Communication in the ICU: Holding a meeting with families and caregivers

INTRODUCTION — Family meetings facilitate communication between health care providers and caregivers of patients in the intensive care unit (ICU). Effective communication improves family satisfaction, trust in ICU physicians, clinical decision-making, and the psychological well-being being of family members [1-5]. As a result, research agendas and lists of quality indicators are increasingly recognizing the importance of good communication with families [6-11]. Multiple large quality improvement efforts have been undertaken to improve communication with families in the ICU with positive results [12-14].

The benefits of communication and issues related to conducting a family meeting will be reviewed here. Throughout the discussion, "family" refers to any person of importance in the patient's life who might attend a conference related to the patient's care, whether they are related or not. Decision making by surrogates and informed consent are discussed elsewhere. (See "Advance care planning and advance directives" and "Ethics in the intensive care unit: Informed consent".)

IMPORTANCE OF COMMUNICATION — The receipt of honest, intelligible, and timely information is among the primary concerns of family members of patients in the ICU [15-20]. In addition, they need support, comfort, proximity, and reassurance [21]. Families want to feel that there is hope and they generally want to know the prognosis [22].

Unfortunately, the evidence indicates that communication with health care providers often leaves much to be desired [23]. Observational studies have found that communication issues with clinicians are the number one source of complaints among families of deceased patients, with as many as 30 percent of family members feeling dissatisfied with communication in the ICU [24-26]. Contributing factors include inadequate time spent communicating with clinicians, lack of consistent information, and information provided by multiple health care providers [27]. Poor communication can affect family satisfaction, clinical decision-making, and the psychological outcomes of family members.

Guidelines have been issued by the American Thoracic Society for shared decision-making which are in keeping with the issues discussed in this topic review [19].

Family satisfaction — Family satisfaction corresponds to the degree to which the expectations of the family are met. This is affected by a number of factors related to providers, including the quality of communication, the quality of interactions, the level of empathy, and the degree to which family members feel isolated while visiting [28,29]. Family satisfaction improves when multidisciplinary teams are trained to communicate with family members [30].

Additional factors that have been associated with increased family satisfaction include the following [29,31-40]:

●Knowing the role of each health care provider

●Having junior clinicians give information

●Avoiding the perception that contradictory information is being provided

●Involvement of a primary care doctor

●Allocating sufficient time to provide information

●Clinician accessibility and continuity of care

●The clinician's ability to listen, encourage questions, provide emotional support, and pay attention to the patient's values

●The completeness of information provided to the family

●The feeling that the decision-maker was supported through the decision-making process

●Discussion of spiritual needs and satisfaction with spiritual care

●Withdrawing life support, especially when based on a clinician's recommendation

●Expression of the patient's wishes

●The use of empathic statements

●Less time spent with the clinician speaking and more time allowed for the family to speak

●Concern and caring by the ICU staff

●A sense of family control over the patient's care

●Expressions of nonabandonment and assurances of comfort

●Provision of written information

Family satisfaction is also improved by early and frequent interaction between family members and providers [41]. Such communication facilitates earlier recognition of how the illness is affecting the family unit and establishes a cooperative patient-focused style of care. It also offers an opportunity to correct any misunderstanding that the family may have about what is wrong with the patient.

According to an observational study of 102 patients, only one-half of the families understood the patient's diagnosis, prognosis, or treatment after an initial meeting with an attending clinician [42]. Moreover, health care providers often do not recognize poor family comprehension and family members are frequently unaware of their own misunderstandings. In a study of 81 decision-makers, most reported excellent comprehension, but only one-half had a good understanding when objectively tested [43]. Furthermore a number of studies have shown that clinicians rarely ask about the patient’s goals, educate the family about their role as a surrogate, or explain their decision in terms of the patient’s values [44].

Addressing the needs of family members may have an even greater impact on family satisfaction than the medical outcome. In a survey of 539 family members of ICU patients, there were higher levels of satisfaction among the family members of patients who died than among those who survived [45]. This difference was attributed to more family-centered care for patients who died.

