Critical care resources should be allocated in ways that promote high-quality care, defined by the Institute of Medicine as care that is safe, effective, patient-centered, timely, efficient, and equitable. However, utilization patterns in intensive care units (ICUs) often vary from ICU to ICU as well as within ICUs among individual providers. The source of this variation, while not fully understood, is likely undue—reflecting nonevidence-based practice patterns. Given that ICU beds comprise a large proportion of the total number of hospital beds in the United States and that their utilization consumes a disproportionate amount of the gross domestic product compared with other developed nations, numerous stakeholders have an increasingly vested interest in improving critical care delivery by aligning reimbursements with quality metrics. It is therefore of paramount importance both to accelerate the adoption of evidence-based critical care processes that are in line with patient values and to minimize both low-value and wasteful practices. With more attention to resource utilization, critical care outcomes will be improved, and there will be better matching of demand for, and supply of, ICU beds in ways that adequately help prepare them for any strains on ICU capacity.
Are Intensive Care Unit Resources Underused?
Appropriate critical care resource utilization should be informed by evidence-based medicine, patient values, and social priorities. However, there is currently a large amount of variation in critical care utilization despite the existence of high-quality evidence. One obvious way to remove much of this undue variation would be to quickly adopt evidence-based resources and promulgate these practices across the critical care spectrum. However, some suggest that it takes roughly 17 years for evidence to be fully implemented into clinical practice, suggesting that a significant proportion of critically ill patients are at risk for not receiving the standard of care due to a slow and capricious approach to implementation. For the recommendation of specific areas within critical care where increased resource utilization is warranted, it is important to first discuss examples where high-value resources have been slowly adopted and then examine whether structural or organizational changes can lead to improvements in implementation.
Recent decades have seen substantial advancements in the care of many critically ill patients, such as those with septic shock and acute respiratory distress syndrome (ARDS)—two highly prevalent conditions that commonly result in both short-term and long-term mortality among patients admitted to ICUs. Despite these improvements, there has been an obvious lag in the implementation of evidence. Almost 15 years after the seminal work demonstrating an impressive mortality benefit of low-stretch mechanical ventilation in ARDS patients, studies suggest that providers still underuse lung protective strategies. In a study of patients with lung injury, Needham et al. found that only 41% of patients meeting criteria to receive lung protective mechanical ventilation actually received this potentially lifesaving measure. In addition, there was a dose-response effect of adherence such that patients with high adherence had a much lower risk of mortality over a 2-year period compared with those patients with lower rates of adherence. If lung protective mechanical ventilation still suffers from delays in implementation, it is highly likely that other, more recent evidence-based strategies are also underutilized. These would include strategies relevant to almost all patients with respiratory failure, such as the daily interruption of sedation combined with a spontaneous breathing trial and early ambulation, as well as time-sensitive strategies reserved for patients with severe ARDS, such as early paralysis and prone positioning.
It is worth noting that the evidence suggests that hospitals caring for higher numbers of patients undergoing mechanical ventilation have superior outcomes compared with lower volume hospitals—likely because of better adherence to evidence-based practices. This volume–outcome relationship also appears to exist for severe sepsis, which has become a growing public health concern. Many think that the volume-outcome relationship in severe sepsis exists because high-volume hospitals are better at adhering to evidence-based sepsis care, such as the administration of early and appropriate broad-spectrum antibiotics and fluid resuscitation, adherence to lung protective mechanical ventilation, and restricted transfusion practices.
Broadly, high-value resources exist beyond the scope of both ARDS and severe sepsis and extend to almost all patients admitted to ICUs with acute and reversible processes. One resource that may improve outcomes for critically ill patients is a multidisciplinary team led by an intensivist. Much like the evidence for a volume–outcome relationship, the presence of an intensivist-led multidisciplinary team may increase adherence to evidence-based practices and improve patient outcomes, but some experts predict that there will be a growing shortage of board-certified critical care providers. If this proves to be true, then efforts to benchmark ICUs based on the presence of an intensivist may be an ineffective undertaking. With evidence lacking for the 24-hour presence of an intensivist, more work needs to be done to determine how to deploy our projected staffing resources in ways that optimize patient outcomes. This includes an improved understanding of which patients derive the greatest benefit from being cared for by intensivists as well as how to think more broadly about critical care organization and structure. Given that it is unlikely that every critically ill patient will go to hospitals that employ intensivists, one possible solution to improve the overall quality of critical care would be the incorporation of checklists and/or leverage defaults embedded in electronic medical records to help providers adhere to evidence-based processes. Examples include prophylaxis for deep venous thromboembolism and gastrointestinal bleeding, implementation of early enteral feeding, prevention of ventilator-associated pneumonias, early physical therapy, and the timely removal of intravascular and urinary catheters. In addition, it is also possible that expanding the use of telemedicine to provide remote access to intensivists may help to bridge the intensivist-patient gap and add quality to our critical care system.
