Determining the Preliminary Efficacy of the Emergency Department Delirium Screening and Detection Program (ED-DDP)

Determining the Preliminary Efficacy of the Emergency Department Delirium Screening and Detection Program

Improving Delirium Screening and Detection for Older Adults Presenting to the Emergency Department (ED): A Novel ED Delirium Screening and Detection Program

Delirium occurs in up to 20% of older adults presenting to the Emergency Department (ED) and is associated with poor outcomes. Failure to identify patients with ED delirium not only prevents initiation of mitigation strategies, but is also a barrier to advancing the field in terms of evaluating management and clinical outcomes. This project studies the potential of an ED Delirium Detection Program (ED-DDP), developed to address the need for consistent and accurate ED delirium detection. This research will have two objectives: Aim 1 will conduct a pilot stepped wedge cluster randomized trial (SW-CRT) of the ED-DDP across 3 diverse EDs to determine preliminary efficacy of the detection training program, and Aim 2 will use a mixed methods approach to assess RE-AIM implementation outcomes (Reach, Efficacy, Adoption, Implementation, and Maintenance) of the training program. Aim 1 will consist of a multicomponent 1-day delirium champion workshop where the training is delivered, real-time direct observation/training of champions via telehealth, practical training of nurses throughout each ED by champions, and patient chart review. In Aim 2, the investigators will assess implementation outcomes using training logs, tele-observation, interviews with champions and nurses, and electronic medical record screening. The overarching aim of this proposal is to determine the preliminary efficacy of the training program for improving ED delirium screening, detection, and management in older adults, while also evaluating implementation outcomes of the program for champions/nurses. The investigators will use findings from this study to inform a full-scale SW-CRT to evaluate the impact of the program on patient outcomes at Northwell Health. The long-term goal of this study is to implement and disseminate a comprehensive ED-DDP that will improve screening, detection, and management of ED delirium in older adults.

Every three months, with the exception of when a site is in their implementation period, trained study staff members will conduct retrospective in-depth chart reviews on 250 randomly selected eligible patient visits per ED site. These chart reviews will assess presence of delirium as well as any corresponding delirium documentation, including the brief confusion assessment method (bCAM) and delirium mitigation and management strategies employed during the patient's ED visit. Charts will be pulled based on dates the participating site is within their control or intervention period (no chart review will take place during program implementation). A total of N=1,000 patients per ED will be reviewed. Two delirium experts will perform chart checks and adjudicate any uncertain cases.

This study is a pilot stepped wedge cluster randomized trial (SW-CRT) of three diverse emergency department (EDs) randomized to receive the intervention at 3-month intervals over a 15-month period. Time of crossover from the control to intervention period will be unidirectional, and each crossover sequence will allow for a 3-month implementation period where data will not be collected.

Delirium screening on initial assessment of all older adults presenting to the ED is considered standard of care, but not mandated. All health system EDs have access to a validated EHR integrated delirium screening tool (brief confusion assessment method, or bCAM) and all ED nurses have received training on bCAM use. The bCAM tool is an open parameter in the EHR. For positive bCAM screens, a message is displayed, "This patient has delirium, alert medical team, identify/ treat the underlying cause, ensure patient safety, nonpharmacologic interventions as first line, and avoid physical and chemical restraints". The control period will include chart reviews of patient visits occurring before implementation of the ED-DDP program.

ED leadership at each site have identified 5-10 delirium champions (nurse educators and managers, bedside nurses) based on their interest in delirium and commitment to program participation, including training of nurses by champions. The intervention period will include chart reviews of patient visits occurring after implementation of the ED-DDP program.

The intervention arm will have standardized delirium screening by nurses using the brief confusion assessment method (bCAM) in the EHR and will receive the ED-DDP. The ED-DDP will consist of: 1) a multicomponent 1-day delirium champion workshop; 2) real-time direct observation/training via telehealth (tele-training); and 3) training of nurses by champions. Champions will participate in a 1-day workshop, consisting of patient testimonials, small group discussions, delirium and screening tool use (didactics), and role-playing. Following the workshop, champions will receive 3 tele-training sessions; 2 to provide direct observation/feedback while the champion performs a bedside delirium screen, and 1 to provide direct observation/feedback of the champion providing training to the nurse. Once delirium champions complete training, they will provide bedside delirium training for all ED nurses.

Delirium detection is defined as the change in the proportion of documented delirium (i.e., the word delirium appears in the chart or there is a positive brief confusion assessment method [bCAM]) over time from baseline to 9 months (every 3 months), among patients with delirium. To identify the sample of patients with delirium, trained research staff will use a validated approach to conduct retrospective in-depth chart reviews to assess for ED presence of delirium (sensitivity 74%, specificity 83%) among a randomly sampled population of 3,000 patients (1,000 patients per ED or 250 patients per ED per data collection period). Two delirium experts will perform chart checks and adjudicate any uncertain cases. To assess change over time, four timepoints will be collected that each summarize the proportion of documented delirium within a 3-month time period. Each cluster will have at least one timepoint pre-intervention and one timepoint post-intervention, depending on cluster assignment.

Analysis of delirium screening will be defined as proportion of eligible patients screened for delirium using electronic health record (EHR) brief confusion assessment method (bCAM) documentation during their ED visit. Analysis will be conducted on a randomly sampled population of 3,000 patients (1,000 patients per ED or 250 patients per ED per data collection period). To assess change over time, four timepoints will be collected that each summarize the proportion of eligible patients who receive delirium screening within a 3-month time period. Each cluster will have at least one timepoint pre-intervention and one timepoint post-intervention, depending on cluster assignment.

Analysis will be defined as proportion of older adults with any order for safety precautions (fall, aspiration, or wandering) during their ED visit among a randomly sampled population of 3,000 patients (1,000 patients per ED or 250 patients per ED per data collection period). To assess change over time, four timepoints will be collected that each summarize the proportion of eligible patients who receive safety precautions within a 3-month time period. Each cluster will have at least one timepoint pre-intervention and one timepoint post-intervention, depending on cluster assignment.

Analysis will be defined as proportion of older adults receiving benzodiazepines, antipsychotics, diphenhydramine, or restraints for behavioral symptoms during their ED visit among a randomly sampled population of 3,000 patients (1,000 patients per ED or 250 patients per ED per data collection period). To assess change over time, four timepoints will be collected that each summarize the proportion of eligible patients who receive physical and chemical restraints for behavioral symptoms within a 3-month time period. Each cluster will have at least one timepoint pre-intervention and one timepoint post-intervention, depending on cluster assignment.

Inclusion Criteria: Presenting to a participating study site emergency department (ED) during control or intervention periods Survival to ED discharge or to hospital admission Exclusion Criteria: Presenting to a participating study site ED during ED-DDP implementation periods