Use of time-limited trials (TLTs; therapies used a defined period of time) has been shown to improve the quality of family meetings. One quality improvement study that trained physicians to communicate a plan of ICU care using TLTs, reported increased discussions with family members regarding risks and benefits of ICU treatments, values and preferences of the patient, and appropriate identification of clinical markers of improvement, compared with the preintervention period [46]. ICU length of stay was also shorter after the intervention.  

Decision-making — The importance of communication with families extends beyond its influence on satisfaction. Communication problems may also adversely affect clinical decision-making, since clinicians often need to rely on a family member to serve as a surrogate for the patient when trying to determine what care the patient would desire and how to proceed [47]. A study of 230 surrogates of patients on mechanical ventilation at high risk for death showed that poor quality of communication with the ICU physician was associated with surrogates lacking confidence in their role as decision maker. Conversely, high quality communication was associated with more confidence among surrogates and shorter use of life support for patients who died, presumed due to an improved decision-making process [48].

Shared decision-making strategies are endorsed by major societies who encourage collaborative decisions to be made by the physician and the patient or surrogate, taking into account the best level of evidence available and the goals and preferences of the patient [49]. Training of physicians in this process is recommended, however, the optimal decision-making strategy is unknown. Decision aids may not impact patient outcomes. One randomized trial of 277 critically-ill patients receiving mechanical ventilation for 10 days that included 416 surrogates and 427 clinicians, examined the impact of a web-based program on decision-making [50]. The program provided patients and clinicians with individual prognostic estimates and treatment options, and facilitated a decision in the context of the patient values and preferences. Clinicians and patients met a day after the intervention to have a family meeting to discuss the goals of care. Compared to standard care (ie, a face-to-face meeting without prior intervention), this web-based decision aid did not increase the prognostic concordance between the surrogate and clinician or change decisions about goals of care, nor did it or improve other outcomes including surrogate psychological distress or decisional conflict or patient length of stay or mortality.

Psychological impact — Communication between clinicians and families may affect the psychological state of family members. Families of patients admitted to the ICU are understandably stressed. Qualitative and anecdotal studies have described the experience as being characterized by uncertainty, emotional turmoil, and stress. Families typically attempt to resolve the uncertainty by seeking both information about the medical condition and resources needed to cope with the event [51,52].

Psychological conditions are common among the family members of patients in the ICU. This was illustrated by a study of 920 family members, in which the prevalence of anxiety and depression was 69 percent and 35 percent, respectively [31]. In another report, approximately one-third of family members developed symptoms that suggested that they were at moderate to major risk of developing posttraumatic stress disorder (PTSD) [53].

The psychological impact of death appears to be more prevalent when the patient dies without withdrawal of the ventilator [54]. Family members who prefer a passive decision-making role may be more likely to be anxious and depressed [55]. In contrast, family members who are present at the time of death or who perceived that communication was inadequate are at increased risk for PTSD symptoms [54,56,57]. Additional risk factors for PTSD symptoms appear to include a young patient and discordance between the family members' preferred decision-making role and actual decision-making role. In a study of 499 family members, more than half had developed a significant degree of learned helplessness characterized by decreased motivation, depression, and problems with figuring out causality.

Interventions to decrease psychological burden have mixed results. As examples:

●The presence of an advance directive or a DNR order reduced the sense of learned helplessness among family members of critically ill patients.

●In another study, effective involvement in decision-making allowed families to feel less helpless and more in control and help end the uncertainty and suffering they often perceive in the ICU [58].

●Another study reported that a communication facilitator may help reduce depressive symptoms in families as well as reduce costs and length of stay [59].

●An informational booklet on how families can engage in care rituals during an ICU stay was associated with a significant decrease in PTSD symptoms in family members 90 days after the patient’s death [60].

