withdrawing versus not offering cardiopulmonary resuscitation: is there a difference?/L'arrêt ou la non-administration de la réanimation cardiorespiratoire : y a-t-il une différence?

Headnote

SJW Oczkowski, B Rochwerg, C Sawchuk. Withdrawing versus not offering cardiopulmonary resuscitation: Is there a difference? Can Respir J 2015;22(1):20-22.

Conflict between substitute decision makers (SDMs) and health care providers in the intensive care unit is commonly related to goals of treatment at the end of life. Based on recent court decisions, even medical consensus that ongoing treatment is not clinically indicated cannot justify withdrawal of mechanical ventilation without consent from the SDM. Cardiopulmonary resuscitation (CPR), similar to mechanical ventilation, is a life-sustaining therapy that can result in disagreement between SDMs and clinicians. In contrast to mechanical ventilation, in cases for which CPR is judged by the medical team to not be clinically indicated, there is no explicit or case law in Canada that dictates that withholding/not offering of CPR requires the consent of SDMs. In such cases, physicians can ethically and legally not offer CPR, even against SDM or patient wishes. To ensure that nonclinically indicated CPR is not inappropriately performed, hospitals should consider developing 'scope of treatment' forms that make it clear that even if CPR is desired, the individual components of resuscitation to be offered, if any, will be dictated by the medical team's clinical assessment.

Key Words: Cardiopulmonary resuscitation; Life support; Medical ethics; Medical law; Palliative care

L'arrêt ou la non-administration de la réanimation cardiorespiratoire : y a-t-il une différence?

Les conflits entre les mandataires et les travailleurs de la santé à l'unité de soins intensifs sont souvent liés aux objectifs des traitements de fin de vie. D'après de récentes décisions judiciaires, un consensus médical quant à l'absence d'indications cliniques pour poursuivre un traitement ne suffit pas pour justifier l'arrêt de la ventilation mécanique sans le consentement du mandataire. La réanimation cardiorespiratoire (RCR), comme la ventilation mécanique, est un traitement de survie qui peut susciter un désaccord entre les mandataires et les cliniciens. Contrairement à la ventilation mécanique, dans les cas où l'équipe soignante ne la juge pas indiquée sur le plan clinique, aucune loi explicite ni aucune jurisprudence canadiennes ne dictent d'obtenir le consentement des mandataires avant d'arrêter la RCR ou de décider de ne pas l'administrer. Sur le plan éthique et juridique, les médecins peuvent donc arrêter la RCR ou ne pas l'administrer, malgré le désaccord des mandataires ou des patients. Afin de s'assurer de ne pas effectuer une RCR non indiquée sur le plan clinique, les hôpitaux devraient envisager de préparer des formulaires sur « la portée du traitement », afin d'établir clairement que, même si une RCR est souhaitée, les divers éléments de la réanimation qui seront entrepris, s'il y a lieu, dépendront de l'évaluation clinique de l'équipe soignante.

Situations in which individuals request life-saving interventions for their family members, despite perceived futility, make many physicians uncomfortable. What are doctors required to offer in these circumstances? Is there a difference between withdrawing versus not offering life-sustaining interventions? What are the legal precedents in such cases? The present article addresses these important issues, which all hospital-based clinicians must be comfortable discussing.

Conflict between substitute decision makers (SDMs) and health care professionals (HCPs) is commonly related to treatment at the end of life ( 1 ). SDMs often request for 'everything to be done', even when clinicians consider aggressive treatments to be of no benefit to the patient. Conflicts can usually be resolved through shared decision making, focusing on the patient's previously expressed wishes, values and best interests (2). If such efforts fail, a request for legal resolution may be made. Although often invoked, the term 'futility' has no clear legal definition, limiting its applicability (3).

