Frequency of Preoperative Advance Care Planning for Older Adults Undergoing High-risk Surgery: A Secondary Analysis of a Randomized Clinical Trial

Key Points

Question  Do older adults undergoing high-risk surgery receive preoperative discussion or documentation of advance directives (ADs) or advance care planning?

Findings  Among 213 patients with at least 1 comorbid condition who were included in this exploratory analysis of data collected in a randomized clinical trial testing a patient-mediated intervention to improve preoperative communication, only 13 preoperative visits included discussion related to ADs, the patient’s health care proxy, or preferences for treatment limitations. One hundred forty-one patients (66%) did not have an AD on file before major surgery.

Meaning  Despite the consensus that preoperative advance care planning is important for older adults undergoing high-risk surgery, these findings suggest that it is not routinely performed.

Importance  For patients facing major surgery, surgeons believe preoperative advance care planning (ACP) is valuable and routinely performed. How often preoperative ACP occurs is unknown.

Objective  To quantify the frequency of preoperative ACP discussion and documentation for older adults undergoing major surgery.

Design, Setting, and Participants  This secondary analysis of data from a multisite randomized clinical trial testing the effects of a question prompt list intervention on preoperative communication for older adults considering major surgery was performed at 5 US academic medical centers. Participants included surgeons who routinely perform high-risk surgery and patients 60 years or older with at least 1 comorbidity and an oncological or vascular (cardiac, peripheral, or neurovascular) problem. Data were collected from June 1, 2016, to November 30, 2018.

Interventions  Patients received a question prompt list brochure with 11 questions that they might ask their surgeon.

Main Outcomes and Measures  For patients who had major surgery, any statement related to ACP from the surgeon, patient, or family member during the audiorecorded preoperative consultation was counted. The presence of a written advance directive (AD) in the medical record at the time of the initial consultation or added preoperatively was recorded. Open-ended interviews with patients who experienced postoperative complications and family members were conducted.

Results  Among preoperative consultations with 213 patients (122 men [57%]; mean [SD] age, 72 [7] years), only 13 conversations had any discussion of ACP. In this cohort of older patients with at least 1 comorbid condition, 141 (66%) did not have an AD on file before major surgery; there was no significant association between the presence of an AD and patient age (60-69 years, 26 [31%]; 70-79 years, 31 [33%]; ≥80 years, 15 [42%]; P = .55), number of comorbidities (1, 35 [32%]; 2, 18 [33%]; ≥3, 19 [40%]; P = .62), or type of procedure (oncological, 53 [32%]; vascular, 19 [42%]; P = .22). There was no difference in preoperative communication about ACP or documentation of an AD for patients who were mailed a question prompt list brochure (intervention, 38 [35%]; usual care, 34 [33%]; P = .77). Patients with complications were enthusiastic about ACP but did not think it was important to discuss their preferences for life-sustaining treatments with their surgeon preoperatively.

Conclusions and Relevance  Although surgeons believe that preoperative discussion of patient preferences for postoperative life-sustaining treatments is important, these preferences are infrequently explored, addressed, or documented preoperatively.

Trial Registration  ClinicalTrials.gov Identifier: NCT02623335

Patients trust their surgeon to make treatment decisions on their behalf yet often have strong preferences about prolonged life-sustaining treatments.¹ This discrepancy can cause conflict among clinicians, patients, and families.^2,3 Routine preoperative discussion of preferences for life-sustaining treatments allows patients to discuss and document the care they would not want if they became too sick to speak for themselves. This discussion can mitigate conflict and help the surgeon and family know how to proceed if complications arise or major changes in cognitive or functional status occur. Moreover, preoperative advance care planning (ACP) may reduce unwanted treatment at the end of life and emotional distress during decisions to withdraw life-sustaining treatment.

In a national survey of cardiothoracic and vascular surgeons and neurosurgeons,⁴ 80% of respondents reported routine discussion of patient preferences for life-sustaining treatments, and 52% reported asking patients about advance directives (ADs) preoperatively. However, the frequency with which preoperative ACP occurs is unclear. One single-center study⁵ showed that 26% of older adults undergoing high-risk surgery had preoperative documentation of an AD after an automated prompt alerted clinicians to patients 65 years or older who lacked documentation. Discordance between reported practice and evidence about the frequency of preoperative ACP raises concerns that surgeons caring for older patients facing major surgery have overlooked an important component of preoperative communication.

