Anesthesiologist to Patient Communication: A Systematic Review

Key Points

Question  Do anesthesiologists or other anesthesia professionals engage in discussions with patients regarding decisions with implications beyond the operating room?

Findings  In this systematic review of the literature on communication between patients and anesthesia professionals, limited data were found on communication regarding perioperative decisions with implications that reach beyond the operating room. These data suggest that communication between patients and anesthesia professionals during preoperative encounters is dominated by discussion of anesthetic planning and perioperative logistics, with variable discussion of risks vs benefits and infrequent discussion of postoperative care or elicitation of patient values and preferences.

Meaning  These findings suggest that patients who become critically ill following scheduled surgical interventions are unlikely to have had discussions with their anesthesiologist regarding values and preferences for navigating complex postoperative care decisions, such as prolonged invasive ventilation, protracted hospital stay with incomplete recovery, or end-of-life care.

Importance  Many patients are admitted to the intensive care unit following surgery, and some of them will experience incomplete recovery. For patients in this situation, preoperative discussions regarding patient values and preferences may direct care decisions. Existing literature shows that it is uncommon for surgeons to have these conversations preoperatively; it is unclear whether anesthesia professionals engage with patients on this topic prior to surgery.

Objective  To review the literature on communication between patients and anesthesia professionals, with a focus on discussions related to postoperative critical care.

Evidence Review  MEDLINE and Web of Science were searched using specific search criteria from January 1980 to April 2020. Studies describing encounters between patients and anesthesia professionals were selected, and data regarding study objectives, study design, methodology, measures, outcomes, patient characteristics, and clinical setting were extracted and collated. The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline was followed.

Findings  A total of 12 studies including 1284 individual patient encounters were eligible for inclusion in the review. These studies demonstrated that communication between patients and anesthesia professionals related to postoperative care is rare: only 2 studies reported communication regarding adverse postoperative events, and this communication behavior was reported in only 46 of 1284 consultations (3.6%) across all studies. Additional findings were that communication during these encounters is dominated by anesthetic planning and perioperative logistics, with variable discussion of perioperative risks vs benefits and infrequent elicitation of patient values and preferences. Some data suggest that patients wish to be involved in perioperative decision-making but are often limited by an incomplete understanding of risks and benefits.

Conclusions and Relevance  This systematic review found that communication in anesthesia is dominated by anesthetic planning and discussion of preoperative logistics, whereas postoperative critical care is rarely discussed. Most patients who are admitted to an intensive care unit after a major operation will not have had a discussion regarding goals of care specific to protracted recovery or prolonged intensive care with their anesthesiologist.

Communication with patients about therapeutic options and care plans is a critical component of shared decision-making and is particularly important when a decision may result in a major or permanent change in a patient’s health status. This situation is relatively common for patients undergoing major surgery. Surgeons and anesthesiologists are the principal clinicians with the opportunity and, arguably, the responsibility to elicit values and preferences about postoperative care from surgical patients to inform care decisions if patients become critically ill and lose decisional capacity postoperatively. Previous work suggests that surgeons uncommonly elicit patient preferences regarding postoperative critical illness preoperatively, even for high-risk patients.^1,2 Anesthesiologists also have the opportunity to elicit patient values and preferences preoperatively, and some members of the specialty have an interest in expanding anesthesiologists’ role in perioperative medicine.^3-5 Knowledge and communication of medical and surgical complications after surgery, as opposed to complications of the anesthetic, are essential to this role. However, the extent of anesthesiologists’ responsibility and their ability to perform this role is not clear, and there are likely variable professional expectations for patient-anesthesiologist communication in different health care systems and settings.

There are few data on communication during anesthesia consultations. Although studies on anesthesiologist-patient communication have been narratively reviewed,^6,7 there is no systematic review on this topic. In this review, a systematic search strategy was used to extract and collate data on communication between anesthesia professionals and patients, and the methodological quality of existing studies was assessed. A synthesis of the data focused on communication about postoperative critical illness is presented.

