In September 2015, the College of Physicians and Surgeons of Ontario (CPSO) released a new policy, Planning for and Providing Quality End-of-Life Care. 1 The new policy is more accurate in its consideration of the legal framework in which physicians practice and more reflective of the current understanding of ethical issues that arise for frontline clinicians who provide end-of-life (EOL) care. Nevertheless, it poses a significant ethical and legal dilemma because, in situations of conflict with patients or substitute decision-makers (SDMs), physicians must provide cardiopulmonary resuscitation (CPR) even if such resuscitation would fall outside the medical standard of care. In this case, a college policy subsumes the medical standard of care under the ethical and legal concept of consent—in other words, consent now “trumps” the standard of care in these situations. While the policy is based in Ontario, it is precedent setting and has implications for clinical practice across Canada as other colleges review their own policies and decide to accept or reject the approach taken by the CPSO.

This paper explores the rationale for the inclusion of mandated CPR in the new policy statement, clarifies the impact of the policy on the medical standard of care, and discusses strategies for both coping with the new CPSO recommendations and using the policy as an opportunity to enhance clinical practice in EOL care.

Health Professions Appeal and Review Board case: EJW v. MGC

This case is the catalyst that generated a change in the CPSO policy on CPR—the College was twice taken to task by the Health Professions Appeal and Review Board (HPARB). In September 2008, the patient was a resident in the Veteran’s Wing at Sunnybrook Health Sciences Centre. He was admitted to the Emergency Department at the hospital because of the progression of ischemia in his legs and resulting sepsis, and ultimately, both limbs were amputated above the knee on September 17, 2008. Unfortunately, his condition continued to deteriorate, and he was evaluated by the consulting Critical Care Rapid Response Team (CCRT) physician (M.G.C.). In consideration of his comorbidities and the severity of his current illness, a judgment was made that CPR and life-sustaining treatments (LSTs) would be unlikely to succeed if his condition did not stabilize with the current treatments. A do-not-resuscitate (DNR)Footnote 1 order was thus written on September 28. The patient’s condition continued to deteriorate and he died later that day.2

An appeal to the CPSO was first heard in January 2011, and a second appeal was heard in August 2011. The problem underlying the complaint and appeals was that the patient’s daughter, who was his SDM, did not consent to the DNR. As far as she knew, her father was “full code”, pursuant to a chart entry made by a medical resident at the SDM’s behest.

The CCRT physician’s position was that he did not think the patient’s condition would deteriorate as quickly as it did, and though he had intended to talk to his patient’s SDM—having called and left a message for her—he had not spoken to her before issuing the DNR order. She arrived at her father’s hospital room to find him in respiratory distress with a physician and respiratory technician present but not initiating resuscitation. The physician’s position at the College and on appeal was that “families are to be consulted, but they do not have the authority to determine the medical decision in question”, which seemingly reflected hospital policy in force at that time.

On appeal, the HPARB may set aside an “unreasonable” decision. Nevertheless, if a decision can reasonably be supported by the information before it and can withstand a somewhat probing examination, the HPARB may not set aside the decision only on the basis of disagreement. At the CPSO Complaints Committee (the Committee) hearing, the Committee held that the physician acted reasonably because the DNR order was “clinically and ethically appropriate in the circumstance”. The Committee also concluded that the SDM was not in a position to demand that her father receive such treatment that his physicians did not judge to be medically appropriate.

The HPARB found this conclusion to be unreasonable—i.e., the Committee did not address whether it was appropriate to “place a DNR order on a patient’s chart” but rather to “execute it in light of the fact that the SDM did not consent to it”. The HPARB returned the matter to the CPSO Committee for reconsideration in accordance with their decision, in other words, to consider the physician’s actions taking into account the Health Care Consent Act (HCCA),3 CPSO policy, and hospital policies. The CPSO, as instructed, reconsidered the complaint and decided to take no action. The CPSO ruled that the physician used “good clinical judgment” because “extreme measures to preserve life should not be attempted in that they would only further exacerbate suffering” as death was inevitable. Furthermore, the ruling stated that the physician’s actions were indeed in compliance with existing policies.