The objective of this study was to quantify the frequency of preoperative ACP discussion and documentation for older patients undergoing major surgery in a national sample. We also aimed to characterize how patients and their family members considered ACP after postoperative complications.

We performed an exploratory analysis of data from a multisite randomized clinical trial testing a question prompt list intervention designed to improve preoperative communication about treatment options, postoperative expectations, and serious complications (eMethods 1 in the Supplement). This study showed that the question prompt list intervention had no effect on patients asking questions about options, expectations, or risks.⁶ The 5 study sites targeted distinct US geographical areas and comprised the University of Wisconsin Hospital and Clinics (UWHC), Madison; the University of California, San Francisco, Medical Center (UCSF); Oregon Health & Science University (OSHU), Portland; the University Hospital of Rutgers New Jersey Medical School (Rutgers), Newark; and the Brigham and Women’s Hospital (BWH), Boston, Massachusetts. This study was approved by the institutional review board at each site and was registered at ClinicalTrials.gov. All participants provided written informed consent. Patients received $55 and family members received $35 for completing all study-related procedures. Surgeons did not receive incentives. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

We enrolled 40 surgeons who perform oncological or vascular (cardiac, peripheral, or neurovascular) surgery in older patients. Using purposeful sampling to achieve variation in surgeon specialty, we randomly selected 1 or 2 surgeons within a particular specialty at each institution. We excluded surgeons who exclusively perform low-risk procedures.

We enrolled patients 60 years or older with at least 1 comorbidity and an oncological or vascular problem that could be treated with high-risk surgery by a study-enrolled surgeon. We defined high-risk surgery for older patients as operations having a 30-day in-hospital mortality rate of 1% or greater.⁷ We invited 1 family member per patient to participate. We excluded people who lacked decision-making capacity or proficiency in English or Spanish. We excluded study-enrolled patients who did not have surgery because they would not need preoperative ACP.

We audiorecorded 1 conversation among the surgeon, patient, and family member during the primary decision-making conversation, as indicated by the surgeon. We recorded postoperative treatments received and complications that occurred within 6 weeks of surgery. We also performed open-ended interviews roughly 6 weeks after surgery with a subset of patients who experienced significant postoperative complications, including prolonged hospitalization (>8 days) or stay in the intensive care unit (>3 days).^8,9 Interview questions focused on treatment decisions, postoperative experiences, and interpersonal relationships among patients, families, and clinicians and between clinicians (eMethods 2 in the Supplement). Patients and family members were specifically asked whether the patient had an AD and how they had anticipated using it postoperatively. Data were collected from June 1, 2016, to November 30, 2018.

One of 6 coders (including A.S.B. and M.L.S.) reviewed all transcripts from the audiorecorded consultation to identify statements associated with ACP from the surgeon, patient, or family. This includes any mention of an AD, health care power of attorney/proxy, or preference for limitations of life-sustaining treatments. Two investigators (E.L.K. and A.K.) reviewed all flagged transcripts to confirm the presence of any ACP discussion. Using a standardized medical record abstraction form, we recorded any written AD, including do-not-resuscitate orders, and noted whether it was present at the time of initial consultation, filed between the initial consultation and the date of surgery, or added postoperatively.

We used descriptive statistics to describe the frequency of preoperative discussions and documentation of ADs. We used the Fisher exact test to test the association of patient factors, hospital site, and type of procedure with AD documentation before surgery. We used R, version 3.6.2 (R Project for Statistical Computing) for all quantitative analyses and considered a 2-sided α < .05 to indicate statistical significance.

We used qualitative content analysis to analyze interviews with family members and patients.¹⁰ Each interview transcript was independently reviewed and analyzed using an inductive approach to generate codes to describe key concepts in the text. We used NVivo software, version 11 (QST International–Melbourne), to catalog codes and group observations from a team of 3 investigators (A.K., A.S.B., and M.L.S.) who met to adjudicate each transcript and recorded consensus and dissent about codes. To facilitate higher-level analysis, we generated construct tables to organize concepts, identify relationships, and further characterize content related to ACP and use of life-sustaining treatments postoperatively. We used data triangulation to identify connections between preoperative and postoperative communication.