A systematic review of the literature on communication between anesthesia professionals and patients was performed to address the following question: in preoperative anesthetic encounters, what are the patterns and content of communication between anesthesia professionals and patients as evaluated by qualitative or mixed methods? Reporting is consistent with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.⁸

A MEDLINE search was performed (from 1980 to April 2020) to retrieve any studies with a focus on communication between patients and anesthesia professionals (eAppendix 1 in the Supplement). A 1-generation, forward-and-backward search on Web of Science was then performed using each of the included studies from the MEDLINE search to identify additional relevant studies.

Only studies with data describing specific encounters between patients and anesthesia professionals were included. Studies with a primary focus other than communication, studies on communication during anesthesia procedures, and studies examining communication with children were excluded (see eAppendix 2 and eAppendix 3 in the Supplement). In addition, studies that developed or evaluated communication interventions were excluded because these studies prescribed communication strategies instead of evaluating established communication practices. The search was limited to studies published in English, which generally gives a sufficient assessment of a given topic,^9,10 and to studies published after 1980. Three reviewers (M.J.T., S.H., and M.C.S.) performed title screening, and 1 reviewer (M.J.T.) retrieved the full text of relevant titles, selected studies, and extracted data.

One reviewer (M.J.T) assessed the quality of all studies using the previously validated Critical Appraisal Skills Program (CASP) tool for Qualitative Studies.¹¹ A second reviewer (N.K.J.A.) verified these assessments.

Individual study results and quality reviews are presented, and overall results are synthesized descriptively. Variables and outcomes extracted from individual studies were too diverse for quantitative synthesis. Continuous data are expressed as means with SDs or as medians with interquartile ranges (IQR). No statistical testing was conducted.

The Figure shows an overview of study selection. Thirty full-text articles from the search were reviewed, of which 20 studies were excluded (see eAppendix 4 in the Supplement). Seventeen of these studies were excluded because they did not include any data about anesthesiologist-patient communication during routine encounters. Three studies were excluded because they were about communication during procedures. The remaining 10 studies were included, and the Web of Science search returned 2 more studies, resulting in 12 studies for review.^12-23

Study characteristics are summarized in Table 1. All studies included descriptive statistics, and 5 studies^13-15,20,23 performed some statistical modeling. Ten studies collected raw communication data on clinical encounters by audiotaping,^{15,17-20,22,23} videotaping,¹⁶ or direct observation with an experienced observer.^12,13 One study collected data using questionnaires only,¹⁴ and another used semistructured interviews.²¹ Of the studies that performed qualitative analyses, only 1 study¹⁵ specified a qualitative analysis approach and framework²⁴ for data coding.

The summary of the methodological quality is shown in Table 2. Only 4 studies^14,16,18,20 used previously validated tools to collect or code data, and 1 study²¹ created and validated a survey. Eight studies^12,15-20,22 used 2 or more assessors to code recorded data. Eleven studies^12,13,15-23 were evaluated on all CASP criteria, with a median (IQR) score of 4 of 5 (3-5). One study¹⁴ was only evaluated on 4 of the CASP criteria and scored 3 of 4. Methodological issues and assessment of quantitative analyses for studies that conducted statistical modeling are shown in Table 3.

Only 2 studies reported communication regarding adverse postoperative medical events, and this communication behavior was reported in only 46 of 1284 consultations (3.6%) across all studies. An a priori decision was made to specifically evaluate papers for communication data in the following categories: (1) discussion of therapeutic options including informed consent, patient comprehension, and risks/benefits, (2) elicitation of values and preferences, (3) shared decision-making, and (4) communication about postoperative care. These categories were chosen because they highlight communication that is central to patient-physician consultations around major interventions. Because of the broad types of data found in the review, the second and third categories were collapsed into a single category, and other data was added as a category to capture data that did not fit into previously defined categories. Study results are summarized in Table 4.

