Posted on January 13, 2016 at 3:43 PM
by David Magnus, PhD and Norm Rizk, MD
This issue’s target article by Kirby (2016) raises an incredibly important and challenging set of issues: Whether, when, and how should limits be placed on patient access to intensive medical care? What are limits of shared decision making? Is bedside rationing ever appropriate? Kirby’s move away from bedside rationing to a mesolevel approach is novel and interesting. However, as some of the commentaries note, the question of whether there are limits to what will be offered to patients and their families often has to be made at the bedside.
Consider the following cases:
- An 84-year-old man with altered mental status, severe aortic stenosis, congestive heart failure, coronary artery disease, and chronic kidney disease presents with increased difficulty breathing. Standard medical management of the patient’s heart failure is no longer working. His desperate family asks for a surgical consult to consider whether there are any surgical interventions (including a ventricular assist device) that would give the patient a chance to live longer. The cardiology team and the cardiothoracic surgeons all agree that there is a small chance (different physicians vary in their estimate, but it ranges from 5 to 20%) that the patient will have his life prolonged by the surgery, but a very high chance he will die either during surgery or from postoperative complications. Considering his background state of health, likely quality of life, and the risks and benefits, the surgeons (with support from the cardiologists) decide that the burdens of treatment outweigh the benefits and decide that the patient is not a surgical candidate.
- An 84-year-old man with altered mental status, severe aortic stenosis, congestive heart failure, coronary artery disease, and chronic kidney disease presents with acute renal failure. His desperate family asks for the team to do anything to prolong the patient’s life. Given his general state of instability (fluid to improve his renal function is likely to produce respiratory failure) and that he is too hemodynamically unstable for hemodialysis, the only option to prolong the patient’s life is continuous renal replacement therapy (CRRT, which can only be done in the intensive care unit [ICU]). Considering his background state of health, likely quality of life, and the risks and benefits of CRRT, the renal service decides that the burdens of treatment outweigh the benefits and decide that the patient is not a candidate for dialysis.
- An 84-year-old man with altered mental status, severe aortic stenosis, congestive heart failure, coronary artery disease, and chronic kidney disease presented with difficulty breathing. Treatment in the ICU that included diuresis and bi-PAP successfully stabilized the patient. The patient has now reached a point of stability where enough fluid is taken off to avoid respiratory failure, while his renal function is a stable. The patient is off the bi-PAP and only requires a nasal cannula. The patient is discharged to the floor, with the expectation that he will soon be discharged either to home or to a skilled nursing facility. The family is very upset, saying that they want a nurse in the room at all times, constant monitoring, and, in general, the same level of care that they received in the ICU. Since the ICU is the only unit that meets their view of the patient’s needs, they demand that the patient be sent back to the ICU. The ICU notes that the patient is stable and does not meet criteria for requiring ICU level care and refuses the request.
- An 84-year-old man with altered mental status, severe aortic stenosis, congestive heart failure, coronary artery disease, and chronic kidney disease presents in multi-organ system failure. He is in acute respiratory distress, and acute renal failure, but since the patient is not a candidate for any surgical interventions, nor for CRRT, there is no way to successfully balance the patient’s respiratory failure and renal failure (which they have been attempting to do on the floor). The desperate family demands that the patient be sent to the ICU and intubated. The patient will likely pass away without intubation, but the team explains to the family that in its judgment, the patient will never survive to discharge from the ICU and the team would merely be prolonging the dying process. Considering his background state of health, likely quality of life, and the risks and benefits of intubation, presser support for his falling blood pressure, and other aspects of ICU care, the ICU decides not to accept him, arguing that ICU-level care would not constitute a benefit to the patient.
It is not clear that any of these decisions represent bedside rationing. If bedside rationing is denying a patient a treatment that is (on balance) beneficial to that patient, then in the judgment of each of the teams involved in the four cases, the treatments requested by the family would not provide a net benefit to the patient. But that judgment is subject to all of the limitations discussed by Kirby (limits of evidence, prejudice, and value judgments). So the concept of bedside rationing is perhaps not the right lens for evaluating these cases.
What is most striking about these cases is that cases like numbers 1–3 happen routinely and with little controversy. Surgeons, anesthesiologists, interventional radiologists, and other consulting interventionists are routinely asked whether they are willing to take on a patient. And it is very common for them to make an independent judgment on whether to say yes. In cases where a requested intervention is denied, families can appeal to other specialists in the service (if there are any available). But rarely is futility or other legal, ethical, or policy mechanism invoked. There may be (and should be) an attempt at something resembling shared decision making (SDM). But this SDM is in the form of explaining to the families why something is not an option for the patient (and hence no informed consent or refusal is being sought).
In contrast, in case 4, a team would normally need to invoke futility or some other similar policy mechanism in order to refuse to admit such a patient to the ICU. In practice, this might mean offering the ICU-level care, at least for a time. It is not obvious why case 4 should differ so markedly from cases 1–3. In all four cases, the team’s professional judgment is that the treatment is on balance nonbeneficial. In all four cases, the family is demanding the treatment. While appeal to better communication in all four cases seems appropriate, the nature of that communication is framed very differently if a requested intervention is not an option. In our experience, families are usually understanding of clinician refusals based upon their clinical judgment that something will not help their loved one.
More ethical analysis is needed to explain why these types of cases are treated so differently and whether there is a sufficient rationale to continue currently quite divergent practices. As Kirby points out, there are simply not enough ICU beds to provide them to everyone who might benefit, much less to those who likely will not. The mesolevel approach might help for cases like number 4, but more needs to be done to account for why something similar is not needed for the other cases. Alternatively, perhaps, case 4 should and could be decided at the bedside by intensivists, similar to decisions made by the other clinicians in the first three cases. We simply do not have consistent guidelines and standards for when clinical decisions need to be made by committee or by individual practitioners. Justice requires that similar cases should be decided similarly.