●A randomized trial examined the effect of an intervention delivered by the ICU team to surrogates of 1420 critically-ill patients who were at high risk of death or poor functional outcomes. The intervention was led by critical care nurses who received 12 hours of training, met with the family daily, prepared surrogates for clinician-led family meetings (within 48 hours of enrollment then every 5 to 7 days thereafter), attended meetings to provide emotional support and make sure essential content was discussed and debriefed with the family after the meetings. The intervention did not affect the median surrogate score on the Hospital Anxiety and Depression Scale (HADS score), although the surrogates perceived that the quality of communication and patient-and family-centeredness of care was better [61]. Additionally, in the intervention group the ICU length of stay was shorter (6.7 versus 7.4 days) and, while in-hospital mortality rate was higher (36 versus 29 percent), the six-month mortality rate was no different (60 versus 55 percent). This suggests that although the intervention did not improve, surrogate psychological burden it did improve decision-making and helped surrogates transition to comfort-focused care when that is aligned with patient values. There is also a suggestion that the intervention was associated with a shorter dying process in patients who are gravely ill, but that conclusion would require further exploration.

THE FAMILY MEETING — The purpose of a family meeting is to give health care providers an opportunity to meet the family, inform them of the patient's condition and treatment, answer questions, and include the patient's surrogates in the decision making process. The optimal method of conducting a family meeting is uncertain [15,62]. However, a growing body of literature has identified many key behaviors. Recognition that the skills necessary to conduct a family meeting can be taught has led to the development of methods to train clinicians to better communicate during family meetings. Such training tools are becoming increasingly available [63-67]. Educational resources, such as "Intensive Talk," a course designed to help faculty develop ICU communication skills training sessions at their home institutions, have been collected and organized on the IPAL-ICU website (http://www.capc.org/ipal/ipal-icu) by the Center to Advance Palliative Care (CAPC).

Purpose of a meeting — There is increasing evidence that family meetings can improve clinical outcomes for patients [1,4,68-70]. One study compared outcomes before and after the institution of a strategy of early (within 72 hours of admission) multidisciplinary meetings plus follow-up meetings with families of patients whose goals were not being achieved with advanced supportive care [1]. Length of stay decreased following the intervention, while mortality was unchanged. These effects persisted for four years [69]. A larger multicenter trial of this same communication structure, however, was ineffective at reducing length of stay [71]. The authors cited multiple possible explanations for the negative result, the most significant of which may be the fact that the intervention only mandated a structure for when meetings should occur and what topics should be addressed, but did nothing to address the communication skills of the clinicians running the meetings. Observational studies have found that either ethics consultation or inclusion of a palliative care consultant in family meetings was associated with shorter ICU lengths of stay and unchanged mortality [68,70]. A study of 79 liver transplant service patients in the surgical ICU found that integrating palliative care processes, including an interdisciplinary family meeting within 72 hours of admission to the SICU, increased goals of care discussions, DNR orders, and withdrawal of life support. DNR orders were placed earlier in the patient's course and SICU length of stay decreased by three days all while mortality remained unchanged, again suggesting an improved decision-making process for patients who die [72].

Family meetings may also improve psychological outcomes for families [2,34,45]. This was illustrated by a trial that randomly assigned 126 surrogates to receive either routine interactions with health care providers or interactions based upon a communication strategy called VALUE plus a bereavement brochure [2]. VALUE describes a communication system whose components include valuing and appreciating what surrogates communicate (V), acknowledging their emotions with reflective summary statements (A), listening carefully (L), understanding who the patient is as a person by asking open-ended questions (U), and eliciting questions (E). In the trial, the family members in the intervention group had less depression, anxiety, and PTSD symptoms. Many of the process measures of this intervention continue to improve at three year follow-up [73]. These results were not replicated in a separate trial of mandatory palliative care-led meetings for patients with chronic critical illness which showed no change in rates of anxiety or depression and the possibility of increased PTSD symptoms [74]. These results should be interpreted with caution, however, as the intervention did not represent usual palliative care practice. In this trial the palliative care physician or nurse practitioner usually met with the family without other palliative care team members and without ICU clinicians. Symptom management was generally not provided and prognosis and goals of care were almost always discussed on the initial meeting without significant longitudinal follow-up.