Ten studies^{12,13,15-17,19-23} included data on these topics. Two studies^16,23 examined communication in general without a specific focus. One study of patient and anesthesiologist utterances during consultations¹⁶ identified a mean of 23% of utterances as being related to patient counseling (exact proportion not provided); however, the coding method used suggests that most utterances coded as counseling were likely related to technical and logistical aspects of care. A similar result was seen in a study of anesthesia consultations with standardized patients,²³ which used mock patient scenarios and 2 different standardized patients. In this study, only a mean of less than 1 minute was spent making a plan in each encounter.²³

Seven studies^{13,15-17,20,22,23} contained data about risk and benefit discussions; 3 studies^13,17,22 specifically evaluated communication of risks. One study¹⁷ found that during 91 clinical encounters with parents of children undergoing anesthesia, in 27 consultations (29.6%) no risks were discussed, and in a further 23 consultations (25.3%), only a general statement of risk was included. Serious risks were only discussed in 4 encounters (4.4%). In adults undergoing elective surgery, another study¹³ found that during 40 routine encounters, only 31 preoperative consultations (77.5%) included discussion of at least 1 risk. Where risk was part of consultations (n = 151), patients were almost always satisfied and not distressed by the discussion (146 of 151 consultations [96.7%]). Conversely, in consultations where no risks were discussed (n = 115), most patients (96 [83.5%]) believed that there was no risk to anesthesia at all.¹³ A small study on epidural insertion²² found a similar degree of variability, where the number of risks discussed in consent conversations varied from 0 to 11 per encounter. In studies with a focus other than risk communication that had ancillary data about risk discussions, there was a similar degree of variability.^15,16,20,23 Most risks specifically evaluated in these studies were minor, short-term risks. Global assessment of informed consent was evaluated in only 1 study,¹⁵ which found that in conversations with parents of children undergoing anesthesia, the minimum requirements for informed consent were included in 68 of 97 cases (70.1%). Only a minority of conversations (12%, exact proportion not provided) included all 7 aspects of fully informed consent as defined by the authors.

Data related to patient comprehension of information communicated by anesthesiologists were extracted from 4 studies.^12,15,19,21 Among studies with objective measures of patient comprehension, patient understanding of risks and benefits of various anesthetic options was poor. For example, 1 study showed that many parents recalled a description of the anesthesia planned for their child (96.2%, exact proportion not provided) and plans for postoperative pain control (81.2%, exact proportion not provided), but follow-up questions suggested very few parents fully understood risks, benefits, and complications (28 of 263 parents [10.6%]).²¹ In another study, parents frequently reported understanding risks, benefits, and the anesthetic plan (88%, 96%, and 96%, respectively; exact proportions not provided).¹⁵ However, this study only included self-reported parental comprehension. When considering specific words used in consultations, 1 study¹² demonstrated that although patients misunderstood a minority of technical terms used by anesthesiologists (49 of 484 terms [10.1%] misunderstood across all encounters), there was at least 1 instance of patients misunderstanding in 32 of 68 individual encounters (47.1%).¹² Another aspect of communication relating to patient comprehension was evaluated by a study that measured the amount of information given to patients preoperatively by anesthesia professionals.¹⁹ This study found that patients’ information storing capacity was consistently exceeded in preoperative encounters.¹⁹

Five studies^14-16,20,23 had data about eliciting patient preferences and shared decision-making. In the 2 studies^16,23 that evaluated communication generally, elicitation of patient preferences and values was uncommon. In 1 study,²³ anesthesiologists spent less than 1 minute obtaining patient perspectives during encounters that were a mean (SD) of 15.9 (4.9) minutes long. Another study¹⁶ showed no utterances eliciting patient preferences during consultations. Across 21 encounters in this study that required a shared decision, the Observing Patient Involvement Scores (OPTION scores²⁵) were poor, with elicitation of patient input categories receiving the lowest scores.¹⁶ There were similar findings in a study of informed consent in pediatric anesthesia,¹⁵ which showed that elicitation of parental preferences was uncommon (18% of consultations, exact proportion not provided). Two studies^20,14 examined shared decision-making. In 1 study of shared decision about neuraxial vs general anesthesia,²⁰ OPTION scores showed that anesthesia professionals rarely explained the benefits and risks of anesthetic options and did not elicit or make adequate attempts to integrate patient preferences into decision-making. Another study¹⁴ had similar findings: most patients (>90%) wanted to be involved in decisions about their care, and anesthetists tended to underestimate patients’ desire for shared decision-making.

Discussions about postoperative care were rare: this type of communication was described in 5 studies,^{13,16,17,21,23} and postoperative pain control dominated these discussions. Only 2 studies^13,17 presented data on communication about specific adverse outcomes. In these studies, there were 4 instances of communication about postoperative events across 91 interviews (4.4%) in 1 study,¹⁷ and death or severe permanent harm discussed in 20 of 272 interviews (7.4%) and 22 interviews (8.1%), respectively, in another study.¹³ None of the studies had any data about elicitation of patient preferences regarding direction of care in the case of serious adverse events.

