Scenario Planning Approach to Adapting in the COVID Era

Abstract

Rationale and Objectives

The COVID-19 pandemic has caused much uncertainty and disruption in healthcare resulting in many challenges for strategic planning. Scenario planning is a tool that allows healthcare leaders to plan healthcare delivery strategies by incorporating the uncertainties into the analysis and planning process.

Materials and Methods

Variables were identified which will have major impact on the future, but whose future direction is uncertain. The extremes of these drivers were used to generate multiple scenarios. A subset of scenarios was used to evaluate potential tactics to determine which may be high yield in the face of uncertainty.

Results

Unlike traditional strategic planning, scenario planning does not develop a single future with a path to that future. Scenario planning evaluates tactics to determine which would be helpful in specific scenarios, multiple different futures or under specific conditions.

Conclusion

We present a scenario planning model which can be used to determine specific tactics to accommodate the uncertainty due to variable healthcare delivery needs in the COVID-19 era.

INTRODUCTION

Much has been learned in the radiology community during the COVID-19 pandemic and predicting the next steps, as well as anticipating short and long term needs and consequences, involves many variables and much uncertainty (1). From the initial waves in 2020 to the subsequent waves of pandemic as COVID variants emerged, the resulting disruptions impacted all aspects of society including the workforce, supply chain as well as individual and population health. The impact varied with each wave and was felt differently by different communities. As of October 2022, the Kaiser Family Foundation estimates global death toll over 6.5 million deaths (2). Given the variation in vaccine adoption among different populations, the continued emergence of variants and the unknown long term health consequences on COVID 19 infection, there is much uncertainty that needs to be considered in any healthcare organizational planning.

In a world of limited resources, how should you decide how to best prepare for the uncertainty that is healthcare in the COVID era? Scenario planning is a planning tool that allows organizations to consider multiple possible futures called “scenarios,” permits organizations to plan and prepare for several different specific scenarios as well as identify the elements in common across several scenarios. The technique originated with Herman Kahn and his RAND Corporation work for the US military, helping the United States plan for uncertain shifting military needs during the Cold War. Royal Dutch/Shell used scenario planning in the 1970s prior to the gas crisis (3). Rather than viewing the oil industry as on a continued trajectory of high availability of oil and low oil prices, Royal Dutch/Shell recognized that major disruptions in the oil industry were possible and could be planned for. This frame shift informed strategic decisions and investments and allowed Royal Dutch/Shell to be better prepared for the oil crisis than its competitors. Scenario planning has been used in Radiology by the Society of Chairs in Academic Radiology in the early 2000s (4).

One advantage of scenario planning is the ability to determine the necessary or high yield strategies and tactics despite the uncertainty of the future. Scenario planning does not predict the future. It identifies tactics that will be helpful for the scenarios studied as well as tactics that might be beneficial in several scenarios or under specific conditions. This is distinctly different from traditional strategic planning that decides on a desired future and builds a strategic plan to get to that single desired future.

In this paper, we apply principles of scenario planning to healthcare delivery in the COVID era in the United States. Rather than planning a single desirable future, we use the scenario planning approach to determine the likely drivers, develop several scenarios and how best to prepare for individual scenarios and the common elements shared by multiple possible futures. The advantage of this approach is that it acknowledges the multiple uncertainties while focusing on the things that would be most useful for healthcare organizations to do without committing to a single defined future.

MATERIALS AND METHODS

In fall 2021, in the midst of a COVID-19 pandemic wave, in an attempt to balance the immediate needs with the longer-term needs, a scenario planning exercise was initiated. There are many possible drivers for the future success including organizational mission, local and national markets/trends, technology changes, etc. An axis of uncertainty is a variable where experts cannot agree and have opposing predictions. Will there more or less COVID cases? It is unknown what direction the virus will take as it continues to evolve with new strains and how will this impact the health of our communities. Will patients return to hospitals for their care all at once or will people continue to avoid coming to hospitals? Of all potential drivers and variables reviewed, three were chosen to be the axes of uncertainty: COVID clinical demand, non-COVID clinical demand and healthcare workforce (Fig 1 ). Other drivers were considered. For example, it is hard to imaging financial success without clinical volume. Yet clinical volume is a balance between the demand for non-COVID care and COVID care. Financial margins are a balance between expense and revenue with human resources being a major expense. Uncertainties in clinical demand make staffing needs unpredictable, introducing more uncertainty and risk.