The HPARB2 once again concluded that this was an unreasonable decision since the DNR order was made despite the offer of resuscitation and the “full code” instructions given by the SDM. The change occurred without prior discussion or consent and precluded any objection that the SDM may have had. The HPARB further ruled that CPSO policies and the hospital policy were inconsistent with the HCCA, which requires consent to withhold or withdraw treatments that have been offered and are being provided. In its ruling, the HPARB emphasized that the law must take priority over any policies.

By the time of this second HPARB decision, the Supreme Court of Canada’s ruling in Cuthbertson v. Rasouli had been released.4 The physician’s lawyer argued in E.J.W. that this case should not be considered because the judgment followed the incident and was therefore unavailable to guide the physician. The HPARB rejected this argument, holding that the case did not change the law but only confirmed it.

A few important points should be made before leaving the E.J.W. case. First, the patient was “full code” with the consent of his SDM as agreed with a resident physician and duly charted. In other words, resuscitation had already been offered, which presupposed that it fell within the standard of care for the patient and was therefore subject to consent. Under the HCCA, consent is required to withhold or withdraw treatment, and its wording does not distinguish whether the treatment it addresses falls outside the standard of care. The use of the word “withholding” presupposes that treatment is being offered—in other words, that it is within the standard of care. It is important, therefore, to be clear in the use of language and for physicians to specify that treatments are not being offered when these fall outside the standard of care.

The issue under review by the CPSO or the HPARB was not whether treatment was appropriately offered or even whether it was accurately explained in the context of the patient’s overall state of health. The CPSO and the HPARB effectively saw the DNR order as a change in the patient’s treatment plan, and that is why consent was required. Generally, there is a difference between “withholding or withdrawing” a treatment and, on the other hand, “not offering” to start treatment that falls outside the standard of care. The former (withholding or withdrawing) requires consent as these actions speak to treatments that fall within the standard of care. The latter (not offering) does not require consent as these treatments fall outside the standard of care. Nevertheless, the new CPSO policy negates this distinction. The CPSO still recognizes the standard of care, as the new policy states that CPR should not be offered under the following circumstances: 1) if restoring circulation would not be possible; 2) if circulation is restored and the patient would not survive the subsequent stay in the intensive care unit (ICU); 3) if, in the rare situation the patient survives, quality of life would be extremely poor (as determined by the patient or SDM); or 4) if there are no further treatment options to cure or stabilize the patient’s state of health.1 Yet, if a conflict arises when CPR isn’t being offered, CPR still has to be provided until such conflict is resolved. All “no-CPR” orders require consent, whether or not they represent a change from an existing treatment plan to provide CPR.1

The CPSO and “must do CPR”

In recent cases, the courts have placed great weight on consent and the right to self-determination as the key considerations in adjudicating some of the most important issues in peoples’ lives, such as withdrawal of LST and access to physician-assisted death.4,5 Even in situations where continuing LST would fall outside the medical standard of care, the Supreme Court has mandated that such treatments cannot be withdrawn in Ontario without consent.4 Likely this ruling will have implications across Canada. In response to E.J.W. v. M.G.C., the CPSO has stated that, in their view, the law is unclear regarding the consent requirements for no-CPR orders.1 The CPSO therefore states1:

A decision regarding a no-CPR order cannot be made unilaterally by the physician. Where a physician is of the opinion that CPR should not be provided for a patient and that a no-CPR order should be written in the patient’s record, the College requires physicians to discuss this with the patient and/or substitute decision-maker at the earliest and most appropriate opportunity, and to explain why CPR is not being proposed. This discussion must occur before a no-CPR order can be written.

If the patient or substitute decision-maker disagrees and insists that CPR be provided, physicians must engage in the conflict resolution process as outlined in Section 8 of this policy which may include an application to the Consent and Capacity Board. Physicians must allow the patient or substitute decision-maker a reasonable amount of time to disagree before a no-CPR order can be written.

While the conflict resolution process is underway, if an event requiring CPR occurs, physicians must provide CPR. In so doing, physicians must act in good faith and use their professional judgment to determine how long to continue providing CPR.