The recent Supreme Court of Canada (SCC) decision, Cuthbertson v. Rasouli, states that clinicians must obtain consent to withdraw mechanical ventilation even when it is perceived by the medical team to be 'futile'. The SCC ruled that withdrawing mechanical ventilation requires consent under the Health Care Consent Act (HCCA) because it is 'treatment', something "...done for a therapeutic, preventive, palliative, diagnostic, cosmetic or other health-related purpose" (4). This ruling reinforces the distinction between withdrawing and withholding/not offering life-sustaining therapy, previously considered to be legally and ethically equivalent (5,6). The withdrawal of life-sustaining therapy is now clearly defined as a treatment requiring the consent of SDMs, whereas decisions about whether such therapies should be offered or withheld in the first place remains under the purview of physicians. The SCC's legal distinction between withdrawing and withholding has implications for forms of life-sustaining therapy other than mechanical ventilation, including chest compressions and defibrillation, jointly referred to as cardiopulmonary resuscitation (CPR). In the present article, we review the major clinical, ethical and legal considerations for withdrawing and withholding CPR.

THE LIMITS OF FUTILITY

Clinicians often invoke the concept of futility to justify limiting treatment. However, 'futility' is a contested term and, in its broadest sense, depends on the values and goals of individual patients (7,8). When HCPs refer to 'futility', they are often referring specifically to whether a treatment will provide a realistic chance of treating the medical condition for which it is intended. Rather than judgments about quality of life, the HCP's definition of 'futility' is, thus, based on the evidence for an intervention and the perceived applicability of that evidence to the patient at hand. For the remainder of the present article, we therefore refer to whether treatments are 'clinically indicated' rather than 'futile' to emphasize the role of medical judgement in such cases.

The SCC ruling in Cuthbertson v. Rasouli demonstrated that even consensus medical opinion - that ongoing life support is not clinically indicated - cannot legally justify withdrawal of mechanical ventilation without consent. This majority opinion was based on two grounds. First, the act of withdrawing ventilatory support fulfills a health-related purpose, namely ending life and is, therefore, a treatment requiring consent under the HCCA. Second, withdrawal of life support in anticipation of death implies the initiation of palliative care, which itself requires consent (4).

Although the absence of an ongoing clinical indication for mechanical ventilation cannot legally justify its withdrawal without consent, it remains a common rationale for not offering the treatment in the first place. Treatments are generally offered only if they are consistent with the current standard of care and can improve or maintain the patient's health. Such determinations are not a question of patient wishes or values, but a matter of what is clinically indicated for a given situation, although all capable patients have the right to opt out of any proposed treatments.

WHAT IS CLINICALLY INDICATED IN CARDIAC ARREST?

The American Heart Association guidelines state that in the event of cardiac arrest, CPR should be started unless there is an existing valid 'do not resuscitate' order (9). Given the potential life-saving benefits and time-sensitive nature of CPR, it has become the 'default' treatment in the event of cardiac arrest. This has led to the unique distinction that CPR, in contrast to other medical interventions, requires a medical order to not perform. Because consent is required for writing other medical orders, it appears to follow that the act of not offering CPR also requires consent from patients/SDMs. We suggest that this is not necessarily the case, specifically when CPR is judged not to be clinically indicated.

The decision to terminate CPR is integrally bound to clinical judgement. The American Heart Association guidelines state that the decision to terminate resuscitation efforts, once started

... rests with the (physician) and is based on consideration of many factors, including witnessed versus unwitnessed arrest, time to CPR, initial rhythm, time to defibrillation, co-morbidities, pre-arrest state, and whether there is (return of spontaneous circulation) (9).

By similar reasoning, physicians who have assessed a patient and the potential benefits and harms of CPR can also determine whether it would be clinically indicated in the event of cardiac arrest. This is consistent with the Canadian Medical Association's statement on life-saving and life-sustaining interventions, which states clinicians should not offer a treatment if they judge that it "offers no reasonable hope of recovery or improvement or because the person is permanently unable to experience any benefit" (5). For many intensive care unit patients, the benefits of CPR are minimal, with those requiring vasopressor support having an estimated survival rate of 6% and a chance of favourable neurological outcome <2% (10). For many critically ill patients, CPR is unlikely to provide benefit, is not clinically indicated and should not be offered.

ARE HCPs ETHICALLY OR LEGALLY OBLIGATED TO PROVIDE NONCLINICALLY INDICATED CPR?