Some of the benefits derived from family meetings may be related to the opportunity that they provide for family members to express their feelings and concerns about the patient and the effect of the patient's illness on the family. This is important because families are less likely to experience conflict if they feel that the care provided was respectful and focused on the physical and emotional needs of the patient and the family. If care planning is viewed by the family as being patient-focused and if the family feels that all the planning was a cooperative effort between the family and the providers, it is more likely that the family will come away with a sense of certainty in the decisions that they made [35].

Such opportunities for expression should not have to wait until a family meeting. Communication between providers and the family should be a continual process. Less formal meetings with individual family members can occur more frequently to ensure that the family is kept informed about the patient's condition [75]. Many ICUs are inviting families to morning rounds [76], or developing web-based mechanisms to keep family informed [14]. Data suggest that family rounds are valued by families and lead to increased satisfaction [76,77]. How these different mechanisms are best integrated is not yet determined [77].

When is a meeting appropriate? — It is generally worthwhile to schedule an initial family meeting within the first 72 hours after admission. Early discussions appear to increase family satisfaction, reduce the length of ICU stay, and increase the likelihood of achieving consensus between the providers and the family about when care should be limited or withdrawn. (See 'Family satisfaction' above.)

Scheduling family meetings should not be dictated solely by the need to make an important health care decision. Family meetings may also be of value in the following situations:

●When the patient is first admitted to the ICU

●When there is a change in the patient's medical status

●When conflict is present among family members, between the patient and family, or between the family and providers

●When the health care team believes that the goals should be transitioned to a solely palliative approach

●When the family requests a meeting or providers believe a meeting would be helpful

Who should be invited? — If a previously competent patient expressed prior wishes that certain people not be included, these wishes should be honored. In the absence of such restrictions, the meeting generally should be open to all family members. Invitations to certain family members when there is a family conflict may be viewed as taking sides and may decrease the ability of providers to serve as impartial patient advocates. Moreover, having all family members hear the same information decreases misunderstandings and allows all the often varying views [78] of family members to be heard. At the very least, the legal decision-maker (if there is one) and all family members who have visited or called should be offered the opportunity to attend. Family members should also be asked if there are any friends or religious counselors they would like to attend the meeting.

It is important to invite all relevant health care providers to the family meeting. Both the primary care clinician and the ICU attending should be present, if possible. Clinical and prognostic information should be disclosed by a physician with a supportive role for members of the healthcare team (eg, nurses, social workers) to reinforce the delivered information and help facilitate emotional processing when the news is bad [71].

The evidence suggests that bad news is better received when it is delivered by the primary care clinician, unless the specialist has a prior relationship with the patient [79]. It is unclear whether this applies in an ICU where the critical care attending is involved in the day-to-day care of the patient and typically has more expertise with the patient's illness. However, involving a primary care clinician who knows the family and the patient has been shown to increase family satisfaction in this setting [31].

The nurse caring for the patient should always be invited to the meeting, since nurses frequently establish a close relationship with the family. The nurse can also reinforce information to the family between meetings and help support them in their decision-making [80]. Nurses see their involvement in family meetings as very important to the quality of patient care but they also identify barriers to being involved including a need for more training and not being asked for their perspectives [81]. A randomized study found that proactive participation of the nurse in physician-led family conferences resulted in less family anxiety and depression and their involvement was valued by the families [82].

A social worker and/or chaplain may be valuable additions to a family meeting. Social workers can be helpful with issues of family dynamics and individual coping mechanisms. Discussions between providers and the social worker prior to the meeting may help the providers better understand the family relationships and beliefs [83]. Chaplains can help with family support and discussion of spiritual, existential, and religious issues, topics that are often challenging and require advanced communication skills [84]. Having chaplains involved with families in the ICU, including discussing end of life wishes and preparing for a family conference, has been associated with family satisfaction with decision-making [85].

While it can be helpful to include the whole team in meetings, more needs to be learned about how to get all team members involved. Data suggest that physicians talk for about 80 percent of the duration of both pre-meeting huddles [86] and the actual meetings [87]. Additionally, while physicians rate collaboration as good, other team members do not always agree [86]. Models for training team members how to hold a meeting together exist [88], although further testing of their impact on ICU meetings is needed.