Of the 446 patients enrolled in the primary study, 213 patients underwent major surgery (122 men [57%] and 91 women [43%]; mean [SD] age, 72 [7] years) (Figure). Of these, 109 patients (51%) were in the question prompt list intervention group. Patients ranged in age from 60 to 90 years. One hundred two patients (48%) had 2 or more comorbid conditions and 172 (81%) had a family member present at the surgical consultation (Table 1). One hundred sixty-eight procedures (79%) were oncological (eTable in the Supplement). Twenty-four patients had serious postoperative complications, and there were 5 deaths. Fifty-one patients were eligible for a postoperative interview, and we completed 20 interviews (Table 2).

Thirteen of 213 preoperative visits included any discussion related to ADs, the patient’s health care proxy, or preferences for treatment limitations. Of these 13 visits, 2 patients had preexisting ADs and 3 patients had subsequent documentation of ADs preoperatively. Given the low number of events, we could not determine whether there was a significant difference between patients in the question prompt list intervention group (n = 9) and those who had usual care (n = 4). Of these 13 visits, 4 remarks about ACP were initiated by patients, and the remaining 9 remarks were initiated by 3 surgeons. Three patients asked about appointing a decision-maker or informed the surgeon of an established health care power of attorney. One patient raised concerns about treatment limitations: “I’m gonna fill out a form which outlines my desires for care . . . I do want CPR [cardiopulmonary resuscitation] if the situation warrants and . . . I want limited care. I don’t want to go on a respirator though. I don’t want heroic measures.” The surgeon affirmed these wishes would be followed but did not explore the inconsistency of a request for cardiopulmonary resuscitation without intubation or how to manage a serious complication when use of life-sustaining treatments might be temporary.

Remarks initiated by surgeons were specifically about the existence or creation of an AD (eg, “this [educational material] is a good booklet to go through”) and formal documentation of a decision-maker (eg, “but if you were to get really sick during the recovery period in some way . . ., do you have assigned 1 person to help make decisions for you? Like health care power of attorney, and you have the paperwork and stuff all ready”). These surgeons initiated conversations about ACP but did not investigate whether patients had preferences to limit treatments potentially needed after surgery, such as mechanical ventilation, or explore health states patients would prefer to avoid, such as losses in cognitive or physical function.

One hundred forty-one patients (66%) in this cohort of older patients with 1 or more comorbid conditions did not have an AD on file before having major surgery. Of 72 patients who had documentation of an AD in the medical record preoperatively, 28 had an AD in place at the time of the initial consultation, whereas 44 had an AD added between their preoperative visit and the date of surgery.

On bivariate analysis, there was no association between study arm (QPL 38 of 109 [35%] vs usual care 34 of 104 [33%]), patient age (60-69 years, 26 of 83 [31%]; 70-79 years, 31 of 94 [33%]; ≥80 years, 15 of 36 [42%]), sex (44 of 122 men [36%] and 28 of 91 women [31%]), or race/ethnicity (White, 61 of 177 [34%]; Black/African American, 1 of 11 [9%]; Asian/Pacific Islander, 2 of 8 [25%]; other, 8 of 18 [44%]; Hispanic/Latino, 2 of 11 [18%]) and the presence of an AD in the medical record preoperatively (Table 3). There was no association between the type of surgery or the number of patient comorbidities and documentation of an AD. We did find a significant association between study site and the preoperative presence of an AD. Patients at BWH (20 of 40 [50%]) and UWHC (27 of 61 [44%]) were more likely to have documentation compared with OHSU (7 of 42 [17%]), Rutgers (1 of 22 [5%]), and UCSF (17 of 48 [35%]) (P < .001). Of the 27 patients at UWHC who had documentation of an AD preoperatively, more than half had it in the medical record at the time of the initial consultation. In contrast, at BWH these documents were added between the surgical visit and surgery.

None of the participants in postoperative interviews had a preoperative discussion with their surgeon about ADs. Three patients had an AD in the medical record at the time of the initial surgical consultation, 6 patients had an AD added between their preoperative visit and surgery, and 2 patients had an AD added postoperatively.

