Closer to Home: A Quality Improvement Initiative to Optimize Satellite Site Bed Capacity Free

This improvement study took place at the satellite campus of our freestanding, quaternary care children’s hospital. The satellite campus is an inpatient and outpatient standalone children’s facility with a 42-bed inpatient unit (LA4) that admits patients to multiple services, including hospital medicine (HM), adolescent medicine, pulmonary, gastroenterology, pediatric surgery, and oncology. The satellite ED sees an average of 54 000 patients annually, with almost 4000 admissions for inpatient care. The satellite HM service admits an average of 1800 patients annually, with 93% originating from the satellite ED. The satellite campus, which is located 22 miles from the main hospital, does not have an onsite ICU; however, medical response team activation and consultation with the ICU at the main hospital are performed via in-room telemedicine capabilities. Patients requiring ICU-level care must be transferred to the main hospital. At the start of the COVID-19 pandemic, because pediatric patients with COVID-19 infections were grouped at the main hospital and admission rates decreased, beds at the satellite campus remained empty, with associated staffing adjustments; beds were not used for adult patients.

In December 2020, we convened a multidisciplinary team of nurses, ED physicians, HM physicians and nurse practitioners, pediatric residents, patient flow coordinators, and family relations representatives. Using the Institute for Healthcare Improvement Model for Improvement, the team created a simplified failure modes and effects analysis⁷  and mapped the ED to inpatient admission process and identified reasons for transfer, which informed key drivers (Fig 1) and the SMART (specific, measurable, actionable, relevant, and time-bound) aim. Interventions were designed to address identified key drivers.

The multidisciplinary team identified the knowledge of resources and capabilities at the satellite as a primary key driver. A lack of clarity regarding current eligibility criteria for admission to LA4 and the availability of resources (eg, subspecialty consultation, radiology study capabilities) led to nonstandard application of admission criteria and unnecessary transfers. The LA4 medical and nursing directors, with input from HM leadership and nursing leaders, reviewed and revised outdated admission guidelines for the satellite campus in March 2021. These were shared with other satellite inpatient subspecialty services, ED providers, radiology leadership, and the chiefs of staff and then revised further on the basis of feedback. The revised admission guidelines (Supplemental Information) were disseminated broadly to HM, ED, subspecialty, and operational leaders across the institution, as well as community physicians in the spring and summer of 2021. Initially we faced challenges identifying an appropriate centralized and easily accessible location to store the living document that could be adapted as our system changed over time. Subsequently, the guidelines were stored electronically on the institution’s intranet Web site.

We identified a need for a shared mental model between the ED and admitting providers regarding who could be cared for safely on LA4, specifically in relation to patient acuity given the distance from an ICU and changing admission policies amid the COVID-19 pandemic. This led to the development of an ED assessment bundle to increase the shared understanding of patients’ clinical status and care needs between the ED and inpatient providers and nursing. The bundle also allowed for documentation of decision-making and clinical reasoning for those who subsequently cared for the patient. The ED assessment bundle consisted of the inpatient team (provider, nurse, and respiratory therapist, if applicable) visiting the ED to physically evaluate the patient and subsequently engaging in a collaborative discussion with the ED provider to determine the next steps (eg, admission to LA4 vs ICU assessment). The inpatient provider would then document the decision and its rationale in the medical record for future reference. This bundle was piloted, revised on the basis of feedback, and implemented with HM providers, because they are the largest admitting service at the satellite, in April 2021. Additionally, because HM providers do not refer their own patients to the ED as a subspecialist may, they frequently have less involvement with ED care and admission decisions, representing a potential area for improved communication between providers. Feedback about the time required to document these assessments led to development of a brief note template, which, after piloting, was added to the electronic medical record to standardize HM documentation of the shared decision-making between the HM and ED providers.

Through chart reviews of all patients presenting to the satellite ED and admitted to the main hospital (excluding ICU admissions), we identified patients who could have been cared for at the satellite based on their acuity and inpatient resource needs. Documentation of the ED assessment bundle discussions and decision-making aided in understanding reasons patients may not have been admitted to LA4. Beginning in June 2021, an HM or ED team member, who also served as medical director for their respective units, provided feedback via e-mail to individual providers and sought the identification of additional barriers to admission to LA4.

