Participants were not assigned to an intervention and as such this was not considered a clinical trial.

There is a fundamental gap between the discovery of proven-effective intensive care unit (ICU) sedation, mechanical ventilation, mobility, and symptom management strategies and approaches that can equip ICU providers with the skills necessary to reliably adopt these interventions in everyday practice. Until this gap is filled, the millions of patients with heart, lung, and blood disorders admitted to ICUs annually will remain at risk for avoidable physical, mental, and cognitive health impairments that may persist for months to years after hospital discharge. In the proposed study, the investigative team will continue their partnership with the Society of Critical Care Medicine's (SCCM's) ICU Liberation Collaborative. Guided by the Consolidated Framework for Implementation Research, the overall objective of the Determinants of Implementation Success Coordinating Ventilator, Early Ambulation and Rehabilitation Efforts in the ICU (DISCOVER-ICU) study is to develop multilevel implementation strategies to enhance sustainable adoption of the ABCDEF (Assess, prevent, and manage pain and delirium, both spontaneous awakening and breathing trials, choice of sedation, early mobility, family engagement) bundle in routine ICU practice. Using a multiphase, sequential, mixed-methods design, this study has three specific aims: 1) estimate the effects of patient-level characteristics on ABCDEF bundle adoption; 2) examine unit-level variation in ABCDEF bundle adoption and associated provider- and organization-level characteristics; and 3) determine which implementation strategies result in the greatest adoption of the ABCDEF bundle. Existing deidentified data will be obtained from >15,000 patients, >5,000 interprofessional ICU team members, and 68 hospitals participating in the ICU Liberation Collaborative to achieve specific aims 1 and 2. For specific aim 3, data collection will be extended using interprofessional ICU team surveys, a modified Delphi process, and concept mapping to achieve greater understanding of implementation strategies that prove most effective for ABCDEF bundle adoption. Results of this work will directly lead to the development of implementation strategies that are adaptable, responsive to community needs, and account for the cultural and organizational factors necessary to increase ABCDEF bundle adoption. These implementation strategies will then be tested in a future cluster randomized hybrid II implementation effectiveness trial.

Patients with heart, lung, and blood disorders severe enough to require an intensive care unit (ICU) stay frequently experience profound physical, mental, and cognitive health impairments that may persist for months to years after hospital discharge. These long-term impairments are commonly acquired in the ICU and are often initiated or exacerbated by sedation, mechanical ventilation, and symptom management decisions. Indeed, ICU-acquired pain, anxiety, delirium, and weakness are associated with numerous adverse health outcomes including prolonged mechanical ventilation, post-traumatic stress disorder, depression, functional decline, new institutionalization, and severe neurocognitive impairment. Considering millions of adults face the challenges of ICU survivorship annually, it is essential findings from high-quality research are reliably adopted and sustained in everyday clinical practice.

As outlined in "Clinical Practice Guidelines for the Management of Pain, Agitation, and Delirium (PAD) in Adult Patients in the ICU," a robust and growing body of evidence demonstrates clinical outcomes are improved when integrated, interprofessional approaches to mechanical ventilation and symptom management are applied during the course of critical illness. One such approach is the ABCDEF bundle, an evidence-based, multicomponent strategy that incorporates many PAD guideline recommendations. While previously shown safe and effective when applied in routine ICU practice, the investigators preliminary data demonstrate that this bundle has yet to be widely adopted. Pain and delirium remain underdiagnosed, and ICU clinicians often fail to maintain patients at a light level of sedation, use a protocolized mechanical ventilation discontinuation approach, get patients out of bed, or accept flexible bedside family presence. There is, therefore, a clear need for studies that use validated implementation frameworks and outcomes to assess strategies for improving ABCDEF bundle adoption and the removal of long-held and potentially harmful ICU practices.

In the proposed study, the investigators will continue their three-year partnership with the Society of Critical Care Medicine's ICU Liberation Collaborative, a national quality improvement (QI) network comprised of 68 ICUs. Guided by the Consolidated Framework for Implementation Research, the overall objective of this T4 research is to develop multilevel implementation strategies to enhance sustainable adoption of the ABCDEF bundle in routine ICU practice. The investigators are particularly interested in discovering how various patient-, provider-, and organization-level characteristics and implementation strategies effect ABCDEF bundle adoption. Using a multiphase, sequential, mixed-methods design the "Determinants of Implementation Success Coordinating Ventilator, Early Ambulation and Rehabilitation Efforts in the ICU (DISCOVER-ICU)" study has three specific aims:

