Study Characteristics
Forty-eight studies were included in the review, with a majority of studies conducted in the United States (n=27). Other countries represented were Norway (n=4), the Netherlands (n=5), Sweden (n=4), Germany (n=2), as well as n=1 each from Canada, Taiwan, Japan, Chile, and Australia. One study was conducted across Norway, Sweden, and the Netherlands. More than half of the studies were conducted within a hospital setting (n=27). Three studies identified the setting as a pediatric hospital. Ten studies took place in adult intensive care units, and one study took place in a neonatal intensive care unit. Additional study sites included aged care, community health and care, psychiatric outpatient clinics, mental health care institutions, in-home care settings, mobile health clinics, and transgender care. Two studies occurred across multiple settings, and one did not identify a clinical setting (Table 1).
The predominant study design was quality improvement (n=13), followed by mixed method (n=10), qualitative (n=9), randomized controlled trial (n=4), and retrospective/prospective or a combination (n=10). Standalone methods included economic analysis and observational study. Twenty-eight studies identified a time period; among these the average length of study was 26.75 months. The majority of studies (n=33) were published in 2016 or earlier; 15 studies were published within the past 5 years, with a greater proportion being conducted in European countries. The primary study population was healthcare professionals (n=29). Few studies focused on patients (n=7), even fewer on family members or substitute decision-makers (n=1), and 11 studies combined two or more of these groups (i.e., patients, family members and/or health care professionals) (Table 1).
Delivery Context
The CEC intervention was most often identified as a clinical / ethics / ethics committee consultation or service (n=34). Six studies used the term moral case deliberation. Other descriptions referenced ethics consultation system, ethics case reflection sessions, ethics rounds, clinical ethics support, structured multidisciplinary medical-ethical decision making, ethics intervention, and clinical ethicist involvement. Differences in terminology reflected nuances in understanding and delivery of CEC, including format, timing of the intervention (e.g., proactive or reactive), and processes (see Table 2). Despite the different descriptors, studies nevertheless fell under the general category of clinical ethics consultation or moral case deliberation given the similarity in structure, delivery, and purpose.
Twenty-nine interventions were described as involving an ethicist(s) or ethics consultant(s), six involved an ethics committee, and 13 involved a facilitator or healthcare professional with no formal ethicist title. Of the studies that described credentials or qualifications, the majority of CEC deliverers were described as having some training, education, or certification in ethics, ethics consultation, and/or moral case deliberation (n=9). Other experience or training included postgraduate degrees (n=5), with some specifying doctorate (n=3) or master’s degree (n=1), fellowship (n=4), clinical experience (n=5), training in medicine (n=3), ethics teaching experience (n=2), non-ethics specific skills training (n=3), and “familiarity”, “knowledge,” or “expertise” with ethics (n=4). Some studies only identified professional background (e.g., lawyer, philosopher). Twenty-five studies (52%) did not discuss or were not clear with respect to the training, credentials, or experience of the ethics consultant(s)/facilitator(s).
Twenty-six studies reported that the intervention began upon request. Physicians were most frequently identified as requesting the consult (n=15), followed by relatives and family members (n=10), nurses (n=8), patients (n=8), and others (n=6). Twenty-two studies did not identify the requestor, or a requestor was not applicable given the study design or CEC delivery. Some studies described the purpose of the interventionist as delivering a recommendation, whereas others described the primary role as facilitating a discussion without offering a particular solution. For example, while Molewijk et al. described the role of the ethicist as, “that of a facilitator who does not give substantial advice and does not morally justify or legitimize a specific decision” [67], Wocial et al. described one of the goals as making specific recommendations [81]. Additionally, some studies described the process as involving stakeholder engagement, whereas a few studies described a more solitary process of deliberation. For example, Orr and Moon referred to the intervention as involving a consultant who discusses the case with the requestor and individually with members of the care team, patient, family, and others involved in the conflict [68]. Smith et al. identified the process as involving a written memorandum containing an ethical analysis and opinions about the case with no external deliberation [75]. For a full description of the interventions, see Table 2.
Outcome Domain Reporting
The top three outcome domains that studies reported on were quality (n=31), process factors (n=23), and clinical factors (n=16). Few studies assessed healthcare utilization (n=4) and economic outcomes (n=4). The majority of studies examined multiple outcome domains. All six outcome domains were multidimensional and included a variety of subthemes (see Table 3).
