Emergency and Disaster Logistics Processes for Managing ORs Capacity in Hospitals: Evidence from United States

Hospitals and the concept of a healthcare delivery system are practically synonymous. Surgical services, emergency and disaster services, and inpatient care are the three main services they offer. Outpatient clinics and facilities are also available at some hospitals, where patients can receive specialty consultations and surgical services. What challenges do hospital administrators face when attempting to balance the supply and demand for medical services while preserving service quality and maintaining low prices? In what ways and to what extent has OM literature contributed to the resolution of these issues? What are the current trends in practice? What additional opportunities and challenges do they present to OM students? This study is an attempt by the authors to answer the aforementioned questions. Although service capacity could be measured by the number of physical and human resources employed, we chose to focus on the three primary types of services provided by hospitals.


INTRODUCTION
We show how these services are interconnected and the issues that develop as a result of mismatches in one area having a cascading effect on all other hospital operations. Our goal is to bring essential concerns to light from practitioners' perspectives, use representative data to highlight difficulties that can be modelled utilizing OM techniques, describe state of the art in modeling such challenges, and recommend future research prospects.
Emergency logistics refers to the logistics actions triggered by emergencies, including restricting emergency logistics demand and emergency logistics supply activities to fulfill these logistics needs, all to improve time efficiency and reduce disaster losses. In emergency logistics, Healthcare expenditures are increasing as a percentage of GDPs in the 34 OECD nations with publicly available data (OECD, 2013). In 2012, ten countries spent between 10% and 11.8 percent of their GDPs on healthcare. With 16.9%, or nearly 2.8 trillion dollars, the United States (US) was an anomaly. Furthermore, inpatient care, the subject of this research, accounts for almost 30% of all healthcare spending (OECD, 2013;. Hospital beds have been steadily declining in the OECD countries over the last two decades, according to official data. It varies from country to country (e.g., McKee (2004) and from year to year (Roberge et al. 2010), but one major factor has been the control of hospital costs (Kroneman & Siegers, 2004). Based on Centers for Disease Control and Prevention, the average number of acute-care hospital beds per 100,000 people decreased from 410 in 2000 to 330 in 2012, and the average hospital stay decreased from 8.4 to 7.7 days. A similar increase in hospital discharges occurred during this time period, with the rate rising from 15,160 per 100,000 people in 2000 to 15,800 in 2010. (2015). (OECD, 2015). It is estimated that approximately 80 percent of inpatient beds are occupied on average. According to the Organization for Economic Cooperation and Development (OECD), Because of these trends, which include higher rates of bed utilization and shorter average lengths of stay, it has become increasingly important to match inpatient demand with available beds.
For five OECD countries, the data in Figures 1 and 2 summarize data on acute hospital bed supply, total hospital discharges, lengths of stay, and use of acute care beds (OECD, 2013;. The top panel of Figure 1 depicts a decreasing supply of available beds. Aside from Canada, the discharges in the bottom panel of the same image are either flat or increasing in size (demand). Consistent or mixed patterns of acute care bed utilization are depicted in Figure   2 at the top, with decreasing average lengths of stay depicted at the bottom.
Specifically, the data in Figures 1 and 2 are significant because they demonstrate how lower bed supply, shorter periods of stay, and higher demand have all had an impact on bed occupancy rates in various countries. Despite the fact that the United States and United Kingdom have approximately the same number of acute care beds and discharges per capita, the average length of stay in the UK is approximately two days longer than in the US. These disparities are consistent with the United Kingdom's counted bed occupancy rate of 85 percent compared to the United States' 65 percent. It may appear that there is an excess of hospital beds in the United States, but because the OECD computation is based on the number of beds that a US hospital is licensed to staff, these figures are not accurate; instead, they are misleading (1, 2). Hospitals adjust their staffing levels on a daily basis to meet the demand for available beds.
Because hospitals in the United States frequently staff fewer beds than are licensed to provide care, staffed-bed utilisation is frequently significantly higher than the national average. It is possible that the situation in the United States is comparable to that in other OECD nations. It is clear from these observations that it is critical to understand operational aspects even when interpreting aggregate data. available is sufficient, and better bed management is required (Allder, Silvester, Walley, 2010).
It is critical to comprehend, shape, and respond to the demand for acute care services, regardless of one's point of view. The fluctuation in demand and availability for such services is a significant problem for hospital executives.

