Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
The delivery of futile care is harmful to other patients Michael S. Niederman, MD; Jeffrey T. Berger, MD Objective: Intensive care units (ICUs) in different parts of the world provide care to patients with advanced age and terminal illness at different rates and in different patterns. In the United States, ICU beds make up a disproportionate number of acute care beds. Nearly half of all patients who die in U.S. hospitals have received ICU, some of which may be futile. The objective of this study was to examine ways in which the delivery of futile care in the ICU can cause harm to patients other than those receiving the futile care. Design: Review of available studies of patient and family attitudes about cardiopulmonary resuscitation and other supportive modalities, including antibiotic therapy, and the relationship of the delivery of such care to the outcomes of others treated in the ICU. Patients: Those treated in ICUs and those receiving futile care. Measurements and Main Results: Compared with younger patients, the elderly in the United States use more ICU care, at higher cost, have more serious comorbidities, and have a higher mortality rate. Certain populations demand ICU care more than others and often with less benefit than less-demanding populations. In a I ntensive care units (ICUs) in different parts of the world provide care to patients with advanced age and terminal illness at different rates and in different patterns. In the United States, ICU beds make up a disproportionate number of acute care beds. Nearly half of all patients who die in US hospitals have received ICU care, some of which may be futile. The use of ICU care for the elderly at the end of life The elderly comprise a large percentage of patients treated in ICUs throughout the United States with patients ⬎65 From the Department of Medicine (MSN, JTB), Division of Pulmonary and Critical Care Medicine (MSN), and the Division of Geriatrics and Palliative Medicine (JTB), Winthrop-University Hospital, Mineola, NY; and and Department of Medicine (MSN, JTB), SUNY at Stony Brook, Stony Brook, NY. The authors have not disclosed any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2010 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0b013e3181f1cba5 S518 situation of unlimited resources, the provision of ICU care, even when futile, has been viewed as an individual patient decision with no harm to others within the hospital. However, even with unlimited resources, the use of antibiotics for those who are receiving futile care can be considered unethical by egalitarian theory because it can lead to antibiotic resistance that may make the treatment of other patients impossible. In the setting of limited resources, like in pandemic influenza, or with the potential limiting of resources, in a pay-for-performance environment, the provision of futile care can also harm the hospital population as a whole. Conclusions: The delivery of futile care is not only an individual patient decision, but must be viewed in a broader context. Societal awareness of this problem is necessary, and better scoring systems to identify when ICU care has limited benefit are needed to address these difficult and challenging realities. (Crit Care Med 2010; 38[Suppl.]:S518 –S522) KEY WORDS: futile care; antibiotic use; antimicrobial resistance; pay for performance; pandemic influenza; end-of-life care; elderly; scoring systems; nonmaleficence; egalitarian theory yrs accounting for 50% of all ICU care, although they constitute ⬍15% of the population (1). In one study, those aged ⬎60 yrs used the ICU in 60% of hospital admissions compared with 30% in those ⬍60 yrs (1). The highest admission rate was in those aged 70 –79 yrs, but the elderly had a higher mortality rate, longer length of stay, and higher cost of care than younger patients. Thus, the elderly consume a large amount of resources and present with some of the most challenging and complex illnesses. Currently, there is no clear consensus about how to best allocate critical care resources to this population. Practices vary widely in different countries, but the focus of this discussion is the premise that the use of resources for elderly individuals can impact on the outcome and care of other populations. This becomes a particular problem when we provide futile care, a prospect that often arises in the elderly, particularly with the use of antibiotic therapy for patients with no meaningful chance of recovery. The use of ICU resources for the elderly is a practice that varies from country to country. In comparing the use of ICU in western Europe with that in the United States, investigators observed that ICU admissions per 100,000 population were higher in the United States with the exception of Germany (2). However, unlike