Chapter 77: The ICU in the Global Hospital Environment

While the “intensive care unit” (ICU) is a distinct location to which patients with critical illnesses are transferred, in order to optimally manage the ICU and care for the critically ill, it is important to understand the ICU as it functions within the global hospital environment. ICUs in the United States originated during the polio epidemic of the 1950s as a place to cohort patients in need of mechanical ventilation.¹ Over time, however, patients classified as critically ill have become far more heterogeneous and the potential to care for them has increased dramatically.² As a result, the role of the ICU in the hospital environment has necessarily changed. Currently, ICUs cannot be viewed as standalone units that operate in silos. Instead, they must be well integrated into the hospital system to maximize the benefit of what is often a sparse resource—the ICU bed.^3,4,5

In this chapter, four aspects of ICU management—(1) the administration of critical care, (2) staffing for critical care, (3) the structure of critical care, and (4) resource utilization for critical care—will be discussed. Some of these topics are addressed in detail in other chapters; herein, I will highlight how they are integral to situating the ICU well within the broader hospital environment.

Critical care medicine has, traditionally, been a specialty without a home. ICU physicians have come from multiple backgrounds (eg, internal medicine, surgery, anesthesiology) and, in larger institutions, the units in which they work have, historically, been separately managed by their individual clinical departments.⁶ To make matters more complicated, the administration of nonphysician ICU staff (eg, nurses, respiratory therapists, nonphysician providers) is further fragmented. Instead of having a department unto themselves, critical care practitioners have been scattered throughout the organization. This type of fragmented or “siloed” structure has known drawbacks; silos can detract from teamwork and worsen communication with patients and families.^7,8 An alternate strategy that is gaining favor is that of “service lines” where all aspects of care pertaining to a given clinical entity are grouped under one organizational unit. This organizational methodology has proven beneficial in widely varying clinical settings.^9,10,11,12 Administering critical care as a service line is of potential benefit to patients and the hospital as a whole.¹³

Creating a critical care service line can be accomplished through the development of a separate critical care division or department or through the appointment of a director of critical care who helps to oversee a critical care administration that remains housed within distinct clinical departments. While the former is a more drastic change for an institution in which the traditional model is in place, it allows for a more centralized system and, potentially, economies of scale and scope¹⁴ and flexibility in resource use.^15,16 The latter option may not reap the same benefits, yet can make standardization of protocols—known to improve critically ill patient outcomes¹⁷—more feasible. In either case, all aspects of critical care delivery are managed together. With this model, the mission of the service line may be made clearer and resources may be more easily optimally allocated to achieve consistent goals.

This strategy of collaborative critical care administration has the potential to be beneficial to the hospital. First, if all providers operate under the same administration, collaboration among them is simplified. If a patient in a hospital with more than one ICU is critically ill and in need of an ICU bed at a time when none is available in the ICU most well suited to him/her, coordinated management may facilitate the admission of that patient to another ICU. Additionally, transfer of that patient to the better suited ICU may be more feasible when bed availability permits. In the extreme, ICUs operating under this paradigm can have beds that are flexibly allocated to different care teams (eg, medical or surgical) and/or different care levels, depending on the day-to-day needs of the hospital.^15,16 Second, there is a single person/administration who is responsible for addressing all issues pertaining to critical care delivery. For example, in New York State, a regulation mandating compliance with and documentation of adherence to sepsis bundles of care has been recently passed.¹⁸ With a centralized critical care administration, one single sepsis protocol can be developed and rapidly disseminated. Third, integration across clinical departments within critical care may foster integration outside of it as well. Lines of communication among providers within medicine and surgery, for example, may be bolstered through the existence of an integrated critical care department.

Staffing has become a topic of much focus in intensive care medicine recently. Leapfrog recommendations and Society of Critical Care Medicine (SCCM) guidelines require high-intensity staffing models where intensivists are available and/or onsite 24 hours a day.^19,20 These mandates are difficult to meet for many institutions. Strategies to improve critical care coverage are discussed in detail in the chapter “Alternate Staffing Models in the ICU.” As staffing pertains to the ICU as a part of the global hospital environment, two main questions for staffing arise: Is the responsibility of an intensivist only to care for patients in the ICU? Are there advantages and/or disadvantages to having physicians from other service areas be primarily responsible for the care of ICU patients?

