Stakeholder perspectives on factors influencing acute care patient outcomes: A qualitative approach to model refinement

Background: Health systems have long been interested in the best practices for staffing in the acute care setting. Studies on staffing often focus on registered nurses and nurse-to-patient staffing ratios. There are fewer studies on the relationship between interprofessional team members or contextual factors such as hospital and community characteristics and patient outcomes. This qualitative study aimed to refine a causal model by soliciting hospital stakeholder feedback on staffing and patient outcomes. Methods: We conducted a qualitative study using semi-structured interviews and thematic analysis to understand hospital stakeholder perspectives and their experiences of factors that affect acute care inpatient outcomes. Interviews were conducted in 2022 with 38 hospital stakeholders representing 19 hospitals across Washington State. Results: Findings support a model of characteristics impacting patient outcomes to include the complex and interconnected relationships between community, hospital, patient, and staffing characteristics. Within the model, patient characteristics are nested into hospital characteristics, and in turn these were nested within community characteristics to highlight the importance of setting and context when evaluating outcomes. Together, these factors influenced both staff characteristics and patient outcomes, while these two categories also share a direct relationship. Conclusion: Findings can be applied to hospitals and health systems across the globe to examine how external factors such as community resource availability impact care delivery. Future research should expand on this work with specific attention to how staffing changes and interprofessional team composition can improve patient outcomes.


INTRODUCTION
Health systems face ongoing challenges in recruiting and retaining staff to meet the needs of their patients.Best practices in acute care sta ng have long been a topic of interest for organizations around the world, with more recent attention in the United States by legislative entities focused on ensuring the availability of a healthcare workforce that can meet the needs of their communities.Frequently, sta ng studies focus exclusively on registered nurses (RNs) and show that a higher nurse-to-patient ratio is associated with improved patient outcomes. 1However, few studies have addressed the relationship between sta ng of other interprofessional team members and patient outcomes, and even fewer have addressed contextual factors in organizations and communities that may in uence sta ng, patient outcomes, or both.A better understanding of the relationships between these factors is needed to ensure that policies and practices support optimal patient outcomes and to provide healthcare work environments conducive to quality care.
In 2021, the Washington state (WA) legislature passed a bill focused on transparency in healthcare. 2 The bill directed the state Department of Health to commission an interdisciplinary team to examine the relationships between the acute care workforce and patient outcomes by systematically investigating how the number, type, education, training, and experience of staff affects patient mortality and other patient outcomes, utilizing scienti cally sound research methods and input from stakeholders throughout the state. 2Our team, led by the University of WA School of Nursing in collaboration with researchers at the Institute for Health Metrics, was selected to conduct this study.To conduct this research, we established partnerships with hospital leaders, healthcare associations, and union representatives.The process included: (1) reviewing studies assessing the impact of workforce characteristics on patient outcomes, (2) developing a preliminary model and analysis plan based on the review, key informant input, and available data sources, (3) re ning the model using stakeholder feedback, and (4) completing a quantitative analysis utilizing anonymized state-and hospital-collected healthcare data guided by the re ned model.
The scoping review we completed in phase 1 identi ed that outside of RNs, healthcare team sta ng is seldom quanti ed in health services literature, and contextual factors at the hospital and community level are rarely included when examining patient outcomes. 3The subsequent phases of our study aimed to understand what those missing factors were and how they played out in various acute care contexts, including impacts on operations, work environment, quality, and outcomes for patients and workers.The purpose of this paper is to present the perspectives of hospital stakeholders across WA on factors that impact patient outcomes and to describe the development of the causal model used in the nal phase of the study.

Study Design
We conducted a qualitative study using semi-structured interviews and thematic analysis to understand hospital stakeholder perspectives and their experiences of factors that affect acute care inpatient outcomes.As part of our causal modeling, qualitative research methods were chosen to explore relationships between variables, including context, mechanisms, and outcomes. 4This exploratory approach also acknowledges the signi cance of subjectivity in the data and allows for inductive inquiry. 5he ndings from this study served as a foundation for developing a comprehensive model we called the 'Washington Acute Care hospital Characteristics and patient Outcomes model' (WACCHO), which considers community, hospital, and patient characteristics that interact with sta ng to affect patient outcomes.This study was granted exempt status by the University of WA and the WA Institutional Review Boards (STUDY00013975).

