High-Intensity Physician Staffing

In a 2002 systematic review, Pronovost and colleagues demonstrated that high-intensity physician staffing was associated with reduced ICU and hospital length of stay and lower ICU mortality (relative risk [RR], 0.61; 95% confidence interval [CI], 0.50 to 0.75) and hospital mortality (RR, 0.71; 95% CI, 0.62 to 0.82). Since then, many studies have confirmed these findings or uncovered additional benefits of high-intensity physician staffing in the ICUs of different types in various settings. As a result, the high-intensity approach is considered to be the staffing model of choice in most ICU settings. It is now being applied worldwide and in atypical settings. Institution of high-intensity physician staffing in an Army hospital ICU deployed in Afghanistan was associated with decreases in mortality, the duration of mechanical ventilation, and the incidence of ventilator-associated pneumonia. Likewise, high-intensity physician staffing in a mixed ICU serving a regional nonteaching medical center was associated with a decrease in hospital length of stay, better compliance with evidence-based practices, and a significant increase in survival from sepsis. However, Levy et al. performed a retrospective analysis of 101,832 ICU patients entered into the Surviving Sepsis Campaign database and compared outcomes in ICUs where critical care physicians provided more than 95% of the care with those in ICUs where intensivists managed less than 5% of patients. Even when data were adjusted for severity of illness and patients were matched with propensity scores, intensivist-led care was associated with a higher standardized mortality ratio (RR, 1.09; 95% CI, 1.05 to 1.13 vs. RR, 0.91; 5% CI, 0.88 to 0.94) and resulted in more interventions. These results stand in stark contrast to numerous investigations and reports that showed an association between high-intensity physician staffing and outcome in ICUs of all types. Although the authors acknowledge the study’s significant limitations (unclear definition of “critical care physician,” significant gaps in the dataset, unmeasured confounders), the large number of patients and the magnitude of the dataset cannot be ignored. Since the publication of the report by Levy and colleagues, a retrospective cohort study of medical ICU patients (n = 107,324) that compared high-intensity physician staffing and multidisciplinary care teams with low-intensity physician staffing without multidisciplinary care teams reported that the former was associated with lower 30-day mortality (odds ratio [OR], 0.78; 95% CI, 0.68 to 0.89). In addition, a recent meta-analysis that included the study described previously indicated that high-intensity physician staffing was associated with lower ICU (RR, 0.81; 95% CI, 0.68 to 0.96) and hospital mortality (RR, 0.83; 95% CI, 0.70 to 0.99). These results also suggested that surgical and combined medicosurgical ICUs received most of the benefits of high-intensity physician staffing.

Despite the overwhelming body of literature (>30 studies) touting the superiority of high-intensity physician staffing, the processes that link this approach to improved outcomes remain obscure. It seems reasonable to postulate that consistent and reliable delivery of care using evidence-based standardized protocols by experienced and trained personnel contributed. Indeed, daily rounds by a multidisciplinary team were associated with a reduction in adjusted mortality, and high-intensity physician staffing was associated with increased compliance with evidence-based practices. It has been suggested that other aspects of care delivery, such as interprofessional communication, are positively altered by high-intensity staffing, but supporting data are lacking.

Nighttime Intensivist Staffing

In an effort to further improve patient care in the ICU, some institutions have attempted to move beyond the Leapfrog standards and toward 24/7 intensivist coverage or nighttime intensivist coverage. A survey of ICU program directors at academic medical centers in the United States indicated that one third (37%) of the respondents’ ICUs were covered 24/7 by board-certified or board-eligible in-house intensivists. More than half of the respondents thought that 24/7 coverage is associated with better patient care and improved education for training fellows, although they did raise concerns about reductions in autonomy and the opportunity to make independent decisions.