An intensivist is a physician with board certification or extensive experience in critical care medicine.^20,21,22 Traditionally, such a physician may be involved in triage for the critically ill before ICU admission but cares, primarily, for patients within the confines of an ICU. Recently, however, the role of the intensivist has been expanded under the rubric of an “ICU without walls.”^23,24,25 Under this paradigm, the intensivist is expected to consult upon and provide care to critically ill patients prior to, following, and, sometimes, in the absence of, an ICU admission. Rapid response/medical emergency teams are the most commonly used method of delivering critical care expertise to non-ICU patients who become critically ill.²⁶ While there is no standard structure for these teams, many include an intensivist and/or a critical care nurse.^27,28,29 In essence, therefore, these teams can be viewed as expanding the ICU. Data on their impact is mixed,^30,31,32,33 yet ward staff satisfaction is improved^34,35 and these teams are being used with increasing frequency.³⁶

Rather than waiting for a clinical deterioration that merits activation of such emergency response teams, some institutions provide critical care consultation by intensivist-led teams for patients who are either not critical enough to be transferred to an ICU or have been recently transferred out of one.^25,37 These services can create a degree of critical care coverage that, while often lacking in the nursing and monitoring capabilities of an ICU, can benefit less critically ill patients. Staffing these services requires a commitment of an expanded workforce, however. Depending on the specifics of a given institution—available staff, ICU capacity, ICU occupancy, etc—staffing such teams may improve care for the critically ill throughout the hospital.

Bringing the intensivist out of the ICU may be one way to improve care throughout the hospital and another may be having appropriately trained clinicians from other departments work in the ICU. Most ICU patients in the United States do not receive care by an intensivist.^21,38 Instead, many are cared for by physicians with no training in critical care medicine; care often falls to hospitalists^39,40,41 and emergency medicine physicians.⁴² Recently, the SCCM and the Society of Hospitalist Medicine issued a joint position paper on the potential merits of accelerated training of experienced hospitalists in critical care.^40,41 Such training is not currently available, but may be possible in the near future. Over the past several years, emergency medicine physicians have been granted permission to enter critical care medicine fellowships and sit for the critical care medicine certification examinations administered by the American Boards of Internal Medicine, Surgery, and Anesthesiology.⁴³ Together, these actions promote an integration of critical care medicine with other disciplines. Hospital administrators can capitalize on the impact of this expanded pool of providers by hiring physicians trained in critical care, whose background makes them well suited to work in non-ICU environments as well. A critical care trained hospitalist will improve care delivery to newly critically ill patients on the general ward; additionally, if he/she also works in the ICU, he/she may enhance communication between the ICU staff and providers in noncritical care areas. Similarly, a critical care trained emergency medicine physician whose responsibilities extend into both environments can bridge care gaps and improve interactions between the ICU and the emergency department. Finally, in times of lower staffing (eg, overnight²¹), having someone available who is willing and able to cover an ICU and another clinical care area may be cost-effective. Critical care staffing can, if done thoughtfully, improve patient care throughout the hospital.

In planning for potentially more diversified ICU staff, hospital administrators must be mindful of competing responsibilities pulling intensivists away from the ICU, however. Most American ICUs have an “open” staffing structure in which there is no unified critical care team of providers looking after all patients in the ICU.^6,39,44 The most common ICU staffing model practiced in other places in the world is a “closed” model in which the primary responsibility for care of all ICU patients is transferred to an intensivist for the time that the patients are in the unit.^45,46,47 Leapfrog and SCCM advocate for high-intensity staffing, which is more often met by a closed ICU structure, but can be attained using an open model with mandatory critical care consultation. Data on the impact of a high-intensity staffing model on patient outcomes is mixed.^48,49 Most intensivists split time between ICU and other responsibilities.³⁸ In an open model, almost by definition, the primary physician caring for each patient has simultaneous non-critical care responsibilities—to care for either non-critically ill hospitalized patients or outpatients in the office setting. This set-up is often the case even for practitioners in closed model units. A potential disadvantage to having intensivists who have multiple additional skills is that they may be asked to multitask across disciplines; in so doing, needed focus may be inadvertently drawn away from ICU patients. Care must be taken to balance this potential negative consequence with the benefits of diversified staff.

ICU occupancy is often high^4,5 and patients who are in need of an ICU bed may experience delayed transfers, which are known to result in poor outcomes.^50,51 As touched upon in the section on “Staffing for Critical Care,” expansion of critical care services outside of the walls of the ICU itself may be desirable to improve the presence of critical care within the hospital environment. As aforementioned, one method for accomplishing this goal is the creation of rapid response/medical emergency teams or consultation services; others may include telemedicine and/or regionalization of critical care.

