Rethinking Operating Room Management: Why Clinical Directors Should Embrace Complexity

THE CURRENT PARADIGM: THE FAILURE OF REDUCTIONISM Managers are not confronted with problems that are independent of each other, but with dynamic situations that consist of complex systems of changing problems that interact with each other. I call such situations messes…managers do not solve problems, they manage messes. —Ackoff1 Health care delivery and its role in the broader economic system is increasingly complex.2 Complexity theory may serve as an ideal platform for operating room (OR) management. OR governance structures have been modeled after administrative leadership structures, in which management decisions are centered on planning, budgeting, organizing, staffing, and problem-solving. These initiatives are all directed at creating predictable, stable processes. In truth, a perioperative system represents an ecosystem of surgeons, nurses, anesthesia health care providers, and patients, interconnected to influence the delivery of surgical care.3 For many health care organizations, perioperative services consume a relatively large share of resources and generate significant revenue streams. Traditionally, OR management decisions are subdivided into 3 categories: strategic, tactical, and operational (Figure 1).3 Strategic decisions focus on establishing a niche or equilibrium in the environment and offer a long-term perspective. Strategic analyses usually aim to identify contextual “boundaries” established by institutional constraints, the local and regional environment, and the organizational mission. Conversely, tactical decisions focus on the utilization of current resources in the near future. Strategic decisions may include building a new service line or expanding an ambulatory surgery center so hospital administrators can expand market share or create operational efficiencies.4 Tactical decisions, in contrast, focus on block allocations, the perioperative nurse skill set needed for the surgical workload, and necessary anesthesia services.Figure 1.: Traditional framework for operating room management decisions showing how strategic and tactical decisions inform operations over time in a unidirectional fashion.Each year, as we attempt to manage the health care system with a series of rules and regulations, the health care system in the United States continues to create tremendous inefficiencies, leading to little progress toward reducing the cost of care. Reeves et al5 recently argued that taming complexity requires the following efforts: creating a modular structure; using simple, common operating principles; embedding a bias for change; relaxing control; optimizing globally; and fixing, repairing, and pruning. Presumably, by adopting the above framework, perioperative managers should be able to manage the perioperative services with better-informed strategic, tactical, and operational processes. Today, much of the current literature is based on the assumption that ORs resemble manufacturing plants. Here, lean manufacturing and Six Sigma approaches are used to reduce variability in clinical processes and minimize inefficiencies to create value.6 With this analytic or reductionist approach, the complex parts are reduced to smaller components. However, we believe that this decomposition usually leads to a loss of information, especially when it comes to downstream consequences, both financial and operational.7 In many respects, these efforts resemble the short-sighted “lean” management style of the past, of a culture that lasts because it works at the present time, even though there is evidence, as we will show, of its inferiority and unsustainability. Previously, Mahajan et al8 noted that perioperative systems behave like complex adaptive systems.9 For example, there are numerous expectations regarding block management, case scheduling, management of add-ons and emergent cases, and individual patient variability. This inherent variability occurs not only at a single institution but also across institutions. Further, complex adaptive systems are characterized by Volatility, Uncertainty, Complexity, and Ambiguity (VUCA; “volatility” represents the systems propensity for sudden and dramatic change; “uncertainty” speaks to the ability, or inability, to predict the incident, frequency, amplitude, or timing of the change event; “complexity” encompasses the multitude of factors, control nodes, and confounding variables that creates a “chaotic,” dynamic, every evolving landscape; and “ambiguity” represents the “fog of war” that robs one’s ability to clearly delineate the relationships, causes, and effects, leading to a potential for assumptions and biases that color decision-making and blur judgment).9 VUCA represents a framework that encompasses the multiple and diverse user interactions in any complex system to seek innovations in a dynamic environment and evaluate these through deliberate transitioning processes. The diversity of activities, the myriad of equipment, persisting uncertainty, patient and provider vulnerability, and modes of delivery are