33 Priority setting
Dr. B is on the seventh day of his rotation as medical director of the intensive care unit (ICU) when he receives a referral call about a patient in emergency who needs ICU admission for ventilation support. Dr. B examines his ICU census and notes that not only are there no ICU beds available but there is also a request from a thoracic surgeon for an ICU bed for a patient currently in the observation room, and there is a request from a nearby hospital to transfer one of their patients to Dr. B’s ICU.
Dr. C, a pediatrician, has been asked to chair her hospital drug formulary committee to examine new drugs and determine which ones should be provided from the hospital budget. She is aware that these decisions are complex and often controversial and is unsure how to proceed.
What is priority setting?
Priority setting involves deciding which resources to allocate to competing needs. It is a key component of every health system because, whether wealthy or poor, no system can afford to provide every service that it may wish to provide. Both publicly and privately funded systems have the challenge of delivering quality care within the limits of government budgets or enrollee and employer contributions.
Within health systems, priority setting occurs at each decision level: micro (at the bedside or in clinical programs), meso (in hospitals or regional institutions), and macro (at the system-wide level). Clinicians are directly involved in priority setting at the micro level and often involved at other levels; however, decisions at each of these levels are inter-related. In the context of clinical programs, such as in the first case opening this chapter, the ICU director must decide which patient gets an ICU bed. This decision itself is related to decisions in the critical care program about how many beds to keep open and how many nurses to maintain on staff, and these are related to the hospital level decision about the importance of critical care in that institutions and how much of the hospital’s budget flows there, and these are related to both funding for that hospital and system-wide funding for critical care.
Why is priority setting important?
Health system sustainability is related to the effectiveness of the priority setting decision making within the system. The costs of health-related services are constantly increasing, with drug costs leading the rise. Demands for health services are increasing as a result of new technologies; increased public awareness fueled by the Internet; aging populations in Western democracies; pandemics such as HIV/AIDS, malaria and tuberculosis; and an alarming increased prevalence of non-communicable diseases, such as cancer and cardiac disease, in the developing world. The current growth in healthcare expenditures is unsustainable and is limiting the ability of governments to fund education, infrastructure, and other priorities. Therefore, setting priorities regarding what we will and will not provide is vital to the sustainability of any and all health systems.
Ethics
Justice requires that like cases should be treated alike and that the benefits and burdens of health services be allocated equitably across patients. Knowing what decisions to make would be quite simple if we could agree on the criteria to guide equitable allocation of resources. However, at the crux, priority setting decisions involve choosing among a complex cluster of criteria (e.g., clinical factors, patient values, system goals) that may be morally relevant to any one specific decision. Moreover, reasonable people may disagree about how these criteria should be applied and which values should be emphasized, particularly in the context of clinical uncertainty; competing patient, program, or system goals; and multiple stakeholder interests. For example, when deciding which patient gets the bed in the intensive care unit (ICU), should the ICU director emphasize benefit, and give the bed to the patient with the longest most productive life ahead, or emphasize need, and give the bed to the patient who is the most vulnerable? The conflict widens when the choice involves cost differences. For example, should patients be given a very costly drug that may keep them out of the ICU? Or, in the context of drug-funding decisions, such as in the second case opening this chapter, should we prefer a very costly drug that would provide a large benefit to a few patients or a less costly drug that will provide a lesser benefit to many patients?
No consensus exists about an overarching moral theory to help to resolve differences between conflicting values. Therefore, the goal must be to make these decisions in an environment where the conflicting values can be explicitly identified and deliberated upon in a morally acceptable manner. In other words, the goal is fairness.
One of the most helpful advances in priority setting has been the development of “accountability for reasonableness,” an explicitly ethical framework that provides guidance for decision makers who want to implement fair priority setting (Daniels and Sabin, 1997; Daniels, 2000; Daniels and Sabin, 2002). It is theoretically grounded in justice theories emphasizing democratic deliberation (Rawls, 1993; Cohen, 1994). “Accountability for reasonableness” specifies conditions that operationalize the ethical concept of fairness. A fair priority setting process meets four conditions: relevance, publicity, revisions/appeals, and enforcement (Table 33.1). Recently, we proposed a fifth condition – empowerment – arguing that power differences between individuals within healthcare institutions, which prevent some from fully participating in decision making, may negatively influence the fairness of priority setting and so must be attenuated by leaders within the institution (Gibson et al., 2005).