European countries, the number of ICU beds in the United States did not parallel the number of acute care beds. The United States had a disproportionately higher percent of acute care beds devoted to ICU care and spent more money than other countries per ICU bed. Another recent study compared practices in the United States with those in England (3). Age-adjusted acute hospitalization rates were 110.5 per 1000 population in England vs. 105.3 in the United States with similar mortality rates. In England, 50.3% of deaths occurred in the hospital, but only 5.1% of deaths involved ICU care. In the United States, 36.6% of deaths occurred in the hospital, whereas 17.2% of deaths involved the ICU. Thus, overall, 47.1% of hospital deaths in the United States involved ICU care compared with 10.1% in England. In the United States, the elderly received ICU care far more often; in those aged ⬎85 yrs, 31.5% of medical deaths and 61% of Crit Care Med 2010 Vol. 38, No. 10 (Suppl.) surgical deaths in the United States involved ICU use compared with 1.9% and 8.5%, respectively, in England. Advanced age may contribute to ICU mortality and older patients often have more severe illness than younger patients. However, age alone does not define risk of death. An Austrian study found that 52% of all ICU care was given to patients aged ⬎65 yrs and 5% to those ⬎85 yrs. Age contributed to mortality with those aged ⬎85 yrs having an odds ratio for mortality that was 1.8-fold greater than those aged ⬍65 yrs (4). Although mortality rose with age, it was also influenced by severity of illness; when age was removed from severity scoring, severity of illness was still an important determinant of mortality. However, the relationship with aging is complex because the older patients generally had more severe illness. Similar data were reported by a US study that showed increased ICU use and mortality with age (5). ICU admission rates increased from 0.36 per 1000 person-years in those aged 18 – 44 yrs to 4.0 per 1000 person-years in those aged 65–74 yrs and to 18.5 per 1000 person-years in those aged ⬎85 yrs. ICU use was also more common as the number of chronic illnesses rose. Nonsurvivors had any chronic condition more commonly than survivors. Overall, 26.7% of all ICU days were used by people in their last 12 months of life, whereas 68% of ICU days in those aged ⬎85 yrs were used in the last 12 months of life. Cardiopulmonary resuscitation (CPR) is commonly used in the elderly with a recently reported survival rate in the United States of 18.3% for those aged ⬎65 yrs (6). In that study, CPR use was noted to have increased over time in the elderly, particularly among black patients. Success rates were lower for nonwhites, for men, for those with chronic illness, and for those coming from nursing homes. Among black patients, CPR was performed twice as often as in white patients. The lower survival of black patients after CPR was not easily explained, although many of the black patients were treated in hospitals with lower CPR survival rates. Many of the survivors did not return to their homes but were discharged to chronic care facilities. Current medical practice in the United States, with its seemingly unlimited resources, allows physicians to provide, and patients to receive, treatment of marginal or even remote medical benefit. The roots Crit Care Med 2010 Vol. 38, No. 10 (Suppl.) of this practice can be traced to an overresponse to excessive physician paternalism several decades ago, the rise of the consumer movement in the 1970s, and an American cultural emphasis on “rugged individualism.” For most of these medical treatments, the benefits to the patient are typically assessed against the harms of treatment to that patient. If a patient makes a calculation that a treatment of marginal or symbolic value is desirable, it is often provided, particularly if the likelihood of harm is low. However, the impact to other patients in the ICU from having provided this therapy is generally not considered. In the setting of plentiful resources, physicians have little imperative to limit even marginally beneficial treatments. Certain therapies such as antibiotic use, particularly in the hospital and ICU setting, are atypical in this regard because antibiotic use can create bacterial resistance, which makes subsequent infections, in other patients, potentially more lethal. Therefore, the purported individual benefits of antibiotic treatment, particularly if its use is in the context of futile care, can lead to potential harm to others. We submit that certain types of futile care such as continued antibiotic use in which it no longer provides tangible medical benefit cannot be justified because of the potential harm to other patients. Furthermore, we contend that physicians and patients have ethical obligations to