Figure 1.

Axis of uncertainty in scenario planning.

Nine scenarios (Table 1 ) were created by taking combinations of the extremes of high and low values of the three axes. By narrowing the focus to just three potential scenarios, which include all three variables at both extremes, we have a comprehensive set of possible futures which has incorporated the uncertainty into the model. There is no correct subset of scenarios to study. When choosing scenarios, one could try to choose scenarios that seem to be the most likely futures and develop specific plans for them. Alternatively, one could choose a combination of probable futures balanced with scenarios that ensure that all extremes of the axes are represented. One value of the scenario planning approach is to disrupt the mental model of how the future is viewed for an organization. Adding a fourth scenario to such an analysis would not add additional variables or extreme considerations and would likely have little incremental yield in the identification of high yield tactics in the face of uncertainty.

Table 1.

Scenarios

Scenarios	Non-COVID clinical demand	Healthcare workforce	COVID clinical demand
Clinical system over run	High volume	High vacancy	High volume
No volume but staff	Low volume	Low vacancy	Low volume
Clinical system over run by routine care	High volume	High vacancy	Low volume
Tight balancing act	High Volume	Low vacancy	High volume
Taking care of COVID	Low volume	Low vacancy	High volume
No volume, no staff	Low volume	High vacancy	Low volume
Taking care of non-COVID	High Volume	Low vacancy	Low volume
Clinical system over run by COVID	Low volume	High vacancy	High volume

RESULTS

The Axes

Axis 1: COVD Clinical Demand

There are 6093 hospitals in the United States with a combined staffed inpatient bed capacity of 90,531 with over 75,000 critical care beds and 33,356,853 annual admissions (5). The Center for Disease Control estimates 5,319,921 cumulative confirmed COVID hospital admissions between August 1, 2020 and October 22, 2022 (6). COVID surges impacted the need for inpatient and critical care beds differently in different communities which in turn varied with each surge. The advent of vaccines led to lower infections and less severe/fewer admissions. In the United States, 68.2% of the population has been vaccinated with the complete primary COVID vaccine primary series, but fewer have gotten the first or second booster (7).

Another uncertainty is how much of future COVID care will require inpatient management, including the use of hospital or ICU beds as opposed to outpatient management alone. Some of the COVID care is going to be nonacute and may be managed in outpatient or urgent care settings if vaccines and boosters are more widely deployed. In a similar fashion, we must consider the COVID volumes that may become chronic diseases as we learn more about “long COVID,” which will add to the health care needs of the population. This could result in a shift of some healthcare from hospital-centric or inpatient-centric COVID care to a sub-acute care and ambulatory model of care.

In addition to a possible increase in chronic care needs, there is evidence that there is likely to be an increase in acute care needs secondary to COVID infection. Preliminary data shows individuals with prior COVID infection have an increased risk for cardiovascular disease including stroke, myocardial infarction and thromboembolic disease (8).

Axis 2: Non-COVID Clinical Demand

Non-COVID clinical volumes were disrupted throughout the pandemic. During the pandemic some routine care was delayed, some avoided and other clinical demand decreased during lockdowns. Cancer screening exams were delayed during the pandemic (9). This allowed redeployment of staff to address other clinical needs. Other care needed to be provided but wasn't; it is unclear how much was through patient choice or changes in access. There was a 38% reduction in United States cardiac catheterization ST elevation myocardial infarction activations (STEMI) (10). Some care no longer needed to be provided. For example, California saw a 50% reduction in car collisions of all types during “the shelter in pace order” (11). Patients also delayed care to avoid contact with COVID. The impact of COVID volumes and workflow changes on the non-COVID radiology volumes varied with different waves with initial reductions as high as 70% reduction followed by rebound and relative stability through subsequent waves (12,13).

