OBJECTIVE: The American Medical Informatics Association (AMIA) Task Force on Diversity, Equity, and Inclusion (DEI) was established to address systemic racism and health disparities in biomedical and health informatics, aligning with AMIA's mission to transform healthcare. AMIA's DEI initiatives were spurred by member voices responding to police brutality and COVID-19's impact on Black/African American communities. MATERIALS AND METHODS: The Task Force, consisting of 20 members across 3 groups aligned with AMIA's 2020-2025 Strategic Plan, met biweekly to develop DEI recommendations with the help of 16 additional volunteers. These recommendations were reviewed, prioritized, and presented to the AMIA Board of Directors for approval. RESULTS: In 9 months, the Task Force (1) created a logic model to support workforce diversity and raise AMIA's DEI awareness, (2) conducted an environmental scan of other associations' DEI activities, (3) developed a DEI framework for AMIA meetings, (4) gathered member feedback, (5) cultivated DEI educational resources, (6) created a Board nominations and diversity session, (7) reviewed the Board's Strategic Planning for DEI alignment, (8) led a program to increase diversity at the 2020 AMIA Virtual Annual Symposium, and (9) standardized socially-assigned race and ethnicity data collection. DISCUSSION: The Task Force proposed actionable recommendations that focused on AMIA's role in addressing systemic racism and health equity, helping the organization understand its member diversity. CONCLUSION: This work supported marginalized groups, broadened the research agenda, and positioned AMIA as a DEI leader while reinforcing the need for ongoing transformation within informatics.

OBJECTIVES: The coronavirus disease 2019 (COVID-19) is a resource-intensive global pandemic. It is important for healthcare systems to identify high-risk COVID-19-positive patients who need timely health care. This study was conducted to predict the hospitalization of older adults who have tested positive for COVID-19. METHODS: We screened all patients with COVID test records from 11 Mass General Brigham hospitals to identify the study population. A total of 1495 patients with age 65 and above from the outpatient setting were included in the final cohort, among which 459 patients were hospitalized. We conducted a clinician-guided, 3-stage feature selection, and phenotyping process using iterative combinations of literature review, clinician expert opinion, and electronic healthcare record data exploration. A list of 44 features, including temporal features, was generated from this process and used for model training. Four machine learning prediction models were developed, including regularized logistic regression, support vector machine, random forest, and neural network. RESULTS: All 4 models achieved area under the receiver operating characteristic curve (AUC) greater than 0.80. Random forest achieved the best predictive performance (AUC = 0.83). Albumin, an index for nutritional status, was found to have the strongest association with hospitalization among COVID positive older adults. CONCLUSIONS: In this study, we developed 4 machine learning models for predicting general hospitalization among COVID positive older adults. We identified important clinical factors associated with hospitalization and observed temporal patterns in our study cohort. Our modeling pipeline and algorithm could potentially be used to facilitate more accurate and efficient decision support for triaging COVID positive patients.

OBJECTIVE: To propose an algorithm that utilizes only timestamps of longitudinal electronic health record data to classify clinical deterioration events. MATERIALS AND METHODS: This retrospective study explores the efficacy of machine learning algorithms in classifying clinical deterioration events among patients in intensive care units using sequences of timestamps of vital sign measurements, flowsheets comments, order entries, and nursing notes. We design a data pipeline to partition events into discrete, regular time bins that we refer to as timesteps. Logistic regressions, random forest classifiers, and recurrent neural networks are trained on datasets of different length of timesteps, respectively, against a composite outcome of death, cardiac arrest, and Rapid Response Team calls. Then these models are validated on a holdout dataset. RESULTS: A total of 6720 intensive care unit encounters meet the criteria and the final dataset includes 830 578 timestamps. The gated recurrent unit model utilizes timestamps of vital signs, order entries, flowsheet comments, and nursing notes to achieve the best performance on the time-to-outcome dataset, with an area under the precision-recall curve of 0.101 (0.06, 0.137), a sensitivity of 0.443, and a positive predictive value of 0. 092 at the threshold of 0.6. DISCUSSION AND CONCLUSION: This study demonstrates that our recurrent neural network models using only timestamps of longitudinal electronic health record data that reflect healthcare processes achieve well-performing discriminative power.