LITERATURE REVIEW
The availability of hospital beds varies significantly depending on where you live in the United States (Dartmouth College, 2012). Discharge rates per 1,000 Medicare beneficiaries 1 . Payment innovations are intended to reduce or eliminate redundant or superfluous services from the payment landscape. After a cost-plus algorithm was implemented in the mid-1980s, the Medicare reimbursement for hospitalization was replaced with a Diagnosis-Related-Group (DRG) episode-payment mechanism. A decrease in the availability of hospital beds resulted in a reduction in the average length of hospital stay, while an increase in the use of post-acute transitional care was seen as a result (Qian et al., 2011;Fischer et al., 1989;Kosecoff et al., 1990). Medicare recently imposed payment penalties on hospitals that had a 30-day readmission rate for patients with specific diseases that exceeded a predetermined threshold, according to a press release (Medicare Payment Advisory Committee, 2014).
Medicare and commercial insurers are also experimenting with new compensation systems in order to reduce the demand for hospitalizations and outpatient care even further. These approaches to compensation place a greater emphasis on quality than on quantity (Ryan & Press, 2014). Conrad et al. (2013) describes several reimbursement innovations that aim to promote integrated care by aligning clinical and financial incentives among all providers involved in a patient's episode of care (for example risk-adjusted global payments, Pay-forperformance, and bundled payments).
Numerous tactics used by healthcare delivery systems to improve their performance in response to reimbursement changes can be found outside of the hospital setting. These programs, which emphasize prevention and wellness in the outpatient setting, aim to reduce the use of emergency departments and hospitalizations. In the case of the patient-centered medical home (Schoen et al., 2011;Nutting et al., 2011), for example, interprofessional physician-led teams provide care to small groups of patients. Another example is the provision ISSN:  (Sutherland & Hayter, 2009;Lion et al., 2014;Mathews, 2014). Physician compensation models are also being implemented that place a greater emphasis on patient outcomes, such as specific biomarkers, rather than the number of patient contacts made by the physician (Latham & Marshall, 2015;Bailit et al., 2015).
Until all insurers transition to an outcome-based payment system, many health-care organizations will be forced to compete with one another for reimbursement. Value-based reimbursement accounts for only a small proportion of many systems' reimbursement streams; the vast majority are still fee-for-service. As a result, when companies use case management solutions, they run the risk of penalizing their bottom line (lower fees), while also incurring more extraordinary production expenses.

WHAT KIND OF SERVICES DO HOSPITALS PROVIDE?
The majority of acute-care hospitals offer at least three major service categories. Figure 3 shows a diagram of the three types of care available: emergency treatment (provided by emergency departments, also known as EDs), surgical services (provided by operating rooms, also known as ORs), and inpatient care (provided by inpatient beds). Throughout the rest of this section, we will discuss hospital beds in order to describe the institutional characteristics that make demand-supply matching difficult. Other types of services face challenges that are similar to these. The patient may be admitted to an inpatient bed following a visit to the emergency room or surgery, or following a physician's request for direct admission to the hospital, as illustrated in Figure 3. Patients are typically moved from one intensive care unit to another several times during their hospital stay, either to a lowerintensity unit if their condition improves or to a higher-intensity unit if it worsens. PATIENT FLOW is the term used to describe the movement of patients from emergency departments, operating rooms, or direct admission through one or more inpatient wards. There has been a lot of discussion about it in the occupational medicine literature.

Demand Variability
Because many patients require a combination of treatments, patient flow creates a random Discharges, on the other hand, are more common in the afternoon. This results in a systemic imbalance of hospital bed demand and supply. The term "census" refers to how hospitals keep track of the number of beds they have. Figure   5 depicts the typical census trend for a 500-bed hospital by the time of day and day of the week.
Because the previous day's patients are still in the hospital and new patients arrive, the daily census peak shifts later in the day as the week progresses. The variability around the mean census increases later in the week. The goal is to discharge eligible patients by Friday afternoon to reduce the need for nurse staffing over the weekend, resulting in a lower weekend census.
Many surgeons prefer to operate earlier in the week to avoid the need for weekend hospital rounds. On weekends, operating rooms only perform urgent or emergency surgical cases, resulting in a lower patient intake at inpatient units.