The nature of care also changed with a significant portion of ambulatory clinic visits shifting to virtual visits during the initial surge. While these virtual exams had limited physical exams, they had reduced travel time for patients. Virtual visits may present opportunities for specialist care to be brought into more rural environments, possibly increasing both access and clinical volumes.

Axis 3: Workforce

The current state of the healthcare workforce is concerning. The great resignation saw approximately 18% of healthcare workers quit their jobs during pandemic (14). On a 2021 McKinsey survey, 22% of nurses indicated they may leave their current positions (15). The replacement pipelines are not prepared to produce new graduates at this scale. An American Association of Colleges of Nurses report stated over 80,000 nursing school applicants were turned away from baccalaureate and graduate programs due to lack of qualified faculty, clinical study sites, classroom space and budget constraints in 2020 (16). The American Society of Radiologic Technologists listed 53.6% of radiologic technology schools at full enrollment and 32% of nuclear medicine technology schools at full enrollment in 2020 (17). A 2021 Doximity Survey estimated that during the pandemic approximately 1% of the physician workforce retired before expected (18). The same report noted that three quarters of physicians reported being overworked with 22% considering early retirement because of overwork. Given continued physician burnout/moral hazard, additional loss of physician workforce is likely.

The nature of our workforce is changing. There has been a multiyear decline in the percent of 16–64 year-old males in the US workforce with the most recent value at 67.7% (19). Our workforce is more mobile and may not be on-site. Remote work (from home) reduces the need for office space and parking and reduces the environmental impact. An estimated 60% of Americans working hours were from home in May 2020, falling to 40% in October 2021; this compares to a pre-pandemic February 2020 baseline of 5% (20). The mobility of the workforce extended beyond working at the same job from home. Licenses and registrations were fast tracked during the pandemic and scopes of service were expanded during the initial COVID surge. The nursing shortage during the pandemic lead to increased use of traveler nurses with an associated significant increase in the cost. Increased hourly rates charged to hospitals for traveler nurses by staffing companies have increased over 200% compared to prepandemic levels. In January 2022, hospitals spent a median of 38.6% of their total nursing labor expenses for nursing traveler contracts compared to prepandemic 2019 levels of a median of 4.7% (21). Hospitals have also increased their use of temporary allied healthcare staff from pre-pandemic levels of 25% to 30% to cover the staffing vacancies(22). Hospitals responded to the workforce shortage with increased compensation for existing staff and paying higher costs for travelers. This resulted in the increase in labor expenses per adjusted discharge between 2019 and March 2022 by 37% on average (23).

The increasingly mobile nature of the healthcare workforce and the turnover in staff has highlighted challenges in rapid on-boarding and standardization of workflow that have been obtained as part of quality improvement projects for example, central line-associated bloodstream infection , catheter-associated urinary tract infection. The result is a loss of institutional knowledge and nationwide shortage of healthcare workforce which existing pipelines are inadequate to rebuild. In addition, the higher personnel expenses for healthcare compared to pre-pandemic levels have no sustainable new revenue source to cover this incremental expense in a low margin industry. When combined with lower volume of discharges and increased length of stay, the net effect is negative hospital margins (24).

Scenarios

None of the three variables is static. How these variables change over time introduces significant uncertainty which can hinder hospital and practice planning. By converting these variables to axis with extreme ranges, the uncertainty can be modeled. Nine scenarios were created using these three axes (Table 1). A subset of scenarios was chosen which included the extremes (high and low) for all of the variables. The three scenarios chosen were 1) Clinical system overrun, 2) No volume, but staff and 3) Tight balancing act.