OBJECTIVE: There are signals of clinicians' expert and knowledge-driven behaviors within clinical information systems (CIS) that can be exploited to support clinical prediction. Describe development of the Healthcare Process Modeling Framework to Phenotype Clinician Behaviors for Exploiting the Signal Gain of Clinical Expertise (HPM-ExpertSignals). MATERIALS AND METHODS: We employed an iterative framework development approach that combined data-driven modeling and simulation testing to define and refine a process for phenotyping clinician behaviors. Our framework was developed and evaluated based on the Communicating Narrative Concerns Entered by Registered Nurses (CONCERN) predictive model to detect and leverage signals of clinician expertise for prediction of patient trajectories. RESULTS: Seven themes-identified during development and simulation testing of the CONCERN model-informed framework development. The HPM-ExpertSignals conceptual framework includes a 3-step modeling technique: (1) identify patterns of clinical behaviors from user interaction with CIS; (2) interpret patterns as proxies of an individual's decisions, knowledge, and expertise; and (3) use patterns in predictive models for associations with outcomes. The CONCERN model differentiated at risk patients earlier than other early warning scores, lending confidence to the HPM-ExpertSignals framework. DISCUSSION: The HPM-ExpertSignals framework moves beyond transactional data analytics to model clinical knowledge, decision making, and CIS interactions, which can support predictive modeling with a focus on the rapid and frequent patient surveillance cycle. CONCLUSIONS: We propose this framework as an approach to embed clinicians' knowledge-driven behaviors in predictions and inferences to facilitate capture of healthcare processes that are activated independently, and sometimes well before, physiological changes are apparent.

OBJECTIVE: Pressure injuries are common and serious complications for hospitalized patients. The pressure injury rate is an important patient safety metric and an indicator of the quality of nursing care. Timely and accurate prediction of pressure injury risk can significantly facilitate early prevention and treatment and avoid adverse outcomes. While many pressure injury risk assessment tools exist, most were developed before there was access to large clinical datasets and advanced statistical methods, limiting their accuracy. In this paper, we describe the development of machine learning-based predictive models, using phenotypes derived from nurse-entered direct patient assessment data. METHODS: We utilized rich electronic health record data, including full assessment records entered by nurses, from 5 different hospitals affiliated with a large integrated healthcare organization to develop machine learning-based prediction models for pressure injury. Five-fold cross-validation was conducted to evaluate model performance. RESULTS: Two pressure injury phenotypes were defined for model development: nonhospital acquired pressure injury (N = 4398) and hospital acquired pressure injury (N = 1767), representing 2 distinct clinical scenarios. A total of 28 clinical features were extracted and multiple machine learning predictive models were developed for both pressure injury phenotypes. The random forest model performed best and achieved an AUC of 0.92 and 0.94 in 2 test sets, respectively. The Glasgow coma scale, a nurse-entered level of consciousness measurement, was the most important feature for both groups. CONCLUSIONS: This model accurately predicts pressure injury development and, if validated externally, may be helpful in widespread pressure injury prevention.

OBJECTIVE: The objective of this paper is to share challenges, recommendations, and lessons learned regarding the development and implementation of a Patient Safety Learning Laboratory (PSLL) project, an innovative and complex intervention comprised of a suite of Health Information Technology (HIT) tools integrated with a newly implemented Electronic Health Record (EHR) vendor system in the acute care setting at a large academic center. MATERIALS AND METHODS: The PSLL Administrative Core engaged stakeholders and study personnel throughout all phases of the project: problem analysis, design, development, implementation, and evaluation. Implementation challenges and recommendations were derived from direct observations and the collective experience of PSLL study personnel. RESULTS: The PSLL intervention was implemented on 12 inpatient units during the 18-month study period, potentially impacting 12,628 patient admissions. Challenges to implementation included stakeholder engagement, project scope/complexity, technology/governance, and team structure. Recommendations to address each of these challenges were generated, some enacted during the trial, others as lessons learned for future iterative refinements of the intervention and its implementation. CONCLUSION: Designing, implementing, and evaluating a suite of tools integrated within a vendor EHR to improve patient safety has a variety of challenges. Keys to success include continuous stakeholder engagement, involvement of systems and human factors engineers within a multidisciplinary team, an iterative approach to user-centered design, and a willingness to think outside of current workflows and processes to change health system culture around adverse event prevention.