desist from marginal and symbolic use of futile interventions, as can be exemplified by antibiotics, because of quantifiable collective harm. Discriminating between futile and marginal care In clinical care, it is important to distinguish between futile and marginal treatments. Futile treatments are those that hold no prospect of achieving the intended outcome. The bioethics community has largely abandoned the notion of futility after efforts to develop a clinically serviceable definition proved fruitless (7, 8). The reality is that in the clinic, few treatments hold zero prospect of achieving the desired result, whereas those that do rarely lead to clinical dilemmas. To illustrate, consider a ventilated patient on 100% FIO2 and 15 cm positive end-expiratory pressure who remains hypoxemic. Because, in its most narrow sense, the goal of mechanical ventilation is to achieve satisfactory oxygenation and the ventilator has proven unable to achieve that goal, it may be considered futile. These situations of physiological futility are not typically the troubling ones in the ICU. Rather, the challenging cases are those in which the effect of the intervention is uncontested (for example, achieving an adequate PaO2) but the benefit of the intervention is unclear. Here, it is sometimes useful to assess the benefit of treatments against a general standard of practice. For example, cardiologists in general may decline to place an intraaortic balloon pump in a patient with cardiogenic shock not because it will fail to raise blood pressure for a period of time (effect), but because no definitive remediation of the pump failure is possible (limited benefit) or because the patient has a limited benefit because of poor clinical status (anoxic encephalopathy) after a cardiac arrest. Futility should be viewed in the overall context of the patient and not just from a physiological perspective. In general, doctors want to withhold certain therapies if the predicted benefits are limited. In a survey of care of patients with pneumonia and dementia in The Netherlands, three-fourths of doctors would withhold antibiotics if the predicted mortality rate was between 75% and 90% (9). However, although physicians often wish to limit interventions that do not benefit the overall well-being of a patient, some ethicists have argued that this type of behavior is not acceptable. For example, one author argued “It is the patient’s life to lead, and death to die.… It is not for the physician to say that a patient— one out of one hundred, but still a patient— who would have lived will instead die, because of his, and not the patient’s, decision” (10). Legal support for the failure to provide futile care is also needed in the United States as evidenced by the arrest (but not indictment) of Dr. Anna Pou who provided end-of-life comfort care to do-not-resuscitate patients who could not be evacuated from a nursing home hospital that was flooded and lost power during Hurricane Katrina (11). Harms, benefits, and why antibiotic use in the ICU is different For treatments that squarely fall into the realm of medical appropriateness, the usual approach to questions of benefit is to clearly define the likely and expected S519 positive and negative effects of an intervention on the patient and to have the patient or surrogate make a personal judgment about its value. The most challenging cases are those that fall between frank physiological futility, in which decision making is within the domain of the clinicians, and situations of limited but defined benefit, in which there is no clear consensus on medical inappropriateness and where decision making tends to default to the patient or surrogate. Patients are given this latitude because, among other reasons, resources are plentiful in the United States and because it is the patient who bears the burdens of treatment in the hope of enjoying its benefits. However, hospital-based use of antibiotic therapy, particularly in the ICU, presents a somewhat novel paradigm for decision making. Antibiotic use not only contributes to microbial resistance in the patient receiving the medication (the typical direct linkage of benefit and harm), but it also increases the prospect of resistant organisms in the entire ICU, which can lead to lethal infections in other patients within the unit (an uncoupling of benefit and harm). Therefore, antibiotics represent a treatment that holds the potential to harm nonrecipients. In this case, the provision of futile or marginally beneficial care may cause limited harm to the treated patient but creates greater harm to other patients in the ICU. Allocation of antibiotics is different from typical rationing dilemmas because the issue is not a direct shortage of the material resource; the same antibiotic may very well be plentiful. Rather, antibiotic use