The past 50 years have witnessed the emergence and evolution of the modern pediatric ICU and the specialty of pediatric critical care medicine. ICUs have become key in the delivery of health care services. The patient population within pediatric ICUs is diverse. An assortment of providers, including intensivists, trainees, physician assistants, nurse practitioners, and hospitalists, perform a variety of roles. The evolution of critical care medicine also has seen the rise of critical care nursing and other critical care staff collaborating in multidisciplinary teams. Delivery of optimal critical care requires standardized, reliable, and evidence-based processes, such as bundles, checklists, and formalized communication processes.
Key points
- •
Over the past half century, health care delivery has evolved such that ICUs, including specialized pediatric ICUs (PICUs), have become a vital element in the health care delivery system.
- •
During this time, both PICUs and the field of pediatric critical care medicine have evolved into their current forms.
- •
Critical care is now provided to a diverse patient population in a wide array of settings by multidisciplinary teams of physicians, nonphysician providers, nurses, pharmacists, nutritionists, and other ancillary staff members who specialize in pediatric critical care and coordinate with members of other medical and surgical disciplines.
- •
As the field of pediatric critical care medicine evolves, it seeks to improve so that patients receive reliable, efficient, and evidence-based care with optimal outcomes.
- •
As the patient population changes, providers adapt, staffing models adjust, technology evolves, therapies advance, and outcomes are transformed.
- •
The evolution of pediatric critical care over its first 50 years has been tremendous, but the next 50 years will assuredly bring even greater change.
Introduction
ICUs are a key element in the overall delivery of health care services. Many patients receive care in an ICU (or in some cases, by an ICU team outside an ICU) at some point during their hospital stay. ICU care also accounts for a significant proportion of total health care costs. For example, recent studies suggest that ICU care represents between 17.4% and 39.0% of all hospital costs and between 0.56% and 1% of the United States’ gross domestic product. These statistics highlight the importance of providing safe, effective, patient-centered and family-centered, efficient, timely, and equitable care in an ICU. One of the challenges faced by every ICU today is to minimize costs of care while maintaining quality (ie, increasing the value). This article provides a brief overview of the current state of pediatric critical care delivery and discusses some of the recent changes in the overall evolution of pediatric critical care delivery in North America, starting with a brief historical description of the field of pediatric critical care medicine as an important backdrop to where the field is likely to move in the future.
A brief history of pediatric critical care medicine
Evolution of the modern PICU and evolution of the specialty of pediatric critical care medicine have occurred approximately in tandem over the last 50 years. The roots of pediatric critical care medicine, as a specialty, include adult respiratory care, neonatology, pediatric general surgery, pediatric cardiac surgery, and pediatric anesthesiology. ICUs first developed in response to the global epidemic of polio, which was a universally fatal disease until the development of the iron lung in the late 1920s. The knowledge gained in artificial ventilation was then generalized to other conditions, and pediatric critical care medicine began to develop as a specialty to provide care for other types of patients, including neonatal ICU graduates.
The first known PICU was established in Sweden in 1955, whereas the first US PICU opened in 1967 at the Children’s Hospital of Philadelphia, a 6-bed unit with a separate nursing staff and 24-hour in-house physician coverage provided by pediatric anesthesia fellows. PICUs began to spread throughout the United States and by the mid-1970s were located in most hospitals that had a pediatric residency program. The Pediatrics Section of the Society of Critical Care Medicine and Section on Critical Care of the American Academy of Pediatrics were established in 1981 and 1984, respectively. In 1987, the first American Board of Pediatrics certification examination for pediatric critical care medicine was offered (182 of the 242 first-time takers passed). The first pediatric critical care medicine textbook was also published that same year. Pediatric Critical Care Medicine , a journal devoted exclusively to the specialty of pediatric critical care, began publication in 2000.