Eight studies^{13-16,18,20,21,23} had some additional data about patient satisfaction or perception of the quality of the encounter following anesthesia consultations. Satisfaction was generally high, regardless of which specific components were included in interviews,^15,20 and satisfaction may have a positive association with degree of patient involvement in care decisions¹⁴ and with more experienced anesthesia professionals.²³

This systematic review of the literature on communication between anesthesia professionals and patients found only 12 studies that met inclusion criteria. The studies had an overall moderate level of methodological quality. The main finding is that communication about postoperative care was rarely described in preoperative consultations with anesthesia professionals; the literature had no data describing anesthesiologist-patient communication addressing protracted ICU stay, protracted ventilation, and end-of-life care in the setting of postoperative incomplete recovery. These findings are consistent with a previous narrative review on patient-anesthesiologist communication⁷; however, this review contributes a more robust summary of the evidence by using a systematic search strategy, extracting and qualitatively collating data from superior data sources, assessing the quality of each study using an established evaluation tool, and including 11 studies that, to our knowledge, have not been summarized in any previous review on this topic. These data are also similar to previously published data on surgeon communication, showing little elicitation of values or preferences regarding these issues in surgical consultations.^1,2 Therefore, most patients who undergo major surgical interventions do not have preoperative discussions about values, preferences, or goals of care that address the scenario of protracted or incomplete recovery from surgery. Several other findings emerge from the data. First, informed consent, including discussion of risks and benefits, is highly variable, and patient comprehension of risks, benefits, and therapeutic alternatives is frequently poor when measured objectively. Second, anesthesia professionals frequently give patients unmanageable amounts of information, and communication is often focused on technical and logistical aspects of care. Lastly, anesthesiologists infrequently engage in elicitation of patient values and shared decision-making, despite patients’ apparent desire to be involved in decision-making.

Professional society guidelines in anesthesia recommend that “anesthesiologists should include patients, including minors, in medical decision making that is appropriate to their developmental capacity and the medical issues involved.”²⁶ However, there are many barriers to discussions of patient values, preferences, and goals of care in the preoperative setting. In many North American surgical centers, anesthesiologists only become involved in the care of surgical patients after they have made the decision to proceed to the operating room with their surgeon. Therefore, it is likely that many anesthesiologists focus on getting the patient through the operation and may see this kind of communication and patient value exploration as not a part of their job. Second, there are large financial incentives to proceed to the operating room, for surgeons as well as anesthesiologists, putting additional emphasis on moving the patient through the operating room. Third, many anesthesiologists lack the specific expertise to speak to perioperative issues that reach beyond the operating room. Lastly, given the volume of patients seen at anesthetic clinics, anesthesiologists likely feel tremendous time pressure and probably feel they do not have adequate time for (and are not adequately compensated for) protracted discussion of perioperative values, preferences, and goals of care. In cases where there are nontrivial risks that may result in a significant change in a patient’s health status or prolonged burdensome care (for example, ventilator dependence after a postoperative stroke), then anesthesia consultations without discussion of postoperative care and elicitation of patient preferences may represent a missed opportunity to raise these issues. Identifying patients at risk for postoperative complications, such as prolonged mechanical ventilation, weakness, and postoperative delirium, can provide an important perspective on perioperative decisions. Literature on communication from outside anesthesiology suggests that patients often agree to a plan of care that is inconsistent with their values and preferences, including undergoing surgery.²⁷ Informed consent for major surgery that explores these factors is often possible in a 20- to 30-minute clinical encounter.²⁸ Therefore, it seems both feasible and valuable for anesthesiologists to engage in balancing risks and benefits in the context of the patient’s values during preoperative consultations, especially when anesthesiologists are involved in postoperative care.