Telemedicine involves the use of technology to allow for patient monitoring, review of data, order entry, and, at times, clinical examination using robotic assistants by clinicians not onsite.^52,53,54 In the ICU, this technology has taken off as a method for expanding coverage of critical care patients nationwide.⁵⁵ Numerous studies have been conducted to evaluate the impact of this strategy; a recent meta-analysis demonstrated that while implementation improves ICU mortality and length of stay, hospital mortality and length of stay are unchanged by the introduction of telemedicine.⁵² This strategy is often considered as a means of bringing critical care expertise to underserved populations. From a hospital’s perspective, however, there are additional potential benefits to offering telemedicine ICU (tele-ICU) services to outlying hospitals. First, ICU patients are often transferred to regional centers from smaller community based institutions when expertise for care is not available at the smaller facility.^56,57 Sometimes the lacked expertise is a subspecialist’s opinion or intervention, which can only be provided at the larger center; such transfers likely result in favorable patient outcomes and can be financially attractive to a referral hospital.⁵⁸ When patients are transferred simply for the provision of high-quality critical care, the transfers may overburden the referral center and are less likely to be financially beneficial to it.^59,60,61,62 Creation of a tele-ICU service may minimize the need for these latter transfers. Second, despite inconsistent data on its benefits,^{63,64,65,66,67} many hospitals have adopted the Leapfrog standard of 24-hour onsite intensivist coverage. For hospitals employing such a model, there may be significant downtime for the in-hospital intensivist in a less busy ICU environment. Consequently, having such providers additionally providing tele-ICU care to outlying hospitals may be a revenue-generating proposition for the hospital. The cost-effectiveness of such a strategy has not been studied, however, and would likely vary greatly across institutions.

Regionalization of critical care is a topic being discussed at the national level.^68,69,70,71 Often compared to the regional trauma systems in place in the United States,⁷² the purpose would be to identify clear critical care referral centers to which appropriate patients would be triaged by emergency medical personnel prior to hospital arrival. In a survey of physicians who provide acute care, a majority felt such a system would improve patient care (52%) and efficiency (66%); however, 66% also felt that it would place strain on families.⁷³ Concerns have also been raised about such a system at the national level over (1) the financial implications for referral centers and smaller hospitals as well as (2) the fact that critical illness (unlike trauma) often develops in patients already admitted to the hospital.^62,72 A hospital system may benefit from regionalization of critical care on the more local or even system level, however. In fact, this type of system-based regionalization exists in some instances.⁷⁴ Hospital system administrators may be able to use interhospital transfers for critical care expertise as a means to provide better patient care (through optimization of volume-outcome relationships^{75,76,77,78,79}) and achieve financial optimization.

Resources for critical care include personnel and equipment. Strategies to optimize personnel use are discussed in the section on “Staffing for Critical Care.” Borrowing from economics, equipment can be broken down into “fixed” and “variable” resources where “fixed” resources are those that are permanent/semipermanent and independent of patient volume and “variable” resources are those that are used only when patients are in need of them.⁸⁰ Put simply, “fixed” resources can be viewed as ICU beds and large equipment (eg, mechanical ventilators). ICU bed utilization is discussed in detail in the chapter “Utilization of ICU Beds.”*** Suffice it to say here that it is important to the hospital’s care delivery, operational efficiency, and financial success to consider how many ICU beds and in what configuration (eg, in one mixed specialty unit, spread amongst specialized units, etc) is best to meet the needs of the hospital as a whole. Similarly, the purchase of large mandatory equipment (eg, mechanical ventilators) must be similarly thought through to optimize the return on investment while providing appropriate care to the patients in need.

Allocation of variable resources can often be adjusted more frequently than fixed resources and the potential impact for the hospital environment can be significant. First, decisions must be made about what types of resources to accrue; for instance, there are advanced care devices (eg, continuous veno-venous hemodialysis/filtration, extracorporeal membrane oxygenation) that can be expensive to acquire, but may reap clinical and financial benefits for an institution. Specifically, the possession of such technologies may funnel patients with higher level care needs to the hospital.^81,82,83 Second, decisions about the availability of “critical care resources” outside of the ICU can impact the hospital as a whole. For example, hospitals with intermediate care/step-down units that allow for higher monitoring than on general wards may improve patient flow out of ICUs.^84,85,86 In so doing, care for patients in need of ICU beds and costs of patients too ill for the general floor and not sick enough for the ICU may be optimized. Also, there is variability between hospitals in the need for ICU admission for certain illnesses.^87,88,89 Decisions to determine resource availability throughout the hospital (eg, the ability to administer continuous insulin infusions for patients with diabetic ketoacidosis in settings other than the ICU) can help maximize critical care bed availability for others more in need. The up- and downsides of these strategies may be different for each hospital and should be evaluated at each institution prior to resource allocation decisions.

ICUs are often viewed as well-circumscribed units. A simplified model of critical care focuses only on the care delivered by traditional critical care clinicians within the confines of an ICU’s walls. Envisioning critical care in this limited fashion, however, benefits neither the critically ill nor the hospital as a whole. Innovative administrative models, staffing strategies, structural plans, and resource allocation can benefit a hospital’s patients (both those who are critically ill and those who are not), its ICUs, and the hospital itself.