OPERATING ROOM (OR) CAPACITY MANAGEMENT
Operating Room (OR) capacity management falls under the purview of multiple stakeholders.
The key players are (1) the OR directors, (2) the service line or surgical group directors, (3) the nursing directors, (4) the surgeons, (5) the anesthesiologists, and (6) the patients. As we explain later in this section, the objectives of these stakeholders are often not aligned. In contrast, others keep a significant proportion of their capacity unassigned. Typically, rules are in place to determine how unassigned capacity is allocated to arriving requests for OR time.
The overall effectiveness of OR capacity matching depends on all these rules and policies.
A different way of thinking about variability is to consider the extent to which it can be anticipated or predicted. Common wisdom suggests that artificial variability should be more predictable because stakeholder actions are observable, often several days in advance. In practice, hospitals may not collect that information and/or may not leverage advanced knowledge of stakeholder actions to lower supply-demand matching costs.
The problem of matching supply and demand (red alert) 1 for OR time is sometimes viewed via the prism of time scales. Considerations such as the sorts of surgical operations performed, 1 Major incident situation happened, how CCUs, and specialist staff act? Between (5 percent -15 percent) of patients presenting to hospital following a bomb blast or terrorist-associated mass casualty incident will need intensive care 2021;Adam & Osborne, 2005). 2021;Adam & Osborne, 2005). The demands on critical care sources for other incidents, like large fires or natural disasters, are less well recorded (Avidan et al., 2007). (Avidan et al., 2007). However, following the aftermath of Hurricane Katrina in New Orleans, the Charity Hospital had to manage enhanced demand without the opportunity to evacuate for many days (Rubinson et al., 2008). (Rubinson et al., 2008).
In 2007 the US Taskforce for Mass Critical Care offered that CCU plan to represent emergency mass critical care at three times the existing capacity for up to ten days (deBoisblanc, Bennett, 2005;. (deBoisblanc, Bennett, 2005;. However, CCU beds and specialists/staff are a limited source usually fully utilised; in a major incident, releasing beds or expanding the source needs good planning and organisation. As critical care sources will often represent the major limiting element encountering large numbers of casualties, it is also essential that early connections with different hospitals be made to transfer more stable ISSN: 2815-9330 (Online) VOL. 1, NO. 1 2022 www.iessociety.org the number of operating rooms (ORs), and how each OR is equipped are long-term decisions in this paradigm (typically made once every few years). Decisions concerning block allocations to surgeons, how many ORs to plan to staff each weekday, and staffed-OR start and end times are medium-term decisions (typically made annually). Finally, surgical-case booking decisions patients in the most secure way possible. Based on this, critical care managers/leaders must be involved in planning  for mass casualties along the healthcare communities.
The CHEST Task Force for Mass Critical Care has suggested different levels of capacity expansion needs via different levels of casualty (Christian et al., 2014;Hendrich et al., 2004).
Critical care sources require to enhance for: (a) Conventional response-able to expand immediately via at least 20% above baseline incentive care unit Max capacity; (b) Crisis response-able to expand via at least 200% above baseline incentive care unit Max capacity via regional, local, national, and international agencies; (c) Contingency response-able to expand rapidly via at least 100% above baseline incentive care unit (ICU) Max capacity via accessing local and regional resources; The Min requirements for critical care offered via the EMCC taskforce (Rubinson et al., 2008)  It is offered that a tiered response shall be set up, allowing increasingly more high-risk management practices to be viewed as the effect of the incident enhances. Ultimately, triage of patients for accessibility to scarce critical care sources would also be included.
Predict and manage such an incident contains: -Training and education of specialist/staff -A level of stockpiling or recognition of alternative resources of equipment (e.g., use of anaesthetic ventilators to supply ventilatory support, and NIV machines) -Recognition of Specialist/staff via transferable skills like recovery nurses, respiratory nurses and previous critical care nurses; When people still coming in but there is no way to put them out, we go to the red alert; In this condition basically, the number of coming in patents is excessed the number going out, and there is nowhere to put them, and we haven't got any spare capacity (bed), we get to the point where we are genuinely thinking can we treat these sick people? A red alert demonstrates there is a significant risk to the patients safety. In this condition, all non-emergency surgeries should be canceled, and we need to search for a free bed to start to discharge the beds until the flow moves forward. Whenever the hospital is on the alert, all wards are under pressure to discharge the patients. What we require to do is follow the patients from inter to leave.
When all wards are full, and there is no bed in any wards; at the moment, the emergency admission system is running at its absolute limit; it is starched. The point where if one person, one key person doesn't turn on for whatever the reason, then whole things just can fall apart. Just like a real domino, just one person comes out the whole thing just collapses. Because the flow stops, you get blockage and back right away up to the front door. Everything just becomes impossible. Whenever the hospital is on alert, all of the wards are under pressure to discharge the patients  (including start times and planned case lengths), scheduling of urgent and emergent cases, and handling of either delays or cases that surgeons may add to their block schedules are short-term decisions (typically made either a few days prior or on the day surgeries take place). On the one hand, the hierarchical division of decision problems makes it possible to decompose a complex problem into tractable models to aid decision-making within each hierarchy. On the other hand, this decomposition may lead to sub-optimal decisions. Note that OR capacitymanagement decisions directly impact the need for staffed beds, which we discuss in the last section.
In this section and others that follow, we proceed as follows. We begin by providing an institutional background and a discussion of performance metrics. Some evidence from data follows this. Data analysis aims to reveal operational inefficiencies, which provide a natural segue into the next section focusing on operational challenges. After that, we describe the state of the practice and current OM approaches. Finally, we close the section with a discussion of opportunities for OM researchers to contribute to this aspect of hospital operations.