Some centers are exploring programs in which team members are specifically employed to support family members during meetings and ICU decision-making. One center gave a nurse additional training in order to be able to provide emotional, communication, decision, and anticipatory grief support to families. This intervention was well received by physicians and families and physicians believed it improved the timeliness and quality of the communication, particularly in how it kept communication patient-centered and focused on patient values [89]. Another site trained "communication facilitators" and found, in a randomized trial, that facilitators were associated with decreased family depressive symptoms at six months and lower cost and length of stay, especially among patients who died [59].

It can also be helpful to involve palliative care consultation teams, especially early in the decision making process. Sometimes palliative care is called in only after the decision has been made to withdraw support. Palliative care specialists receive specific training in advanced communication skills and can be particularly helpful when there is conflict or other factors that are making decision-making problematic. The team also provides multidisciplinary support for patients and families to help address physical, psychological, spiritual, and social suffering. A recent review of the literature concluded that involving palliative care early through the use of specific criteria helps reduce ICU length of stay without changing mortality which is thought to be due to facilitation of communication and decision-making [90]. Other studies have shown similar results [91].

Meeting before the family meeting — Family conferences bring together a number of different health care providers. It is important for the providers to be on the same page before the meeting as conflicting information can confuse families and leave family members feeling that the health care professionals are not communicating with each other [31,92]. One study found that 25 percent of families perceive that they are receiving inconsistent information from the various providers and this made decision making more difficult and affected satisfaction [92].  

It may be useful for the health care team to have a meeting before the family conference to discuss and come to agreement on the major medical facts, the prognosis, and possible treatment plans. Data collected by the social worker, chaplain, or nurse regarding how the family perceives information, their coping mechanisms, and any specific communication styles can be discussed at this time. When there is a language barrier and a translator is needed, it can be helpful to meet with the translator before the conference, especially if the conference will involve end of life issues because translators find these discussions often do not go well [93]. It can be particularly helpful to discuss the appropriate balance of interpretation and cultural brokering [94]. This may reduce the likelihood of serious mistranslations [95].

Uncertainty is common in ICU care, as are differences of opinion about prognosis and what should be done for the patient. However, the appearance of provider consensus at a family meeting is usually an important aspect of achieving family satisfaction and family consensus [96]. Thus, all providers should strive to reach a consensus position that can be communicated to the family. Any differences of opinion about prognosis or management that persist should be communicated in a balanced and straightforward manner that shows respect for all of the providers.

Clinicians need to consider their own beliefs and emotional reactions to the patient and the family prior to the family meeting. The evidence suggests that clinicians' unacknowledged fears of loss and death are a barrier to emotional engagement with patients facing death and nearly all providers are uncomfortable discussing death [96,97]. The longer that a clinician knows a patient, the more likely it is that the clinician will be overly optimistic and the less likely it is that therapies will be forgone. Additionally, a clinician belief that continuing life prolonging care is most appropriate has been associated with not even offering the family the option of comfort care, even an average of two weeks into an ICU stay in a population with a >70 percent mortality [98]. Self-awareness of these beliefs and feelings may be a helpful way to keep these issues from affecting the family meeting.

Finally, a facilitator should be chosen prior to the family meeting. Continuity of care is important for family satisfaction [99]. For this reason, one should attempt to have the same provider serve as the facilitator if there is more than one family meeting. No studies have compared physician facilitators with other health care providers, but usually the primary care clinician or the ICU attending serves in this role. All measures should be taken to ensure that the meeting is not interrupted, such as adjusting pagers to vibrate.

Where should the meeting be held? — The meeting should be held in a comfortable, private room. Tissues should be available. Seating should be arranged to allow all family members to sit in the front row. An effort should also be made to spread providers throughout the room rather than having the providers and family on opposite sides of the room. Providers should be seated during the conference as this creates the impression that one has time to talk and listen.

