1. 
   

   How did intensive care units (ICUs) and critical care medicine develop in the United States?

   
   
2. 
   

   Who currently provides critical care in the United States?

   
   
3. 
   

   What is an intensivist?

   
   
4. 
   

   What are the evidence-based recommendations for ICU staffing?

   
   
5. 
   

   What are some of the challenges of nationwide ICU staffing?

   
   
6. 
   

   What is the role of hospitalists in future critical care provision?

Current estimates indicate that approximately 6000 adult intensive care units (ICUs) provide complex care for 55,000 to 70,000 patients per day in the United States. The number of ICUs continues to rise despite a nationwide decrease in hospitals. ICU beds account for over 10% of all US hospital beds with over 4.4 million patient admissions annually. Costs associated with ICUs account for 1% of the US annual gross national product and approximately 20% of individual hospital expenditures. These national trends are the likely results of medical advances compounded by the surge of baby boomers reaching more advanced age and the growing obesity epidemic.

The United States faces current and future ICU critical care physician staffing shortages. The most recent estimates suggest that only one-third of critically ill patients receive medical care from a physician with specialized critical care training, commonly referred to as an “intensivist.” The intensivist shortage coupled with the explosive growth of hospitalists over the past decade has led to more than 75% of hospitalists providing critical care as part of their clinical practice. A brief historical review of ICU development and of relevant medical literature will provide insight into current US physician staffing practices. While their role is not yet clearly defined, hospitalists will certainly continue to play a major role in future critical care provision.

The first American ICU is attributed to Dr. W. E. Dandy, a surgeon who created a three-bed unit in 1926 dedicated to the postoperative care of neurosurgical patients at The Johns Hopkins Hospital in Baltimore. In response to the nursing shortages after World War II, patients requiring intensive postoperative care or mechanical ventilation were located together within hospitals to improve medical care efficiency. The polio epidemic in the late 1940s and subsequent advances in mechanical ventilation in the early 1950s fostered the expansion of ICUs. An anesthesiologist and pioneer of cardiopulmonary resuscitation, Dr. Peter Safar, started the first 24-hour, physician-staffed ICU in 1958 at Baltimore City Hospital. Since the 1960s, most hospitals have maintained at least one ICU. There are presently more than 65,000 adult ICU beds and 20,000 pediatric ICU beds distributed unevenly among hospitals in the United States.

In 1970, a group of subspecialists with shared interests established the Society of Critical Care Medicine (SCCM) and defined critical care medicine as “the multidisciplinary healthcare specialty that cares for patients with acute, life-threatening illness or injury.” SCCM remains the largest multispecialty medical group in the United States today with 14,000 current members. In 1979, the American Board of Medical Specialties (ABMS) proposed to make critical care a multidisciplinary subspecialty of four existing primary boards—anesthesia, internal medicine, pediatrics, and surgery. These boards failed to reach consensus on training requirements; therefore, in 1986 the ABMS approved separate critical care board certification in concurrence with fellowship program development in these fields. Each discipline has its own board-eligibility requirements and subspecialty exams in critical care which remain separate today. Fellowship programs span from 1 to 2 years for anesthesia or surgery and 2 to 3 years for critical care medicine, combined pulmonary-critical care medicine or pediatric critical care. There are less than 1000 critical care fellowship positions offered each year between the four disciplines of anesthesia, internal medicine, pediatrics, and surgery. Almost 80% of adult intensivists initially train in internal medicine while less than 10% train in anesthesia and surgery. The most current estimates report there are almost 11,000 intensivists who provide care in the ICU setting as part, or all, of their patient responsibilities and approximately half maintain their critical care board certification.

A few critical care medicine fellowship programs have accepted emergency medicine residency graduates but have not granted them critical care board eligibility upon fellowship completion. Instead, many of these graduates have completed the European Society of Intensive Care Medicine examination to document competency. In 2009, the ABMS and the boards of emergency medicine and internal medicine agreed to accept emergency medicine residency graduates in all 2-year critical care medicine fellowships and to grant critical care board eligibility upon completion.

The medical literature on ICU physician staffing is scant and relies heavily on small, observational trials that often use historical controls. One of the first published studies was led by Li et al in 1984 and examined mortality outcomes with the use of onsite physicians. Specifically, two internists were hired to provide care exclusively for hospitalized patients at one institution, replacing two physicians who had concurrent inpatient and outpatient responsibilities. This study showed decreased ICU and hospital mortality correlating with the use of onsite physician staffing. The greatest limitation of this study was its small size, but its design is characteristic of subsequent larger studies by other investigators. Despite its limitations, this published article was well received by the medical community. Li’s study is often inaccurately referred to as the first one to examine intensivists; under closer inspection, the two physicians studied are actually hospitalists by today’s definition.