Before discussing the implications of this statement, it is essential to understand what the policy does not do. The policy does nothing to change the medical standard of care regarding when CPR should not be offered. Furthermore, it does nothing to change current standard of disclosure of such non-offers. This statement refers to the dispute resolution process which involves clear communication of diagnosis, prognosis, treatment options and an evaluation of these options, clarification of any misconceptions, explanation of the availability of support from social work, spiritual care (among other services), and the availability of palliative care resources.1 In addition, the CPSO recommends referral and facilitation of a second opinion and consultation with an ethicist or ethics committee as appropriate and available.1 If the dispute remains intractable, the physician is to take “reasonable steps” to transfer the patient to another facility as a last resort.1 The policy therefore does nothing to change existing dispute resolution processes.

Existing dispute resolution processes are limited

The problem is that these standard dispute resolution practices actually do little to solve conflicts that arise from an entrenched desire to seek cure and stabilization when such goals cannot scientifically be achieved. Physician or SDM requests to transfer patients only serve to confirm, through an external second opinion process, that the request for ongoing aggressive treatment falls outside of the standard of care. Based on our clinical experiences, patients in such situations are never actually accepted in transfer as there is usually consensus that transfer will not improve patient care.

The CPSO advises to seek legal advice where appropriate regarding mediation, adjudication, or arbitration processes, which would include the Consent and Capacity Board (CCB) since it places such issues under the rubric of consent rather than acknowledge the key role of the medical standard of care in such decisions. Decisions in the hospital setting about not offering CPR often need to be made within a relatively short period of time, since patients for whom CPR falls outside of the standard of care are usually seriously ill and their condition may deteriorate quickly. The literature has revealed that the CCB adjudication process can be cumbersome and take longer than portrayed in recent court rulings.6–8 The condition of an acutely ill patient with end-stage disease may very well deteriorate before the CCB meets and/or before it rules. Moreover, CCB rulings can be appealed to the Superior Court and subsequently to the Court of Appeal, and these appeals make take months to schedule. These timelines would be incompatible with the expedient clinical decision-making needed regarding CPR at the EOL. In fact, there has been less recourse to the CCB regarding even withdrawal of LSTs since the Rasouli ruling, with only three EOL cases heard in 2014-2015.9,10 The nearly inescapable result of this new policy is that more patients at the EOL will receive CPR unless something changes.

The CPSO could have taken a lead role in prospectively evaluating whether, in situations of conflict, CPR would actually still fall within the medical standard of care. This would have created a new approach and used consideration of both the standard of care and informed consent to fulfill the CPSO’s mandate of both setting the standard of care and protecting the public. What better body to provide guidance to physicians on such issues, particularly in cases of uncertainty? And who better to protect patient rights to access healthcare? Unfortunately, the emphasis on informed consent and the failure to give weight to the standard of care in the new policy will very likely result in more people being harmed by medicine—instead of being helped through quality EOL care.

Strategies for change: ensuring quality of EOL care

Understand the help-hurt line

The foremost principle for clinicians caring for patients in the end stages of illness is to try not to violate the line between helping and hurting—“first do no harm.” At the EOL, cure is not possible. The medical standard of care asks physicians to consider which, if any, treatment options can stabilize a patient’s state of health and well-being and which treatment options will help alleviate pain and distress. Cardiopulmonary resuscitation will do neither and should not be offered. Yet, if care is not taken in how this is communicated, such a harsh blunt message often leaves patients and SDMs feeling abandoned. Empathy and gentleness are needed, as is a focus on what actually can be done to help. Most people will always fear death. Such fears will never go away. They are protective in nature and deeply engraved into our psyche as part of our “fight or flight” response to any actual or perceived danger to survival. Everyone is therefore “a fighter”, as SDMs frequently convey to the healthcare teams. Messaging that death is a part of life only succeeds as an abstract concept that few people will grasp as applying to themselves or to someone they love. Nevertheless, the great majority of people would not insist on CPR if they understood that it would only hurt them and they would “live” (often only in the most primitive sense without awareness of self or others) to die another day.