Although the ethical principle of beneficence suggests that HCPs should do everything they can to help their patients, there are limits to those obligations. HCPs should encourage patient autonomy by eliciting the goals of treatment from patients/SDMs; however, such goals must be realistic and tempered by other ethical principles. The autonomous right of patients to self-determination does not include a right to coerce HCPs into offering care against their professional judgment. Offering nonclinically indicated treatments is contrary to the ethical principles of nonmaleficence (do no harm) and justice (the judicious allocation of resources). Withholding such treatments, even at patient request, has itself been considered a fiduciary obligation for caregivers (11). When CPR is unlikely to benefit a patient, withholding it is ethically sound practice.

There is no explicit law in Canada dictating that withholding CPR requires the consent of SDMs (12). Importantly, in the Cuthbertson v. Rasouli ruling, the SCC clarifies that consent is not required for all forms of treatment withdrawal, otherwise:

...patients could arguably compel the continuation of any treatment, regardless of its medical implications... common sense suggests that many withdrawals of treatment... must be excluded from the definition of 'treatment' under the HCCA (4).

We argue that CPR fails to fulfill the two criteria that mandate consent for withdrawal. First, unlike the withdrawal of mechanical ventilation, it does not involve an act performed for a health-related purpose. A patient in cardiac arrest is, essentially, dead and would remain so without intervention. The absence of CPR to reverse that process cannot reasonably be defined to constitute a treatment. Second, unlike the withdrawal of mechanical ventilation, withholding CPR is not integrally linked to the need for palliative care. Cardiac arrest patients experience no demonstrable suffering and, therefore, do not require enactment of palliative measures. The need for consent, thus, never arises (as noted by the Canadian Medical Association's statement on life-saving and-sustaining interventions, the withholding of life sustaining measures such as CPR should not preclude the offering of other treatments, including palliative care [5]).

This legal reasoning was supported in the case of Cefarelli v. Hamilton Health Sciences (13). The treating physician determined that a patient was unlikely to benefit from CPR in the event of arrest, and placed a medical order for 'no CPR' in the patient's chart, despite the presence of an advanced directive form stating the patient was 'full code'. The SDM disagreed with the order and sought to rescind it, filing for a court injunction, which was heard at the Ontario Superior Court and, subsequently, at the Court of Appeals. The final ruling stated that:

... the (treatment plan) (which was consented to) gives the responsible physician discretion regarding which components of (CPR) are to be used... the contested order was simply one available to the doctor within that plan. It cannot be said to be a withdrawal of treatment from that treatment plan. No question therefore arises of the need for consent (pertaining to CPR) (12).

This ruling supports that withholding CPR does not require consent because it is not a question of withdrawal of treatment but rather of not offering treatment - and determining which treatments are to be offered is a matter of medical judgment.

IMPLICATIONS FOR PHYSICIAN ORDERS AND DO NOT RESUSCITATE STATUS

Under this interpretation, 'full code' represents a patient's request for a physician to use any clinically indicated medical intervention(s) to save that patient's life. It does not specify which elements of resuscitation must be used. 'Full code' status does not mandate chest compressions, defibrillation or open cardiac massage, unless such interventions are clinically indicated. Advanced directives provide an opportunity for patients/SDMs to set limits on the interventions to be provided. They are a form of negative consent, necessary for CPR because of its status as a 'default' treatment option during cardiac arrest. Institutional advanced directives forms should make it clear that even if 'full code' status is desired, the decision to use individual elements of the resuscitation, including CPR, is up to physician judgement. For example, in our institution's 'Physician Ordered Scope of Treatment' form, the option for 'full code' is worded as: "Provide CPR as clinically indicated". This assists HCPs in working toward the stated goals of patients and SDMs, but without mandating the use of specific treatments that may not be clinically indicated.

IMPLICATIONS FOR CLINICIANS

Physicians need be clear with patients and SDMs about which elements of resuscitation are not clinically indicated and will not be offered. Honesty and realistic expectations are the key to a healthy therapeutic alliance that will maximize patient outcomes. Judgements about whether CPR is clinically indicated can be complex due to considerations of a patient's comorbidites and prognosis, alternative treatments and local practice patterns. A second medical opinion should be sought early to corroborate the assessment if there is clinical uncertainty. If there is consensus that CPR is not clinically indicated, an order can be written in the chart indicating that it is not to be performed. This should be periodically re-evaluated, especially if the patient's condition changes.