Participant Recruitment
Participants were purposively sampled through announcements to state-wide hospital email listservs.Hospital executives and administrators were also directly emailed to increase participation and ensure participants represented diverse hospital types and geographic settings.Site contacts were asked to invite any hospital representative who could provide perceptions of sta ng's impact on patient outcomes.Participants were unknown to the research team prior to the interviews.

Data Collection
The preliminary model from study phase 2 was used to develop a semi-structured interview guide, which was tailored for each site with a few hospital-speci c data points and shared with participants prior to the interview.Open-ended questions explored factors, mechanisms, contextual elements, and additional variables that could potentially in uence patient outcomes in 4 main categories: hospital characteristics and external factors, patient characteristics, sta ng characteristics, and patient outcomes.Interviewers also presented facility-speci c data, asked participants for their perceptions of accuracy, and discussed the basic analysis plan for the quantitative portion of the study.A list of interview questions and prompts is provided in an additional le (Additional File 1).Interviews were conducted between January and June 2022 via video conferencing by 2-5 members of the research team (SI, AF, NBS, NH, AP).Each participant was interviewed once, and all interviews were audio recorded and transcribed for analysis.The research team introduced themselves and explained the purpose of the study.Upon obtaining oral consent from participants, 2 members of the research team (SI, NH) took detailed eld notes.Transcripts were uploaded to ATLAS.ti (version 9).

Analysis
Both deductive and inductive methods were used in the thematic analysis of data. 5An initial codebook was created based on our initial model and interview notes, and emergent codes were added inductively during analysis. 5Five team members (NH, JZ, ED, KN, KB) contributed to coding.They met weekly to review codes, ensure a uniform interpretation and application of the coding framework, and address any discrepancies.A portion of each transcript was coded by at least 2 researchers to ensure consistency.
Once coding was completed, codes were iteratively organized into main themes and subthemes to capture the range of narratives. 5Saturation was determined when no new themes were identi ed in nal interviews.We followed the consolidated criteria for reporting qualitative research guidelines (COREQ) to ensure comprehensive reporting. 6

Limitations
This study had several notable limitations.First, the timing of interviews during the COVID-19 pandemic made it challenging for hospital representatives to participate.Frontline staff were often unavailable, and leaders were frequently supporting patient care activities during surges in admissions.This resulted in less robust representation from some healthcare team members and more prominent representation of executive and administrator perspectives, which may have focused the discussion on nurse sta ng structures rather than perceptions of patient needs.Second, as the study focused on experiences before the pandemic, participants were asked to remember past perceptions, which challenged their focus and could have led to limited recall bias.Third, our team's innovative approach to expanding sta ng models may have challenged participants to think beyond their current perceptions and narrowed their feedback.Finally, as our causal model was iteratively developed throughout the interview period, interview questions were not static and discussion may have focused on elements that stakeholders felt more strongly about, in uencing the quantity of participant feedback on speci c elements of the model.

Participants
A total of 20 interviews were conducted with 38 participants from 19 hospitals in 8 out of 9 regions across WA.Participants worked at 3 main types of hospitals: acute care (23/38), critical access (11/38), and sole community hospitals (4/38).While the de nitions of hospital types may vary in some literature, the Centers for Medicare and Medicaid Services (CMS) o cially designates critical access and sole community hospitals as speci c types of acute care hospitals, where critical access hospitals are typically smaller and located in rural settings. 7,8Participants included a broad range of executives and administrators (23/38), directors and managers (10/38), and interprofessional care team members (5/38) from all 3 hospital types.Mean interview length was 61 minutes.

Causal Model
The nal model shows the primary factors and drivers impacting patient outcomes as agreed upon by participants (Fig. 1), which includes reorganized categories from the initial model to better represent ndings.Hospital characteristics and external factors were divided into two distinct categories, with external factors renamed as community characteristics.The reorganized model nested patient characteristics into hospital characteristics, and in turn these were nested within community characteristics to highlight the relationships between the three categories.The new community characteristics category, inclusive of hospital characteristics and patient characteristics, impacts sta ng characteristics and patient outcomes, and sta ng and patient outcomes are directly connected.Findings are presented here by model category and Table 1 summarizes themes and their frequency.

