The role of AI-driven communication in delirium prevention, detection, and care for critically ill ICU patients: A systematic review with inductive thematic synthesis.

To identify, describe and synthesise what is known about family carers' experiences and perspectives of engagement in delirium prevention and care for adults in hospital. Systematic review and synthesis of qualitative evidence. Comprehensive literature search within PubMed, CINAHL, EMBASE, Scopus, Cochrane Central and PsycInfo databases to August 2022. Peer reviewed original qualitative research published in English. Data were extracted using Covidence systematic review software. Methodological quality was reviewed against the Critical Appraisal Skills Program (CASP) Qualitative Checklist. Thematic synthesis was used to develop analytical themes. Confidence in the Evidence from Reviews of Qualitative Research (GRADE-CERQual) framework was applied to findings. Of 3429 records identified, 29 met criteria for final inclusion. Studies included a range of settings (intensive care, medical and surgical wards) and family carer types (spouse, daughter, parent and friend). Three primary themes Shared Caregiving, Partnership and Support, Communication for Shared Understanding and three cross-cutting contextual themes Care Transitions, Family Carer Context and Hospital Context were identified. Family carers have mixed experiences of involvement in delirium care and prevention. Family carer engagement in care impacted perceptions of care quality and their own self-efficacy. Where person-centred care expectations were not met, some family carers adopted compensatory or care contingency strategies. Information sharing and timely support from hospital staff who understood the carer context supported carer involvement. Supportive physical environments and addressing power imbalances allowed greater engagement. Family carers often wish to be involved in delirium prevention and care, but need to be recognised as individuals, listened to, informed and supported to optimise their contribution. The review findings can guide health professional and decision makers to optimise family carer involvement in delirium care programs. PROSPERO [CRD42020221854]. ENTREQ. No Patient or Public Contribution.

SummaryIntroductionDelirium is common in patients who are critically ill and associated with increased mortality and long‐term cognitive impairment. The objective of this study was to explore the barriers and facilitators to the use of existing ICU delirium identification tools and implementation of care protocols.MethodsSemi‐structured interviews were conducted with healthcare professionals working in UK NHS ICUs. Purposive sampling was utilised to recruit a diverse range of participants to account for profession; user status of delirium identification tools or care packages; size of unit; and UK location. Thematic analysis was undertaken using the Theoretical Domains Framework.ResultsTwenty‐one ICU healthcare professionals were interviewed from 20 hospitals. Participants included consultants (n = 9); nurses (n = 8); advanced critical care practitioners (n = 2); and allied health professionals (n = 2). Five major barriers to effective delirium care emerged: lack of prioritisation; lack of structured ICU delirium care protocols; inability to implement interventions due to physical space constraints and/or lack of resources; loss of experienced staff; and changes in ICU nursing culture. Facilitators included the presence of clear protocols; continued staff training; increased awareness; delirium champions; post‐ICU follow‐up clinics; family engagement; effective communication; and the use of digital prompts as a component of mandatory electronic documentation.DiscussionAlthough ICU delirium is acknowledged as being important by clinical staff, management is often hindered by systemic and cultural barriers. Healthcare professionals highlighted the need for protocol‐driven care, enhanced training and awareness, and the inclusion of families in care processes. These findings will inform the design of a multicomponent care package to improve delirium care in the ICU.

