Outcomes in the ICU

Over the last 3 decades, critical care has matured into an independent field within medicine. Physicians practicing intensive care receive specialized training, irrespective of whether their background is in internal medicine, anesthesia, or surgery. Reflecting this, there are pathways for formal certification in critical care. Select disease states are also viewed as falling under the purview of the intensivist. Namely, septic shock, acute lung injury, and care of the high-risk postoperative patient are felt to be in the specific domain of the critical care physician. However, despite the maturation of critical care, the field remains relatively young. Unlike other established medical and surgical subspecialties that are now well into their adulthood, critical care perhaps is best considered as moving into late adolescence. It would behoove us, therefore, to pause and consider the trajectory of our subspecialty. Such introspection cannot proceed, though, without some initial assessment of what we as intensivists offer, and more importantly, what we accomplish.

This issue of Seminars in Respiratory and Critical Care Medicine helps to illustrate where we are as a field with respect to health care delivery. Many times advances in health care seem to progress and become adopted out of sheer inertia. Little thought or dialogue transpires before new treatments and new strategies are put in practice. When such a dialogue does occur it is often limited and fails to focus on one key issue: What is the value of what we can do? By value, I mean more than simply financial or economic end points. Value as a concept comprises issues about quality health care delivery, both quality and quantity of life, and aspects of morbidity. “Value” also of course begs the question of value to whom and for whom?

Any discussion of value, like most policy questions, is preferably based on a data-driven appreciation of the current state of affairs. This is particularly true in critical care medicine. For it is in critical care medicine that several important issues intersect, which, in turn, makes questions about value particularly acute. First, critical care is often delivered to patients at high risk for death. End of life issues are fraught with ethical conflict. Second, critical care resources are scarce. The potential demand for intensive care unit (ICU) care will always outstrip the supply. In light of continuing critical care physician and nursing shortages this will likely become a more acute problem in the future. Because of this, we, both as physicians and as a society, require a fair means for allocating and triaging these scarce and precious resources. Third, critical care remains expensive. Irrespective of the organizational form of health care delivery (single payer system, USA-like system, etc.) a day in the ICU is orders of magnitude more costly than a day spent on a general hospital ward. In the United States, some researchers estimate that nearly 1% of the gross domestic product is consumed in the ICU.

Because of these factors, outcomes research becomes a key endeavor in critical care medicine. Broadly, outcomes research describes a form of health care services research that addresses the final output of care delivery. In other words, outcomes research deals with a fundamental question, What do we actually accomplish and achieve for our patients in the ICU? Although broad conceptually, outcomes research ties together aspects of medical science, economics, and other social sciences to help patients, physicians, and policy leaders make more informed decisions. Methodologically, formal techniques and approaches exist to facilitate outcomes research. Examples of methodological tools in outcomes research include cost-effectiveness analysis, quality-of-life assessments, and epidemiological studies. Because all of these topics are often found in the critical care literature, the first two articles of this issue provide a formal overview of these topics. In the first article, Dr. Carson discusses methodologies of outcomes research studies and reviews the strengths and weaknesses of specific study designs. Observational studies include descriptive or analytic (retrospective or prospective), case-control or cohort studies, and randomized, controlled trials. Interpretation of studies must take into account study designs, bias, confounders, and interactions of multiple variables. Complex multivariate models may be required to assess prognosis, risk, and impact of therapeutic interventions. The purpose here is to demystify for the nonexpert some of the language regarding outcomes research and to make more transparent the techniques employed. Just as with any research in medicine, there are both high- and low-quality publications. This article should therefore help general readers become better consumers of the outcomes research found in critical care. In the second article Dr. Zilberberg discusses methodologies for evaluating health care effectiveness and costs. Analysis of cost-effectiveness and cost-utility is critical to appropriately allocate limited and often very expensive resources. In the third article, Drs. Kollef and Micek review diverse strategies employed in the ICU to optimize medical care and outcomes. These include paper-based and electronically based protocols for disease (e.g., severe sepsis) or process of care (e.g., weaning from mechanical ventilation) management, national guidelines, and targeting clinician education (with or without periodic feedback) to assure best medical practices. In the fourth article, Drs. Vincent and Bruzzi de Carvalho elegantly describe various severity of illness scoring systems developed over the past 2 decades. These tools exist to facilitate both prognostication in the ICU and research. Severity of illness scoring tools are also crucial instruments for efforts at benchmarking. It is futile to try to determine if outcomes are changing either over time or between centers if one cannot control for potential confounding due to severity of illness. Tied to attempts to improve outcomes in the ICU, irrespective of the ICU type or patient syndrome, is the evolution of protocols to guide care. Protocols help to standardize and routinize care in the ICU. Protocols exist for many aspects of ICU treatment ranging from liberation from ventilation to sedation. Because the need to improve patient outcomes motivated the development and then testing of protocols in clinical trials, this first overview section of this issue addresses protocols for care in the ICU.

The next series of articles deal with specific disease states that intensivists grapple with on a daily basis. The first set of these diseases are ones that are often preventable. The fifth and sixth articles discuss in depth deep vein thrombosis (Drs. Chan and Shorr) and ventilator-associated pneumonia (Drs. Combes, Luyt, Trouillet, and Chastre). These conditions share several key features. Effective prevention strategies exist for each. More importantly, however, these diseases are expensive and lead to substantial morbidity. As such, they all adversely affect outcomes, and rates of these conditions are considered to be surrogate markers for assessing the quality of care delivery. Following this discussion of ICU-acquired complications, readers find articles devoted to unique disease states found in the ICU. Specifically, acute lung injury/acute respiratory distress syndrome (Drs. Wilcox and Herridge), severe sepsis (Drs. Sandrock and Albertson), and fungemia (Dr. Vazquez) are encountered on a daily basis in the ICU. Each of these articles contains a state of the art update on these syndromes. Rather than simply presenting a general review, though, these sections emphasize the outcomes aspects of these topics. For example, the article on acute lung injury/acute respiratory distress syndrome thoroughly discusses the long-term patient outcomes for this disease. It discusses physical functioning and also presents data regarding neurocognitive outcomes.

Building on this theme, the final article by Drs. King, Render, Ely, and Watson explores the emerging field of neurocognitive outcomes in critical care. For all too long we have devoted efforts to short-term outcomes such as in-hospital and 28-day mortality. Although we can alter these, do we actually do any good? This final contribution to the issue highlights that bedside interventions can have long-term repercussions. How we manage sedation and delirium clearly affects the functional outcomes for those who survive their ICU care. Failing to grasp this key issue precludes one from appreciating the need to question all interventions and to subject them to rigorous evaluation. The explosion of research in this area of ICU outcomes research emphasizes how much we have learned but simultaneously underscores the vast limitations in our understanding of ICU outcomes.

In summary, we hope this issue of Seminars will educate and enlighten critical care physicians across the wide spectrum of our field. Through discussing, summarizing, and synthesizing the most relevant literature and research in ICU outcomes research, this issue can serve as a resource for clinicians. Finally, it can sensitize us to the need to concentrate more resources for outcomes research.