Change CPR to an “opt-in” treatment: Reconsider the institutional presumption of “full code”

In view of the invasiveness of CPR and the need for subsequent LST and its implications for patient outcomes, consent to initiate CPR should be required in all hospital policies. The provision of CPR should no longer be an automatic default. Medicine has come a long way in its understanding of the potential outcomes and harms of CPR—to continue to view it as a default treatment from which all must “opt out” is indefensible in today’s day and age. Changing the presumption that every patient is “full code” on admission will require changes to hospital policies and procedures as well as ensuring that patients and their SDMs understand the changes. Nevertheless, it is possible to change the “full code” presumption in a manner that puts CPR more clearly in the category of “not offered” rather than in the category of “withheld or withdrawn”. Treatments “not offered” generally do not require consent because they fall outside the standard of care. Nevertheless, the CPSO policy in Ontario remains problematic—specifically, if conflicts arise once the non-offer of CPR is disclosed, CPR would still have to be provided pending conflict resolution. Such changes in hospital policies and physician practices to an opting-in approach to CPR would not be in vain. Quite the contrary, they remain crucial to facilitate the development of more realistic treatment goals and the framework needed for a better understanding of when medicine can help and when it can only hurt. As such, these changes could promote much better EOL care across hospitals in Canada.

Change the language used when discussing cardiac arrest once and for all

In current clinical practice, all physicians are expected to engage capable patients and SDMs in developing realistic plans for EOL care in the form of either advance care planning (where plans are made in advance of acute illness) or “goals-of-care” discussions (where plans are made in the setting of acute illness). Currently, some physicians are not sufficiently skilled in such conversations, and some do not accurately place such discussions within the context of a patient’s current situation. Vague questions—such as “Would you like everything done?”—fail to explain what “everything” entails and imply that the sole purpose of medicine is to keep a person “alive”. Such questions do nothing to inform decision-making about the probable outcomes of CPR and other LST in the patient’s context, and they frame CPR an “opt-out” treatment decision. Increasingly in clinical practice, people insist on being “full code”, yet they rarely understand what this entails in terms of either the resuscitation or its consequences. In such cases, language that makes “no CPR” seem like the alternative rather than the standard treatment reinforces this misperception.11

With the courts giving the weight to self-determination in their interpretation of the Ontario HCCA,3 the need to improve the discussion of feasible treatment goals and to clarify the role of the standard of care in such discussions is now more urgent than ever. If, in the context of any patient’s illness, CPR could not offer a future opportunity to cure or stabilize the medical condition, physicians should disclose that CPR will not be offered and provide an explanation. The difference in approach is the difference between a statement (disclosure of a non-offer) and a question (asking if the patient would want to be resuscitated). Framing CPR as an “opt-in” treatment appropriately emphasizes the minimal medical benefits (i.e., its inability to cure or even stabilize a patient at the EOL and its ability to cause rather than alleviate pain and distressing symptoms) it can provide and the significant harms (i.e., high risk of anoxic brain injury, multisystem organ failure, and death a few days later in the rare situations where return of spontaneous circulation [ROSC] is achieved) it will cause for patients at the EOL.

Understand the implications for perioperative and periprocedural patient care

Patients who are severely or critically ill and those approaching the EOL may need surgical or interventional radiology procedures that, in such circumstances, would generally carry higher risks than in young healthy patients. Surgeons, interventional radiologists, and anesthesiologists need to discuss the risk of death as part of the consent process for such interventions. If the surgery/procedure is very high risk and is performed urgently in a desperate attempt to save a life, then a cardiac arrest may be the final outcome of a fatal illness. Whether a patient undergoing surgery should be resuscitated in the event of a cardiac arrest needs to be carefully evaluated and discussed in collaboration with the patient, SDM, surgeon, and anesthesiologist. Ethically and legally, consent for CPR (an invasive aggressive treatment with a guarded prognosis, even if arrest occurs intraoperatively) and postoperative LST is required. A unilateral lifting of a “no-CPR” order (i.e., without discussing with patients or SDMs) is not ethically or legally justifiable since these events must be anticipated. Furthermore, in these situations, resuscitation cannot be considered an emergency for which consent may not be required.12 A failure to discuss CPR in the perioperative setting does not respect patient autonomy, violates the principles of beneficence and non-maleficence, and precludes informed consent.