This case report applied principles from the data visualization (DV) literature and feedback from nurses to develop an effective report to display adherence with an evidence-based fall prevention program. We tested the usability of the original and revised reports using a Health Information Technology Usability Evaluation Scale (Health-ITUES) customized for this project. Items were rated on a 5-point Likert scale, strongly disagree (1) to strongly agree (5). The literature emphasized that the ideal display maximizes the information communicated, minimizes the cognitive efforts involved with interpretation, and selects the correct type of display (eg, bar versus line graph). Semi-structured nurse interviews emphasized the value of simplified reports and meaningful data. The mean (standard deviation [SD]) Health-ITUES score for the original report was 3.86 (0.19) and increased to 4.29 (0.11) in the revised report (Mann Whitney U Test, z = -12.25, P < 0.001). Lessons learned from this study can inform report development for clinicians in implementation science.

Learning representations of clinical notes poses challenges in handling complex content that necessitates preprocessing steps to make the data more suitable for data mining. An important issue, addressed here, is that of temporal expressions, where cues indicate the time when clinical events occur. We present a three-step data reconstruction algorithm for transforming similar clinical entities (e.g., symptoms, complications) into sequential data through unsupervised annotation of temporal expressions. First, the data reconstruction algorithm detects if an expression has temporal intent. Second, it decomposes and rewrites the expression into non-temporal sub-expression and temporal constraints. Finally, it clusters similar non-temporal sub-expressions by using unsupervised sentence embedding under the modified K-medoids paradigm. We experimented with our proposed algorithm on clinical notes associated with chronic obstructive pulmonary disease (COPD). Visualizing reconstruction results of cardiology reports for a longitudinal cohort of patients with COPD demonstrated that this algorithm is feasible.

Objective: Representing nursing data sets in a standard way will help to facilitate sharing relevant information across settings. We aimed to populate nursing process and outcome metrics with electronic health record (EHR) data and then compare the results with event reporting systems. Methods: We used the "eMeasure" development process of the National Quality Forum adopted by the American Nurses Association. We used operational definitions of quality measures from the American Nurses Association and the US Institute for Healthcare Improvement and employed concept mapping of local data elements to 2 controlled vocabularies to define a standard data dictionary: (1) Logical Observation Identifiers Names and Codes and (2) International Classification for Nursing Practice. We assessed feasibility using the nursing data set of 7829 and 8199 patients from 2 general hospitals with different EHR systems. Using inpatient falls as a use case, we compared the populated measures with results from the event reporting systems. Results: We identified 17 care components and 118 unique concepts and matched them with data elements in the EHRs. Including suboptimal mapping, 98% of the assessment concepts mapped to Logical Observation Identifiers Names and Codes and 52.9% of intervention concepts mapped to International Classification for Nursing Practice. While not all process indicators were available from event reporting systems, we successfully populated 9 fall prevention process indicators and the fall rate outcome indicator from the 2 EHRs. We were unable to populate the falls with an injury rate indicator. Conclusions: EHR data can populate fall prevention process measure metrics and at least one inpatient fall prevention outcome metric.

As part of an interdisciplinary acute care patient portal task force with members from 10 academic medical centers and professional organizations, we held a national workshop with 71 attendees representing over 30 health systems, professional organizations, and technology companies. Our consensus approach identified 7 key sociotechnical and evaluation research focus areas related to the consumption and capture of information from patients, care partners (eg, family, friends), and clinicians through portals in the acute and post-acute care settings. The 7 research areas were: (1) standards, (2) privacy and security, (3) user-centered design, (4) implementation, (5) data and content, (6) clinical decision support, and (7) measurement. Patient portals are not yet in routine use in the acute and post-acute setting, and research focused on the identified domains should increase the likelihood that they will deliver benefit, especially as there are differences between needs in acute and post-acute care compared to the ambulatory setting.