Patients and families reported feeling unprepared for serious complications. During surgical consultation, they focused on details related to their disease and treatment; the risks of surgery seemed inconsequential. Some anticipated the outcomes of surgery would be binary (“either you make it, or you don’t”). They did not consider that complications could affect their life or their family members. They noted that it was difficult to translate a list of discrete complications, such as bleeding or heart attack, into a vision of what their life might be like after a serious complication. Looking back at their preoperative conversations, we found wide variation in how surgeons described risks or unwanted outcomes. Patients whose surgeons described complications extensively or provided a long list of risks still reported feeling blindsided when complications occurred.

Respondents viewed ACP favorably and reported having discussed or documented treatment preferences with other clinicians or family in the past (Table 4): “I insisted that we, for whatever reason, decided to have the power of attorney for health care and so we got that taken care of and I am glad we did.” However, they had varied interpretations of ACP. Some respondents misunderstood the distinction between a living will and a legal trust conferring distribution of property. Despite this, respondents consistently expressed a desire to have an AD, power of attorney, or preferences for care documented. They endorsed the documents as an important tool for families to make decisions on their behalf, relieving their loved ones of guilt from having to guess their end-of-life preferences. For example, “If I would have had complications, my wife would have had to deal with it . . . [so] we talked about DNR [do not resuscitate] and basically the directive going into this last operation . . . was if I’m in a vegetative state . . . then it’s DNR.”

Respondents did not consider how their desire to limit life-sustaining treatments could change perioperatively or how events might occur for which their AD was ambiguous or contradictory to their goals for surgery. Instead, they saw their AD as an undisputable, protective document that could be used when needed to ensure that their preferences would be followed. They held these preferences for treatment limitations as personal and private, not something to discuss with their surgeon. Instead, they believed these documents could be presented in-the-moment to their surgeon, who would follow a directive to cease life-sustaining treatments. Respondents were confident that an undisclosed AD would ensure they would receive care according to their wishes: “I had an advance directive lined up but didn’t have it on file at the hospital . . . if it was needed, they [family] would show it to the surgeon.”

In this multisite study of 40 surgeons across the US who perform major oncological or vascular surgery, we found that preoperative ACP rarely occurs. More than two-thirds of patients 60 years and older with comorbid conditions did not have an AD on file before undergoing high-risk surgery. This makes it challenging for surgeons and families to navigate the difficult terrain around the use of postoperative life-sustaining treatments, unwanted outcomes, and goals of surgery when patients can no longer speak for themselves. Surgeons agree that patient preferences for limitations of life-sustaining treatments are important to discuss, but there is a vast disconnect between the stated value of this communication and how often it occurs. Although surgeons believe they are having these conversations preoperatively,⁴ they are not. Compounding this gap, patients may be unlikely to share these documents with their surgeon because they view their AD as personal or not relevant to surgical care. This study sheds new light on an important source of postoperative conflict about the use of life-sustaining treatments and has significant implications for surgeons, patients, and their families.

For surgeons who perform high-risk procedures, particularly in older adults who are vulnerable to serious complications,^11,12 this is an opportunity to reflect on past events and reevaluate whether preoperative consultation could be dispositive for directing postoperative care. Surgeons describe the need for surgical buy-in—that is, a mutual, informal contract about duration or use of life-sustaining treatments.^3,13 When surgeon and patient cannot forge agreement, a decision is made to not operate, because surgeons report they would not proceed with major surgery without permission to use postoperative life-sustaining treatments.³ Given that patients may not volunteer their preferences about life-sustaining treatment preoperatively, postoperative conflict likely arises from the surgeon’s belief that buy-in has occurred even though these preferences have not been discussed or explored.

In the wake of serious complications, families may produce an AD that contradicts the patient’s goals for surgery. This sudden opposition to intensive treatments can be distressing for families, surgeons, nurses, and intensivists making decisions about end-of-life care.¹⁴ In addition to harming interpersonal relationships, intensive care unit conflict prolongs length of stay, contributes to adverse events, detracts from patient safety, and can negatively affect health-related quality of life for surviving surrogate decision-makers.^15-17 Patients have the right to refuse medical treatment at any time, and surgeons have a deep commitment to patients to reach their goals. Improved clarity around what patients would and would not tolerate to reach specific outcomes has the potential to reduce conflict when the burdens of treatment are no longer tolerable or the initial goals of surgery are no longer possible.

