishing, 2013).

Method Retrospective audit of case notes of 100 terminally ill people admitted to two hospice inpatient units between 16/8/23 and 13/11/23 (n=50 per site). Audit standards focused on 4AT delirium detection tool use and diagnosis.

Results Retrospective review using DSM-5 criteria suggested delirium prevalence was 28% (28/100). However, only 11% (11/100) had the diagnosis ‘delirium’ documented in their case notes. Of the 28 patients with delirium, according to DSM-5 criteria, 50% (14/28) were not screened using the 4AT delirium detection tool, nor had ‘delirium’ been documented in their case notes. Overall, 25% of patients (25/100) were screened for delirium using the 4AT - though practice varied between sites, with 34% screened in one hospice and 16% in the other. Of the 75% of patients not screened using the 4AT, explanation was only documented for 3% (2/75), when patients were reported as untestable.

Conclusion The audit showed limited and inconsistent delirium detection tool use, under-diagnosis of delirium, and inadequate use of the term ‘delirium’, as documented in case notes of patients with delirium. Use of the term ‘delirium’ is the first step in ensuring optimal care through adherence to best practice.