Patient and doctor attitudes and beliefs concerning perioperative do not resuscitate orders: anesthesiologists’ growing compliance with patient autonomy and self determination guidelines

It is reported that up to 15% of patients undergoing surgery have pre-existing DNR orders on record. First described in 1960 to treat witnessed intraoperative events, closed cardiac massage became increasingly utilized to treat all patients suffering cardiac arrest [1–3]. In 1983, the President’s Commission for the Study of Ethical Problems in Medicine suggested favoring implicit consent to CPR, thereby requiring an order for resuscitation to be withheld [1]. In 1988, the Joint Commission on Accreditation of Healthcare Organizations required that hospitals develop policies on resuscitative efforts [3, 4]. Following passage of the Patient Self-Determination Act in 1991, hospitals receiving Medicare funds were required to provide patients with written information regarding their right to make health care decisions, including advance directives [5, 6].

The surgical setting has been described as “the last bastion of resistance to acceptance of DNR orders [7].” The operating room (OR) environment offers challenges to following pre-existing DNR orders [3, 8]. OR personnel - particularly anesthesia providers - are trained and equipped to provide resuscitation efforts. Limiting this skill causes discomfort to some providers [1, 9]. Furthermore, resuscitation in this monitored and controlled environment is often more successful and providers may find it difficult not to intervene while a patient physiologically deteriorates [5, 10–15].

Despite potentially conflicting goals and interests, decisions involving perioperative resuscitation remain largely patient choices. In 1993, the American Society of Anesthesiologists (ASA) published guidelines stating that automatic suspension of DNR orders infringed upon a patient’s right to self-determination [16]. The 2001 ASA guidelines (reaffirmed in 2008) recapitulated this concern [17]. The American College of Surgeons and the Association of Operating Room Nurses have expressed similar sentiments [16, 18, 19].

Given that several professional guidelines have been in place for nearly 20 years, we explored both patient and doctor views concerning perioperative DNR status.

A majority of patients in our study had previously made a decision regarding perioperative resuscitation, with just over half confirming this in writing. Thus, a substantial number of anesthesiologists are expected to provide care to patients who have pre-existing DNR orders.

Nearly one-third (30%) of doctors in our study stated that they would automatically suspend a patient’s DNR order - lower than the 43% reported elsewhere in 1994 [20]. In a previous analysis, 60% of anesthesiologists, 37% of surgeons and 34% of internists assumed DNR suspension in the perioperative period [20]. Our findings suggest anesthesiologists’ views and practices have changed over 18 years, perhaps due to increased awareness of the published guidelines. A minority of anesthesiologists (18%) would automatically suspend a patient’s DNR order, whereas the rates of suspension by surgeons (38%) and internists (34%) in our cohort were similar to previous reports [20].

Responding to a fictional patient scenario designed to highlight the difficulties associated with preoperative DNR orders 53% stated they were unlikely to follow the order. A majority (64%) of those who would not follow the DNR order considered the non-contemporaneous nature of the DNR order to be of importance. Nevertheless, patients do not renounce their right to decide their care merely because they are entering an operating room [5]. Patients may prefer to have their DNR status retained for several reasons, including concerns regarding a worsened state following successful resuscitation, and that expiring under anesthesia is a more peaceful death [5].

Many doctors were concerned with risk of liability in either following or forgoing a patient’s DNR order during surgery, although risk consideration did not seem to affect decision-making. Others have suggested that a patient’s long- established right to refuse medical care, state statutes reflecting this right, and the small number of cases finding providers liable for following DNR orders contribute to limited risk for liability involving DNR orders during the perioperative period [21]. It should reassure doctors practicing in the United States that rarely will following a patient’s request for refusal of aggressive care given a properly addressed DNR order result in liability [16, 21]. Most cases have involved conflicts with patient or surrogate informed consent to forego resuscitation [21]. However, when resuscitation occurs despite a DNR order, courts have been resistant to provide damages for continued or “wrongful life” [5, 21].

The vast majority of doctors surveyed who would (86%) and would not (93%) comply with the patient’s DNR request were concerned that a pre-operative discussion about DNR status under anesthesia had not occurred. This level is higher than reported previously [22, 23] where only approximately half of anesthesiologists in those studies stated they would discuss automatic DNR suspension with patients. In accordance with Joint Commission policy, our institution requires doctor-patient discussion of DNR orders prior to surgery [24]. One possible outcome includes continuation of DNR status. Given competing interests that accompany a patient with a DNR order to the OR, it is suggested that an automatic reconsideration of patient DNR status take place prior to surgical procedures [1, 5, 20, 21]. An open discussion between patient and doctor may enable patients to better appreciate the likelihood of needing intraoperative resuscitation. Furthermore, this “required consideration” may better enable the health care providers to understand the patient’s wishes with regard to resuscitation efforts [1, 5]. The majority of providers (77%) in our study believed patients can appreciate the idiosyncrasies involved with OR care and make decisions regarding resuscitation efforts. Furthermore, an equal percentage agreed that patients who have sufficient capacity to consent to surgery also have the capacity to refuse or agree to resuscitation efforts during the perioperative period. Our results reflect the importance of provider-patient discussions in informing patients about outcome of intraoperative resuscitation efforts while ensuring patient autonomy in decision-making.