Other approaches, such as economic evaluations, may be helpful: for example, cost-effectiveness analysis when setting priorities among new technologies or drugs, or program budgeting and marginal analysis when deciding among programs. However, these approaches emphasize a narrow range of values (e.g., efficiency), and not the full range of relevant values. Therefore, economic evaluations, like any other technical approach, must be considered within the context of a fair priority setting process (as described above). For example, recent work in Canada and the UK has shown how program budgeting and marginal analysis may be used in conjunction with “accountability for reasonableness” framework (Gibson et al., 2006; Peacock et al., 2006).
Law
Legal frameworks are not clinically precise and are more often helpful in identifying what decision makers may not do, rather than what they should do. In general, the law focuses on the reasonableness of allocation decisions in light of existing legal standards and the salient facts. Legally, physicians have a fiduciary relationship with their patients and are expected to meet a reasonable standard of care. Similarly, as fiduciaries of the hospital corporation, hospital board directors have a duty to act honestly and in the best interests of the corporation and its members as a whole, to exercise due diligence in making decisions on the basis of information reasonably available, and to meet an expected standard of care in discharging these duties. Therefore, in some jurisdictions, the courts have been reluctant to become involved in judicial review of how physicians or hospitals use their resources (e.g., R. v. Cambridge Health Authority, 1995.) To use Canada as an example, the Canada Health Act mandates reasonable access to medically necessary services but does not specify what those services should be, nor does it specify a mechanism for making these difficult and contentious decisions. The Canadian Charter of Rights and Freedoms and provincial human rights codes prohibit discrimination on grounds of race, ethnicity, religion, age, sex, sexual orientation, and physical or mental disability. In case law, a British Columbia judge noted that physicians’ primary duty is to their patients, and that financial considerations cannot play a decisive role in clinical decisions (Law Estate v. Simice, 1994).
Empirical studies
At the micro level, healthcare professionals decide which individuals are cared for first, which patients receive which diagnostic tests and which drugs, which patients are admitted to a hospital bed, and which patients are taken to the operating theater. Micro-level priority setting (also known as bedside rationing) is inevitable because of the increasing gap between the possibilities of effective medical interventions and the available resources (Pearson, 2000). Klein et al. (1998) described six forms of bedside rationing: denial, deflection, deterrence, delay, dilution, and termination. They stated that denial and termination were the most severe forms of bedside rationing and they rarely occurred in industrialized countries with publicly financed healthcare systems.
However, broad characterizations often cover context-specific practices, which often vary. In emergencies, triage conventions require that life-threatening situations be addressed first. But in non-emergency clinical programs, such as critical care, neurosurgery, cardiac surgery, and general medicine, the allocation conventions are unclear and variable and should be examined in context. A number of professional associations have also developed detailed allocation policies for non-emergency clinical programs such as critical care (e.g., Council on Ethical and Judicial Affairs of the American Medical Association, 1995; Carlet et al., 2004). However, the allocation conventions are often variable across policies and it is not always clear how they should be applied in context.
Critical care studies by Mielke et al. (2003), Martin et al. (2003a), and Cooper et al. (2005) determined that ICU admission decisions varied from clinician to clinician. Some prioritized great need, others the potential for benefit. Some prioritized the young, others the elderly; often, referring physicians who pushed the hardest and loudest “found” a bed for their patient. Even in contexts where admission policies exist, the policies only distinguished between broad categories of patients and not between specific patients. Moreover, where ICU admission policies exist, they are typically not well known.
Severe funding cuts and an increased prevalence of severe head injuries challenged neurosurgeons at Groote Schuur Hospital in South Africa. Unsure about what to do and troubled by the enormous moral consequences of these decisions, the clinicians initiated a collaborative effort with the University of Cape Town Bioethics Centre to develop a morally defensible allocation policy. A key feature of the policy was that it allowed all head-injured patients to be fully resuscitated and admitted to the ICU, followed by a full assessment by a neurosurgical team 24 hours later with a view to withdrawing aggressive treatment from those with the worst prognosis (Benatar et al., 2000)
Walton et al. (2007) examined the selection of patients for elective cardiac surgery and described the clinical reasons (e.g., pathology and anatomy) and non-clinical reasons (e.g., social supports, clinician-specific experiences, and remuneration schemes) that surgeons used to decide which patient to take to surgery. Even in jurisdictions where standardized urgency rating scores had been developed to help cardiac surgeons to prioritize patients on waiting lists, the scores were used for record-keeping purposes only, not for allocation decisions; decisions were based on “clinician judgement,” which varied from clinician to clinician.