emblematic of the complexity of perioperative systems. Below, we review how perioperative services behave like complex systems and then apply the lens of complex adaptive systems to understand how to harness the power of complexity. PREDICTABILITY IN AN UNPREDICTABLE WORLD: THINKING IN SYSTEMS Every day, in every OR across the country, various teams of anesthesiologists, nurses, and surgeons decide where, when, and how cases are accommodated and completed, using mostly “rules of thumb” or intuition.9 Most health care systems try to maximize OR utilization rates and maintain high throughput to maximize revenue streams.10 Therefore, operational decisions made on the day of surgery attempt to minimize variability from the perioperative processes.4 Ironically, even though anesthesiologists expect variability in the way patients respond to treatments, anesthesia clinical directors are taken aback and often challenged to understand “illogical responses” of the perioperative system they work within.11 Within any perioperative system, a “stock” (referring to the previously mentioned “manufacturing plant” analogy) can be represented by block allocations. For anesthesia clinical directors, the ability to manage OR throughput is through inflow to and outflow from the OR. Although block allocations may change over time, the dynamic responsiveness inherent in any complex system exists because of many feedback loops.12 These feedback loops, whether positive or negative, occur when human agents disrupt the system by implementing and executing strategic, tactical, and operational decisions. By acting as open systems, block allocations respond to external input and self-evolving feedback loops that are often changing and unpredictable.12 In contrast to fixed rules and regulations, health care organizations are natural systems that are continually changing, and the agents within these systems are constantly adapting to these changes.13 As nonlinear systems, output from a complex system is not proportional to the input generated, as it would be in complicated systems.14 Returning to the previous block allocations example, surgeons may use block allocations to maintain access to perioperative services (Figure 2). Unfortunately, this self-serving mentality increases inefficiency at the systems level and makes it difficult for OR managers to appropriately manage variability in the daily workload. A specific surgeon’s “backlog” is only one of many inputs into the system. However, many other services have patient backlogs, and patients entering perioperative services can come from the emergency room, the inpatient setting, and intensive care units. Simply focusing on patient backlog assumes that the linear model presented above captures the complexity of a perioperative system, limiting the opportunity for predictive load balancing and staffing adjustments to mitigate overutilized time.Figure 2.: A bathtub is analogous to the block allocations of the OR (Figure 1) and represents the time necessary for a surgeon or surgical service to operate despite the changing demands and workload. The major service lines create demand. Release times represent a surgeon’s or a surgical service’s ability to control the system “stock,” and the patient backlog is a consequence of the release time for block allocations from the surgeon’s perspective. Presumably, many surgeons continue to use release time (ie, the point in time when future OR block allocations are made available to the other surgical services) as a method to protect their own clinical operations. However, the perioperative service system is a nested system; the outflows are directly affected by patient access to rehab facilities, inpatient settings, and the clinic. OR indicates operating room.Often, the least obvious part of the system is its function or purpose, which is also often the most crucial determinant of the system’s behavior.14 Consequently, most organizations actively and intentionally create 2 conflicting organizational models: one supports stability, as defined by organizational structure with roles/responsibilities and a governance model that addresses decisions with broad scope and impact; and the other supports agility (eg, interdisciplinary teams focused on decisions that are narrow in scope or exploratory in nature and new processes that address evolving changes in the environment).15 Again, feedback loops are nonlinear within any perioperative system, and unexpected consequences might begin to show up as first-case start delays, longer turnover times, or lack of surgical equipment.12 By understanding the role of positive and negative feedback loops, a perioperative service should be able to design a system that strives to maintain an equilibrium. In terms of a surgical service’s block allocations, it should be recognized that the equilibrium is merely a distribution at a specific point in time. Ultimately, the astute operating room managers should understand the downstream impact of tactical and operational decisions in the context of maintaining an equilibrium that is in alignment with hospital strategy as