may render supplies of the antibiotic useless to the next patient and cause a widespread increase in the clinical impact of hospital-acquired infections. Conceivably, at some future point, current antibiotic use will have created completely resistant bacteria for which no other therapeutic resource exists, thereby leading to high rates of mortality (12). Antibiotic allocation is not a zerosum game as it occurs with typically limited resources. This is potentially exponential. Many studies have documented that mortality from infection in the ICU increases with inappropriate therapy and that such therapy is more likely if patients harbor multidrug-resistant pathogens (13). The use of antibiotics is directly related to the development of multidrug-resistant pathogens; if this use is for futile care, the harm that results is impossible to justify. S520 In limited studies of physician, patient, and family attitudes, antibiotics are among the least refused therapies. Doctors often agree to use antibiotics when they would not provide CPR. Even in patient scenarios when ⬎90% of physicians would not provide CPR, almost half of doctors would still prescribe antibiotics if needed (14). Similarly, in surveys of patients and families, even in the setting of a patient with coma or a persistent vegetative state, 20% to 30% still want antibiotics (14). Therefore, the ethical implications of antibiotic use in certain critical care settings warrant particular consideration. Antibiotics can delay dying, without adding to life quality, and can expand the reservoir of resistant pathogens in the ICU. Certainly, when continued use of antibiotics has merely symbolic value or is done to respond to the emotional needs of a patient or surrogate, it cannot be justified against its serious collateral harm to others. Continued antibiotic use in these situations is an indulgence that antibiotic recipients have little moral right to impose on other patients. Clinicians should seek alternative ways to support such patients. Antibiotic use and application of the principle of justice In situations in which antibiotic use offers patients waning prospects of benefit, the use of such therapy must be assessed against not only the burdens or harms to the patient, but also the harms to all other relevant patients. What moral claims do these patients with unsalvageable illness have on antibiotics, a shared resource, where its use has a low likelihood of benefit and has measurable harms to the community? Who should make the calculation of relative benefit to the patient vs. harm to others? How should this assessment be made? One approach to these assessments involves applying the ethical principle of justice. Of the various theories of justice, the pertinent ones for this analysis are communitarian theory and egalitarian theory (15). Communitarian theory holds that there are mutual obligations and responsibilities between an individual and a community of individuals with solidarity rather than liberty being a central virtue. Communities decide which resources should be apportioned to its members, and individuals accept foregoing some opportunities for the good of the group because the shared standard of just distribution exists. Within egalitarian theories of justice, the idea of “fair opportunity” requires that every individual receive sufficient resources to equalize opportunities to achieve a particular benefit (16). For example, persons who are disadvantaged through no fault of their own (e.g., genetic predisposition causing diabetes) would be entitled to a greater amount of community resources (medical, ophthalmologic, nutritional, etc) than healthy persons so that persons with diabetes can achieve fair opportunities for life expectancy, being productively employed, etc. In the case of antibiotic use, if survival from an infection is the defined benefit, egalitarian theory could support a physician not to provide treatment to patients who have already received a substantial attempt at antibiotic cure to preserve a fair opportunity for the next similarly infected patient to be cured. This fair opportunity concept is particularly appropriate for antibiotic use because providing treatment to one person could diminish the opportunity of others to secure the same benefit. Thus, the principle of nonmaleficence would need to be applied to the community at large and not just to the patient being treated. Of course, for these arguments to hold, empiric evidence is needed for the effects of antibiotic use and nonuse on clinical outcomes for individual patients and for communities. In justice theory, the requirement of an “impartial participant” or a “veil of ignorance” suggests that groups of persons deciding on fair distributions within a society must be plausibly vulnerable to the consequences of such decisions (17). For