Aim 1: Estimate the effects of patient-level characteristics on ABCDEF bundle adoption. Patient-level characteristics are defined as non-modifiable factors exhibited by critically ill adults at ICU admission that may influence ABCDEF bundle adoption. Existing data on ≥ 15,000 critically ill adults who spent at least one full day in a participating ICU Liberation Collaborative unit will be used to estimate the effects of: (1) demographics, (2) body mass index, (3) admitting diagnosis, (4) severity of illness, and (5) preadmission residence/mobility status on ABCDEF bundle adoption. The hypothesis is that the ABCDEF bundle will less likely be adopted into the care of particularly vulnerable populations (e.g., older adults, obese, non-English speaking).

Aim 2: Examine unit-level variation in ABCDEF bundle adoption and associated provider- and organization-level characteristics. Existing data will be obtained from surveys of over 5,000 interprofessional ICU team members and 68 hospitals participating in the ICU Liberation Collaborative. Provider-level characteristics are defined as ICU clinicians': (1) ABCDEF bundle-related knowledge and attitudes, (2) self-efficacy, and (3) perceptions of work environment (both overall health and degree of teamwork/collaboration). Organization-level characteristics are defined as those related to: (1) administrative structure, (2) staffing patterns, (3) ICU team composition, and (4) prior QI experience. The hypothesis is that ABCDEF bundle adoption will be greater in ICUs where providers value and are more familiar with the principles of the bundle, display higher levels of self-efficacy, and report more positive teamwork. Secondly, there will be better bundle adoption in organizations with greater absorptive capacity for change, more resources, and who use interprofessional ICU team models.

Aim 3: Determine which implementation strategies result in the greatest adoption of the ABCDEF bundle. Data collection will be extended by using interprofessional ICU team surveys, a modified Delphi process, and concept mapping to achieve greater understanding of implementation strategies that prove most effective for ABCDEF bundle adoption. The focus on high- and low-ABCDEF bundle-performing units will allow the investigative team to select and tailor potential ABCDEF bundle implementation strategies to meet specific local contextual needs.

With expertise in dissemination and implementation science, biostatistics, critical care, and QI, the interdisciplinary team investigative is ideally prepared to lead this three-year study. The results will directly lead to the development of implementation strategies that are adaptable, responsive to community needs, and account for the cultural and organizational factors necessary to increase ABCDEF bundle adoption. These implementation strategies will then be tested in a future cluster randomized hybrid II implementation-effectiveness trial.

Inclusion Criteria:

• ICU patient who was enrolled in ICU Liberation Collaborative (preexisting data)
• ICU provider or organization who participated in ICU Liberation Collaborative (preexisting data)
• ICU provider who was responsible for ABCDEF bundle implementation efforts associated with SCCM ICU Liberation Collaborative (prospective collected data)

Exclusion Criteria:

• Did not participate in ICU Liberation Collaborative efforts

_Preexisting data on patients, providers, and hospitals participating in ICU Liberation Collaborative

-Prospectively collected data from ICU RN, MDs, Respiratory/Physical/Occupational Therapist, Pharmacist who was responsible for ABCDEF bundle implementation efforts associated with SCCM ICU Liberation Collaborative

Study data will be made available to researchers upon request after analyses of specific aims have been conducted and findings have been published. In order to protect the privacy of individuals and institutions, as well as the confidentiality of hospitals, units and research participants, no identifying information will be collected or shared in this research endeavor. Persons interested in using the data will need to submit a request in writing, stating their intended use. Requirements for sharing will include acknowledgement in all publications of the funding source and of the study authors. Data sets will be accompanied by a data dictionary for all study variables, both derived and raw data. The most cost-effective means for sharing data will be considered after a data-sharing agreement has been reached. Research findings will also be shared through publications and presentations at scholarly research meetings.

when study complete

Study data will be made available to researchers upon request after analyses of specific aims have been conducted and findings have been published. In order to protect the privacy of individuals and institutions, as well as the confidentiality of hospitals, units and research participants, no identifying information will be collected or shared in this research endeavor. Persons interested in using the data will need to submit a request in writing, stating their intended use. Requirements for sharing will include acknowledgement in all publications of the funding source and of the study authors. Data sets will be accompanied by a data dictionary for all study variables, both derived and raw data. The most cost-effective means for sharing data will be considered after a data-sharing agreement has been reached. Research findings will also be shared through publications and presentations at scholarly research meetings.