Perhaps reflecting the continued frustration that doctors may feel when confronted with a patient who has a pre-existing DNR order is that 55% of those in our survey considered it illogical for a patient to undergo anesthesia in the presence of a DNR order. Commonly used anesthetic medications would cause death were it not for the anesthesiologist’s interventions. Doctors may believe that perioperative DNR orders limit providers from “sav[ing] their patients [5].” Seventy percent of those we surveyed agreed that anesthesiologists should be permitted to use their skills to provide best outcome for patients regardless of DNR status perhaps reflecting the increased likelihood of successful resuscitation in the OR [3, 5, 11–16]. Almost half (47%) of those doctors we surveyed agreed that DNR status should be suspended due to the increased likelihood of successful resuscitation in the OR environment. Although our survey did not assess whether doctors would forego offering surgical care to patients who refused to suspend their DNR orders perioperatively, a recent study reported that 54% of surgeons queried would refuse to operate on patients whose directives placed limits on postoperative care [25]. Since it may at times be difficult to distinguish between cardiopulmonary arrest and iatrogenic (anesthesia-related) depression of vital organs, some providers find it difficult to follow a patient’s DNR status intraoperatively [5, 9]. However, an equal number of those doctors we surveyed agreed as disagreed with this concern.

Provider uncertainly and confusion when caring for patients with pre-existing DNR orders may be mitigated through better appreciation for patient expectations concerning their resuscitative care. Over half (57%) of patients felt that preoperative DNR orders should be suspended during anesthesia for an elective surgical case. Despite this, an overwhelming majority (92%) expected doctors to discuss their requests not to be resuscitated. These findings are similar to those reported by Clemency et al. when interviewing terminally ill patients with DNR orders [26]. Two matters appeared to be important to those surveyed by Clemency - “being ready to die” and limiting financial and emotional burdens to themselves and family members [26]. Upon learning how anesthesia care is provided and managed, some in Clemency’s study acknowledged that this was a different circumstance and temporarily suspended their DNR orders [26].

Over half (56%) of those patients we surveyed agreed that the type of surgery should influence whether DNR orders remained during the perioperative period and a large majority (79%) expected that if orders were suspended, they should be reinstated at a predetermined point postoperatively. These results are consistent with the work of others [26]. Our finding further reflect the importance of provider-patient communication given that patients may have different expectations for how their pre-existing DNR orders will be managed during the perioperative period.

The limitations of our study must be recognized. Our respondents were from a single center. Although our institution has a large referral base which is geographically and socio-economically diverse, whether the results can be extrapolated to other patient populations is uncertain. Tertiary referral centers such as ours employ doctors from diverse medical, geographical and cultural backgrounds. It is therefore possible that the diverse staff background may lead to a different mix of social, religious and ethical beliefs than would be seen in other medical centers, potentially limiting the external validity of the results. The patient survey response rate was excellent but the doctor response rate was suboptimal – especially from internal medicine physicians. Nonetheless we obtained a sample of almost 400 doctors. The differences between specialists may reflect a type II error because of non-comparable response rates between physician types. A real difference, assuming it exists, might also be a reflection of differences in awareness of the existence and content of perioperative DNR guidelines from specialist organizations rather than opinions formed on the basis of an individual’s specialty. Our survey assessed the importance patients and doctors place on communication, but we did not gauge the level of respondent appreciation for the necessity of both a thorough understanding of the resuscitative process and the need for bidirectional communication. Because “partial” resuscitation (e.g. chest compressions but no defibrillation) is not consistent with our medical center’s written DNR policy, and because DNR is compatible with maximal medical intervention, we did not ask our patients which resuscitative services would be acceptable in the OR. This policy may further limit the applicability of our results to other institutional practices. Our study did not ask patients to provide information concerning the extent of their upcoming surgical procedure or overall level of health. Both these factors may influence the desire to discuss resuscitation. Although our survey found that 54% of patients reported that they had previously documented their decision on resuscitation, the type of written form (DNR orders, power or attorney, living will) this took is unknown and unverified. Finally, because our study queried patients and providers from a single United States medical center, our results may not be applicable to other health care settings across the globe. Beliefs about the primacy of patient autonomy in the U.S and the legal system that helps to protect this right may be quite different from what exists elsewhere.

In summary, active participation on the part of anesthesiologists, surgeons and internists will help ensure that the underlying aims of patients’ DNR orders are fulfilled during the perioperative period [26]. Although many of the patients we studied may agree that their DNR orders should be suspended perioperatively, this assumption is not universal and should not be presumed. The patients studied expected to be informed of how resuscitation efforts under anesthesia are, and are not, similar to those outside the OR environment. In this study anesthesiologists (versus surgeons and internists) were more likely to report that DNR orders should not be automatically suspended in the perioperative period.