Delirium is a commonly observed problem for adult patients in the intensive care unit (ICU) that is associated with increased mortality, increased hospital length of stay, and long-term disability of ICU survivors. This makes prevention and prompt treatment of delirium imperative goals for critical-care providers. Bedside critical-care nurses are uniquely poised to champion efforts to prevent, detect, and treat delirium but oftentimes are not aware of the most recent evidence-based guidelines or the potential long-term implications for the ICU survivor. Nurses have reported higher levels of confidence in detecting delirium, increased knowledge of delirium, and enhancement of patient care skills after receiving formal didactic educational training. The objectives of this quality improvement project were to develop, implement, and evaluate a nursing education program for critical-care nurses that presented a protocol for the prevention and management of delirium in adult ICU patient populations, as well as improve nurses' familiarity, comfort, and compliance with using a standardized evaluation method for delirium and intervention care bundle. A didactic training program for bedside critical-care nurses was developed and implemented. Upon completion of the educational sessions, a daily bedside delirium screening and care bundle protocol were implemented for all patients in ICUs throughout the facility. Bedside critical-care nurses were invited to participate in the formal teaching sessions. This included nursing staff from each of the 3 ICUs in this facility, as well as supplemental "float pool" staff. Pre-educational and posteducational nursing attitudes and self-perceived confidence levels regarding delirium care were measured using a 5-point Likert scale. Pre-educational and posteducational nursing knowledge was measured using raw scores obtained from a 15-item multiple choice test. Following the implementation of the delirium screening and care bundle protocol, nursing staff compliance with screening and care bundle items was measured, as well as overall incidence of positive delirium screenings among all screenings completed. A sample of 49 nurses participated in the formal educational teaching sessions. All 5 nursing attitude and perceived confidence statements measured before and after the educational sessions showed a significant increase in positive perceptions overall (P < .0001). Overall mean posteducation knowledge test raw scores showed a significant improvement from pre-educational scores (70% ± 12.8% vs 95% ± 6.9%; P < .0001). Once-daily ICU delirium screenings and care bundle interventions were initiated for all ICU patients; overall compliance during the measurement period was 56.3% (598 of 1061 possible delirium screenings and protocols completed). Of all daily patient screenings performed, 20.4% resulted positive for ICU delirium. All patients who received the care bundle interventions received the interventions uniformly, regardless of clinical delirium status. This quality improvement project demonstrates that a formal didactic training program for ICU nurses can result in increased awareness and knowledge of ICU delirium and adequately prepare them for how to properly screen and treat patients. Although overall compliance with implementing daily ICU delirium screenings and care bundle interventions was relatively low, those patients who did receive screening and interventions demonstrated that the incidence of positive delirium screenings at this facility is low compared with reported figures. In addition, patients who did receive the appropriate care bundle interventions did so in a uniform manner, regardless of the presence or absence of clinical delirium.

Purpose: This study aimed to identify the mediating effect of delirium care self-efficacy and the moderated mediation effect of delirium stress on the relationship between ICU nurses' delirium knowledge and delirium care. Methods: Participants included 186 nurses working in the ICUs of two university hospitals located in different cities. Data were collected through an online survey conducted from August 2 to September 9, 2024. Analyses were performed using SPSS/WIN 27.0 and PROCESS Macro version 4.1. Results: Delirium care self-efficacy mediated the relationship between delirium knowledge and delirium care, showing a significant indirect effect (β=.17, CI [0.10, 0.24]). The direct effect of delirium knowledge on delirium care was not significant (β=.02, p=.737, CI [-0.23, 0.33]). The moderated mediation effect of delirium stress on the relationship between delirium care self-efficacy and delirium care was also not significant (β=-.00, CI [-0.01, 0.00]). Conclusion: This study confirmed the complete mediating effect of delirium care self-efficacy in the relationship between ICU nurses' delirium knowledge and delirium care. However, delirium stress did not exhibit a significant moderated mediation effect, suggesting that stress levels did not influence the mediating role of self-efficacy. These findings indicate that enhancing knowledge alone is insufficient to improve delirium care. A multidisciplinary approach is required to strengthen nurses' self-efficacy through participatory educational methods while also implementing systematic strategies to manage and reduce delirium stress in clinical settings.

Background: Delirium is a distressing condition often experienced by hospice in-patients. Increased understanding of current multidisciplinary care of delirium is needed to develop interventions in this setting. Aim(s): To explore hospice staff and volunteers’ practice, its influences and what may need to change to improve hospice delirium care. Design: Qualitative interview study using behaviour change theory from a critical realist stance. Setting/participants: Thirty-seven staff, including different professional groups and roles, and volunteers were purposively sampled from two in-patient hospices. Results: We found that participants’ practice focus was on managing hyperactive symptoms of delirium, through medication use and non-pharmacological strategies. Delirium prevention, early recognition and hypoactive delirium received less attention. Our theoretically-informed analysis identified this focus was influenced by staff and volunteers’ emotional responses to the distress associated with hyperactive symptoms of delirium as well as understanding of delirium prevention, recognition and care, which varied between staff groups. Non-pharmacological delirium management was supported by adequate staffing levels, supportive team working and a culture of person-centred and family-centred care, although behaviours that disrupted the calm hospice environment challenged this. Conclusions: Our findings can inform hospice-tailored behaviour change interventions that develop a shared team understanding and engage staff’s emotional responses to improve delirium care. Reflective learning opportunities are needed that increase understanding of the potential to reduce patient distress through prevention and early recognition of delirium, as well as person-centred management. Organisational support for adequate, flexible staffing levels and supportive team working is required to support person-centred delirium care.