Later studies using critical care trained physicians in lieu of hospitalists consistently suggest improved patient outcomes when intensivists provide care to ICU patients. A meta-analysis by Pronovost et al (2002) of 19 studies, including the original one by Li et al, reported decreases in both hospital and ICU mortality and length of stay with the use of intensivists. All studies compared intensivists to traditional control models—physicians with both hospital and clinic responsibilities—and not to hospitalists. Other limitations of this meta-analysis include the retrospective or observational designs of the included studies, single-center studies, small sample sizes, and an inappropriate outcome adjustment based on proposed study quality in the analysis. Despite these study limitations, the specialized training of intensivists likely improves critical care provision and patient outcomes.

Aside from the study by Li et al, no published trials compare adult hospitalists’ ICU care to intensivists’ (or other clinicians) ICU care since the “hospitalist movement” began in the 1990s and critical care board certification began in 1986. A recent study from Emory University School of Medicine suggests favorable outcomes when ICUs are staffed by hospitalists utilizing selective intensivist consultation (unpublished data, awaiting peer review). To date, one published trial evaluated pediatric patient outcomes with the use of pediatric hospitalists at a single institution. Tenner et al reported decreased mortality and length of stay when pediatric hospitalists provided after-hour care rather than pediatric residents.

A study by Levy et al in 2008 examined the impact of intensivists by utilizing a large, multicenter database. The study produced controversial findings that suggested intensivist-staffed ICUs may have higher mortality rates. Although this retrospective study had a sample size of more than 100,000 patients from multiple medical centers, the authors concluded that confounding factors contributed to their findings. The database relied on the accuracy of diagnosis coding and self-identified specialty by physicians. Hospitalists were included among the nonintensivist physicians but were not evaluated separately, nor compared directly, with the intensivists. This paper received critical editorial review by some intensivist peers but most agreed that further well-designed studies are needed to more accu-rately define the impact of intensivists and hospitalists.

Based on available medical literature and expert opinions, an advocacy group made a recommendation in 2000 to staff urban ICUs with intensivists. This advocacy organization—the Leapfrog Group—is a nonmedical business consortium that strives to improve the quality of American medical care by using employer healthcare purchasing power as a compelling influence. One of its initial “three leaps” to improve patient safety focused on the use of intensivist-led ICU Physician Staffing (IPS) and had four principle components: (1) physicians with board certification in critical care manage or co-manage all ICU patients; (2) physician clinical responsibilities are exclusively to the ICU; (3) physicians are present in the ICU during daytime hours (8 hours per day); and (4) physicians are accessible at all other times with an assistant who can reach the ICU in less than 5 minutes. When the Leapfrog Group proposed IPS, estimates suggested that 10% of American ICUs already used this staffing model. A follow-up study in 2006 by Angus et al estimated that merely 4% of ICUs met these stringent IPS standards. Nearly 25% of ICUs attained “high intensity” staffing, defined as intensivist management or co-management of at least 80% of ICU patients. Several barriers to IPS implementation include the perceived loss of control or potential income by primary physicians and other subspecialists. Interestingly, hospitalists have faced and overcome similar concerns over the past two decades regarding their role in inpatient care. Other barriers to IPS include the significant costs to hospitals for recruitment and retention of intensivists and the limited availability of intensivists for hire. The latter reason has received the most attention.

In 2000, the Committee for Manpower for Pulmonary and Critical Care Societies (COMPACCS) published the first large-scale assessment of the American intensivist workforce. With representation from key professional groups, including the American College of Chest Physicians, the American Thoracic Society, and the Society of Critical Care Medicine (SCCM), this committee carefully examined the current and future supply and demand for adult intensivists. It discovered that most intensivists spend considerable time on activities outside the ICU and on average allocate only 26% of their clinical time to the ICU itself. COMPACCS demonstrated a projected widening disparity between intensivist supply and demand, with a predicted shortfall of intensivist hours that will equal 22% by 2020 and 35% by 2035 (Figure 133-1). This study’s findings suggest the future growth of critical care needs in the United States will outpace the ability to train and retain intensivists.