Take a preventative approach: Engage the CCRT/ICU team early in goals-of-care discussions

Critical Care Rapid Response Teams exist in many hospitals in Canada. The team is interprofessional in nature with a critical care physician, nurse, and/or respiratory therapist. The role of the CCRT is to provide rapid assistance in stabilizing patients on the wards and/or to facilitate their admission to the ICU if needed in order to improve patient outcomes. In hospitals with no CCRT, ICU teams often fulfill a similar role. Critical Care Rapid Response Teams have increasingly been called on to facilitate discussions of non-offers of LST, including CPR.13,14 Since the CCRT and ICU teams provide LST, they are ideally best suited to discuss whether such treatments have any potential to help patients, particularly those near the EOL. Given the sheer volume of work involved and the amount of time required, it would be prohibitive for these teams to initiate proper discussions with all patients in the hospital. Nevertheless, hospitals would be well served to develop a preventative approach and identify select patient groups where the line between helping and hurting is narrow indeed—for instance, patients with progressive terminal illnesses such as cancer, severe dementia, or end-stage organ failure. Once the patient groups are identified, processes could be established that include the CCRT/ICU team in targeted discussions about LST and CPR. Decisions would be better informed and, hopefully, conflicts would be minimized. Palliative care teams are often highly skilled in facilitating these discussions. Ideally, palliative care should be involved early to manage symptoms and begin advance care planning. Then again, consulting palliative care prior to these discussions may risk causing distress, as patients and SDMs may wrongly perceive that the medical/surgical teams have prejudged the decision-making at hand. Decisions about involving palliative care depend on the clinical situation, patient needs, and the availability of resources.

Mentor interested staff outside of the ICU to improve their skills in discussing life-saving/LSTs

Hospitals should engage their ICU team to create a mentorship program for staff interested in improving their skills in placing the discussion of life-saving and LSTs in the context of the patient’s medical realities. This mentorship could include the development of a toolkit that builds on existing advance care planning initiatives. The results of existing advance care planning initiatives have not met people’s expectations simply because they fail to take into account the patient’s state of health and expected illness progression. They are far to general and abstract to be of much use in many clinical situations. Nevertheless, advance care planning can be helpful to patients and SDMs if skilled clinicians can convey what the proposed LSTs entail and the likelihood that the treatments will provide benefit in the patient’s situation. Mentorship programs could include the development of a toolkit that builds on existing advance care planning initiatives, simulation training with standardized patients and SDMs, and shared experience from skilled colleagues. Such training could be made available to teams who currently see the highest volume of patients in the end stages of illness, e.g., general internal medicine, general surgery, orthopedics, and emergency. Those interested in this area could then serve as resources for their colleagues. The joint efforts of skilled specialists in conjunction with CCRT/ICU teams could improve end-of-life decision-making by contextualizing such decisions within a realistic medical framework, thus ensuring that only beneficial treatments are offered to patients and SDMs.

Do NOT entertain the notion of slow codes

The CPSO’s policy1 states:

While the conflict resolution process is underway, if an event requiring CPR occurs, physicians must provide CPR. In so doing, physicians must act in good faith and use their professional judgment to determine how long to continue providing CPR.

The CPSO is clearly calling on physicians to follow standard Advanced Cardiac Life Support (ACLS) algorithms if a cardiac arrest ensues while the conflict is unresolved. In the past, there had been a resort to “slow codes” so that it would appear that some resuscitative efforts had been made; however, the efforts would clearly have been designed to be ineffective.15 Such an approach remains deceptive, unethical, and likely to cause greater harms than subjecting a patient to CPR in the first place. If ROSC is achieved, the resulting anoxic brain injury and multisystem organ failure may be even more devastating than if resuscitation were performed according to ACLS algorithms. Instead, clinicians should provide CPR in keeping with medical standards —recognizing that, in many cases (unwitnessed arrest, pulseless electrical activity, asystole, absence of reversible cause), prolonged efforts would not be medically appropriate.