Quality
Quality was the most frequently reported outcome domain (n=31). This domain captured the quality of the CEC, consultant, and/or overall stakeholder experience as it related to perceived satisfaction, helpfulness, timeliness, accessibility, and overall benefit. Quality-related outcomes were primarily measured by survey (n=22), followed by qualitative interviews or focus groups (n=7) and mixed methods (n=6).
Helpfulness was the most frequent construct measured and reported (n=10). Helpfulness was related to the perceived usefulness of the CEC in informing practice, the perceived overall benefit of the CEC, the extent to which the CEC was beneficial in assisting in patient care, the importance for physician education and medical treatment, whether a CEC would be used in the future, effectiveness in providing emotional support, mediating disputes, clarifying ethical issues, and improving communication (see Table 4). Other prominent quality subthemes included satisfaction (n=8), quality of experience and/or service (n=5), overall experience (n=4), usefulness/practical implications (n=3), and value of experience (n=2), as well as ability to improve practice (n=1) and effectiveness (n=1).
Respondents in almost all studies reported a positive or beneficial encounter with CEC either in terms of overall experience, quality, usefulness, value, helpfulness, or satisfaction (n=28), with healthcare professionals being the most frequently examined group (n=26).[1] Seven studies reported a moderate to high level of satisfaction with the CEC experience. These were described in various ways: an inclination to use the service again or recommend the service to others, the degree of consensus between stakeholders, the ability to meet perceived goals, general satisfaction with the service or intervention, the positive or negative nature of the experience, practical usefulness, and the impression of the intervention. White, Dunn, and Homer reported specifically on low levels of satisfaction, with shared decision making, documentation adequacy, and increased knowledge of ethical issues receiving the lowest satisfaction scores [79]. Only one study by found that the respondents’ overall experience with CEC was negative (n=1) [40], although other studies suggested improvements that could be made [74]. Negative experiences were attributed to factors such as a lack of structure in the consult, a lack of timeliness, or an absence of integration into clinical practice [40, 74].
Among the 10 studies evaluating helpfulness, eight reported CEC to be helpful. These eight studies situated helpfulness with respect to resolving issues or conflict (n=5), identifying, analyzing, and clarifying issues and/or values (n=5), and in providing education (n=3). Some studies concluded that the intervention improved interpersonal and professional qualities such as mutual understanding and cooperation, personal ethical reflection and insight, and ability and confidence to act in practice. Respondents in a few studies would seek CEC again (n=2) or would recommend the service to others (n=4). Insights and cross-comparisons between participant groups who experienced higher or lower levels of each attribute are difficult to ascertain given insufficient statistical power or incomplete reporting on the perspectives of certain stakeholder groups, such as patients and families (Table 4).
Process Factors
Process factors comprised the second most reported outcome domain (n=23). Process factors referred to any set of activities that occurred between the CEC provider and other stakeholders, such as identifying, clarifying, resolving ethical issues, reaching consensus, and facilitating understanding of different viewpoints (Table 5). Process factors were identified by qualitative interviews or focus groups (n=13), survey (n=10), mixed methods (n=3) and record review (n=1).
The most reported subtheme in the process factors domain was purpose and/or impact of the intervention (n=14). Other subthemes included establishing consensus and integration (n=3), identification (n=1), advancing care (n=1), helpfulness (n=2), support (n=1), and clarification (n=1).
With respect to the dominant subdomain, the data demonstrate a diverse understanding within and across stakeholder groups regarding the purpose and impact of ethics intervention. Many study respondents perceived ethics involvement to be most effective with respect to identifying or further elucidating core [ethical] tensions or barriers to resolution (n=8), facilitating the sharing of viewpoints and better communication between stakeholders (n=5), and/or achieving resolution, consensus, or clarity on the problem at hand, plan of action, or desired outcome (n=7). Five studies additionally reported ethics intervention as having an impact by improving individuals’ feelings of safety and support (n=2), trust (n=2), and confidence in decisions (n=1). Two studies reported ethics intervention as increasing mutual understanding and respect within and between stakeholder groups, as well as providing valuable education (n=4) or an objective opinion on the problem at hand (n=3).