Objective: To provide recommendations on how to most effectively implement advanced features of acute care patient portals, including: (1) patient-provider communication, (2) care plan information, (3) clinical data viewing, (4) patient education, (5) patient safety, (6) caregiver access, and (7) hospital amenities. Recommendations: We summarize the experiences of 6 organizations that have implemented acute care portals, representing a variety of settings and technologies. We discuss the considerations for and challenges of incorporating various features into an acute care patient portal, and extract the lessons learned from each institution's experience. We recommend that stakeholders in acute care patient portals should: (1) consider the benefits and challenges of generic and structured electronic care team messaging; (2) examine strategies to provide rich care plan information, such as daily schedule, problem list, care goals, discharge criteria, and post-hospitalization care plan; (3) offer increasingly comprehensive access to clinical data and medical record information; (4) develop alternative strategies for patient education that go beyond infobuttons; (5) focus on improving patient safety through explicit safety-oriented features; (6) consider strategies to engage patient caregivers through portals while remaining cognizant of potential Health Insurance Portability and Accountability Act (HIPAA) violations; (7) consider offering amenities to patients through acute care portals, such as information about navigating the hospital or electronic food ordering.

Objective: To define the types and numbers of inpatient clinical decision support alerts, measure the frequency with which they are overridden, and describe providers' reasons for overriding them and the appropriateness of those reasons. Materials and Methods: We conducted a cross-sectional study of medication-related clinical decision support alerts over a 3-year period at a 793-bed tertiary-care teaching institution. We measured the rate of alert overrides, the rate of overrides by alert type, the reasons cited for overrides, and the appropriateness of those reasons. Results: Overall, 73.3% of patient allergy, drug-drug interaction, and duplicate drug alerts were overridden, though the rate of overrides varied by alert type (P < .0001). About 60% of overrides were appropriate, and that proportion also varied by alert type (P < .0001). Few overrides of renal- (2.2%) or age-based (26.4%) medication substitutions were appropriate, while most duplicate drug (98%), patient allergy (96.5%), and formulary substitution (82.5%) alerts were appropriate. Discussion: Despite warnings of potential significant harm, certain categories of alert overrides were inappropriate >75% of the time. The vast majority of duplicate drug, patient allergy, and formulary substitution alerts were appropriate, suggesting that these categories of alerts might be good targets for refinement to reduce alert fatigue. Conclusion: Almost three-quarters of alerts were overridden, and 40% of the overrides were not appropriate. Future research should optimize alert types and frequencies to increase their clinical relevance, reducing alert fatigue so that important alerts are not inappropriately overridden.

OBJECTIVE: Variation in the use of tests and treatments has been demonstrated to be substantial between providers and geographic regions. This study assessed variation between outpatient providers in overriding electronic prescribing warnings. METHODS: Responses to warnings were prospectively logged. Random effects models were used to calculate provider-to-provider variation in the rates for the decisions to override warnings in 6 different clinical domains: medication allergies, drug-drug interactions, duplicate drugs, renal recommendations, age-based recommendations, and formulary substitutions. RESULTS: A total of 157 482 responses were logged. Differences between 1717 providers accounted for 11% of the overall variability in override rates, so that while the average override rate was 45.2%, individual provider rates had a wide range with a 95% confidence interval (CI) (13.7%-76.7% ). The highest variations between providers were observed in the categories age-based (25.4% of total variability; average override rate 70.2% [95% CI, 29.1%-100% ]) and renal recommendations (24.2%; average 70% [95% CI, 29.5%-100% ]), and provider responses within these 2 categories were most often clinically inappropriate according to prior work. Among providers who received at least 10 age-based recommendations, 64 of 238 (27%) overrode ≥ 90% of the warnings and 13 of 238 (5%) overrode all of them. Of those who received at least 10 renal recommendations, 36 of 92 (39%) overrode ≥ 90% of the alerts and 9 of 92 (10%) overrode all of them. CONCLUSIONS: The decision to override prescribing warnings shows variation between providers, and the magnitude of variation differs among the clinical domains of the warnings; more variation was observed in areas with more inappropriate overrides.

We implemented a web-based, patient-centered toolkit that engages patients/caregivers in the hospital plan of care by facilitating education and patient-provider communication. Of the 585 eligible patients approached on medical intensive care and oncology units, 239 were enrolled (119 patients, 120 caregivers). The most common reason for not approaching the patient was our inability to identify a health care proxy when a patient was incapacitated. Significantly more caregivers were enrolled in medical intensive care units compared with oncology units (75% vs 32%; P < .01). Of the 239 patient/caregivers, 158 (66%) and 97 (41%) inputted a daily and overall goal, respectively. Use of educational content was highest for medications and test results and infrequent for problems. The most common clinical theme identified in 291 messages sent by 158 patients/caregivers was health concerns, needs, preferences, or questions (19%, 55 of 291). The average system usability scores and satisfaction ratings of a sample of surveyed enrollees were favorable. From analysis of feedback, we identified barriers to adoption and outlined strategies to promote use.