Introduction
ICUs are a key element in the overall delivery of health care services. Many patients receive care in an ICU (or in some cases, by an ICU team outside an ICU) at some point during their hospital stay. ICU care also accounts for a significant proportion of total health care costs. For example, recent studies suggest that ICU care represents between 17.4% and 39.0% of all hospital costs and between 0.56% and 1% of the United States’ gross domestic product. These statistics highlight the importance of providing safe, effective, patient-centered and family-centered, efficient, timely, and equitable care in an ICU. One of the challenges faced by every ICU today is to minimize costs of care while maintaining quality (ie, increasing the value). This article provides a brief overview of the current state of pediatric critical care delivery and discusses some of the recent changes in the overall evolution of pediatric critical care delivery in North America, starting with a brief historical description of the field of pediatric critical care medicine as an important backdrop to where the field is likely to move in the future.
A brief history of pediatric critical care medicine
Evolution of the modern PICU and evolution of the specialty of pediatric critical care medicine have occurred approximately in tandem over the last 50 years. The roots of pediatric critical care medicine, as a specialty, include adult respiratory care, neonatology, pediatric general surgery, pediatric cardiac surgery, and pediatric anesthesiology. ICUs first developed in response to the global epidemic of polio, which was a universally fatal disease until the development of the iron lung in the late 1920s. The knowledge gained in artificial ventilation was then generalized to other conditions, and pediatric critical care medicine began to develop as a specialty to provide care for other types of patients, including neonatal ICU graduates.
The first known PICU was established in Sweden in 1955, whereas the first US PICU opened in 1967 at the Children’s Hospital of Philadelphia, a 6-bed unit with a separate nursing staff and 24-hour in-house physician coverage provided by pediatric anesthesia fellows. PICUs began to spread throughout the United States and by the mid-1970s were located in most hospitals that had a pediatric residency program. The Pediatrics Section of the Society of Critical Care Medicine and Section on Critical Care of the American Academy of Pediatrics were established in 1981 and 1984, respectively. In 1987, the first American Board of Pediatrics certification examination for pediatric critical care medicine was offered (182 of the 242 first-time takers passed). The first pediatric critical care medicine textbook was also published that same year. Pediatric Critical Care Medicine , a journal devoted exclusively to the specialty of pediatric critical care, began publication in 2000.
Current state of pediatric critical care medicine
Number of Pediatric ICUs
The exact number of pediatric critical care beds in the United States is unknown, but several studies have attempted to clarify the issue. A 2004 report estimated that there were 350 PICU units, although more than half of them had fewer than 8 beds. A 2003 survey of PICUs in the United States demonstrated 337 PICUs with a total of 4044 beds, a median bed number of 12 beds, and 58 admissions per bed. Although the number of PICU beds in the United States varies based on which survey data are used, the number is lower than the estimated 67,357 adult critical care beds and 20,000 neonatal intensive care beds.
Dedicated pediatric cardiac ICUs (CICUs) have been a more recent development in larger children’s hospitals, along with creation of educational conferences and a professional society devoted to this field (Pediatric Cardiac Intensive Care Society). It is estimated that approximately half of pediatric cardiac surgery centers operate dedicated CICUs, although the overall impact on patient care outcomes and pediatric critical care education is uncertain. Specialized PICUs, such as CICUs, are discussed later.
Pediatric Critical Care Medicine Specialists
In a 2008 survey, slightly more than 10,000 physicians reported their practice as adult critical care medicine or pediatric critical care medicine. This study emphasized, however, that the actual number of physicians practicing full-time critical care was likely significantly less than what was reported. The American Board of Pediatrics reported 1881 board-certified pediatric intensivists in the United States in 2011, with an average age of 49 years. Of the board-certified pediatric intensivists, 25% were 56 years of age or older. Significant geographic variation exists for board-certified pediatric critical care physicians, resulting in wide variation in available specialist-to-child ratios across the United States ( Fig. 1 ). The exact number of physicians practicing pediatric critical care medicine is unknown due to both board-certified physicians no longer practicing pediatric critical care and non–board-certified practitioners, such as pediatric anesthesiologists and pediatric cardiologists, among others, who do.