For patients and families, it is important to know that ADs are useful in some settings but likely fall short postoperatively. Advance directives document preferences that are concrete, explicit instructions about the use of specific treatments (eg, feeding tubes). Surrogate decision-makers often focus on treatments, such as life-sustaining machines, that are objectionable without regard to circumstance or duration. Existing ADs may not reflect patient preferences in the surgical setting when such treatments are needed to reach the goals of surgery. There is little public narrative about the limits of ACP when these directives are rendered ambiguous or are contrary to surgical goals.

In this study, BWH and UWHC had significantly greater preoperative documentation of ADs. We believe this finding reflects regional initiatives to improve ACP in Wisconsin¹⁸ and an institutional initiative at BWH to increase AD documentation in the preanesthesia clinic.^19,20 Given the difficulties navigating these documents postoperatively and evidence that they were not discussed preoperatively, we worry that these interventions may give patients a false sense of assurance that their wishes will be followed. The American College of Surgeons Geriatric Surgery Verification Program requires preoperative review of existing ADs by the surgeon.²¹ It recommends that patients who lack an AD have an opportunity to establish one and that a medical proxy be documented.²¹ We agree with this innovative quality standard, noting that verification is the first step to supporting patients’ perioperative values and goals. As a starting point, surgeons might consider asking patients if they have someone to speak for them should a serious complication occur.

This study cohort included a large and diverse collection of preoperative surgical conversations with patients who would most benefit from preoperative ACP; regardless, the study itself has some limitations. Although we made a great effort to confirm with study-enrolled surgeons that we had recorded the primary decision-making conversation—which for most covered robust discussion of risks and expectations—patients may have had additional visits with the surgeon or an advance practice clinician in which ACP was addressed. Our estimation of the percentage of patients with a preoperative AD is restricted to patients for whom this is recorded in the medical record. Thus, we likely missed some preexisting ADs. This reflects a major problem with ADs: they often cannot be found when needed.

Although surgeons believe preoperative ACP is important, it rarely occurs. In this large national cohort study, the frequency and quality of preoperative discussion of patient preferences for postoperative life-sustaining treatment was sparse and therefore seems ineffective. For patients with an existing AD, there was no integration of the patient’s preferences with their goals for surgery. Surgeon engagement in preoperative ACP could help patients and families prepare for difficult decisions in the setting of serious complications and decrease postoperative conflict.

Accepted for Publication: February 14, 2021.

Published Online: May 12, 2021. doi:10.1001/jamasurg.2021.1521

Corresponding Author: Margaret L. Schwarze, MD, MPP, Department of Surgery, University of Wisconsin–Madison, 600 Highland Ave, Room K6/134 Clinical Sciences Center, Madison, WI 53792 (schwarze@surgery.wisc.edu).

Author Contributions: Dr Schwarze had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Kalbfell, Kata, Brasel, Mosenthal, Schwarze.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Kalbfell, Schwarze.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Kalbfell, Marka, Schwarze.

Obtained funding: Cooper, Schwarze.

Administrative, technical, or material support: Buffington.

Supervision: Buffington, Mosenthal, Schwarze.

Conflict of Interest Disclosures: Dr Brasel reported receiving grants from the Agency for Healthcare Research and Quality during the conduct of the study. Dr Finlayson reported receiving grants from the National Institute on Aging and the National Cancer Institute during the conduct of the study and founding Ooney, Inc (from which she does not receive any money), outside the submitted work. Dr Schwarze reported receiving grants from the National Institutes of Health (NIH) and Greenwall Foundation to address clinician-patient communication outside the submitted work and ownership interest by spouse in MezLight LLC, a light for the operating room. No other disclosures were reported.

Funding/Support: This study was supported by a grant from the Patient-Centered Outcomes Research Institute (PCORI) for design and conduct of the study, data collection, management, analysis, and interpretation of the data (all authors); award CDR1502-27462 from PCORI (Dr Schwarze); grant R21AG055876-01 from the NIH (Dr Schwarze); training award 7T32AI0125231-01 from the NIH (Dr Kalbfell); and grants from the National Institute on Aging (Dr Kata).

Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.