BACKGROUND: Experts suggest that formulary alerts at the time of medication order entry are the most effective form of clinical decision support to automate formulary management. OBJECTIVE: Our objectives were to quantify the frequency of inappropriate nonformulary medication (NFM) alert overrides in the inpatient setting and provide insight on how the design of formulary alerts could be improved. METHODS: Alert overrides of the top 11 (n = 206) most-utilized and highest-costing NFMs, from January 1 to December 31, 2012, were randomly selected for appropriateness evaluation. Using an empirically developed appropriateness algorithm, appropriateness of NFM alert overrides was assessed by 2 pharmacists via chart review. Appropriateness agreement of overrides was assessed with a Cohen's kappa. We also assessed which types of NFMs were most likely to be inappropriately overridden, the override reasons that were disproportionately provided in the inappropriate overrides, and the specific reasons the overrides were considered inappropriate. RESULTS: Approximately 17.2% (n = 35.4/206) of NFM alerts were inappropriately overridden. Non-oral NFM alerts were more likely to be inappropriately overridden compared to orals. Alerts overridden with "blank" reasons were more likely to be inappropriate. The failure to first try a formulary alternative was the most common reason for alerts being overridden inappropriately. CONCLUSION: Approximately 1 in 5 NFM alert overrides are overridden inappropriately. Future research should evaluate the impact of mandating a valid override reason and adding a list of formulary alternatives to each NFM alert; we speculate these NFM alert features may decrease the frequency of inappropriate overrides.

OBJECTIVE: Currently, the processes for harmonizing and extending standards by leveraging the knowledge within local documentation artifacts are not well described. We describe a collaborative project to develop common information models, terminology bindings, and term definitions based on nursing documentation systems, and carry the findings through to the adoption in standards development organizations (SDOs) and technical implementations in clinical applications. MATERIALS AND METHODS: Nursing flowsheet documents from six large organizations were analyzed to generate a common information model and terminologies that fully expressed documentation across all systems, and were sufficient for evidence-based decision support, reporting, and analysis. RESULTS: Significant gaps in existing standards were identified. The models and terminologies were submitted to and incorporated by SDOs, are published, implemented, and now serving as a foundation for an eMeasure. DISCUSSION: There are few examples in the literature of success working through the standards development process from a bottom-up perspective. Subsequently, standards do not yet fully address the need for detailed clinical data that enables, for example, decision support as well as a range of reporting and analytic requirements. Recommendations from this project include transparent processes within SDOs, registries that make models and associated terminologies freely available, and coordinated governance processes. CONCLUSION: We demonstrated the feasibility of using documentation artifacts in a bottom-up approach to develop common models and sets of terms that are complete from the perspective of clinical implementation. Importantly, we demonstrated a process by which a community of practice can contribute to closing gaps in existing standards using SDO processes.

BACKGROUND: Maintaining continuity of care (CoC) in the inpatient setting is dependent on aligning goals and tasks with the plan of care (POC) during multidisciplinary rounds (MDRs). A number of locally developed rounding tools exist, yet there is a lack of standard content and functional specifications for electronic tools to support MDRs within and across settings. OBJECTIVE: To identify content and functional requirements for an MDR tool to support CoC. MATERIALS AND METHODS: We collected discrete clinical data elements (CDEs) discussed during rounds for 128 acute and critical care patients. To capture CDEs, we developed and validated an iPad-based observational tool based on informatics CoC standards. We observed 19 days of rounds and conducted eight group and individual interviews. Descriptive and bivariate statistics and network visualization were conducted to understand associations between CDEs discussed during rounds with a particular focus on the POC. Qualitative data were thematically analyzed. All analyses were triangulated. RESULTS: We identified the need for universal and configurable MDR tool views across settings and users and the provision of messaging capability. Eleven empirically derived universal CDEs were identified, including four POC CDEs: problems, plan, goals, and short-term concerns. Configurable POC CDEs were: rationale, tasks/'to dos', pending results and procedures, discharge planning, patient preferences, need for urgent review, prognosis, and advice/guidance. DISCUSSION: Some requirements differed between settings; yet, there was overlap between POC CDEs. CONCLUSIONS: We recommend an initial list of 11 universal CDEs for continuity in MDRs across settings and 27 CDEs that can be configured to meet setting-specific needs.