The 2 other physician specialties that routinely encounter surgical patients, namely surgery and critical care, have studied communication extensively compared with the findings here. For example, there have been several recent reviews on surgeon-patient communication,^1,29,30 including a systematic review¹ that only included studies with audiotaped or videotaped interactions and at least 1 objective measure of surgeon behavior or communication skills. This review reported data from 21 studies and an additional 13 companion reports. If these selection criteria were applied to this systematic review of anesthesia and patient communication, only 1 study would be included. Similarly, there have been multiple reviews of physician-patient communication in critical care medicine,^31-33 a discipline in which specific types of communication, such as end-of-life communication^34,35 and communication strategies for difficult decision-making,^36-38 have been broadly evaluated.

Based on the data in this review, several hypotheses follow regarding strategies to improve patient-anesthesiologist communication. First, if anesthesia professionals adapt to patients’ individual communication needs, patient participation and satisfaction may improve, although this strategy has not led to measurable improvement in communication about end-of-life issues in ICU.³⁹ Second, while there is tension between providing too much information (risking information overload) and not providing enough information (risking inadequate patient understanding and informed consent), the data suggest that communication may improve if anesthesia professionals identify and emphasize important nontechnical information specific to each individual patient. Lastly, for anesthesiologists involved in perioperative medicine, patients who are at high risk of incomplete recovery may benefit from elicitation of values and preferences regarding postoperative care during preoperative consultations. Shifting the focus of anesthetic care to perioperative medicine and specifically improving preoperative communication about goals of care is likely to be a significant challenge for the specialty of anesthesiology. Several interventions aimed at perioperative advance care planning have been developed and evaluated,^40-42 providing some guidance for anesthesiologists expanding their practice into perioperative medicine.

This systematic review has several limitations. First, the search was limited to studies published in English from 1980 to April 2020. Although additional data may have been published earlier or indexed elsewhere, they are not likely to be relevant to current practice. The search only found 12 studies with different designs, settings, and outcomes, making synthesis challenging. Common limitations for the studies that were reviewed included unavoidable selection bias due to selective participation; the Hawthorne effect in studies that employed direct observation (2 of 12 studies), and the infrequent use of validated analysis or coding tools (only 4 of 12 studies used validated tools). The survey-based studies (3 of 12 studies) were limited by recall bias of patients and health care professionals. Nine studies implemented mitigation strategies for these biases. Lastly, only 3 studies provided data about the risk category of the patients in their analyses, and most patients were considered low risk for complications. Preoperative communication with patients with higher risk may be substantially different compared with the communication patterns found in this review. These limitations make it difficult to draw concrete conclusions about communication in anesthesia and implications for patients who have incomplete recovery.

This systematic review of the literature on patient-anesthesiologist communication found that communication in anesthesia rarely includes discussion of postoperative care or patient values and preferences, but rather is dominated by anesthetic planning and perioperative logistics. These findings, coupled with similar data from surgical literature, suggest that most patients who arrive in the critical care unit following a major operation have not had a preoperative discussion about values, preferences, and goals of care specific to protracted recovery or prolonged intensive care.

Accepted for Publication: August 28, 2020.

Published: November 12, 2020. doi:10.1001/jamanetworkopen.2020.23503

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Tylee MJ et al. JAMA Network Open.

Corresponding Author: Michael J. Tylee, MD, Department of Anesthesia and Pain Management, University Health Network, Toronto General Hospital, 200 Elizabeth St, 3EN-464, Toronto, ON M5G 2C4, Canada (mike.tylee@mail.utoronto.ca).

Author Contributions: Drs Adhikari and Tylee had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Rubenfeld and Adhikari contributed equally in senior authorship positions.

Concept and design: Tylee, Rubenfeld, Sklar, Hussein, Adhikari.

Acquisition, analysis, or interpretation of data: Tylee, Rubenfeld, Wijeysundera, Sklar, Adhikari.

Drafting of the manuscript: Tylee, Rubenfeld, Sklar.

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

Administrative, technical, or material support: Sklar, Hussein.

Supervision: Rubenfeld, Adhikari.

Conflict of Interest Disclosures: None reported.

Funding/Support: Funded academic time provided for Dr Tylee by the Department of Critical Care Medicine, Sunnybrook Health Sciences Centre, and by the Department of Anesthesia and Pain Management, Toronto General Hospital.

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.

Additional Contributions: Henry Lam, MLS (Sunnybrook Health Sciences Centre Library) assisted with the literature search and was not compensated for this contribution.