Finally, in addition to increasing the utilization of evidence-based resources, it is equally if not more important that critical care practitioners effectively elicit patient preferences to ensure that the deployment of these resources is in line with patients’ goals and values. Because of severity of illness, many patients in the ICU setting will die despite being exposed to evidence-based processes. Therefore it is important to emphasize that high-quality critical care should include a timely elicitation of patient values to help promote a patient-centered approach that minimizes unwarranted and overly aggressive care at the end of life. Early meetings with patients or their surrogates to elicit goals and values, strategies that promote incorporating families/surrogates on ICU rounds, and the inclusion of palliative care experts in a multidisciplinary approach to patients faced with a high risk of morbidity and mortality all represent promising interventions in need of rapid study and, if successful, implementation.
Are Intensive Care Unit Resources Overused?
At any hour of the day approximately one third of U.S. ICU beds are vacant and available with roughly one in three beds in use for patients receiving mechanical ventilation. This surplus of U.S. critical care beds outnumbers the total number of ICU beds in many other modern countries. The concept of supply-induced demand, or “if you build it, they will come,” suggests that this excess in ICU beds results in the routine admission of patients that are either too well or too sick to benefit from critical care—signifying that many ICU admissions are likely examples of either low-value or no-value (wasteful) care. One suggestion to help optimize critical care resource utilization is to better determine who truly benefits from an admission to the ICU.
There are some data to suggest that the severity of illness of patients admitted to ICUs is inversely correlated with the number of ICU beds available at the time of admission. This hypothesis begs the question: Do we admit patients who are too well to benefit from critical care when ICU beds are plentiful? This possibility of overutilization is supported by two recent observational studies of low-risk patients. Gershengorn et al. published a retrospective study of 15,994 severity-adjusted patients with diabetic ketoacidosis and found that patients were more likely to be admitted to ICUs in hospitals that had higher overall rates of ICU utilization. Importantly, this greater use of ICU level care was not associated with improved outcomes such as reductions in hospital length of stay or in-hospital mortality. In a similar study, Admon et al. examined the relationship between ICU utilization in 61,249 patients with pulmonary embolism and a variety of outcomes and demonstrated that hospitals varied considerably in ICU admission rates for acute pulmonary embolus. They also found that patients admitted to ICUs in “high utilizing” hospitals were less likely to receive critical care procedures, suggesting that their indications for ICU admission were weaker—a notion supported by the additional finding that there was no relationship between ICU admission rate and risk-adjusted hospital mortality.
These are also data to suggest that there may be a subset of patients who are “too sick” to benefit from critical care. Stelfox et al. performed a prospective study of clinically deteriorating floor-level patients for whom a medical emergency team was activated. They specifically looked at whether the number of ICU beds available at the time of the clinical deterioration was associated with a patient’s time to ICU admission as well as subsequent changes in their goals of care and in-hospital mortality. They found that fewer available ICU beds were directly associated with patients being more likely to have their goals of care changed to comfort-based approach, but there was no association between the number of available ICU beds and the subsequent in-hospital mortality. These findings suggest that relative scarcity may expedite transitions of care toward palliation in a group of patients that is likely to die with or without ICU-level care. Therefore, when examining the possibility that excess ICU beds may result in overutilization, it is worth considering whether restricting the future expansion of our critical care bed supply may paradoxically result in higher quality critical care.
In addition to considering how existing ICU beds are appropriated, it is also worth reflecting on whether there are common critical care practices that are either low value or wasteful. A common theme is emerging within the field of critical care: that less is often more . Recently, there has been a palpable increase in the interest of multiple stakeholders to target inefficiencies in our health-care system. One example of this concerted effort is the American Board of Internal Medicine’s Choosing Wisely Campaign. This campaign collectively embodies a professional societal spirit that leads on identifying areas of low-value or wasteful care by asking health-care providers to determine a list of five specialty-specific services that should not be routinely provided. Given the resource intensive nature of critical care, embracing these efforts may lead to improved quality. Within critical care, the expert consensus suggests that health-care providers should reduce the reflexive use of diagnostic testing such as routine blood draws and chest imaging, the liberal use of blood products and total parenteral nutrition, deeply sedating mechanically ventilated patients without a specific indication, and continuing life support in high-risk patients without offering patients the alternative of a comfort-based approach.