LIMITATIONS

Although the ethical imperative of clinicians to avoid providing non-clinically indicated CPR applies across Canada, the legality of such decisions is based on regional policy and law. The recently updated Canadian Medical Association statement supports the withholding of nonindicated treatments, including CPR (5). Along lines similar to Ceferelli v. Hamilton Health Sciences, courts in jurisdictions outside of Ontario have also supported withholding (as opposed to withdrawing) life-sustaining treatments, including British Columbia (14) and Manitoba (15).

CONCLUSIONS

In light of the SCC's Cuthbertson v. Rasouli decision, the distinction between withdrawing and not offering a medical treatment is increasingly relevant. Because CPR is a 'default' treatment for cardiac arrest, it requires a physician order to be withheld. However, the individual components of resuscitation that are clinically indicated are based on physician judgment and best practices. A physician who has assessed a patient and does not believe CPR to be clinically indicated can write an order to withhold from CPR without patient/SDM consent. This is an ethically defensible course of action and has been legally tested in Ontario. Institutions should word their advanced directives forms to make it clear that even if CPR is desired by a patient or SDM, the individual components of resuscitation to be used, if any, will be limited by what is clinically indicated rather than patient/ SDM preference alone.

ACKNOWLEDGEMENTS: SO and BR conceived the article and wrote the manuscript. CS conceived and reviewed the manuscript. The authors thank Cynthia Clark, Partner, Borden Ladner Gervais LLP for her review and feedback on the manuscript.

References

REFERENCES

1. Danjoux Meth N, Lawless B, Hawryluck L. Conflicts in the ICU: Perspectives of administrators and clinicians. Intens Care Med 2009;35:2068-77.

2. Joosten EA, DeFuentes-Merillas L, de Weert GH, Sensky T, van der Staak CP, de Jong CA. Systematic review of the effects of shared decision-making on patient satisfaction, treatment adherence and health status. Psychother Psychosom 2008;77:219-26.

3. Wilkinson DJ, Savulescu J. Knowing when to stop: Futility in the ICU. Current Opin Anaesthesiol 2011;24:160-5.

4. Cuthbertson v. Rasouli. SCC 53 ed, 2013.

5. CM A Statement on life-saving and -sustaining interventions. 2014. <www.cma.ca/index.php/ci_id/33236/la_id/1.htm.> (Accessed August 9, 2014).

6. Opinion 2.20 - Withholding or Withdrawing Life-Sustaining Medical Treatment. American Medical Association, 1994. <www. ama-assn ,org//ama/pub/phy sician-resour ces/med ica 1-eth ics/code-medical-ethics/opinion220.page.> (Accessed December 30, 2013).

7. Bernat JL. Medical futility: Definition, determination, and disputes in critical care. Neurocrit Care 2005;2:198-205.

8. Schneiderman LJ. Defining medical futility and improving medical care. J Bioethical Inquiry 2011;8:123-31.

9. Morrison LJ, Kierzek G, Diekema DS, et al. Part 3: Ethics: 2010 American Heart Association Guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2010;122:S665-75.

10. Kutsogiannis DJ, Bagshaw SM, Laing B, Brindley PG. Predictors of survival after cardiac or respiratory arrest in critical care units. CMAJ 2011;183:1589-95.

11. Hafemeister TL, Gulbrandsen RM. The fiduciary obligation of physicians to "just say no" if an "informed" patient demands services that are not medically indicated. Seton Hall Law Rev 2009;39:335-86.

12. Handelman M. Consent to Withdrawal of Life Support: What the Supreme Court Said In Cuthbertson and Rubenfeld v. Rasouli.: Whaley Estate Litigation; 2012.

13. Cefarelli v. Hamilton Health Sciences. 2013, ONCA 413.

14. Rotaru v. Vancouver General Hospital Intensive Care Unit. 2008 BCSC 318.

15. Child & Family Services of Central Manitoba v. L. (R.) 1997 154 D.L.R (4th) 409 Man. R. (2d) 135 (C.A.)