Community Characteristics
Community characteristics were de ned as the setting in which the hospital exists.This setting includes factors outside of the hospital's control, such as sociopolitical, geographical, and economic factors, and availability of other healthcare resources.
Location and community resources.Participants often described the di culty of discharging patients to the appropriate level of care, meaning the resources and facilities needed to meet patient needs and including higher acuity care such as transfer to a referral hospital, or subacute care such as discharge to a skilled nursing facility.When resources for the appropriate level of care are limited, the hospital must keep patients in acute care beds, limiting available resources for other patients.These challenges were more pronounced in rural settings with fewer community resources.
Community characteristics also impacted sta ng.Both rural and urban participants described how location and community resources like affordable housing, public transportation, and commute times made it di cult to recruit and retain hospital staff.Additionally, participants in rural locations noted the di culty in recruiting staff when they are not close to or connected with teaching institutions producing new graduates.
Population.Participants discussed their facilities' unique challenges due to the populations they served and health disparities present within the community.Participants listed characteristics such as homelessness, homes without basic utilities (e.g., running water or electricity), health insurance in the community, and transient seasonal populations including migrant workers and summer tourists.

Hospital Characteristics
Hospital characteristics were de ned as the structural and functional qualities of acute care hospitals that in uenced the services they offered and the complexity of patients they served.
Hospital type and access to resources.Participants stated that their hospital type, speci cally size and connection to larger health systems, in uenced their access to resources.Various stakeholders described limited budgets and reduced access to equipment secondary to supply chain constraints.Critical access hospitals identi ed their smaller size and limited resource pool as reasons they must be more particular with capital investments that would enable them to care for more complex patients while simultaneously having the obligation to provide specialty services that are not otherwise available in their communities.
Participants from critical access hospitals also described lack of access to the relationships and shared knowledge within larger health systems, which they felt negatively impacted their e ciency in rolling out new policies and processes.
Hospital leadership structure and culture as foundational to quality.Participants considered sta ng and leadership culture as a product of organizational priorities that in uenced staff satisfaction and quality outcomes.Participants identi ed that characteristics including union status, sta ng strategies, budget, and patient care equipment in uenced sta ng and leadership culture.Stakeholders cited efforts by the organization to ensure adequate equipment and supplies as important to providing quality patient care.
As one acute care hospital administrator described, "Something as simple as an overbed table… When we talked about this at incident command...the answer was no.And then, thank God, our CEO is also a nurse and she's like no, this is basic to taking care of patients and keeping them from falling." Participants noted organizational features that emphasized safety culture, with elements like care quality and improved organizational processes.Multiple participants referenced standardized protocols as a safety tool that contributed to improved patient care.Participants also felt an organizational focus on safety and transparency improved staff satisfaction and quality of care.Comments referenced the importance of continuous quality improvement and a focus on process improvement instead of individual errors.
In uence of organizational culture on staff retention.Participants agreed that the culture of an organization in uenced work environment and staff retention.They described approaches to support and engage with staff which promoted a positive organizational culture.One approach included providing staff with incentives and bene ts such as increased pay, bonuses, parking passes, exible shifts and scheduling.Other examples included programs which covered the cost of nursing education in exchange for commitment to work in a given facility for a period of time.Participants also presented upstream approaches which improved the work environment, such as involving workers in organizational decision making and appropriate sta ng of the interprofessional team.Overall, as one interprofessional care team member stated, "if you're not given the tools to do your job well, anybody with any empathy is going to go nd something else to do." Units vary across and hospitals.When discussing data metrics, participants often discussed the di culty in making comparisons of the same unit between different hospitals and comparisons of units within the same hospital.They expressed confusion with how acuity is de ned, especially when comparing patient care across different facilities.Participants felt it was too di cult to use case mix index, a metric used to identify the diversity and severity of patients cared for at speci c hospitals, to compare outcomes between units within a hospital or across healthcare systems.Participants did not think case mix encompasses all the variables that should be considered when evaluating the complexities of the patient and the care infrastructure.