While most guidelines refer to in-person meetings, geography sometimes requires telephone or video meetings. During the COVID-19 crisis, visitor restrictions have made telephone and video communication much more common. A qualitative survey of ICU clinicians and family members or caregivers about phone and video interactions during COVID-19 visitor restrictions found that in-person communication was preferred by all contributors for important conversations. While clinicians worried about their ability to convey empathy, caregivers felt empathy was successfully conveyed. Suggested strategies for phone/video meetings included frequently checking understanding of individuals involved, including the patient on camera if possible, and offering time for the family member or caregiver and patient to interact without clinicians participating [100].

Understanding the family — Family meetings are similar to one-on-one clinician-patient interactions in some respects [79,101].

However, family meetings also differ from clinician-patient interactions in a couple important ways. First, there is a group dynamic that needs to be considered. Second, such conferences involve families, which are unique systems with their own rules of interaction and their own style of decision-making. A lot can be learned about the family system in advance of the meeting by talking to a social worker or a primary care clinician who knows the family. Paying careful attention to the family during the conference may also yield important clues about how the family interacts and the roles that particular family members play:

●Note where family members sit. The person who sits closest to the clinician may be the informal family spokesperson. Family members who deliberately sit apart may harbor old grudges. How the family negotiates who stands and who sits if there are not enough seats may provide clues about the family hierarchy.

●Observe whether the family members talk about themselves, whether the conversation is respectful, and whether questions focus on a single topic. Notice who is involved in the conversation and who keeps talking when there are interruptions.

●Note whether the family is empathic or dismissive when a family member becomes emotional.

While insight into a family's dynamics can help providers understand family reactions, it is crucial that providers not try to correct what they believe are dysfunctional family dynamics. Similarly, providers should stay neutral in family disputes. The family dynamics predate the patient's illness and reflect lifelong patterns of interaction. While these interactions may seem dysfunctional to health care providers, they may serve useful functions within the family system that are not immediately apparent to those outside of the system [83]. It is unlikely that a clinician who is meeting the family for the first time can influence these patterns of behavior, especially in such a stressful situation. Attempts to do so are likely to be viewed as intrusive, since the family did not request counseling. In the case of disputes, it is reasonable to say something like, "I can see that you all have some disagreements on a number of issues. I wonder if you could put these aside for now, so that we could try to focus on what is going on with your parent."

Starting a meeting — It is important to introduce all of the providers and family members at the start of a family meeting. Introductions should include name, position, role, and the level of involvement in the patient's care. This is particularly important in teaching hospitals where it is hard to distinguish between medical students, residents, fellows, and attendings on the basis of dress or age. Knowing clinicians' roles correlates with family satisfaction [31].

After making sure that everyone is comfortably seated and that all the relevant family members are at the meeting, the meeting's goals should be clarified. As an example, the facilitator might say on the first hospital day, "I want to spend some time talking with all of you about how your dad is doing. I also wanted to make sure that you all understand what we are doing for him and what we will be watching for over the next couple of days. Are there any other things that you want to discuss?" Providers should allow family members adequate time to speak during the meeting and not dominate the discussion [96].

Sharing clinical information — It is helpful to ask family members what they already know prior to explaining the patient's condition. Understanding the family's level of knowledge, emotional state, and degree of education allows for more efficient discussions and has been shown to provide important insights including perceived prognosis and causality [102]. Sample questions that can be used to start a conversation include the following:

●"What have the other doctors told you is going on with your father?"

●"What do you think is going on with your mother?"

●"To make sure we are on the same page, can you tell me what your understanding of this disease is?"

Clinical situation — The next step is to describe the situation to the family. It is essential for health care providers to have reviewed the medical record and test results prior to presenting information to the family or caregiver.