However, some studies were less sanguine about these descriptors. For example, Weidema et al. reported that ethics intervention revealed team conflicts, but maintained that it did not assist in solving them due to the ethical issues being too complex or ambiguous [77]. This finding is corroborated by Vrouenraets et al. who reported that enabling concrete steps to navigate ethically difficult situations was the least common outcome associated with ethics interventions [85]. In some studies, the positive impact of CEC was not consistent across stakeholder groups. Yen and Schneiderman, and Schneiderman et al., for example, found that the positive experience reported by physicians was not experienced by family members, with 75% strongly disagreeing that an ethics intervention was important for resolving ethical issues [72, 83]. Other studies reported dissatisfaction when expected outcomes were not met, such as the expectation of resolution or the formulation of a clear plan of action [47]. In studies that provided recommendations for process-related improvement, respondents prescribed better communication with the health care team [82]. Poor communication, complexity, and ambiguity were also reported as barriers to the resolution of ethical problems in other studies in the process domain category [77].
Clinical Factors
Sixteen studies were coded under the clinical factors domain, which was related to a change in the clinical care of the patient, including but not limited to, adherence with a recommendation, agreement with a recommendation, a decision about care or treatment, concordance between patient wishes and treatment delivered, a change in treatment plan, a reduction of non-beneficial treatment, and/or conflict resolution regarding patient care (Table 6). The majority of studies in this domain utilized record review as the method of data collection (n=13). Other methods included survey (N=6), qualitative interviews (n=1) and mixed methods (n=1).
Studies that reported on clinical factors examined a narrower set of characteristics compared to other domains; patient management and provision of care was the most frequently reported construct (n=8). Patient management and provision of care included the following indicators: the presence of, or recommendation to change, particular orders or decisions (e.g. do-not-resuscitate orders, withholding/withdrawal orders, life-sustaining treatment, limits of care, requests for services, and pain management), provision of palliative care and/or chaplaincy services, agreement with a decision, and a change in treatment plan, plan of care, or patient management. Other subthemes reported in this domain included non-beneficial care (n=3), implementation/adherence (n=2), clinical consensus (n=2), quality of care (n=1), and coercion (n=1).
Constructs, outcomes, and measures varied considerably between the studies examining patient management and provision of care, despite their similarities. For example, Dowdy, Robertson, and Bander measured the presence of particular constituents of care, understood by the term quality of communication index — communication and decision making that resulted in a treatment order between a control and ethics intervention group. These researchers noted a “higher communication score” and significant difference between intervention and control with respect to the presence of do-not-resuscitate (DNR) and life-sustaining treatment decisions [51]. Another study by Cohn et al. examining communication and decision-making, measured the level of agreement with the decision and the degree of change that occurred due to CEC. Healthcare provider respondents were significantly more likely than family members to report a high degree of agreement with the outcome of the ethics consultation, and both healthcare providers and patient/family perceived similar degrees of change to the plan of care following CEC [46]. Other studies measuring patient management considered whether consultations changed patient care, measuring the change rather than the presence (or not) of particular types of care. Outcomes demonstrated that the majority of respondents reported CEC changing patient management in some regard, with most changes occurring due to the consultation prompting the withholding of life support therapies that would have otherwise been used [70]. Thus, despite parallels across the various constructs, differences exist with regards to the terminology (and their related understandings) and measurement of outcomes.
Wide variation also existed in measures and outcomes with respect to non-beneficial treatment, with minimal consensus on what might be considered and measured as non-beneficial, as well as what can be concluded with respect to the impact of CEC on the provision of non-beneficial treatment. For example, Schneiderman, Gilmer, and Teetzel measured the number of ICU days and life-sustaining treatments in patients who died before discharge. They reported a reduction in ICU days, days receiving artificial nutrition and hydration, percentage on ventilation, and days receiving ventilation in patients receiving consultation compared with the control [72]. In later research, which focused on number of ICU days, hospital days, and life-sustaining treatment in patients who did not survive to discharge, Schneiderman et al. found no significant difference between intervention and control; however, a pattern in the reduction of hospital and ICU days with ethics intervention was observed. Additionally, this study justified outcomes as “non-beneficial” because they were perceived to “represent a failure to achieve a fundamental goal of medicine” [73]. Another study by Andereck et al. examined non-beneficial treatment among patients receiving ventilation and/or artificial nutrition. While the intervention and control groups varied by one median day, the authors did not make any causal claim with respect to ethics intervention and reduction in the provision of these treatments [39].
Personal Factors
Personal factors was the fourth most reported on domain (n=11). Studies coded under this domain related to changes in personal state or stakeholder (patient/family/surrogate/health care professionals) perspective or experience; for example, moral distress, enhanced knowledge, patient survival/morality, and/or feeling supported (Table 7). Data collection methods included survey (n=7), record review (n=2) and qualitative interviews (n=1). One study utilized both qualitative interview and record review.