Patient Characteristics
Patient characteristics are de ned as individual demographic, social, and health characteristics of patients admitted to the hospital that may impact the level of care needed.
Underlying conditions impact the intensity of care.Participants used the term 'care intensity' to describe how patient care needs impacted work demands on staff, with agreement that the care intensity is not always directly tied to the patient's admitting diagnoses or assigned acuity.Participants reported this disparity between acuity and care intensity as a challenge to accurately predict sta ng needs.They noted that speci c health conditions with higher care intensity included aggressive behavior, traumatic brain injury, obesity, substance use, and dementia.Participants described different strategies to account for care intensity variations, such as having a centralized sta ng o ce or a prede ned team who coordinated activities to accommodate rapid and uctuating changes in sta ng needs.In addition to increased care intensity and inpatient sta ng demands, patients with certain underlying conditions were di cult to discharge due to the availability of appropriate care in the community or required social support, such as individuals needing guardian assignment.
Social and economic characteristics impact on status.In addition to descriptors of the population served the community characteristics section, the demographics, social determinants of health (SDOH), and insurance status of patients in uenced their care needs.Factors such as access to routine care, prior healthcare utilization, and comorbidities impacted care intensity and needed resources.

Sta ng Characteristics
Sta ng characteristics are de ned as acute care team members, their roles, and aspects of sta ng which in uence how facilities provide staff and deliver patient care.
Interprofessional acute care team composition and the central role of nurses.When considering the relationship between sta ng and patient outcomes, participants discussed team members who contribute to the care team and work in tandem to provide patient care.Participants mentioned roles in multiple professions including physicians, advanced practice providers, RNs, certi ed nursing assistants, occupational and physical therapists (PT), pharmacists, workers, dietary aides, environmental service workers, billing/coding staff, students, and others.Care team members were generally categorized as either clinical, non-clinical, or temporary roles.There was a lack of agreement around the types and quantities of roles included in the acute care team.However, participants discussed state mandated annual RN sta ng plans and nurse-to-patient ratios, highlighting the central role and value placed on RNs in acute patient care and interprofessional care teams.
In uence of sta type on work environment.Participants emphasized the importance of differentiating between temporary (e.g., contract, agency, or travel) and permanent RNs when examining how sta ng impacted patient outcomes.Stakeholders expressed that temporary workers may be less familiar with facility policies and may not have the same unit-speci c training as permanent staff.
Additionally, facilities with a larger proportion of rotating temporary workers may not have an established culture of communication and support, which diminishes the quality of the work environment and negatively in uences patient outcomes.
RN absorption of non-nursing resulting in the dilution of nursing care roles.Although facilities submit annual nurse sta ng matrices, participants frequently spoke to the need to deviate from planned models, highlighting variation in direct and indirect patient support staff which make nurse-to-patient ratios in one setting incomparable to the same workload in another setting.Participants also presented instances when facilities had di culty lling sta ng roles, so RNs absorbed responsibilities, diluting the scope of nursing practice.For example, one sole community hospital administrator stated, "If you're short PT assistants or PT aids, that falls back on the RN and the nursing assistant.If you don't have case management or social work, that also falls on the RN.Everything falls on the RN, if... the rest of the team is missing." Education, training, and experience.Discussions education, training, and experience centralized around nursing staff and largely focused on the nuances of the term 'experience'.Participants agreed that RN experience was complex and di cult to capture, quantify and standardize.Various metrics for measuring experience were presented and considered, such as years of RN or inpatient experience and unit tenure.Participants also quanti ed RN experience with standards such as a novice to expert or years since licensure.Degrees, licenses, and certi cations were discussed as components of education, with several participants stating that nurse training was not well documented except in human resource records.Participants noted that overall training and experience on the unit in uenced the ability to staff appropriately for patient acuity and diagnosis.When units had higher numbers of staff with more training and experience, the unit could manage more complex patients, yet in many locations, the limited number of experienced staff made patient assignments di cult.