In general medicine, Kapiriri and Martin (2007) found that in Uganda, where drugs are extremely scarce, publicly funded drugs were routinely denied to patients with sufficient resources to purchase them privately. In addition, those patients who received a first course of treatment often did not receive a second course, so that the drug could be given to another patient who had not yet received any treatment.
In a hospital drug formulary in Canada, funding decisions were based on a complex cluster of factors, including benefit, quality of evidence, toxicity, number of patients requiring the drug, comparison with alternatives, cost, and an informal assessment of cost effectiveness (Martin et al., 2003b). Significantly, though often perceived as the accepted approach to drug evaluation, several studies of actual practice provide evidence that cost effectiveness analysis plays a relatively minor role in drug formulary decisions (Luce and Brown, 1995; Sloan et al., 1997; Foy et al., 1999; Martin et al., 2001; PausJenssen et al., 2003).
How should I approach priority setting in practice?
At the micro level, clinicians are forced to act as gatekeepers for the health system, though they are neither trained nor inclined to perform this burdensome task (Carlson and Norheim, 2005). Consequently, they often fall back on clinical guidelines. But guidelines are typically based on the narrow range of values inherent in “evidence-based medicine” and not on the entire range of values relevant to these difficult allocation decisions (Norheim, 1999). Clinicians often struggle with these complex allocation decisions without support or guidance. They find themselves torn between the position that “physicians are required to do everything that they believe may benefit each patient without regard to costs or other societal considerations” (Levinsky, 1984), and the view that “the physician’s obligations to the patient … [must] be weighed against the legitimate competing claims of other patients, of payers, of society as a whole, and sometimes even of the physician himself” (Morreim, 1995). Sabin (2000) argued that the ethical physician should embrace both the values of fidelity and stewardship. Moreover, the role of clinician expertise has been viewed by the public as essential to priority setting (Cookson and Dolan, 1999).
Ultimately, the way forward for clinicians making priority setting decisions at the micro level is to form collaborations with supportive managers, patients, and others to develop admissions policies and elective treatment guidelines (as was described in the South African situation above). Such decisions can be made using a fair process guided by “accountability for reasonableness” that encompasses the views of all relevant stakeholders and makes the policies/guidelines accessible and known. In addition, the experiences of clinicians are vital to priority setting at other decision points in the health system, including the meso (institutional) and macro (system) levels.
This usually involves determining the substantive criteria for allocation decisions, as well as the processes that will be followed in such decisions. An example of criteria and processes for priority setting, in the context of hospital strategic planning, is provided in Box 33.1 (Gibson et al., 2004). Analogous criteria and processes could be developed for other types of priority setting decision.
Criteria
- Strategic fit
- Alignment with external directives
- Academic commitments: education, research
- Clinical impact
- Community needs
- Partnerships with external institutions or organizations
- Interdependency between programs within the same institution
- Resource implications
Process elements
- Confirm the strategic plan
- Clarify program architecture, including program groupings and definitions
- Clarify board/management roles and responsibilities
- Determine who will make priority setting decisions and what they will do
- Engage internal/external stakeholders
- Define priority setting criteria and collect data/information
- Develop an effective communication strategy
- Develop a decision review process
- Develop process monitoring and evaluation strategies
- Support the process with leadership development and change management strategies
The cases
Deciding which patient to admit, or not admit, without support is “a damned if I do, damned if I don’t” distressing situation. Dr. B should evaluate the alternatives and clearly articulate the criteria being used; discuss the criteria with others for feedback and transparency; make a decision; communicate the decision and the reasons to all relevant staff and the patients involved; then be open to responding to new information or different arguments. Once the presenting situation is dealt with, Dr. B should immediately initiate a process to develop an admission policy that meets the conditions of “accountability for reasonableness,” and which includes a dissemination strategy to ensure that all critical care staff participate, buy-in, and use the policy. A regularly scheduled policy review that examines people’s experiences with the policy, which acts as a quality-control mechanism, will help to ensure that the policy is “fairness in action.”
Dr. C must develop an environment of fair deliberation concerning decisions about which drugs should be included in the formulary, guided by the four conditions of accountability for reasonableness (Table 33.1). This priority setting process should be characterized by inclusiveness (i.e., seeking honest deliberation about competing values advanced by different types of people, including managers, clinicians, and patients), transparency (i.e., ensuring that decision criteria are communicated throughout the hospital, and even to the hospital’s community), and responsiveness (i.e., providing a vehicle for others to contribute to, or even challenge, the committee’s reasoning, as a quality improvement mechanism).