Finally, critical care providers must continually reevaluate their practices and de-adopt practices that are no longer rooted in evidence. Within critical care, it is not uncommon for diagnostics and therapies to initially demonstrate a positive outcome only later to be shown to lack efficacy or be harmful. Examples would include the routine use of Swan-Ganz catheters, the routine placement of central venous catheters in an algorithmic approach to sepsis, and overly tight glucose control. Although the notion of de-adoption is in the same vein as minimizing low-value and wasteful care, it merits acknowledgment given the evolving nature of critical care clinical trials. Our patients deserve a critical care system that praises the stewardship of resources as opposed to a system that promotes haphazard utilization patterns and wasteful practices. Thankfully, more high-quality research is being done to help us better understand how best to deploy our resources and has only served to strengthen the notion that a conservationist approach to critical care resource utilization is often the correct approach.
What Is Needed and What Should Be Eliminated?
We discussed that focusing on correcting the underutilization or the slow adoption of evidence-based practices while simultaneously eliciting patient values would vastly improve the quality of critical care provided. To do this, we recommend focusing on four major areas. First, given the large degree of “negative” critical care trials, there needs to be a concerted effort among critical care researchers to improve the quality of critical care trial design to ensure that future trials are relevant, have patient-centered outcomes, possess high-quality statistical methodology that maximizes the likelihood of achieving enrollment and statistical power, and segue to comparative effectiveness research.
Second, the lion’s share of health-care research dollars currently targets “bench-to-bedside” or “Translational 1” research centered on technologic and pharmaceutical innovation while only a paucity of health-care dollars are set aside to improve health-care delivery, integration, and quality. Therefore more emphasis must be placed on “beside-to-policy” or “Translational 2” research that focuses on both disseminating knowledge across the critical care spectrum and advancing implementation science to ensure that knowledge dissemination is integrated into routine practice frameworks.
Third, the transition from fee-for-service to value-based purchasing reimbursement mechanisms has helped to stress the importance of patient safety and quality improvement by ensuring that best practices are sought. Given that cost-awareness and appropriate resource stewardship is an important component of quality, formal training structures must be put in place at the graduate medical education level that serve to train future providers how to conduct quality improvements and incorporate cost-awareness into their practice. Much of how physicians ultimately practice stems from their early training environment so it is possible that focusing on quality improvement by distinguishing high-value from low-value care may translate into long-lasting improvements in critical care resource utilization.
Fourth, it is worth conceptually exploring the notion of a “net ICU benefit” by improving our understanding of who benefits most (or not at all) from receiving critical care. ICU admissions are costly and should be reserved for severely ill patients with reversible disease processes. Efforts to help further elucidate who best benefits from critical care may require advancements in predictive models that generate probabilities of in-hospital mortality based on the presenting diagnosis and various comorbidities. In turn, the promotion of standardized ICU admission guidelines may then help to implement these findings. For those who may not be sick enough to benefit, hospital organization and staffing efforts to redistribute these patients to lower acuity settings would undoubtedly improve critical care utilization without compromising patient care. For those patients that are too sick to benefit from critical care, rather than act primarily as obstructionists or ICU gatekeepers, we believe a more long-lasting benefit may be achieved by fostering improvements in the end-of-life decision sciences, educating high-risk patients and their surrogates on the risks and benefits of ongoing critical care, and promoting the timely and early elicitation of patient values.
In an effort to avoid overutilization of resources, critical care providers should be circumspect in adopting the newest technology or drug unless it is rooted in high-quality evidence. Although every new technology, device, and drug is marketed as truly innovative, what enters into practice may often be pseudoinnovative. Pseudoinnovative critical care services increase costs without clearly improving patient outcomes. These technologically advanced but low-value services then compete with evidence-based practices and threaten the delivery of high-quality critical care. One suggestion for optimizing resource utilization would be for hospitals to establish ICU resource utilization review committees. These committees would continually appraise critical care evidence, leverage electronic medical records and incorporate defaults to make it more difficult for providers to adopt pseudo-innovation, streamline the adoption of best practices, and prompt providers to de-adopt services later proven to lack evidence. In addition to aligning reimbursement mechanisms with quality and performance metrics and increasing the transparency of provider practice patterns, focusing on optimizing resource utilization as we have outlined previously will help maximize the provision of high-quality critical care.