Patient Outcomes
Patient outcomes describe metrics pertaining to characteristics of a patient's stay at a hospital and the time immediately following discharge, which are a collection of quality and safety metrics tracked by the hospital and the state.
Impact of sta ng on patient outcomes.When asked about patient outcomes, participants described some measures as more sensitive to sta ng than others.Participants characterized sta ng-susceptible outcomes as being dependent on care team composition and sta ng type rather than number of staff or staff-to-patient ratios.Participants speci cally mentioned falls and pressure ulcers as sta ng-sensitive outcomes, with one hospital administrator noting that, "one of the things...making a signi cant impact on patient outcomes or patient satisfaction and sta ng is the number of non-hospital nurses that we have here.So, we have 72 travelers, and we have FEMA [Federal Emergency Management Agency] staff, and so our fall rates increased, our HAPIs [Hospital-Acquired Pressure Injuries] have increased, complaints have increased." Cumulative impact of themes on patient outcomes.Participants discussed how some patient outcomes length of stay or readmission are signi cantly impacted by model categories outside of sta ng.One example was length-of-stay, de ned as the number of days a patient is cared for in an acute care facility.Several participants described the in uence of community and patient characteristics on length-of-stay related to needs for social support or availability of a skilled nursing facility bed.One critical access hospital administrator stated, "it happens, where we cannot get a patient out.We don't have a receiving hospital or we don't have EMS... that's the challenge of being... rural."

DISCUSSION
This study produced critical ndings on factors in uencing sta ng and patient outcomes in the acute care setting.While some ndings reinforce existing knowledge such as the importance of adequate RN sta ng, others identify gaps in both knowledge and theory.While the discussion highlights the gaps in each of the categories in our causal model, the gaps are often interconnected and responsive to dynamic changes in other model components.For example, changes in hospital leadership may impact both hospital and sta ng characteristics in ways that subsequently change patient outcomes.We also note a need to expand future work to examine the impacts of the pandemic, as this work focused on prepandemic experiences.

Community Characteristics
Participants practiced in a wide array of settings and consistently brought forward the need to account for different contexts when considering sta ng and related policy.Findings suggest that a 'one size ts all' approach to sta ng is undesirable, instead emphasizing the need for individual organizations to account for their communities and settings when establishing sta ng standards.This viewpoint is consistent with implementation science theories such as the Consolidated Framework for Implementation Research, 9 which emphasizes the inclusion of contextual factors when planning, developing, implementing, and evaluating a practice or policy change.Accounting for community contexts allows organizations to attend to the populations they serve and the resources available in their settings.For example, communities with lower demand for inpatient beds and more limited access to skilled nursing facilities may need the exibility to provide a lower level of care (e.g., a higher patient to nurse ratio) when a patient ready for skilled nursing is still physically present in the hospital.

Hospital Characteristics
While organizational culture has been linked with workforce outcomes such as RN turnover and retention, 10 participants indicated that elements of culture are also vital to conversations about sta ng and patient outcomes.An organizational emphasis on safety and just culture provides opportunities for workers to provide input on sta ng needs and challenges.Within just culture, transparency and psychological safety work bidirectionally to ensure that staff can bring forward concerns without penalty and that management and administration share information on their own challenges and progress related to sta ng. 11veral proven strategies for approaching this type of culture are Magnet® designation, which emphasizes the involvement of RNs in hospital administration, policy, and practice, 12 and American Association of Critical Care Nurses' Healthy Work Environment, which identi es 6 critical elements to a just unit culture. 13Accounting for features of organizational culture using an established framework such as these would help provide additional information and clarity into organizational practices around sta ng, which may be an important predictor or mediator of the relationship between sta ng and patient outcomes.
Hospital environment and culture impact patient care in other ways.For example, one participant's recollection of a discussion about bedside tables shows how a leader with bedside experience recognized the importance of a piece of equipment in promoting patient safety.In addition to these administrative types of decisions, structural and logistical features of hospitals impact sta ng and workload.For example, when patient care supplies were not readily available, nurses or other direct care staff had to leave the unit to retrieve them, taking time and focus away from patient care.