OBJECTIVE: As healthcare systems and providers move toward meaningful use of electronic health records, longitudinal care plans (LCPs) may provide a means to improve communication and coordination as patients transition across settings. The objective of this study was to determine the current state of communication of LCPs across settings and levels of care. MATERIALS AND METHODS: We conducted surveys and interviews with professionals from emergency departments, acute care hospitals, skilled nursing facilities, and home health agency settings in six regions in the USA. We coded the transcripts according to the Agency for Healthcare Research and Quality (AHRQ) 'Broad Approaches' to care coordination to understand the degree to which current practice meets the definition of an LCP. RESULTS: Participants (n=22) from all settings reported that LCPs do not exist in their current state. We found LCPs in practice, and none of these were shared or reconciled across settings. Moreover, we found wide variation in the types and formats of care plan information that was communicated as patients transitioned. The most common formats, even when care plan information was communicated within the same healthcare system, were paper and fax. DISCUSSION: These findings have implications for data reuse, interoperability, and achieving widespread adoption of LCPs. CONCLUSIONS: The use of LCPs to support care transitions is suboptimal. Strategies are needed to transform the LCP from vision to reality.

OBJECTIVES: The purpose of this study was to evaluate the adequacy of the International Classification of Nursing Practice (1) (ICPN) Version 1.0 as a representational model for nursing assessment documentation. DESIGN AND MEASUREMENTS: To identify representational requirements of nursing assessments, the authors mapped key concepts and semantic relations extracted from standardized and local nursing admission assessment documentation forms/templates and inpatient admission assessment records to the ICNP. Next, they expanded the list of ICNP semantic relations with those obtained from the admission assessment forms/templates. The expanded ICNP semantic relations were then validated against the semantic relations identified from an additional set of admission assessment records and a set of 300 randomly selected North American Nursing Diagnosis Association defining characteristic phrases. The concept coverage of the ICNP was evaluated by mapping the concepts extracted from these sources to the ICNP concepts. The UMLS Methathesaurus was then used to map concepts without exact matches to other American Nursing Association (ANA) recognized terminologies. RESULTS: The authors found that along with the 30 existing ICNP semantic relations, an additional 17 are required for the ICNP to function as a representational model for nursing assessment documentation. Eight hundred and five unique assessment concepts were extracted from all sources. Forty-three percent of these unique assessment concepts had exact matches in the ICNP. An additional 20% had matches in the ICNP classified as narrower, broader, or "other." Of the concepts without exact matches in the ICNP, 81% had exact matches found in other ANA recognized terminologies. CONCLUSIONS: The broad concept coverage and the logic-based structure of the ICNP make it a flexible and robust standard. The ICNP provides a framework from which to capture and reuse atomic level data to facilitate evidence-based practice.

OBJECTIVE: The use of health information technology (HIT) for the support of communication processes and data and information access in acute care settings is a relatively new phenomenon. A means of evaluating the impact of HIT in hospital settings is needed. The purpose of this research was to design and psychometrically evaluate the Impact of Health Information Technology scale (I-HIT). I-HIT was designed to measure the perception of nurses regarding the ways in which HIT influences interdisciplinary communication and workflow patterns and nurses' satisfaction with HIT applications and tools. DESIGN: Content for a 43-item tool was derived from the literature, and supported theoretically by the Coiera model and by nurse informaticists. Internal consistency reliability analysis using Cronbach's alpha was conducted on the 43-item scale to initiate the item reduction process. Items with an item total correlation of less than 0.35 were removed, leaving a total of 29 items. MEASUREMENTS: Item analysis, exploratory principal component analysis and internal consistency reliability using Cronbach's alpha were used to confirm the 29-item scale. RESULTS: Principal components analysis with Varimax rotation produced a four-factor solution that explained 58.5% of total variance (general advantages, information tools to support information needs, information tools to support communication needs, and workflow implications). Internal consistency of the total scale was 0.95 and ranged from 0.80-0.89 for four subscales. CONCLUSION: I-HIT demonstrated psychometric adequacy and is recommended to measure the impact of HIT on nursing practice in acute care settings.