DATA ANALYSIS
This section aims to illustrate key drivers of OR performance with the help of data from typical community hospitals. We obtained data from three Hospitals. Nearly all of the cases performed at this hospital finished within 10 hours.
The natural variability in volume and length of different procedures and the variability induced by the surgery booking process often result in a situation in which service lines with blocks may underutilize their block allocation and simultaneously use more than their allotted time for non-urgent cases. We illustrate this phenomenon in Figure 12, for the top five service lines, by total minutes of OR time used over 12 months at Hospital 1. The vertical axis in Figure 12 shows the OR minutes. There are three vertical bars for each service. The first bar, which is stacked, shows demand in OR minutes for non-urgent and urgent cases. Non-urgent cases are also referred to as scheduled demand. The second bar for each service line shows the allotted block minutes, and the last vertical bar shows the minutes of block time utilized. Some rooms may have been dedicated to the exclusive use of certain service lines, and block allocations may have been made based on the perceived importance of some service lines.
More interesting to note is that none of the top five service lines fully utilizes its allotted block minutes. This is true even for the first two service lines; whose scheduled demand exceeds the allotted block minutes. This phenomenon reveals the complexities of OR capacity management.
In US hospitals, surgical procedures are booked one at a time and must fit into the length of the block. The underlying problem is an online bin-packing problem, which remains an open challenge. The authors also found that hospitals may not provide clear guidelines to schedulers about using non-block time by block surgeons, which can contribute to the issues exemplified in Figure 12.
We highlight a different problem in the following analysis, which focuses on staffed OR utilization. This analysis was performed using data from a hospital that uses two-shift lengths -8 hours and 12 hours. The hospital staffed different numbers of ORs by day of the week.
Moreover, more 8-hour shift ORs were staffed each day than 12-hour shift ORs. This may partly be because of the difficulty and cost of continuously staffing ORs for 12 hours. All shifts started at 7 AM, and the first half-hour was reserved for preparation. That is, ORs opened for surgery at 7:30 AM. Figure 13 shows differences in utilization of staffed ORs between 7:30 AM and 3 PM compared to their utilization between 3 PM and 7 PM. We observe that staffed OR utilization regularly exceeds 100% after 3 PM. This happens because surgeries scheduled in rooms that are supposed to close at 3 PM do not completed by 3 PM. Staff who work 8-hour shifts and need to stay beyond 3 PM must be paid overtime, while they may have been idle earlier in the day. The annual costs for opening ORs longer than planned for the duration of this data were of the order of hundreds of thousands of dollars. This example shows that hospitals need to Evidence from our analysis of data uncovers a series of problems that could potentially be modeled and solved using OM methodology. To summarize, the key issues are as follows: 1. there is significant variability in the volume of surgeries by type, 2. case length durations within each surgery type are also highly variable, 3. block utilization is often low, even when surgeons' total case lengths far exceed their allotted block size, 4. and ORs typically run in overtime, especially the 8-hour shifts, because cases are not finished by 3 PM.
In the next section, we describe the operational challenges different stakeholders face.