To address the staff concern that an inpatient transfer from LA4 to the main hospital because of an escalation of care that could not be provided at the satellite (eg, ICU transfer) represented a patient care failure, we sought to reframe these transfers as the appropriate identification of clinical changes in patients whose care needs progress over the course of hospitalization. Staff often expressed that this concern centered on patients and families having a negative experience if a transfer occurs unexpectedly. To facilitate proactive communication with patients and families about the capabilities of the satellite campus, we worked with family relations representatives and our institution’s communications department to develop written information for families that was distributed on admission in August 2021 (Supplemental Information). This information sheet indicated that although the satellite campus can accommodate many of the same care needs as the main hospital, occasionally, needs arise that warrant a transfer, and the family would not be responsible for any associated transfer costs. The sheet was initially provided as an insert to the family information packet that patients and families receive on admission and was subsequently permanently added to the packet. Additionally, the timing of when families received this information changed from being in the inpatient room to distribution while in the ED awaiting admission to an inpatient bed in October 2021 to optimize the patient and family’s time before the busy admission process.

We used plan-do-study-act cycles to design and learn from interventions.⁷  We followed LA4 bed capacity data, reporting the average percentage of filled beds monthly for a baseline period of 2 years (January 2019–December 2020). Additionally, we tracked inpatient transfer rates from LA4 to the main hospital and average length of stay (LOS) during this time. We followed satellite ED admissions monthly and, via chart review, identified admission disposition (main hospital vs LA4) of non-ICU admissions and reasons for admission to the main hospital. Baseline admission data were collected for 1 year (January 2020–December 2020) before the project initiation. Data were subsequently followed prospectively for 16 months (10 months of active QI work and the following 6 months).

Measures included (1) the outcome measure of monthly average bed capacity usage on LA4 (defined as the average number of occupied beds over total beds on the unit at 6 am daily) and (2) the process measure of the average monthly percentage of all patients presenting to the satellite ED admitted to LA4 over the total patients requiring admission. Balancing measures included inpatient LA4 to main hospital transfer rates and average LOS for HM patients admitted to LA4 (as the largest admitting service and because several interventions primarily targeted this population).

Given the overall stable maximum bed capacity and concern for possible associated Type 1 error,⁸  we analyzed the primary outcome measure on a run chart depicting average monthly bed capacity (by percentage bed utilization) over time. The process measure of percentage of satellite ED patients admitted to LA4 was also followed on a run chart given a smaller number of baseline data points. Run chart rules for identifying special cause were used for centerline shift (eg, 8 points above or below the center line or a trend of 6 points in the same direction).⁷  Balancing measures were followed on a statistical process control p-chart (inpatient transfer rate) and x-chart (LOS), with special cause delineated by points outside of the control limits (defined as 3 standard deviations above or below the mean).⁹ 

With this work, we aimed to improve the local quality of care and did not intend to generate generalizable knowledge. This study was not human subject research and was not reviewed by our institutional review board per institutional protocols.

Over the first 6 months of improvement work, the average LA4 bed capacity usage increased from baselines of 60% (pre-COVID-19) and 45% (during COVID-19) to 69% (Fig 2). This increase was sustained for 1 year. The average percentage of patients admitted from the satellite ED to LA4 initially increased from 76% to 84% (Fig 3). In the setting of an unanticipated surge of respiratory illnesses in summer 2021, our measure decreased to 77%. However, our process measure ultimately increased back to 84% and was correlated with sustained increased LA4 bed capacity utilization.

Our balancing measures of inpatient LA4 to main hospital transfers and average LOS for HM patients admitted to LA4 did not change over the course of the study period (Fig 4).

Using the model for improvement,⁷  we increased bed capacity utilization at our hospital’s satellite site. Although our process measure of the percentage of patients presenting to the satellite ED and admitted to LA4 was variable, increased bed capacity usage was sustained. Th strengths of this work include leveraging our multidisciplinary team, including nonclinical team members, partnering between the ED and HM teams, and focusing on collaborative communication among medical providers and between families and providers. Because of regionalization, pediatric inpatient care has been increasingly consolidated at tertiary care centers, leading to greater distances between where patients live and receive care, especially for rural communities.^10,11  Other hospital systems may model QI initiatives on this as one step toward reversing this phenomenon and, thus, easing the burden of travel on pediatric patients and their families who require inpatient admission.