The assumption underlying the move to 24/7 staffing is that the intensity of physician staffing (“dose”) will improve patient outcomes (“response”). It is essential that this dose–response relationship remains sufficiently positive so that the benefit is worth the additional cost. A randomized trial in an academic ICU running under a high-intensity physician staffing model compared the addition of nighttime in-hospital intensivists with a model in which the nighttime intensivist (often the same one who covered in house during the day) provided coverage via telephone. The study did not demonstrate a difference in any of the selected outcome variables—ICU or hospital length of stay, mortality (OR, 1.08, P = .78), or readmission within 48 hours. A retrospective cohort study showed that adding a nighttime intensivist to an ICU using a low-intensity physician staffing model during the day resulted in a reduction in mortality (OR, 0.62, P = .04). Thus the current evidence is not sufficient to justify 24/7 intensivist coverage in ICUs with daytime high-intensity physician staffing. One could argue that the retrospective arm of the previous study indicates that nighttime coverage may be of benefit in ICUs operating under a low-intensity daytime model. In addition, there are other reported benefits associated with 24/7 intensivist coverage (e.g., earlier decision making regarding end-of-life care, improvement in the quality of end-of-life care ) that have not been subjected to evidence-based investigation.

Coping with Shortage of Intensivists

The demand for intensivists has been increasing and is projected to continue to do so. There is an ongoing shortage of intensivists, though, that has been foreseen for some time. The projected shortages have engendered strategies to provide enhanced coverage without a need for additional personnel. Models under evaluation include telemedicine for remote or underserved ICUs and deployment of alternative providers—nonintensivist physicians (hospitalists) or nonphysicians (nurse practitioners, physician assistants).

Intensive Care Unit Telemedicine

The initial report by Rosenfeld and colleagues that showed a significant reduction in mortality, length of stay, and costs demonstrated that ICU telemedicine is a potential solution to ICU workforce shortages. Several publications have addressed the issue. A systematic review and meta-analysis using a preobservational and postobservational study design showed that telemedicine was associated with lower ICU (RR, 0.79; 95% CI, 0.65 to 0.96) and hospital mortality (RR, 0.83; 95% CI, 0.73 to 0.94) as well as a significant decrease in ICU and hospital length of stay when compared with standard care. However, a more recent study that evaluated ICU telemedicine using a pre- and postcomparison and a concurrent control ICU in a network of Veterans Affairs hospitals failed to show any improvement in ICU, hospital, or 30-day mortality or in ICU or hospital length of stay. A similar nonrandomized, unblended, preassessment and postassessment of ICUs involving 118,990 patients in 56 U.S. ICUs showed that ICU telemedicine was associated with lower adjusted ICU and hospital mortality, as well as a reduction in ICU and hospital length of stay that was particularly pronounced in patients with very long ICU courses. ICU telemedicine was also associated with higher adherence to clinical practice guidelines.

As is often the case in medicine, the clinical benefit to patients managed with ICU telemedicine is not a product of the technology alone. In an observational, before-and-after comparison of ICU telemedicine use in six ICUs within a single health-care system, ICU telemedicine was not associated with any discernible benefit. Proposed explanations, including minimal delegation to the telemedicine team, a lack of access to clinical notes, and computerized physician order entry, suggest that the degree of integration of the telemedicine team into the ICU is an important determinant of efficacy. The presence of an ICU culture dedicated to outcome improvement and the impact leadership models have also been touted, but they have not been sufficiently investigated.

Although ICU telemedicine may provide clinical benefits, it is an expensive undertaking that mandates careful financial consideration. Capital expenditures and maintenance costs for ICU telemedicine are not trivial, and to date the care in the United States is not subject to direct reimbursement. Thus, use of ICU telemedicine will expand only if enhanced ICU use can offset the cost or if changes in health-care delivery and reimbursement (i.e., bundled payments) translate into an enhanced margin.

Coping with Shortage of Intensivists with Nonintensivists: Hospitalists as Intensive Care Unit Workforce

Hospitalists have a primary focus on the general medical care of hospitalized patients. They have increasingly been asked to care for patients with a lower severity of illness in the ICU and in step-down units. A single small, prospective, observational study found no significant differences in ICU or hospital mortality or length of stay between medical ICU patients cared for by a hospitalist team and those cared for by an intensivist-led team.