Pediatric Critical Care Medicine Trainees
Completion of a 3-year pediatric critical care fellowship (after completion of a pediatric residency or an internal medicine–pediatric residency) at 1 of the 70 Canadian or United States accredited training programs is currently the only pathway to become board certified by the American Board of Pediatrics. The Royal College of Physicians and Surgeons of Canada has less-restrictive requirements for both fellowship admission and board certification. The mid-1990s showed a decreasing number of critical care trainees (adult and pediatric), but more recent trends have been encouraging, with an overall increase in critical care trainees of 11% during the period from 2001 to 2006 and an increase of 18% in pediatric critical care trainees. This upward trend has continued ( Fig. 2 ). Pediatric critical care training programs, like all training programs, have had to adjust to changing requirements and duty hour regulations, and the ideal curriculum and method of delivering this content remain uncertain.
Models of critical care delivery
Avedis Donabedian, considered the founder of the study of health care quality and outcomes research, proposed that the evaluation of health care delivery be conceptualized in 3 dimensions—structure, processes, and outcomes ( Fig. 3 ). Structure refers to the setting in which care is delivered, whose elements can be easily recalled by P’s and T’s . The 2 P’s refer to the people in ICUs (patients and providers). The types of patients cared for in a particular ICU (trauma, solid organ transplant, bone marrow transplant, neurointensive care, cardiac intensive care, and so forth) as well as the providers working in an ICU (background, level of education, training, and certification of individuals who provide care in ICUs) have an impact on ICU outcomes. The 2 T’s refer to the technology (ventilators, physiologic monitors, and so forth) and therapy. Processes refer to how care is provided in ICUs, which encompasses how different providers interact and work together to take care of critically ill patients. Outcomes refer to the endpoints of care, encompassing survival and quality of life as well as other important outcomes, such as complications (adverse drug events, unplanned extubation, hospital-acquired infections, and so forth), length of stay (LOS), patient/family experience, staff satisfaction, and costs. Donabedian suggested that processes are effective only when the appropriate structures are in place to support them and when outcomes are measured, so that processes can be evaluated for effectiveness and modified to produce better outcomes.
Structure
Patients
As discussed previously, an important structural consideration driving outcomes is the type of patients who receive care in a particular ICU. Critical illness stems from a diverse range of conditions that can affect just one or even multiple organ systems. This diversity in the type and severity of critical illness (ie, case mix) is one of the most commonly cited reasons why it is so difficult to compare outcomes among multiple different ICUs. The particular case mix is an important consideration even in specialized organ-specific or disease-specific ICUs (eg, neuro-ICUs and CICUs). These ICUs take advantage of economies of scope and scale to care for critically ill patients with a specific disease type or organ dysfunction. Although previously limited to adult centers, there are now several specialized ICUs or critical care teams in cardiac intensive care and neurointensive care at several large children’s hospitals in North America. Whether or not these specialized ICUs or critical care teams improve outcomes is a matter of debate.
Over the past few decades, a new population of patients has emerged in the PICU. Children with diseases once considered uniformly fatal are surviving into late adolescence and early adulthood. These children are surviving but develop complex, chronic medical conditions that frequently require admission to the hospital. The number of children with complex, chronic medical conditions requiring PICU care is increasing. Advances in intensive care are also enabling children to survive acute critical illness in greater numbers, many of whom develop long-term dependence on technology to support chronic organ dysfunction (eg, long-term mechanical ventilator support and renal replacement therapy). Many of these patients may remain dependent on intensive care for sustained organ dysfunction that persists for weeks to months, even before discharge to home. These patients have been described as chronically critically ill and will continue to pose challenges for critical care medicine and for the overall system of health care delivery.
Providers
Staffing shortages in ICUs
The critical care team is one of the key structural elements driving quality of care in an ICU. Unfortunately, there is a growing shortage of adequately trained critical care physicians, nurses, respiratory therapists, and pharmacists in the United States. These shortages are especially relevant with regard to physician staffing in ICUs, as demonstrated by studies showing that no more than 1 of every 3 adult ICU patients is managed by an intensivist. The situation becomes even more worrisome because of the trend of early retirement of current intensivists, because less than 1% of medical school graduates are choosing to enter the field of critical care medicine, and because an increasing number of patients will require ICU care. Early retirement is a real concern, with results of studies revealing burnout as a common occurrence in critical care medicine. As Krell states in his essay on the critical care workforce, “finding new drugs or devices to treat the critically ill will have little effect compared with solving our manpower problems.”