In Clinical system over run scenario, both the COVID and non-COVID clinical demand and volumes were high. The workforce staffing vacancy levels were high as well. In this first scenario adequate healthcare workforce is not available and COVID care demands continue to be high in addition to non-COVID care demands. This is the environment of care rationing, expanding scope of practice and likely high mortality. Examples of a similar scenario were seen in Italy early in the pandemic and in India in 2022 when the health systems were literally overwhelmed and unable to meet the demand for care to COVID patients.

In the second scenario, No volume but staff, the other extreme in clinical demand would occur. There would neither be on-going severe COVID surges nor unmanageable clinical demand from the deferred non-COVID care or routine care. There would be low vacancy rates in the healthcare workforce with an adequate staff to manage the care and meet demands of both the COVID and non-COVID demand. This may lead to overstaffed hospitals with increased cost and decreased clinical volume and revenue, resulting in staff and service reductions.

In the final scenario, Tight balancing act, both COVID and non-COVID clinical demand is high, but the healthcare workforce vacancy is low. This most likely requires healthcare systems to manage the capacity by doing most non-COVID routine care, while maintaining the capacity for COVID admissions. This results in deferring of some routine/nonemergent non-COVID care during COVID surges. Difficulties arise in defining what routine care can be deferred and for how long. This scenario was seen in some communities in the later surges in the United States with hospitals trying to load balance between COVID and non-COVID care needs.

Next Steps

The next step would be to identify possible tactics that would be helpful in individual scenarios. A comparison of those three lists would yield tactics which would be helpful in all three scenarios. In other words, what we will we need to do regardless of which scenario happens? This approach allows planning for specific scenarios as well as across several scenarios. There is no one correct answer because each local healthcare environment must be considered as well as the healthcare organization(s) in that community. Tip O'Neil said “all politics are local”. During the pandemic, each healthcare organization has evolved and matured and developed (and lost) some organization capacity and capabilities, but those vary from organization to organization. Thus, a tactic that one organization might need to develop, another organization may already have in place or might not even be helpful in that local environment.

A systematic review of people, process and physical plant at a hospital/health system/practice or community level for each scenario would reveal tactics that need to be locally evaluated. For people/human resources, considerations might include cross training, flexible work hours, staff development, wellness programs or remote work options. One approach to addressing the work force shortage might include the recruitment of international physicians and nurses. The American Board of Radiology has a pathway for International Medical Graduates which includes completing four years of radiology residency training, fellowship, and/or full-time faculty appointment in one institution with a diagnostic radiology training program and allows the candidate to take the board certification exams (25). In prior nursing shortages, nurses have been recruited from the Philippines. The Philippine government has supported overproduction of nurses in the Philippines to create a surplus of bachelors’-trained nurses who would work internationally and send money home regularly (26).

For process/workflow, considerations might rely on adaptations from the pandemic surges or focus on tasks deferred during pandemic, automation of manual tasks or improved efficiency of rate limiting step processes. A review of physical plant and infrastructure might increase the flexibility of space or explore off-campus locations possibly closer to where patients or staff live. In radiology, one modality or site may be in one scenario while another modality or site within the same organization is experiencing a different scenario.

Many of these tactics will seem familiar and were explored during the pandemic. So, what is different? In the short term, the benefits of improving workforce pipelines seem obvious. Retention of workforce, reducing burnout and increasing joy in work would also likely be beneficial in multiple environments. These tactics are the reaction to an acute staffing crisis and lack the refinement of a more complex time horizon. Workforce is likely to be both an acute problem as well as a long-term challenge. The question is how will this evolve over time? Looking at the three scenarios, two of them would benefit from increased workforce, but one of them has excess workforce for the low volumes. While having too many staff seems unimaginable at this time, scenario planning highlights the need to balance the workforce to the uncertain clinical demand as demand changes. Considering the increased labor costs per adjusted discharge, staffing at the current cost is unsustainable and will likely lead to reduced services and staff layoffs in some markets if no other changes are made. Developing this insight further raises the question of whether there is a flexible staffing model that fits all three scenarios. (Figure 2 ) What new infrastructure would be needed to have a more flexible staffing? What is the correct balance between full-time and part-time workforce. How can staff performance and productivity be improved? What tasks can be automated? How many hospitals administrative processes are inefficient? Which workflows could be improved so that staff time is not wasted, thus increasing the efficiency of the staff? Alternatively, if no other changes are made the unsustainable higher cost per discharge could result in reduced services and lay-offs. Reduced services by one practice may be an opportunity for another practice if they have more flexible capacity or can provide the care more efficiently at a lower cost.

