An algorithm to be used nationally for delirium screening and treatment in acute hospital wards: lessons from the development process

Abstract

Introduction: Delirium (acute confusion) affects 1 in 8 acute hospital inpatients and is highly distressing for patients and families. It is associated with increased length of hospital stay, risk of falls, mortality and institutionalisation. Early detection improves outcome, whereas established delirium (present for a few days) is difficult to treat. Thus screening for delirium in at-risk patients is key to early detection and treatment to improve outcomes. Description of policy context and objective: NICE guidelines recommend daily delirium screening for high risk patient groups. However, delirium screening and assessment is poorly performed in Irish hospitals (Irish National Audit of Dementia, 2014). We aimed to develop a simple algorithm for delirium screening on hospital wards, helping staff to identify those patients requiring daily delirium screening and outlining the action to be taken when a patient screened positive for delirium. Targeted population: Hospital in-patients on any general medical or surgical ward (i.e. excluding Intensive Care Unit or orthopaedic/neurosurgery wards), with ward nurses performing the daily delirium screening and hospital doctors acting on the abnormal result. Highlights: (innovation, Impact and outcomes) The interdisciplinary national algorithm development team firstly developed an evidence-based algorithm. Two versions were developed- one using the 4AT test (a well-recognised short delirium screening test) and the other using a quicker, deconstructed form. Both versions were piloted in three hospital sites, in a single ward at each site, with extensive user evaluation of the algorithm. Further modification was performed, followed by a second pilot in a single hospital on four wards. The version using the regular 4AT was preferred by users. Staff required pre-use training particularly on how to score patient refusals and using the tool in a patient with severe dementia. A Frequently Asked Questions guide, and an online training tool were developed, to support staff use of the algorithm, based on the pilot studies. This algorithm is being rolled out nationally in 2017 so impact and outcomes can’t be assessed as yet. Comments on transferability: The delirium algorithm was developed for use in all acute hospitals in Ireland. It is highly transferrable to other countries. Conclusions: Any new policy or guideline requires extensive user engagement at all stages of development. Different hospitals have different resources and training needs. Extensive piloting of our proposed algorithm identified areas that required targeted training. This allowed resources to be developed to facilitate adoption of the national policy, targeted to specific end-users needs.