Telemedicine has been proposed as one potential solution to the growing shortage of trained critical care providers. With merely 5500 intensivists practicing in the United States and only one-third of ICU patients cared for by an intensivist, telemedicine makes it possible for more patients to be managed or triaged by intensivists, especially in underserved areas. The use of telemedicine in ICUs is reviewed extensively in the accompanying article by Dr. Marcin in this volume. Regionalization of services is another solution. In examining staffing of PICUs, in particular, Randolph and colleagues demonstrated that resources for PICUs in the United States continue to be “spread out across diverse PICUs, with highly variable staffing and subspecialist availability patterns.” This reality, coupled with the high cost associated with critical care medicine, has prompted many policy experts to advocate for regionalization of PICU care. This position is supported by evidence that centralization of PICU resources in tertiary centers corresponds with improved health outcomes.
High-intensity versus low-intensity physician staffing
Another important structural consideration driving outcomes in ICUs is the critical care providers themselves. The American College of Critical Care Medicine Task Force on Models of Critical Care Delivery published a summary of the relevant literature on this topic in 2001 and suggested that critically ill patients should be cared for by a multidisciplinary team of critical care providers led by a physician specifically trained and certified in the field of critical care medicine (ie, an intensivist). The Task Force further suggested that a “closed” model of ICU organization (now more commonly referred to as high-intensity staffing ), in which this dedicated team is directly responsible for all aspects of care in ICUs, is superior to the more traditional and ubiquitous “open” model of ICU organization (now more commonly referred to as low-intensity staffing ), in which patients are admitted to an ICU and are cared for by their primary care or subspecialty physician, often with little or no input from the critical care team. Multiple studies have demonstrated better outcomes—including improved mortality—in ICUs with high-intensity physician staffing. Based on these studies, the Leapfrog Group published a report in 2007 that upheld ICU staffing as the most important safety standard by which to measure a hospital’s quality of care. The group outlined a standard for adult ICU and PICU staffing that included an intensivist dedicated to an ICU to manage the care of patients through direct, on-site care during daytime hours every day of the week and as needed via page at other times. Unfortunately, more recent studies suggest that a large percentage of ICUs across the United States are unable to comply with the Leapfrog Group recommendations.
24/7 Intensivist staffing
Several hospitals and organizations are taking the Leapfrog Group’s recommendations a step further by suggesting that ICUs should be staffed by an in-house intensivist 24 hours per day, 7 days per week (24/7 intensivist staffing). Historically, the vast majority of hospitals have not staffed their ICUs in this manner. For example, a survey of 2900 ICUs at more than 1700 US hospitals in 1991 found that only 6% of hospitals provided attending coverage with 24/7 intensivist staffing. On average, ICUs were only staffed for 6 hours per day by an intensivist. A pediatric study published in 2001 showed that less than 1 in 5 (17%) of all PICUs had 24/7 intensivist staffing. In this particular study, larger PICUs (>20 beds) were more likely to have 24/7 intensivist staffing compared with smaller PICUs. In a more recent study, limited to ICUs at academic medical centers, only 33% of those responding stated that they provided 24/7 intensivist staffing; 24/7 intensivist staffing was more common in PICUs and surgical ICUs.
Unfortunately, the results from studies comparing outcomes in ICUs with 24/7 intensivist staffing versus those without are mixed. For example, Arias and colleagues retrospectively reviewed their single-center experience and found a higher risk-adjusted mortality rate for patients admitted to the PICU during evening hours compared with those admitted during daytime hours. Although staffing was not directly assessed in this study, one of the concerns is that differential staffing from day to night may have correlated with and contributed to this significant difference in outcomes. Peeters and colleagues, however, did not find any significant differences in risk-adjusted mortality during off-hours admission to their PICU. Coverage at night in this PICU was provided by a resident physician, although all of the intensivists lived within 15 minutes of the hospital. Numa and colleagues reported similar findings in their retrospective, single-center review of a 10-year period. Subsequently, in their retrospective review of approximately 6000 admissions to a single PICU with 24/7 in-house staffing of pediatric intensivists, Hixson and colleagues found no significant differences in outcomes between weekend and/or evening admission versus weekday admissions. Finally, Nishisaki and colleagues reported that duration of mechanical ventilation and ICU LOS significantly decreased, although risk-adjusted mortality did not change, after introduction of 24/7 intensivist staffing in their single-center PICU. Unfortunately, single-center studies and retrospective, multicenter cohort studies are not adequate to address this issue, due largely to the heterogeneity in physician coverage models across the country, as suggested by a recent systematic review and meta-analysis based on 10 adult cohort studies that failed to show a significant differences in outcome between daytime versus nighttime admission to an ICU.