example, a group of male health benefit administrators would never be subject to the consequences of their decisions about allocating resources for gynecologic care. For decision making, this principle would require specific involvement from constituencies vulnerable to infectious critical illness in the development of policies limiting antibiotic use. This last point raises an important distinction between macroallocation and microallocation decisions. Physicians are not ethically empowered to make bedside rationing judgments or to make assessments about duration of antibiotic therapy where the therapy continues to have defined but low benefit to the patient simply because of concerns that its use Crit Care Med 2010 Vol. 38, No. 10 (Suppl.) also harms the community. These sorts of decisions are vulnerable to violating the ethical requirement of fairness because of interphysician variability in making such judgments (18). Furthermore, bedside rationing creates an untenable ethical position for the physician who, at once, is divided between the role of patient advocate and steward of societal resources. Physicians may justifiably engage in allocation activities when processes and procedures for allocation have been authorized by appropriate and publicly accountable bodies. Other examples of futile care creating harm to others Although antibiotic therapy provides a tangible example of how providing futile care, even in the absence of limited resources, can create harm to those not involved in the decision to provide such care, other examples also exist. In some settings, these examples relate to how the provision of futile care, in the setting of actual or potentially limited resources, may create adverse consequences. Thus, if a patient and family demand the delivery of mechanical ventilation and pressors for a patient with no meaningful chance of recovery, this decision will have no immediate harm to others if there are unlimited ICU resources available. However, this type of care may be unjustified and unethical if it means that in the midst of an influenza pandemic, a young, potentially salvageable patient with respiratory failure cannot get an ICU bed or a ventilator because these resources are already being used by others, including some with no likelihood of personal benefit. Of course, this situation becomes less clear if care for both patients is not futile and if, for each, there is a potential benefit and resources are limited, particularly if this leads to comparing the “value” of one life vs. another. Such a situation has been discussed in New York State in anticipation of pandemic influenza. In a publicly circulated draft document from a Department of Health Working Group, the concept of ventilator allocation during a flu pandemic was discussed (19). In this document, which was not adopted because of a lack of legal and legislative support for physicians making these difficult decisions, the group considered the idea that a limited supply of ventilators should be used for those most likely to benefit. To make such decisions, we need new alloCrit Care Med 2010 Vol. 38, No. 10 (Suppl.) cation scoring systems, which, at present, do not exist. The New York document stated that certain principles should be followed, including: expanding resources before developing a triage system for allocating ventilators when demand exceeds supply; categorizing all patients for triage; having consistency among hospitals with regard to the triage criteria used; using clinical assessment with objective criteria, but applying support in a time-based trial, in which priority could change with the clinical evolution of the patient; that decision making should be made by supervising doctors and not the primary physician caring for the patient; offering palliative care for those who do not receive ventilator therapy; having an appeal process regarding triage decisions; and communicating to the community at large about the triage system. One scoring system approach to this problem that was discussed in the New York document is the Ontario Health Plan for an Influenza Pandemic, which is based on the Sepsis-related Organ Failure Assessment score (20). The triage scoring is based on inclusion criteria (patients can get a physiological benefit from critical care), exclusion criteria (no severe, terminal illness), and an assessment of minimum qualifications for survival. The Sepsis-related Organ Failure Assessment score is assessed in treated patients in a serial fashion, and care may be withdrawn if the patient fails to improve in a specified time period. In this system, the highest priority for critical care is the patient with severe illness, but only with a single organ failure and showing a good response to treatment. Those with a high probability of death, or who deteriorate during