Tiered approach on opening an observation unit: a quality improvement case study using PDSA (Plan-Do-Study-Act)

Dillon W. Casey, MD
CDU Medical Director High Point Medical Center , Atrium Health Wake Forest Baptist
Lisa Duncan, RN BSN
CDU Section Administrator High Point Medical Center, Atrium Health Wake Forest Baptist

Introduction

This piece serves as a reflection on the quality improvement of a staggered approach to opening a new observation unit. Our Emergency Medicine Clinical Decision Unit (CDU) at Atrium Health Wake Forest Baptist-High Point Medical Center was opened in March 2023 using this phased strategy. Initial data and findings were presented in September 2023 at the Observation Medicine- Science and Solutions Conference, now in March 2025, we have two years of data and trends to analyze, further confirming the viability of this approach.

Traditionally, new hospital observation units are launched all at once, often placing a significant strain on staff, resources, and workflow adaptation. At Atrium Health Wake Forest Baptist-High Point Medical Center, we implemented a staggered opening for our new Emergency Medicine Clinical Decision Unit (CDU), a model that unfolded over several months and demonstrated significant advantages in workflow refinement, resource allocation, staff adaptation, and patient care. This case study highlights the benefits of this approach and its potential implications for future observation unit openings.

Background

Our CDU is a 12-bed, individual-room unit staffed by two 12-hour APP shifts (8 AM–8 PM and 10 AM–10 PM), with two hours of physician rounding each morning from 9 AM to 11 AM. There are no dedicated APPs covering the unit from 10 PM-8AM. Overnight, the patients continue to receive protocol driven care and the nightshift ED attending is available to cover any urgent concerns that arise. New patients can be placed in the CDU overnight from the emergency department using our CDU protocol order sets for patient care.

Our CDU was modeled after an established observation unit and was designed as a protocol-driven, closed unit within the emergency department. Instead of launching at full operational capacity, the CDU's opening was executed in a phased manner over four months:

• Phase 1:Open three days per week (Monday-Wednesday)
• Phase 2:Expanded to five days per week (Monday-Friday)
• Phase 3:Increased to six days per week (Monday-Saturday)
• Phase 4:Full operational capacity (seven days per week)

This tiered approach allowed for progressive workflow adjustments, real-time staff feedback, and the identification of operational hurdles before the unit reached full capacity. By structuring the rollout in this manner, we ensured a measured, thoughtful approach to patient care and operational readiness, avoiding many of the pitfalls seen with full-scale, immediate openings.

Methods

During each phase, we closely monitored key performance metrics, including:

• Patient volume
• Length of stay (LOS) in the CDU
• Inpatient conversion rates
• Provider and protocol utilization

Additionally, multidisciplinary meetings were held weekly to discuss observations, challenges, and necessary modifications before progressing to the next phase. This iterative process provided an opportunity to refine protocols, adjust staffing needs, and ensure that both efficiency and patient care quality remained at the forefront.

Results

Over the course of the phased rollout, a total of 517 patients were seen in the CDU. The findings from each phase highlighted key improvements:

• Phase 1:46 total patients, median LOS of 15.03 hours, inpatient conversion rate of 28%
• Phase 2:88 total patients, median LOS of 15.29 hours, inpatient conversion rate of 18%
• Phase 3:135 total patients, median LOS of 15.68 hours, inpatient conversion rate of 15%
• Phase 4:248 total patients, median LOS of 15.05 hours, inpatient conversion rate of 18.9%

We observed a consistent and manageable increase in patient volume while maintaining relatively stable LOS and improving inpatient conversion efficiency. Importantly, the staggered approach allowed for iterative improvements and problem-solving between phases. Each transition phase allowed staff to adapt to increased demand while troubleshooting logistical and operational barriers before fully scaling up.

Since adopting this staggered approach, our unit has now surpassed the two-year mark in operation. This model has demonstrated its feasibility and sustainability. With two years of data available, we can now analyze trends that further confirm the effectiveness of this phased implementation.

Last 12 month averages:  

• 188 patients per month, median LOS of 17 hours, inpatient conversion rate of 22%

Qualitative Benefits

In addition to quantitative improvements, the phased opening yielded several qualitative benefits:

1. Enhanced Workflow Refinement
   • Progressive adjustments minimized disruptions to the emergency department flow and throughput.
   • Staff had time to adapt to new workflows before the next expansion.
   • Real-time feedback from multidisciplinary meetings led to iterative improvements.
   • Unexpected inefficiencies were identified early, allowing for proactive solutions rather than reactive fixes.
2. Improved Staff Adaptation
   • Stakeholders experienced a gradual learning curve rather than abrupt changes.
   • The phased approach facilitated stronger team cohesion and confidence in new protocols.
   • Training new staff became more manageable by onboarding them in phases rather than overwhelming them with a fully operational unit on day one.
3. Maintained Patient Care Standards
   • Controlled expansion reduced the risk of patient care disruptions.
   • Gradual increases in patient volume ensured quality standards remained intact.
   • The staggered approach allowed for a focus on quality improvement initiatives, ensuring patient safety was prioritized at all times.
4. Early Identification of Operational Hurdles
   • Logistical challenges with consulting services (e.g., Neurology, Cardiology) were detected early.
   • Solutions were implemented between phases, preventing bottlenecks at full-scale operation.
   • Challenges related to interdisciplinary communication were worked through in real-time, ensuring that patients were not impacted by administrative delays.

Discussion and Future Implications

The staggered opening model provided a structured yet flexible approach to implementing a new observation unit. By allowing staff to refine workflows, adapt incrementally, and troubleshoot potential barriers before full-scale operation, we were able to achieve a smoother and more efficient launch that has led to sustained success over the last two years.

This approach may be applicable across various clinical settings and unit types, particularly in high-acuity areas like observation medicine where workflow disruptions can significantly impact patient care and total length of stay is such a key metric for hospital administrators. Future research should explore the scalability of staggered openings in other institutions and specialties. Additionally, it would be beneficial to examine patient satisfaction metrics in units that utilize a staggered opening model versus traditional launch methods.

Conclusion

The phased approach to opening our CDU resulted in tangible benefits in workflow efficiency, staff adaptation, and patient care quality. This case study supports the idea that staggered implementations can enhance operational success and should be considered as a best practice for launching new observation units in emergency medicine. With two years of data confirming its sustainability, this model serves as a valuable framework for other institutions seeking to optimize quality improvement and patient outcomes in their observation units.