Wallace and colleagues recently conducted a multicenter, retrospective cohort analysis that attempted to address the issue of heterogeneity of coverage models in the United States. These investigators used the Acute Physiology and Chronic Health Evaluation clinical information system (2009–2010 data) combined with a survey of ICU staffing practices, which combined data from more than 65,000 critically ill adults admitted to 49 ICUs at 25 different hospitals. In ICUs with high-intensity daytime staffing, nighttime (ie, 24/7 intensivist staffing) was not associated with improved outcomes. Conversely, among ICUs with low-intensity daytime staffing, nighttime intensivist coverage significantly improved outcomes. These findings are similar to the available pediatric data, but as of yet, there are no multicenter pediatric studies exploring this issue.
Physician assistants, nurse practitioners, and hospitalists
The shortage of critical care physicians and coverage demands (discussed previously) have led to the growing use of so-called physician extenders (ie, physician assistants [PAs] and advanced practice nurses) in critical care medicine. The proper term for such providers (physician extenders, midlevel providers, affiliated providers, or licensed independent providers), however, is under debate, with the term, physician extenders , currently the most commonly used term by health care organizations. A national survey of PICUs in 2005 found that nurse practitioners (NPs) worked an average of 50 hours per week, with responsibilities including direct patient management, coordination of care, education, research, and consultation. NPs also performed procedures, including venipuncture, intravenous placement, endotracheal intubation, central venous and arterial catheterization, and peripherally inserted central catheter insertion. In that same year, Mathur and colleagues described their experience with PAs in their PICU in New York. After completing a 6-month to 1-year orientation, PAs primarily serve as providers of bedside care. Like NPs, they also manage a variety of other responsibilities, including assistance with procedures, education, research, and quality assurance.
NPs and PAs bring strengths different from those of resident physicians. Although residents typically possess a greater fund of knowledge and understanding of pathophysiology, PAs (and likely also NPs), with increasing experience in ICUs, possess knowledge of unit-specific policies, practices, and common diagnoses. If managed well, these 2 roles may complement each other to the benefit of the individuals, the patients, and the unit as a whole. Current literature, although limited, suggests that retention of NPs and PAs has proved a challenge to units that seek to use them and requires thoughtful attention given the labor-intensive nature of orientation and the disruption to the function of the unit by turnover. The trend toward using NPs and PAs in pediatric critical care continues, with anticipation of greater numbers in the future. As a result, the body of literature that examines the preparation, use, and retention of NPs and PAs is mounting.
In recent years, the specialty of pediatric hospital medicine has evolved rapidly, and the hospitalist has become an increasingly prominent figure. Increasing numbers of patients admitted to hospitals receive care from hospitalists—or house physicians—rather than from primary care physicians, who provide care in both outpatient and inpatient settings. In many hospitals, in particular those in less well-served areas, a hospitalist cares for patients not only in general wards but also in low-level ICUs. For example, Tenner and colleagues demonstrated improved outcomes with hospitalists providing after-hours coverage instead of residents in a PICU. A single-center, prospective, cohort study suggested that adjusted mortality and LOS were not significantly different between a hospitalist-led ICU versus intensivist-led ICU. LOS was shorter and mortality was better, however, when mechanically ventilated patients with intermediate severity of illness were cared for by an intensivist-led team. Pediatric critical care is just beginning to wrestle with how best to train, use, and incorporate hospitalists into the evolving discipline of pediatric critical care.
Critical care pharmacist
The critical care pharmacist is a key member of an ICU team. Several studies show that the addition of a critical care pharmacist to an ICU team reduces the incidence of adverse drug events, improves antibiotic stewardship, and improves compliance with evidence-based medication practices, all of which are associated with improved outcomes. Ideally, a critical care pharmacist should join bedside rounds with the rest of the ICU team.
