© 2006 American Thoracic Society doi: 10.1164/rccm.200609-1388ED
Reducing the Cost of Critical Care: New Challenges, New SolutionsUniversity of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Réanimation Médicale, Hôpital Cochin, Paris, France Reports on the rapidly increasing cost of health care are now commonplace in the medical literature. Total per capital health care expenditures have increased 560 percent since 1980, and health care costs comprise over 16 percent of the United States gross domestic project (1). A large portion of these costs are directly attributable to the intensive care unit (ICU) (2). The ICU is a resource-intense environment where new drugs, expensive technologies, and specialized clinical care all contribute to dramatic health care expenditures. Reducing the costs of health care in general, and intensive care in particular, is a priority for physicians, hospital administrators, and policy makers. One important approach to reducing the cost of critical care is to diminish the unnecessary variation in care that exists across regions, hospitals, and providers (3). Usually variability is viewed in terms of unequal access to care and subsequent health care disparities (4). Variability can also be viewed directly in terms of costs. This concept is highlighted by Garland and colleagues (pp. 1206–1210) in this issue of the Journal (5). Using clinical and administrative data from an academic medical intensive care unit staffed by multiple physicians, the authors demonstrate extensive provider variability in adjusted variable costs. Physician identity accounted for over two percent of the variability in costs in a multivariate linear model. This may seem like a small amount. However, severity of illness accounted for only seven percent of the variation costs, and in econometric research it is frequently the case that the entire model may account for only ten percent or less of the variation in the cost of health care (6). Several important findings in this study are especially noteworthy. First, although the identity of the physician was associated with variation in resource utilization, it was not associated with concomitant variation in adjusted mortality or length of ICU stay. This suggests that increased use of discretionary resources, such as laboratory, radiology, pharmacy and blood bank services, is not necessarily associated with improved patient outcomes. This interpretation does require some caution. It is possible that high-cost physicians used additional resources to compensate for other unmeasured variations in quality. Had these physicians' resource use not have been elevated, perhaps their outcomes may have been worse. Nonetheless, this study provides preliminary evidence that reducing variability of resource use in the ICU can reduce costs without affecting outcome. Second, the variation in cost between physicians was independent of ICU length of stay. Length of stay is tightly associated with costs, and a large portion of the variation in the cost of care may be attributable to variation in discharge practices (7). This appears not to be the case. Significant variation appears to exist, at least in part, due to differences in basic practice patterns. How, then, to effectively reduce variation in practice patterns? As evidenced by Garland and coworkers, simply informing physicians of their resource overuse will not be enough. Physicians in the study were largely aware of their utilization habits and were able to identify themselves as high or low spenders. As the authors themselves note, advice and feedback has not been a reliable method of affecting practice change among physicians (8). Instead, novel solutions are required to both standardize care and improve outcome for critically ill patients. The first of these solutions will involve better standardization of care practice though protocols and care pathways. The era of dismissing protocols as "cookbook" medicine is over. Protocolized care for sedation, analgesia, glycemic control, ventilator management, and liberation from mechanical ventilation have been shown to reduce variation and improve the outcome of critical illness (9). It is likely that broad application of these protocols will also reduce costs, not only by reducing length of stay, but also by reducing the incidence of complications of critical illness, such as ventilator-associated pneumonia. By taking important aspects of care decisions out of the physician's hands, protocol use recognizes the simple fact that harm can result when care is completely left to a single individual's skill and memory. A second method of reducing variability in critical care is to standardize the way ICUs are organized and managed. Considerable evidence exists that staffing the ICU with a multidisciplinary care team under the oversight of a trained intensivist is associated with improved risk-adjusted outcome and reduced cost (10). A dedicated multidisciplinary care team means that fewer routine care decisions are in the hands of a single individual, ultimately reducing unnecessary variability. For example, pharmacists and respiratory therapists can standardize length of antibiotic courses and use of low tidal volume ventilation for patients with acute lung injury (11, 12). At this point, more research is needed to determine which aspects of multidisciplinary care are most effective at standardizing critical care and how to best export these to a broad range of hospitals. Nonetheless, it is likely that the multidisciplinary model offers a great opportunity to reduce inter-physician variation in the cost of patient care. Aside from demonstrating the need for greater use of care protocols and standardized ICU management, Garland and colleagues have also shown that potentially wasteful spending can occur even in high-quality ICUs. The practice group in this study was relatively attentive to issues of quality improvement and cost, the ICU was staffed entirely by trained intensivists, and several care protocols were in place. While protocols and standardized management are important, reducing variability also means developing consistent practice patterns in areas that are not necessarily governed by hard evidence. The frequency of laboratory and radiological tests, the use of generic versus name-brand drugs, and the specific indications for transfusion are all opportunities for physicians to reduce variation in the process, and cost, of care. In the era of health maintenance organizations and "pay for performance," doctors are increasingly asked to provide care that is not only evidence-based, but also value-based. Perhaps the greatest lesson of the Garland study is that cost control is not just the task of the health policy expert or the hospital administrator—it is also task of the individual ICU clinician. It is now more clear than ever that accepting this task is a difficult but necessary part of the practice of critical care in the twenty-first century. FOOTNOTES Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
Related articles in AJRCCM:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