Patient Characteristics
Patients different types of acute care issues have various levels of need, often represented in terms of patient acuity or some measure of nursing hours invested in care. 3In our study, despite consensus across participants that the unique care needs of individual patients are not routinely captured in acuity measures or admitting diagnoses, there was no agreement on a standardized way these needs could be measured or reported.High care needs, particularly those stemming from the intersection of behavioral, mental, and physical health status, require additional work from the care team.Participants indicated that these situations disrupt the unit's work ow and change sta ng needs, even when no additional staff are available.While care quality measures aim to increase inpatient assessment of SDOH, this evaluation may indicate a need for more resources than staff have available to address issues.Overall, a more nuanced understanding of patient needs as they affect the unit workload is necessary when evaluating sta ng practice and policies.

Sta ng Characteristics
Sta ng has been a topic of interest in health services literature for decades, with most data focused on RN sta ng levels. 1 One of the main issues identi ed in our scoping review and reiterated by participants in this study was the lack of interprofessional team member inclusion in sta ng plans. 3Participants brought forward concerns about what work the RN is doing when other staff are missing and how doing that work impacted their availability to perform needed nursing tasks.Diluting RN time with non-nursing tasks means that RNs are not working at the top of their scope of practice, which leads to dissatisfaction and connects to burnout and turnover. 14Similarly, when there are not enough RNs with the training and experience to care for certain types of patients, patient outcomes may suffer. 15tient Outcomes When assessing patient outcomes in health services research, data are often sourced from statewide administrative bodies and include a range of quality metrics such as falls, skin breakdown, length of stay, mortality, and patient satisfaction.While measures like falls and skin breakdown are often labeled as "nursing sensitive", participants indicated that nurses are not the only staff members whose presence or tasks may impact those outcomes.For example, if typical resources such as PT/aides are unavailable to ambulate a patient, the RN may not be able to add that task to their workload, leading to skin breakdown.In this case, the 'nurse-sensitive' indicator may not tell the whole story about sta ng.
Other patient outcomes like length of stay or readmission may be more indicative of community resources.For example, the availability or sta ng levels of residential facilities that care for patients with sequelae of brain injury may mean that patients linger in the acute care setting or get sent back to the emergency room if facility staff are unable to handle symptoms.These types of in uences are rarely accounted for in studies which focus on direct measures of nursing sta ng and patient outcomes in acute care.
Patient outcomes also vary when underlying conditions or characteristics, including SDOH, impact overall health and the ability to provide needed services in the acute care setting.WA state now requires hospitals to report certain data on SDOH to the Department of Health, 2 which will improve the ability to account for these characteristics in future analyses of patient outcomes and provide more conclusive evidence related to health equity in different patients and communities.

Policy Implications
Throughout the study, stakeholders noted the di culty of applying a blanket policy to individual organizations.Instead, they identi ed a need for a more nuanced understanding of individual hospitals when setting sta ng standards.When aiming to ensure safe sta ng levels at the local, state, or national level, policy needs to re ect more than just the numbers of a speci c type of staff at the bedside, instead drawing on a more comprehensive understanding of the communities, settings, and patients served at different facilities.This process may require more robust data collection and commitment to ensuring an adequate supply of workers.

CONCLUSION
Altogether, this study enhanced the initial ndings of our scoping review by providing insight from healthcare stakeholders in various types of acute care hospitals across the state.Findings highlight the complexity and interrelatedness of the categories in the causal model, while drawing attention to critical gaps that must be addressed to better understand how communities, organizations, patients, and sta ng all impact patient outcomes.Our study highlights the need to ensure that RN-centered care teams include appropriate interprofessional sta ng to meet the needs of patients, and that access to community resources is critical both for ensuring that patients receive e cient continuity of care throughout their recovery and that acute care beds and staff are appropriately used.Future research should expand on this study to better understand lessons learned from the COVID-19 pandemic, with speci c attention to sta ng changes and interprofessional team composition that can direct future work to improve patient outcomes.Ensuring optimal sta ng of interprofessional teams also has the potential to decrease burnout, leading to improved outcomes for interprofessional acute care staff and improved retention of this vital workforce.
Abbreviations nurse

Table 1
Summary of themes by model category and frequency