Addressing Variability in Emergency Department Hospitalization Rates: Insights from Daniel Kahneman's Noise

The decision to admit a patient from the emergency department (ED) into the hospital varies significantly across different regions, hospitals, and even among physicians within the same hospital. Such variability, often referred to as noise, poses a significant challenge in healthcare delivery. This concept, explored by Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein signifies the unwanted variability in professional judgments that impact hospitalization rates. This brief highlights the issue of hospitalization rate variability through the emergency department, employing Kahneman's framework to understand its underpinnings and to propose strategic interventions.

Unpacking Noise in ED Hospitalization Decisions

Kahneman and his co-authors describe noise as the variability in judgments that should be identical. Noise can have a significant impact on patient outcomes, despite being often overlooked in favor of more overt forms of error, such as biases. It is evident that noise is not merely a byproduct of the healthcare system’s complexity but a critical factor influencing patient care paths and healthcare resource utilization. Although some providers may experience bias toward admitting more elderly, nursing home patients, or those with a similar clinical presentation to their only malpractice case, much of the variability is simply noise that is not biased in a particular direction.

The Multilayered Variability in ED Admissions

Smulowitz’s research highlights variability in ED admission rates that stretch across several dimensions: regional, hospital, and physician. These variations suggest that factors like institutional norms and individual decision-making styles play a crucial role in the observed discrepancies, for instance, within-area variations—differences across hospitals in the same region and among physicians within the same hospital. Each level introduces its own set of variations, compounding the complexity of addressing this issue, and highlighting the need for a nuanced understanding and targeted interventions that address each dimension. Despite attempts to adjust for patient characteristics and other variables, significant unexplained variations remain, highlighting the pervasive influence of noise.

Kahneman's Framework for Mitigating Noise

To confront noise, Kahneman proposes actionable strategies that are directly applicable to reducing variability in ED hospitalization rates. The implementation of standardized guidelines stands out as a potent tool for aligning clinical judgments. Such ED guidelines have been partially implemented for conditions like low risk chest pain, uncomplicated DVT’s, or community acquired pneumonia but need to be expanded to cover areas requiring more subjective judgment. Time of day or week, business of the Emergency Department, provider shift changes, availability of discharge support, and family caregiver availability post discharge compound to the variability that is observed.

Fostering collaborative decision-making between ED staff and hospitalists may protect against pattern noise, which arises from the varying standards applied by different physicians or specialties. Collaborative approaches ensure that a diversity of perspectives is considered, leveraging the collective expertise of the medical team to arrive at more consistent and justified admission decisions. Adding room in fixed ED admission guidelines for patient specific subjective judgment can assist in making standardized decisions that are applicable to a larger variety of case presentations by incorporating nuances and reducing noise among different hospitals.

Value-based Hospitalists Programs

Many hospital-employed or contracted hospitalists are given bonuses on efficiency based on their Relative Value Units (RVU). Systems that incorporate these bonuses tied to provider efficiency will see providers admit patients into the hospital at higher rates because it systematically takes more time for them to find an appropriate alternative placement for a patient and discharge them from the ED. Although the initial ED provider admit decision variability is considered noise, the hospital’s unwillingness to safeguard against it by putting more tools in disposition and discharge can bias providers to value being efficient over evaluating patient complexity.

Strategic Recommendations

Standardized Admission Guidelines: Develop comprehensive, evidence-based guidelines for most frequent medical admissions (i.e. CHF, COPD, PNA). The ED pathways should include timely treatment and appropriate observation of these patients, as a percentage improve enough to go home in a few hours. Encourage physicians and incentivise hospitals to adhere to evidence based guidelines by rewarding consistency.
Collaborative Decision-Making Processes: Develop systems that lead to joint admit decision and accountability of subsequent patient outcomes between ED and Inpatient teams. Encourage routine review of cases at joint conferences, and build a shared culture between providers. Rotate residents between the ED and Inpatient teams so they can understand the care from dual perspectives, learn how to effectively communicate their recommendations, and make better informed decisions.
Spectrum of Discharge Options: Develop a range of suitable discharge options for patients (i.e. Hospital@Home, ED to SNF), moving away from the binary choice of simply admitting or discharging with minimal support. This approach allows for more nuanced patient care tailored to individual needs and circumstances and allows providers to discharge patients with more confidence in their after hospital care.
Value-based Care: Appropriately compensate hospitalists for the incremental work it requires to collaborate on discharging patients from the ED through RVU adjustments or VBC bonuses. Encourage VBC coordinators and educate administrators on how to use VBC strategies to supplement their existing financial frameworks.
Technology-Enabled Decision Support: Deploy advanced decision-support tools that provide real-time, evidence-based recommendations to ED staff, aiding in the reduction of noise by offering a standardized reference for decision-making. This aids ED staff by giving a third party access to accelerate cases for concern and support the discharge of lower acuity patients while using evidence-based predictive data analytics and patient monitoring.
Case Management and Social Work: Provide extended hours of disposition support for providers who are attempting to reduce the hospitalization noise on evenings and weekends.

Conclusion

The variability in ED hospitalization rates is a multifaceted issue deeply rooted in the noise inherent in clinical judgment. By applying Daniel Kahneman's insights into noise and its effects, healthcare leaders can begin to untangle the complexities of this variability, implementing strategies aimed at reducing noise at its source. Through standardized guidelines, collaborative decision-making, professional development, and technology-enabled support, it is possible to achieve a more consistent, equitable, and efficient approach to hospital admissions from the emergency department.

Written By: Reza Alavi, MD, MHS, MBA; reza@quintupleaim.com