OPERATIONAL CHALLENGES
We propose four main decisions concerning nursing unit operations that affect supply-demand matching. These are (1) unit size and scope decisions, (2) choice of performance metrics, (3) choice of policies concerning assignment of shifts and time-off requests, which we call HR policies (HR stands for human resources), and (4)  We justify our focus on nursing costs while ignoring other care providers because (1) an estimated 80% of the direct care costs in hospitals is payroll related (Heydari et al., 2021), and (2) nursing care plays a dominant role in care provision, and (3) the cost of available nursing hours is high relative to that of other care providers.
Staffed-bed capacity decisions lie within a continuum of choices that hospitals make. These choices fall into a natural hierarchy based on the frequency with which they are made (see Figure 14). At one end of this hierarchy are the least frequently changed parameters, such as the total number of certified beds, the amount of space per floor, and contractual agreements with employee unions. Unit size and scope choices are next in this hierarchy, and they are followed by the choice of the number of nurses and their skill levels. Next, hospital managers develop a work schedule for nurses for 2-4 weeks at a time posted 4-6 weeks in advance. Such staffing plans are also medium-term plans (Latham & Marshall, 2015). Finally, hospitals make start-of-shift nurse assignments (in response to realized nursing care needs and absent/sick calls) and real-time patient movement decisions as requests for beds are received. These decisions lie at the other end of the hierarchy. As shown in Figure 14, all of these decisions are affected by choice of performance metrics. hierarchy. They serve a purpose similar to the cross-training of nurses and the availability of acuity-adaptable beds. Relevant operational decisions concern the determination of rules governing patient movement.

CONCLUSION
Spending on inpatient care across OECD countries accounts for almost 30% of overall healthcare expenditures, with outpatient care consuming another 33% (OECD, 2013).
Although spending growth rates have slowed in recent years, spending is still increasing. It is anticipated that the cost pressures of GDP spending on healthcare will continue to grow as the Efforts to reduce healthcare costs have focused on cutting hospital beds, cutting salaries of healthcare workers, reducing reimbursement to healthcare providers, cutting the healthcare workforce, and increasing patient responsibility for covering their healthcare costs through copayments (Cunningham & Sammut, 2012;Morgan & Astolfi, 2013). But trying to shape supply and demand for healthcare services using these approaches has its limits, and their unintended consequences can be counterproductive. The recent Medicare hospital payment penalties for 30-day hospital re-admission rates (see Matthews, 2014), for example, were implemented to incentivize improved hospital discharge planning and better patient handoffs from inpatient to outpatient care. What has happened instead is that many hospitals are readmitting patients for observation? Stays (Weaver et al., 2015). This reduces the hospital's readmission rate since observation status is not technically a hospital admission. But observation stays can result in higher out-of-pocket costs for patients, especially if they require nursing home care after hospitalization. This is because Medicare does not cover nursing home care unless a patient has been admitted to the hospital for at least 3 days.
Healthcare organizations face many challenges as they adjust to reimbursement changes evolving from fee-for-service to value-related payments. Current and future reimbursement policies will encourage healthcare organizations to focus on managing the care across the entire supply chain to reduce the need for escalation of care to the inpatient setting. These challenges highlight the need for OM models that help improve the efficiency and effectiveness of care for populations of patients across the entire supply chain of care, focusing simultaneously on throughput and outcomes and provider and patient perspectives. OM efforts to improve chronic care management highlighted earlier provide examples of ongoing work needed in designing and evaluating new care models.
Regardless of how care models evolve, there will always be a need for inpatient care. This monograph highlighted recent OM work that models the dynamic inter-related impact of demand-supply matching across the ED, OR, and inpatient units. Healthcare managers and clinicians tend to make staffing and scheduling decisions across these areas independently.
Then, as demand unfolds in real-time, clinicians and managers reactively firefight as best they can to reallocate staffing to where it is most needed at a given point in time to alleviate bottlenecks in patient flow. As OM researchers, we need to develop models that enable healthcare managers to improve simultaneously OR scheduling policies, ED demand prediction, and medium-and short-term staffing plans that take into account the interrelationship of how demand unfolds over time across the ED, OR, and inpatient units.
OM work has fallen short of its full potential in implementing and testing models for real-time decision support tools. In the medical field, major research funding agencies have focused on translational research to more quickly bring new medical knowledge from the bench to the bedside. Some funds are targeted to support such translational research specifically. We recommend a similar focus on the OM fields. Newly developed OM models have the potential to aid healthcare managers in better matching demand and supply. The increasing use of electronic medical records, staffing systems, and scheduling systems contain relevant data to feed these models to support real-time decision-making. To date, we have not systematically focused on the challenges of bringing OM models from bench to bedside and then testing the impact of the models on improving the efficiency and effectiveness of care delivery. To be more relevant, we need to understand and appreciate how to embed OM models for decision support into the computer systems used by clinicians and managers and work collaboratively with them to achieve the full potential that OM has to offer.