Figure 2.

Venn diagram of three scenarios and a tactic that meets needs of all three scenarios.

In considering the human resources needs of the organization for each of the three scenarios, one might apply a framework such as the Baldrige Criteria (27). Such a comprehensive review would allow the reassessment of the processes for functions such as predicting organizational workforce needs, planning for workforce capacity and engaging workforce for retention and high performance would allow the beginning of a discussion. (Table 2 ) Work force is not homogeneous and is segmented in multiple roles including receptionists, technologists, nurses, physicians, and advanced practitioners-each with different needs, skills and pipelines. In academic radiology, the faculty segment needs processes for promotion as well as non-promotion faculty development pathways. Some medical schools may expand the criteria for academic promotion by recognizing activities in clinical work, administration, quality initiatives and education that were undervalued earlier in an effort to improve job satisfaction and presumably retention. Dissatisfied and unengaged staff and faculty may ultimately leave. In general replacement cost is 1.5-2.0 times the salary of the individual who has left further adding to the cost of providing care (28). If they do not leave, dissatisfied faculty may also exhibit disruptive behavior which in turn adds incremental costs which could be as high as $1 million per radiologist annually in a 400-bed hospital (29). Further segmentation of work force needs may cause some organizations to re-evaluate and redesign fundamental processes such as recruiting, on-boarding, staff development, hours and location of work. In academics, virtual interviews are now routine, but can they be improved? Virtual education has exploded but there is still much opportunity in both asynchronous learning and synchronous and interactive learning.

Table 2.

Baldrige Criteria Workforce

Capability and Capacity Needs
New workforce members
Workforce change
Work accomplishment
Workplace environment
Workforce benefits and policies
Drivers of workforce engagement
Assessment of Workforce engagement
Organizational culture
Performance management
Performance development
Learning and development effectiveness
Career development
Equity and inclusion

In all three scenarios being able to flex organizational resources to meet demand and to efficiently use existing resources is advantageous. This is not limited to human resources. Similar to flexible staffing, facility renovations and space utilization may need to be done with more intentional flexibility. This could include site selection for new clinical sites or renovation of existing sites. Processes, such as admission, discharges and transfer can also be redesigned and made more efficient. Reduced length of stay would provide increased inpatient capacity with the same staff and physical plant. Academic processes such as research can also be approached differently. Chesbrough describes an approach called “open innovation” (30) where innovation is not done in secret and in silos. Organizations promote collaboration with outside people and organizations by placing some of its resources outside of its walls through partnership, licensing and allowing others to develop your non-strategic initiatives. This allows both cost sharing (reduced expenses) and potential new revenue generation for existing resources.

Other tactics may only be valuable in a unique set of circumstances. Identifying both the tactics and the unique circumstances sets a sign post for when resources should be directed towards such a tactic. Systemic analysis of the organization in light of the three scenarios and identification of best practices would be needed. As landscapes within each scenario are further developed, secondary scenarios and special cases can be developed. Change is likely to be part of the healthcare landscape for the next generation. What new infrastructure or training should we be putting in place? Should change management training be part of staff orientation, graduate medical education training or faculty development?