therapy, are considered for removal from ventilator support. One unresolved question in this discussion is whether chronically ventilated patients are exempt from evaluation or if they too would be compared with all patients needing ventilator support. A recent study applied this system to all patients admitted to a multidisciplinary ICU to see how it would have affected allocation of resources (21). The triage system was not used to make decisions but the impact of using such a system was evaluated. Of the 255 patients admitted to the ICU, 116 (46%) would have been denied ICU care under the Ontario triage system, yet many of these patients survived. Of those 48 patients categorized as too ill to warrant ICU admission, 69% survived, whereas of the 66 with no significant organ failure, 95% survived (with ICU care). The study concluded that the triage system did not accurately predict benefit from ICU care and that it should not be used. Although we clearly need a better system, the study is limited because the overall survival rate of all ICU patients was 80%, patients managed out of the ICU were not studied, and management in a non-ICU setting for those admitted to the ICU was obviously not evaluated. In the United States, we are practicing medicine in a “pay-for-performance” environment, which creates a situation of potential limiting of resources to certain hospitals that perform less well than the norm. Certain medical events are already classified as “never events” with insurance payers threatening nonpayment for these conditions. The current never events include catheter-associated urinary tract infections, pressure ulcers, and vascular catheter-associated infections. All these events could easily occur in patients whose lives are prolonged by futile care (22). Ventilator-associated pneumonia is not on this list but is an illness that has also been proposed as a “never event.” If this were to occur, it is likely that regardless of efforts at prevention, certain patient populations will have a high frequency of ventilator-associated pneumonia, if mechanically ventilated, because of the seriousness of their underlying illnesses (23). Thus, if a decision is made to ventilate patients with terminal illness, in a setting where ventilation can be defined as futile care, some of these patients will get ventilator-associated pneumonia. If this were to occur, the hospital will earn a reputation as a place with a high infection rate, reimbursements may be then withheld, and so the hospital would have less financial resources to treat other patients without futile illness. This may be a particular problem in hospitals that serve indigent populations and are already limited in their reimbursements and resources. As discussed earlier, certain indigent populations, which have a high mix of nonwhite patients, are more likely to ask for CPR and advanced care, even when such care could be futile, and these patients have a lower survival rate than white patients (6). If these patients receive futile care and they develop ventilator-associated pneumonia and other complications in the process, reimbursement to the hospital may decline further and harm future patients. Unfortunately, pay for performance has been imposed S521 without needed health system and tort reforms to allow hospitals to define and limit marginal treatments. The Texas Advance Directives Act of 1999 offers one legislative approach to this issue (24). with limited benefit of aggressive, intensive care, we need enhanced societal appreciation of the fact that the delivery of such care is not always a private decision with impact on a specific patient and his family, but sometimes a decision with a broader impact on others that are not directly involved in the decision making process. For this reason, we need better tools for limiting certain therapies in the elderly and in those with terminal illness, and we will need to find ways not to offer ICU care and certain other therapies in this setting. To achieve this end, we will need an enlightened society, but we must also equip our ICU physicians with better methods of defining futile care. Invariably, this will require better scoring systems that provide information not only about survival, but also on the quality of that survival. Summary REFERENCES Table 1. Harm to others from the delivery of futile care In the setting of unlimited resources Use of antibiotics to treat infections in patients with no hope of meaningful recovery In the setting of limited or potentially limited resources Providing futile care in the midst of an influenza pandemic Providing futile care in a pay-for performance environment May create future limitation of resources if “never events” occur No member or group of members of a just society has unlimited claims on shared resources. Futile care can lead