Subthemes explored under the personal factors domain included perceived value and outcomes (n=3), patient mortality (n=3), experience (n=2), moral distress (n=2), learning (n=1), confidence (n=1), suffering (n=1), and clarity (n=1). With respect to perceived value and outcomes, understandings consisted of value for self and practice, impact on the individual, impact on values, changes to perspective during the session, changes to perspective in practice, and whether any outcomes were experienced at all during the session and/or in practice. Interestingly, other subthemes did not demonstrate as much variation.
Findings of the studies that examined the perceived value of CEC resulted in a variety of understandings of perceived values and outcomes with respect to ethics interventions. Respondents most frequently reported ethics interventions as valuable for enhancing their understanding and awareness of ethical issues (n=4), developing confidence (n=2), fostering open communication and expression of feelings (n=2), improving mutual understanding and cooperation (n=2), enabling and delivering solutions (n=2), improving skills (n=2), and achieving consensus (n=2). Brännström et al. described the experience as encouraging stakeholders to meet in an ethical “free zone” in which the nature of the intervention created a safe and inclusive forum for stakeholders to express their viewpoints and to be heard in a space where they felt confident and trusted [45]. Wocial, Molnar, and Ott recommended effective communication with the health care team as an area for improvement [82].
Notably, most respondents in the two studies measuring moral distress reported a reduction in levels of distress following CEC intervention. Respondents in these studies consisted of health care providers, such as nurses and physicians. Moral distress levels were not measured among patients, families, or other stakeholder groups [57, 81]. Other studies reported perceived increases in clarity, confidence, and learning among providers [70]. With respect to patient outcomes, no significant difference in mortality rate was observed between participants who received ethics intervention and those who did not [39, 56, 73].
Healthcare Utilization
A few studies reported on healthcare utilization (n=4). Data collection methods consisted entirely of medical record review. Studies coded under this domain evaluated outcomes in terms of service users’ consumption of health-related resources (Table 8). All four studies evaluated outcomes in terms of length of stay and sought to examine whether length of stay was reduced following ethics intervention, both in terms of length of hospital stay (n=4) and length of stay in the ICU, specifically (n=3) pre- and post-intervention. While two studies denoted a shorter length of stay in the intervention arm compared to the control [55-56], one study identified no statistically significant difference [39], and one study found length of stay in relation to discharge status significant [51].
Economic Outcomes
Economic outcomes were reported infrequently (n=4). Studies falling within this domain measured the financial impacts of CEC or used hospital records to compare cases involving ethics interventions to those that did not. Subthemes included measures of cost and cost avoidance, the former including descriptions such as total cost of stay, inpatient stay costs (total service cost + acute inpatient cost + ICU room and board), net cost of consultation, and charges for patients. Cost avoidance included variables such as cost for treatment, expenses avoided or added for patients receiving consultation, and total cost avoidance (variable costs + fixed costs) (Table 9).
Outcomes across these four studies were not congruent with respect to actual and perceived impact of CEC. A study by Meltzer, Heilicser, and Siegler examining cost avoidance through retrospective record review drew a strong connection between ethics intervention and cost, reporting a $288,827.00 total cost avoidance over a six-month period, with savings obtained by decreasing length of stay, costs associated with resuscitation, number of surgical and diagnostic procedures, among other factors. This was compared to the expense of ethics support and resources, which was reported to be $12,000 for each patient who received consultation. Qualitative reporting in this study reflected similar outcomes: 69% of consultations resulted in cost avoidance, and an additional 10% resulted in potential cost savings. Researchers in this study asserted that cost savings would have been greater if ethical recommendations were followed [66]. Similarly, Gilmer et al. demonstrated comparable findings with an estimated annual savings of $157,380.00 related to ethics consultation practice [55]. On the other hand, while reporting a reduction in average charges compared to baseline groups, two studies did not report significant differences between intervention and control with ethics intervention [39, 51]. In fact, Dowdy, Robertson, and Bander specifically asserted that despite a reduction in costs, this difference was not statistically significant enough to demonstrate the efficacy or causal relationship between ethics consultation and cost reduction [51].
[2] This includes quality studies that examine healthcare professionals as a standalone group, or in conjunction with other study populations (e.g. healthcare professionals and families)