DISCUSSION

Unlike traditional strategic planning, scenario planning does not predict or plan for a single future. Our scenario planning model provides scenarios which when considered would yield potential tactics. One advantage of scenario planning is the ability to do pre-work before the actual disruption occurs which allows the organization to react to unexpected situations and conditions more rapidly and with an existing process-based approach. Applications of scenario planning in other industries allowed investments to be made strategically in high yield long term activities. Some tactics represent work we are likely going to need to do in healthcare. Starting some of the pre-work or removing barriers now may enable us to get these done faster than being reactive later or starting later during crisis.

If our workforce pipeline is a small, then addressing it will require time. There are long lead times during which needs will change. Nursing schools and radiology technologist programs have entry requirements that need to be met prior to admission to 4-year programs. Residencies plus fellowship for radiology can be 6 or more years after a 4-year medical school. How do we address the generational differences in the multiple segments of our workforce? We may not be able to train the next generation the same way the last generation was trained. Healthcare may have to adapt to how the next generation of learners wants to be trained and how they optimally learn (31). If a workforce cannot be developed fast enough, an organization may shift its focus from people to changes in process/workflow (eg, reduced length of stay) to optimize the clinical work that can be done with the existing workforce. If a balance is not found between the care provided and its cost with the clinical demands, then services will likely need to be decreased to maintain financial viability of the organization.

The study has several limitations. It could be argued that the axis chosen are not the biggest drivers. While demand for COVID and non-COVID healthcare are likely to be included in any analysis, workforce as an axis has some limitations. We looked at the workforce supply chain problems and the issues of pipelines for workforce. However, workforce could be a surrogate for supply chain. There are multiple examples of non-human resource supply chain/pipeline challenges impacting the ability to provide healthcare. Currently the Food and Drug Administration has an on-line database of drug shortages listing nearly 200 agents listed as currently in shortage or resolved shortage. (https://www.accessdata.fda.gov/scripts/drugshortages/default.cfm) The computer chip shortage limited access to computers which limited replacement and additional computer and medical equipment. Radiology has seen shortages of helium impacting MRI. The shutdown of the Shanghai contrast manufacturing plant during the 2022 summer due to a COVID lockdown resulted in a global reduction of iodinated contrast for over 2 months with extensive changes in clinical practice (32,33). Much of the supply chain issues are external to hospitals but they will impacted by them. Having a systematic approach to managing critical shortages may become an essential skill. Building new or redundant pipelines for both human resources and non-human resources will likely require new approaches.

Other limitations include barriers to implementing scenario planning. Despite these potential benefits, many health centers will not undertake this process. In an acute crisis where, daily staffing is a challenge and monthly volumes and revenue are highly variable, organizational resources are more likely to be focused on the acute needs and looking at longer term may seem an unaffordable luxury. Organizations that remain financially viable initially may have fewer reasons to look for new approaches at this time. In resource constrained organizations, it may not be possible to develop detailed plans for each scenario. Such organizations could look at what tactics are common to several scenarios to be more prepared.

Another potential limitation is that organizations may be focused on local needs and solutions. Tactics might address local needs but they are not restricted only to local resources. Advocacy efforts could result in the development or acceleration of state or federal funding or policy changes. Health and Human Services recently announced a $60 million investment in rural healthcare workforce (34). While this includes $9.7 million to establish new rural residency programs in rural communities, a bigger investment in Graduate Medical Education expansion would be needed to increase the pipeline of physicians.

CONCLUSION

The uncertainty that COVID has brought to planning in healthcare has shifted much of healthcare to adapt to changes as they occur. Scenario planning offers an opportunity to assess the uncertainty and try to anticipate what planning will likely be needed despite the uncertainty. This allows some of these steps to begin earlier. Some of the tactics identified will be generalizable to many healthcare organizations. This may allow collaboration and cooperation between healthcare organizations, professional societies or state/federal government. Potentially some steps, which in the longer term are counterproductive, might be avoided. Scenario planning represents a tool that health systems could use to better prepare for the uncertainty in the COVID era. We don't want to be the last ones to recognize that the world has changed and new approaches are needed to manage an uncertain future.