to harm to other patients (that is, those not receiving the futile care) in a variety of circumstances, including situations in which resources are either limited or not (Table 1). Even in the absence of limited resources, an ethical argument in favor of avoiding antibiotic therapy in patients with marginal benefit, or in those in whom therapy would be viewed as futile, can be made based on its collateral harm to other future patients. A strong argument can be made in support of categorically discontinuing antibiotic therapy when it use is futile, and only of symbolic value, or if it provides the family an emotional benefit but does not offer patient benefit. In situations in which resources are limited or may become limited, the provision of futile care can also be harmful. For example, in a pay-for-performance environment, or in the setting of a surge in ICU demand in the context of limited resources, futile care also can create harm to the community as a whole. However, the balancing of minimal patient benefit against collateral harm can only be done with societal agreement and with transparency. Thus, as we approach the care of the elderly, particularly those S522 1. Frezza EE, Squillario DM, Smith TJ: The ethical challenge and the futile treatment in the older population admitted to the intensive care unit. Am J Med Qual 1998; 13: 121–126 2. Wunsch H, Angus DC, Harrison DA, et al: Variation in critical care services across North America and western Europe. Crit Care Med 2008; 36:2787–2793 3. Wunsch H, Linde-Zwirble WT, Harrison DA, et al: Use of intensive care services during terminal hospitalizations in England and the United States. Am J Respir Crit Care Med 2009; 180:875– 880 4. Brunner-Ziegler S, Heinze G, Ryffel M, et al: ‘Oldest old’ patients in intensive care: Prognosis and therapeutic activity. Wien Klin Wochenschr 2007; 119:14 –19 5. Seferian EG, Afessa B: Adult intensive care unit use at the end of life: A population-based study. Mayo Clin Proc 2006; 81:896 –901 6. Ehlenbach WJ, Barnato AE, Curtis JR, et al: Epidemiologic study of in-hospital cardiopulmonary resuscitation in the elderly. N Engl J Med 2009; 361:22–31 7. Medical futility in end-of-life care: Report of the Council on Ethical and Judicial Affairs. JAMA 1999; 281:937–941 8. Berger JT, Rosner F, Potash J, et al: Medical futility: towards consensus on disagreement. HealthCare Ethics Committee Forum 1998; 10:102–118 9. Van der Steen JT, Helton MR, Ribbe MW: Prognosis is important in decisionmaking in Dutch nursing home patients with dementia 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. and pneumonia. Int J Geriatr Psychiatry 2009; 24:933–936 Wreen M: Medical futility and physician discretion. J Med Ethics 2004; 30:275–278 Okie S: Dr Pou and the hurricane—Implications for patient care during disasters. N Engl J Med 2008; 358:1–5 Boucher HW, Talbot GH, Bradley JS, et al: Bad bugs, no drugs: No ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect Dis 2009; 48:1–12 Niederman MS: Use of broad-spectrum antimicrobials for the treatment of pneumonia in seriously ill patients: Maximizing clinical outcomes and minimizing selection of resistant organisms. Clin Infect Dis 2006; 42:S 72– 81 Marcus EL, Clarfield AM, Moses AE: Ethical issues relating to the use of antimicrobial therapy in older adults. Clin Infect Dis 2001; 33:1697–1705 Courwright AM: Justice, stigma, and the new epidemiology of health disparities. Bioethics 2009; 23:90 –96 Daniels N: Equity and population health: toward a broader bioethics agenda. Hastings Cent Rep 2006; 36:22–35 Rawls JA: Theory of Justice. Cambridge, MA, Harvard University Press, 1971 Sulmasy DP: Physicians, cost control, and ethics. Ann Intern Med 1992; 116:920 –926 NYS Workgroup on Ventilator Allocation in an Influenza Pandemic: Allocation of Ventilators in an Influenza Pandemic. Draft for public comment March 15, 2007. Available at: http://www.health.state.ny.us/diseases/ communicable/influenza/pandemic/ventilator/ docs/ventilator_guidance.pdf. Accessed March 1, 2010 Christian MD, Hawryluck L, Wax RS, et al: Development of a triage protocol for critical care during an influenza pandemic. Can Med Assoc J 2006; 175:1377–1381 Guest T, Tantam G, Donlin N, et al: An observational cohort study of triage for critical care provision during pandemic influenza: ‘Clipboard physicians’ or ‘evidence based medicine’? Anaesthesia 2009; 64:1199 –1206 Rosenthal MB: Nonpayment for performance? Medicare’s new reimbursement rule. N Engl J Med 2007; 357:1573–1575 Uckay I, Ahmed Q, Sax H, et al: Ventilatorassociated pneumonia as a quality indicator for patient safety? Clin Infect Dis 2008; 46: 557–563 Smith ML, Gremillion G, Slomka J, et al: Texas hospitals’ experience with the Texas Advance Directives Act. Crit Care Med 2007; 35:1271–1276 Crit Care Med 2010 Vol. 38, No. 10 (Suppl.)