After embarking on this work, we identified inconsistent communication between providers and with families as a key driver, leading to unnecessary transfers from the satellite ED for admission at the main campus. The lack of clarity regarding satellite campus admission criteria was also a key driver; we addressed this challenge by revising and disseminating admission guidelines because no previous centralized document existed. Input from stakeholders with a knowledge of the resources and capabilities of the satellite was key at this stage. The encouragement of collaborative communication between the ED and inpatient providers through the ED assessment bundle allowed for shared mental models of patient care needs and capabilities of the satellite campus and was generally well-received by both HM and ED providers. The documentation of these discussions in the electronic medical record was initially perceived by HM providers as extra work; however, a standardized note template streamlined this documentation. The dissemination of the revised admission guidelines and the implementation of the ED assessment bundle precipitated an upward shift in the center line for our outcome measure. Although we also endeavored to increase communication with families regarding the potential of future transfer to the main hospital, this did not appear to have a significant impact on our process or outcome measures. However, this transparent communication may have assuaged provider misconceptions about transfer as a failure. Regardless, the impact of this work on the patient and family experience is an area for future study.

The beginning of the COVID-19 pandemic saw a decrease in pediatric hospital admissions nationally.^4,5  Even after our institution expanded clinical services back to our satellite campus, bed capacity use lagged. This was likely multifactorial and due to many of the key drivers we identified in our improvement work. However, tracked bed capacity did not accurately reflect the number of “staffed patient beds,” or the combination of beds and staff availability that represented the true capacity on the unit. The COVID-19 pandemic exacerbated a preexisting national nursing shortage with significant increases in nursing turnover and unfilled positions.^6,12–14  In identifying staffing shortages as a potential explanation for lower-than-expected bed capacity use, even as admissions increased, we performed a post hoc review of our local nurse staffing data at the satellite. We tracked operational nurse vacancy rates (ie, the number of unstaffed nurse positions on the unit) and observed a 3-fold increase from the previous fiscal year in the operational nurse vacancy rate on LA4 during the summer and fall of 2021. Even as pediatric admissions from the satellite ED increased, the discrepancy between available beds and staffed beds limited the impact of our interventions (as noted by the decrease in our process measure during a time of high census during the respiratory surge in summer 2021 without a correlating rise in our outcome measure). Although nurse staffing was a limiting factor on the success of our interventions, implementing this work may ensure that the unit can be used to its best capacity as staffing is optimized.

The previous literature reveals financial and nonfinancial stressors related to the hospitalization of an ill child. Financial stressors may include lost wages from missed work and additional expenses, such as the costs of meals for family members at the bedside, transportation to and from the hospital, and childcare for other children.^1,15,16  Families may spend up to 6 times their daily income on nonmedical costs related to a hospitalization.¹⁵  In addition to these financial stressors, the hospitalization of a child may lead to logistical (eg, juggling work and caring for ill child and other children, child’s missed school) and emotional stressors (eg, worry, overwhelm, isolation from support systems).^1,3,16  Although our work did not formally address these stressors, admitting patients “closer to home” may mitigate long commutes and potential transportation barriers that could add additional challenges to an already stressful time for a family. An investigation of the impact of disposition decisions on stressors related to hospitalization is a crucial next step.

This QI initiative has several limitations. It was completed at a single institution within a context specific to our location which may limit generalizability to other hospital systems; however, our focus on building partnerships between the ED and inpatient services and collaborative communication may be generalizable to other system-focused interventions across specialties. Additionally, our outcome measure depicts total unit bed capacity, which does not accurately reflect true capacity as defined by staffed beds. This discrepancy may underestimate the impact of our interventions. Our process measure of the percentage of admissions presenting to the satellite ED admitted to LA4 relied on manual chart review, and we do not have these data before January 2020; therefore, the baseline time period for this measure differs from the 2-year baseline we were able to obtain for our outcome and balancing measures. Because of the inconsistency in how providers documented interdisciplinary huddles, we were unable to track the frequency of the ED assessment bundle completion as an additional process measure. Our data reflect all patients admitted to the satellite inpatient unit, irrespective of admitting service, although many of our interventions focused on HM as the admitting provider. The expansion of interventions to the other admitting services may have a larger impact on our outcomes. Although our team included a family advocate, our study lacked direct patient and family perspectives, which is important to consider when assessing interventions that impact disposition.

Using QI methods, we successfully increased bed capacity utilization at our satellite campus. Although our interventions revealed a variable impact on our process measure of the percentage of patients admitted to LA4 from the satellite ED, this was likely related to staffing challenges limiting the number of available staffed beds in the setting of high census. Next steps include a closer analysis of the interplay between unit bed capacity use as it relates to staffing. Additionally, the authors of future efforts should examine the impact of this work on front-line providers and the patient and family experience, including whether some stressors related to hospitalization may be mitigated by admitting children “closer to home.”