delineated by executive leadership. ORGANIZATIONAL AGILITY: REAL-LIFE APPLICATIONS Inevitably, clinical directors are confronted with new technology, operational demands, and ever-changing regulatory and financial forces. They must manage a complicated set of rules, policies, and processes impacting clinical delivery in a complex patient ecosystem.13 We will elaborate on the earlier mentioned complexity-taming efforts defined by Reeves et al,5 applying them to real-world scenarios where the consequences of isolated strategic, tactical, and operational decisions have larger ramifications. In contrast, we hope to demonstrate that it is possible to move perioperative services toward self-governing, transparent processes that seek equilibrium. Strategic Versus Tactical At many institutions, there is a lack of congruency when it comes to the implementation of strategic decisions and the allocation of tactical resources. When the University of Vermont Medical Center instituted a minimally invasive surgical service, the Division of Urology led the strategic and operational changes necessary to get the initiative off the ground. The Division of Urology worked closely with the Department of Anesthesiology and the perioperative nurses to communicate the anesthetic and procedural considerations for the patient. However, the Division of Urology elected to put the robotic unit in an OR that had traditionally served as the dominant block allocation for the service. Although the strategic plan was appropriate, tactical errors appeared soon after implementation of the new technology. The tactical decision to place the new equipment, used by multiple services, in a room used by a single division, coupled with an operational metric of defining block time by room rather than service, resulted in the unintended consequence of the Division of Urology working outside of their block allocations any time another service needed the robotic equipment. This corruption of utilization data had a domino effect on other services as well, resulting in an inability to accurately track and manage to clinical needs. In August 2013, the University of Vermont Medical Center was more careful the next time when the Division of Vascular Surgery created a hybrid OR unit. The redesign involved an OR that had traditionally been the daily block allocation for the Division of Vascular Surgery. Here, the tactical committee asked the Chief of Vascular Surgery: Who should control the block allocations: the committee or the division? The surgeon requested that the division manage the block allocations for the hybrid OR (modular structure). To prevent the tactical errors associated with implementation of the robotic unit, all surgical services requesting time for the hybrid OR were required to swap block allocations with the Division of Vascular Surgery (an example of simple, common OR principles). From a complex systems perspective, management of block allocations for the hybrid unit did not necessitate a rigid bureaucratic structure, but rather depended on relaxing central control and giving more responsibility to end users. Finally, this move to a “floating” block utilization facilitated the revalidation of block utilization and helped the Operating Room Block Utilization Committee recognized the need for additional block time for the Division of Urology (an example of fixing, repairing, and pruning). When “managing” a complex perioperative service, embracing complexity means relaxing control, pushing the time and resources to the individuals on the front line, and embracing change that facilitates systems evolution that supports the institutional mission. Tactical Versus Operational Tactical decisions in OR management focus on the allocation of block time and are usually made 3–6 months in advance.4 Block allocations are predicated on the total surgical demand for a future date and the anticipated variability in the workload for the perioperative system. Managing this anticipated workload is a function of the tactical block allocations and staffing patterns. Subsequently, many institutions develop rules on first-case start delays, establish minimums for block utilization rates, and optimize release times. At the University of Pittsburgh Medical Center, the perioperative services implemented a strategic initiative to globally optimize staffing allocation and utilization for all the surgical sites across the network. In January 2017, individual anesthetizing location site limits were set at for each surgical facility based on the 90th percentile of rooms running for the previous quarter, establishing a system-wide cap on locations at 208 as an initial step in controlling maximum anesthetizing locations across the system. This process has evolved to include adjustments based on expected provider efficiency, surgical time divided by total staffed time, with a stepwise reduction in defined total system-wide anesthetizing location to a limit of 194 beginning in February of 2019. To actively manage case volume and labor resources within this cap, 3 days before the day of surgery, the Central