Studies have emphasized the limited availability of data regarding the current practices of nurses in delirium care for patients in acute care settings. Similarly, there is a lack of information regarding the obstacles and facilitators in providing optimal care for the prevention, detection, and management of delirium. Aim To describe nurses’ practices about delirium care in acute care patients and their perceptions about barriers and facilitators regarding optimal care. Method: A two phase, multi method design was used. The quantitative phase utilized a self-reported survey to assess nurses’ knowledge, practice, confidence, and collaboration regarding delirium care. The qualitative phase employed focus groups to complement and explore survey data in-depth. Recruitment took place on nine acute surgical and medical units across two university-affiliated hospitals in Canada, involving nurses in direct patient care. Results 231 nurses reveal diverse insights into delirium care in acute settings and survey participants showed solid knowledge of delirium symptoms and outcomes, yet 23% did not identify the hypoactive form’s prevalence. While 81% reported receiving information on delirium screening, challenges included time constraints and misuse of detection tools. Qualitative themes highlighted communication’s vital role, challenges posed by delirium presentations, the significance of staff support, time constraints, the impact of experienced staff, the role of families, and the importance of additional resources for optimal delirium care. Conclusion Findings align with existing literature, emphasizing the multifaceted nature of delirium care and the need for tailored approaches, education, and collaborative strategies to improve overall care quality.

BackgroundAcute, transient, but sometimes persistent, delirium is characterized by a sharp disruption in attention, consciousness, and cognitive function, and can be caused by many medications and disorders. Delirium occurrence and negative consequences, such as falls and functional decline, can be decreased with multifactorial prevention and timely detection.AimsTo describe current clinical practice in relation to the prevention, assessment, and management of delirium in Irish hospitals; awareness-raising and educational activities; and barriers to good practice.MethodsOn World Delirium Awareness Day (15th March 2023), a global survey was conducted of delirium prevalence and care. A senior clinical staff member on each participating ward reported on delirium prevalence at 8AM and 8PM, and on usual ward practice; this data was entered into an online survey by a data collector (typically a clinician from the site, visiting several wards to record data). This study reports data from Irish hospitals.ResultsIn total, 132 wards from 15 hospitals across Ireland participated. Almost 60% of wards used ‘personal judgment’ for delirium assessment. Having at least one delirium training session in the preceding year was associated with greater use of a formal assessment tool (60.3% versus 18.8%; p < 0.001). Wards reported staff training/education as the main priority to improve care, but 72.7% of wards identified insufficient time to train staff as a key barrier.ConclusionsClinical practice related to delirium care requires improvement. Awareness raising and staff training require more focus and time in busy clinical settings.

BackgroundDelirium is a serious and distressing condition which commonly affects people with life-limiting illness. Around one-third of those admitted to hospices have delirium, and many more experience delirium in the...

The current study aimed to describe the effect of an educational program on RN-initiated efforts to involve family caregivers in delirium care. A descriptive qualitative study was performed. A purposive sample of 12 RNs who participated in a one group, pre-post evaluation of a delirium educational program, and a nominated sample of six family caregivers of patients who had been cared for by RNs in the program participated in individual, in-depth interviews. The qualitative findings indicated that the inclusion of a delirium brochure in ward orientation on admission, with reinforcement during shift changes, and RN-initiated engagement with family caregivers promoted family caregiver involvement in delirium prevention. Further studies in South Korea are needed to determine which family-centered nursing care interventions are culturally appropriate and most effective for RNs and family caregivers in delirium care. [Journal of Gerontological Nursing, 43(12), 44-51.].