Nursing and ancillary staff
The critical care nurse represents the front line in an ICU and is a key member of the critical care team. ICU nursing staff should work closely with the intensivists, pharmacists, and other members of the ICU team (pharmacist, respiratory therapist, dietician, social worker, chaplain, physical/occupational therapists, psychologist, and so forth) to meet all of the needs of critically ill children and their families. All members of an ICU team (including the intensivists) should meet on a regular basis to discuss difficult cases and ethical issues, review current performance on key outcome and process measures, share and discuss best practices, and provide continuous education and training. These regular meetings also serve to build a culture of mutual respect, trust, and camaraderie in ICUs. Multiple studies conducted over the past 20 years suggest that an organizational culture of mutual respect, trust, and teamwork significantly improves outcomes in ICUs. Some investigators have suggested that the proper organizational culture may be more important than the ongoing controversies and issues related to staffing (discussed previously).
Optimal staffing ratios
The optimal physician staffing ratio in ICUs is not known. Dara and Afessa conducted a retrospective cohort study at a single center to determine whether there was an association between the intensivist-to-bed ratio and outcome. Four different staffing ratios were used over time (1:7.5, 1:9.5, 1:12, and 1:15). Differences in intensivist-to–ICU bed ratios were not associated with differences in either ICU or hospital mortality. A 1:15 intensivist-to–ICU bed ratio was associated with increased ICU LOS. This is an important area that has not been adequately studied. Most critically ill children warrant a nurse-to-patient ratio of 1:1 or 1:2, depending on their underlying severity of illness. Several studies have linked nursing workload to nurse retention, burnout, patient safety, and outcomes in ICUs. The qualification and training of the nurses also must be considered when determining staffing ratios.
Technology and therapy
Aside from the physical design and layout of ICUs (not discussed in this article ), the other key structural elements driving quality of care in ICUs are technology and therapy. Technologic advances in monitoring, information systems and clinical decision support, and end-organ support have advanced the care of critically ill patients over the past decade. In addition, translational research continues to bring the latest and most advanced treatments from bench to bedside. Many of these advances are discussed elsewhere in this issue and, therefore, are not discussed further in this article.
Processes
Structural elements (discussed previously) interact with key processes in ICUs to drive improved outcomes. Structural elements, however, cannot make up for lack of an appropriate organizational climate that supports implementation of standardized, reliable, evidence-based processes. For example, the mere presence of an intensivist in an ICU is not likely to improve patient outcomes if the intensivist is not effectively communicating with other physicians, with the nursing staff, and with the patients and their families. Similarly, optimal nurse-to-patient staffing ratios are not likely to reduce medical errors or hospital-acquired infections unless standardized, reliable processes are in place. A retrospective analysis of a large, prospective database (Project IMPACT), involving more than 101,000 critically ill patients receiving care at 123 ICUs in 100 US hospitals, showed that after adjusting for severity of illness and propensity score, hospital mortality rates were higher for those patients cared for exclusively by intensivists compared with patients cared for entirely by nonintensivists. Rather than calling for a moratorium on intensivists, most critical care policy experts suggest that the results of this study put into sharp focus the need for reliable processes in ICUs.
Bundles and checklists
There are several studies that have shown significant reductions in adverse drug events, unplanned extubations, central line–associated bloodstream infections, ventilator-associated pneumonia, catheter-associated urinary tract infections, pressure ulcers, surgical site infections, and other errors in PICUs through the implementation of standardized, reliable processes, or bundles ( Box 1 ). A bundle is a small set of standardized, evidence-based interventions (or, if there is insufficient evidence, best practice interventions) for a specific patient population or disease condition, that, when implemented reliably and consistently, leads to improvements in outcomes. Certain cultural elements (eg, communication, teamwork, leadership, and accountability) are also necessary to assure that these improvements are sustainable for the long term. Many of these cultural elements are found in so-called high-reliability organizations. High-reliability organizations are discussed in greater detail in an accompanying article by Niedner and colleagues.