Scheduling Department sends an electronic announcement to all sites when the total number of expected anesthetizing locations exceeds the defined system cap (another example of using simple, common operating principles). By embedding a bias for change, the individual surgical facilities work in a collaborative manner with the primary tactical goal of consolidating the schedule down to within the total site allocation limits. For example, surgeons with first-case starts with no case to follow condense their case lists into 1 site. Subsequently, the workforce necessary for the final distribution of anesthetizing locations is finalized and deployed for distribution across the system the day before surgery. Again, tactical decisions may mitigate, but not necessarily solve operational issues. It is the role of clinical directors to minimize the implications and consequences of these decisions on the day of surgery. In this type of organizational framework, frontline employees are empowered to identify and address emerging problems and opportunities and exploit any opportunity in real time. In agile organizations, individual teams focus on small parts of a complex adaptive system and work with a list of organizational priorities.15 The responsibility of the upper-level managers is to relax control, shift decision-making farther down the line, and create the urgency to continue refining the process to align operations with the mission and values of the organization. Operational Versus Strategic Each day, operational decisions mitigate or amplify deficiencies in tactical processes up until the day of surgery.4 In 2009, a private ophthalmology group at the University of Vermont Medical Center decided to leave the ambulatory center and build their own surgical center. With this move, the surgical volume at the ambulatory center dropped from 5000 to 3000 cases annually. Faced with a 40% drop in surgical volumes, the organization should revisit its tactical allocations and shorten the staffed hours to match the historical workload (an example of fixing, repairing, and pruning).4 However, in this instance, the institution bewilderingly embarked on an operational efficiency project to improve turnover times at the ambulatory center. The literature is strewn with similar misconceptions of time savings.16,17 Shortening turnover times, an operational metric, to satisfy a strategic initiative will not mitigate a tactical issue. The most important aspect of achieving system efficiency is that, in dynamic, complex systems, optimization of the system must supersede optimization of any individual node.18 Unfortunately, there are ample instances of perioperative decisions made to optimize individual operational metrics for an individual or service line that carries the most clout.19 By embracing complexity, perioperative services can transform and innovate processes at strategic, tactical, and operational levels for perioperative services. As surgeons, nurses, and anesthesiologists gain experience and accumulate systemic knowledge of their administrative and operational systems, they need to work as teams and build “processes that balance freedom and control.”20 Ultimately, a clinical director must have a true understanding of the system’s competing goals, embed a bias for change by strengthening the decision-making power of those working in the system, and remove the organizational obstacles that prevent a perioperative service from adapting to changing conditions and demands. CONCLUSIONS Perfect prediction is only possible under 2 conditions: when nothing changes and when the behavior of a system occurs in “accordance with deterministic causal laws, and that we know perfectly these laws.”1 Managing a complex adaptive system requires the foresight to recognize the limitations of reductionist approaches. The single-minded drive for efficiency does little to create true innovation. Perioperative leaders need to create modular governance structures; develop simple operating procedures; simultaneously invest the resources necessary to create change and relax control; and recognize that in globally optimizing dynamic, complex systems, optimization of the system must supersede the optimization of any individual node or metric. In closing, we need to “design and manage systems so that they can effectively serve their own purposes, the purposes of their parts, and those of the larger systems.”1 DISCLOSURES Name: Mitchell H. Tsai, MD, MMM, FASA, FAACD. Contribution: The author helped create, prepare, and edit the manuscript. Name: Stephen J. Kimatian, MD. Contribution: The author helped create, prepare, and edit the manuscript. Name: James R. Duguay, MD. Contribution: The author helped create, prepare, and edit the manuscript. Name: Mohan R. Tanniru, PhD. Contribution: The author helped create, prepare, and edit the manuscript. Name: Elie Sarraf, MDCM. Contribution: The author helped prepare and edit the manuscript. Name: Mark E. Hudson, MD, MBA. Contribution: The author helped prepare and edit the manuscript. This manuscript was handled by: Nancy Borkowski, DBA, CPA, FACHE, FHFMA.