Background: Delirium is a high prevalent postoperative complication in older cardiac surgery patients and can have drastic consequence for the patient. Preventive interventions, diagnosis and treatment of delirium require specialized knowledge and skills. Objective: To gain insight in the current opinion and beliefs of nurses in hospitals concerning prevention, diagnosis and treatment of delirium in older hospital patients in general and specifically in older cardiac surgery patients. Methods: In a cross-sectional study from February to July 2010, we distributed a survey on beliefs on delirium care among 368 nurses in three hospitals in the Netherlands, in one hospital in all wards with older patients and in two hospitals in the cardiac surgery wards only. Results: Although in literature incidence rates up to 54.9% in cardiac surgery patients in hospitals are reported, with a response rate of 68% (250), half of the nurses believe that the incidence of delirium is not even 10%. Two thirds think that delirium in patients is preventable. Although, the Delirium Observation Scale is most often used for screening delirium, nearly all nurses do not routinely screen patients for delirium. Opinions on delirium of nurses working in cardiac surgery wards did not differ from nurses caring for older patients in other hospital wards. Conclusions: Nurses do have knowledge on delirium care, but there is a gap between the reported incidence in literature and the estimation of the occurrence of delirium by nurses. A two-way causal relationship emerges: because nurses underestimate the occurrence, they do not screen patients on a routine basis. And because they do not screen patients on a routine basis they underestimate the incidence.

Background An interdisciplinary approach is fundamental for effective prevention and treatment of delirium. Pharmacists could play a role in identifying and resolving medication-related delirium. However, little is known about their role in delirium care. Objective The main purpose of this survey was to assess the current practice and opinions of pharmacists concerning their involvement in screening, prevention and treatment of delirium. Setting Pharmacists in public and private hospitals in Australia. Method A cross-sectional survey was conducted using a pilot tested web-based questionnaire which was distributed primarily via a link in the electronic newsletter of the Society of Hospital Pharmacists of Australia. Main outcome measure Number and proportion of respondents answering questions related to the practice and perceptions of pharmacists in delirium management. Results Responses from 106 pharmacists were included in the analysis. Most respondents believed that pharmacists could play a role in prevention (92%) and screening (62%) of patients for delirium. However, in practice only 8% of pharmacists reported that they had ever screened a patient for delirium using a validated tool and 79% indicated that pharmacists were never or rarely involved in delirium treatment. When pharmacists did make recommendations half of the respondents said that pharmacists' recommendations were frequently or always accepted by the delirium treating teams. Conclusion Hospital pharmacists are underutilised in the prevention and management of delirium. Strategies to increase their involvement in the prevention and management of delirium should be implemented.

This issue of JAMIA focuses on natural language processing (NLP) techniques for clinical-text information extraction. Several articles are offshoots of the yearly ‘Informatics for Integrating Biology and the Bedside’ (i2b2) (http://www.i2b2.org) NLP shared-task challenge, introduced by Uzuner et al ( see page 552 )1 and co-sponsored by the Veteran's Administration for the last 2 years. This shared task follows long-running challenge evaluations in other fields, such as the Message Understanding Conference (MUC) for information extraction,2 TREC3 for text information retrieval, and CASP4 for protein structure prediction. Shared tasks in the clinical domain are recent and include annual i2b2 Challenges that began in 2006, a challenge for multi-label classification of radiology reports sponsored by Cincinnati Children's Hospital in 2007,5 a 2011 Cincinnati Children's Hospital challenge on suicide notes,6 and the 2011 TREC information retrieval shared task involving retrieval of clinical cases from narrative records.7 Although NLP research in the clinical domain has been active since the 1960s, progress in the development of NLP applications for clinical text has been slow and lags behind progress in the general NLP domain. There are several barriers to NLP development in the clinical domain, and shared tasks like the i2b2/VA Challenge address some of these barriers. Nevertheless, many barriers remain and unless the community takes a more active role in developing novel approaches for addressing the barriers, advancement and innovation will continue to be slow. Historically, there have been substantial barriers to NLP development in the clinical domain. These barriers are not unique to the clinical domain: they also occur in the fields of software engineering and general NLP. ### Lack of access to shared data Because of concerns regarding patient privacy and worry about revealing unfavorable institutional practices, hospitals and clinics have been extremely reluctant to allow access to clinical data for researchers from outside … Correspondence to Dr Wendy W Chapman, Department of Biomedical Informatics, University of California San Diego, 9500 Gilman Dr, Bldg 2 #0728, La Jolla, California, USA; wwchapman{at}ucsd.edu

Efficacy of simulation-based education on intensive care nurses' knowledge, self-efficacy, and critical thinking in delirium care: A randomised controlled trial.