Do not stop advocating for quality EOL care

The worst thing that physicians and healthcare teams can do in view of this new CPSO policy is to take the position that they “must do CPR”. It is true; it can be exhausting having these conversations with patients and SDMs about the use of LST and CPR at the EOL. Emotions run high. To maintain a therapeutic relationship, physicians need to align themselves with their patients and SDMs and make it clear that they want to care for their patient to the extent of their abilities. Physicians must clarify what can be done to help their patient as well as the very real limits that the patient and physicians are unfortunately confronting. Many disputes are caused by perceptions of lack of caring, judgmental attitudes, “being written off,” and not being heard or seen as a person and also by physicians not taking the time for gently explaining the limits of medical science in achieving cures.

Nevertheless, it is simply not good medicine to avoid challenging conversations or to offer and agree to do CPR blindly in response to the new policy. The medical standard of care for deciding when such treatments should not be offered has not changed, and clinicians should uphold those standards despite an overall reluctance to engage in more dispute resolution practices. The fundamental responsibility of all clinicians is to advocate for their patients so that they receive the highest quality of care possible throughout their entire life. It is never a good idea to allow CPR to happen knowing that it will only harm the patient without changing his/her imminent death and cause distress and burnout among the hospital staff.16,17 Clinicians need to continue to engage in these challenging discussions and conflict resolution practices.

By mandating CPR in cases of dispute over the standard of care and placing any disputes under the rubric of informed consent adjudication, the CPSO policy fails to protect patients at the EOL. Any ethical principle or legal ruling fails to provide adequate protection when it places the importance of consent above the medical standard of care. Consent, while very important, does not and cannot diminish or negate the importance of the standard of care in protecting human rights. Failure to acknowledge the importance of the standard of care diminishes medicine in many ways: 1) It fails to respect the importance of scientific knowledge and research in advancing treatments and care. 2) It negates the importance of the help-hurt line and the role of physicians in protecting and advocating for patients. 3) It shifts responsibility in decision-making from evaluating whether a treatment can help to stating that, if the patient consents, it’s acceptable to “treat” when to do so would solely harm the patient.

All healthcare providers and members of the public need to understand the forces that shape policies and legislation. Society has a collective responsibility and those who draft policy and legislation have an even greater responsibility to have a clear concept of the purpose of medicine in all of our lives—i.e., to help and not to hurt. The CPSO policy frustrates clinicians as it fails to balance the protection of autonomy against that of beneficence/non-maleficence.18 The CPSO policy reflects an interpretation of existing legislation and court rulings that reflect societal values. We suggest that this interpretation is perhaps a sign of a broader problem in our society—that of an overemphasis on rights and an underemphasis on protections.

Be proactive: Advocate for a more responsive legal adjudication process

As discussed earlier, the CCB process and the appeal process from the Board through the courts frequently takes too long. The intent of the legislation was to create an expedient adjudication process. The need to streamline the CCB and appeals process in these conflict situations is equally urgent as the need for physicians to upgrade their communication skills with patients and SDMs and the need for hospitals to revise their “opt-out full-code” policies. Physicians can play an important role in advocating for future change. Nevertheless, the fact that the CCB process can be lengthy and time-consuming does not obviate the obligation to go to the Board when intractable disputes arise between the treatment team and SDMs. The physician’s legal obligation under the HCCA is to obtain substitute consent “in accordance with this Act”. That includes the obligation to ensure that substitute consent accords to the principles set out in s. 21 of the HCCA.Footnote 2

Conclusion

The CPSO’s Planning for and Providing Quality End-of-Life Care is a better policy than that disseminated previously because it clarifies the roles and responsibilities of physicians and more accurately reflects legislation and common law. Nevertheless, it fails to promote quality EOL care effectively by placing the resolution of any conflicts regarding CPR under the rubric of informed consent rather than under the medical standard of care. Physicians, healthcare teams, and hospitals need to develop systems and strategies to improve discussions about LST and CPR, or adhering to the policy will result in many more people undergoing CPR when such treatments can offer no medical benefit.