Automating the identification of risk for developing hospital delirium with models that use machine learning (ML) could facilitate more rapid prevention, identification, and treatment of delirium. However, there are very few reports on the performance of ML models for delirium risk stratification in live clinical practice. To report on development, operationalization, and validation of a multimodal ML model for delirium risk stratification in live clinical practice and its associations with workflow and clinical outcomes. This quality improvement study developed an ML model supported by automated electronic medical records to stratify the risk of non-intensive care unit delirium in live clinical practice using the Confusion Assessment Method as the diagnostic reference standard, with an iterative model update method. Data from patients aged at least 60 years admitted to non-intensive care units at Mount Sinai Hospital between January 2016 and January 2020 were used to train and test the ML model presented. The model was validated in live clinical practice from March 2023 to March 2024. Analysis of the model's associations with workflow and clinical outcomes was conducted retrospectively in 2024, comparing hospitalized patients prior to deployment of any model version (pre-ML cohort) and during model clinical deployment (post-ML cohort). Outcomes of interest were area under the receiver operating characteristic curve, monthly delirium detection rates, median length of hospital stay, and daily doses of opiate, benzodiazepine, and antipsychotic medications administered. The overall sample included 32 284 inpatient admissions (mean [SD] age, 73.56 (9.67) years, 15 157 [46.9%] women). A total of 25 261 inpatient admissions of older patients with both medical and surgical primary diagnoses represented the combined model testing and training cohort (median age, 73.37 [66.42-81.36] years) and live clinical deployment validation cohort (median [IQR] age, 72.11 [62.26-78.97] years), while 7023 inpatient admissions of older patients with both medical and surgical primary diagnoses represented the combined pre-ML (median [IQR] age, 74.00 [68.00-81.00] years) and post-ML (median [IQR] age, 75.33 [68.34-82.91] years) cohorts. The model presented is a fusion of electronic medical record patient data features and clinical note features processed by natural language processing. The results of model validation in live clinical practice included an area under the curve of 0.94 (95% CI, 0.93-0.95). Median (IQR) monthly delirium detection rates of inpatients assessed for delirium with the Confusion Assessment Method increased from 4.42% (95% CI, 3.70%-5.14%) in the pre-ML cohort to 17.17% (95% CI, 15.54%-18.80%) in the post-ML cohort (P < .001). Post-ML vs pre-ML cohorts received lower daily doses of benzodiazepines (median [IQR] 0.93 [0.42-2.28] diazepam dose equivalents vs 1.60 [0.66-4.27] diazepam dose equivalents; P < .001) and olanzapine (median [IQR], 1.09 [0.38-2.46] mg vs 2.50 [1.17-6.65] mg; P < .001). This quality improvement study demonstrates the feasibility of a novel multimodal ML model to automate delirium risk stratification in live clinical practice. The model demonstrated acceptable performance in live clinical practice and may facilitate resource allocation to enhance delirium identification and care.

Efficiency is a paramount concern in neural network research, impacting the scalability and practicality of AI models.This paper delves into the fusion of Natural Language Processing (NLP) and computer vision, focusing on the ground breaking Squeeze BERT model, which strives to bring efficiency principles from the realm of computer vision to NLP.While deep learning has significantly improved language understanding and generation in NLP, it has also led to resource-intensive models.Computer vision has, in contrast, pioneered efficient architectures for image recognition.This paper explores whether NLP can glean insights from computer vision to develop more efficient models.The paper discusses efficient computer vision architectures, such as Convolutional Neural Networks (CNNs), model pruning, and knowledge distillation, as sources of inspiration for Squeeze BERT's design.Key elements include convolutional token embeddings and depth wise separable convolutions, which reduce the model's size and complexity.Experimental results demonstrate Squeeze BERT's superiority over traditional NLP models in various benchmarks, showcasing the potential of this cross-disciplinary approach.It concludes that the integration of computer vision principles into NLP architecture can lead to highly efficient models that cater to resource-constrained environments.This work encourages future exploration of other computer vision techniques in NLP, fostering interdisciplinary collaboration for more efficient and effective neural networks.