The types of information to be discussed generally include diagnosis, prognosis, and therapy. Family members or caregivers may prefer to hear all prognostic information even if their loved one is doing poorly [103,104]. Such information will reduce uncertainty about what is happening or is likely to happen, increase the family's or caregiver’s ability to act as the patient's surrogates in making decisions, and strengthen the relationship with all contributors [105]. It is important that the provider ensure that the information is understood by the family and caregivers. It is recommended that information be provided at a ninth grade level using as little jargon as possible. Providers should modify the way in which they are conveying information based upon a family's or caregiver’s understanding and communication style. Given that family members or caregivers are under stress, their ability to understand large chunks of information at one time is diminished. A rule of thumb is not to give more than three pieces of information without stopping and checking comprehension. Explaining physical manifestations of severe disease at the bedside, written leaflets, pictures, radiographs, and videos often help convey information and increase understanding [104,106,107]. In addition, lists of common important questions may empower families and caregivers and help them know what to ask in meetings [108].

Multiple short meetings may be better than one long meeting. At the first meeting, it may be enough to provide an idea of what is wrong with the patient and the plans for therapeutic interventions. The next meeting may focus on progress since the last meeting and prognosis. A following meeting might address the possibility of transitioning to a solely palliative approach if things are going badly. While some discussion of medical details may help reassure the family or caregiver regarding the medical team's competency, detailed pathophysiologic information is unlikely to increase a family's or caregiver’s understanding of the clinical situation. Long and detailed discussions regarding urine output, ventilator settings, and vasopressors may confuse individuals or lead to a situation where they focus on details to the exclusion of the bigger picture.

Prognosis — Most families or caregivers feel that clinicians are obligated to discuss prognosis and should not try to protect families or caregivers from the truth. Such information helps individuals support the patient and prepare emotionally and logistically for the possibility of the patient's death. While disclosure of prognosis leads to a variety of responses [109], most surrogates do not feel that prognostic information affects their ability to maintain hope [103].

How best to convey prognosis is a matter of debate. While many surrogates prefer numeric estimates to convey prognosis, clinicians feel variably in this regard [104]. How numeric estimates are conveyed affects family perception of risk. Prognosis conveyed as a frequency (ie, 1 in 10 chance) is perceived as higher risk than when it is conveyed as a percentage (ie, 10 percent). If estimates are conveyed qualitatively, discussing chance of death rather than chance of survival results in higher perceived risk [110]. It can be helpful to help families and caregivers "see the prognosis" by showing radiographic images or explaining physical findings [104].

Families occasionally do not want prognostic information, or they may desire more information than the patient wants. Unfortunately, there is no simple way to identify who wants information or how much. Experienced clinicians suggest asking how much information people want before giving information. One might say, "I want to give you as much information as you want about the future, but not overwhelm you. Some people want to hear about what I think may happen, while others do not want this information. Which kind of person are you?" It is important to check in with different family members if there is more than one person at the meeting [111,112].

Despite its importance, many clinicians do not like discussing prognosis because they worry that it will decrease surrogates' hopefulness with one study reporting that most family meetings in the ICU lacked important elements of communication [113,114]. Discussion of prognosis is especially difficult in the ICU because of uncertainty and the rapidity with which the clinical situation can change. This may explain why prognosis for survival is not discussed in more than one-third of ICU family conferences about major end-of-life treatment decisions [115]. Most prediction models focus on prognosis for survival, but discussing the patient's likely functional outcome if they do survive may be equally important [115-117]. The simple act of requiring intensivists to record functional prognosis made them more likely to discuss withdrawal of life support [118]. Seeing the prognosis documented made intensivists more likely to disclose prognosis to a proxy in a simulated encounter compared to when the prognosis was not documented [119].

Families and caregivers tend to be more optimistic about prognosis than clinicians:

●One study reported a 53 percent rate of discordance regarding prognosis between surrogate decision-makers and physicians. Much of the difference was due to misunderstandings by surrogates regarding patient prognosis [120]. The estimates by surrogates of prognosis tended to be less accurate than those of physicians.

●In another study of 169 surrogate decision-makers, 20 percent of surrogates estimated that the patient's prognosis was more than 20 percent more favorable than that estimated by the clinician [121]. Concordance was equally bad whether the clinician used numeric or qualitative terms to describe the prognosis. A greater degree of discordance was associated with less trust in the clinician.

●In a study of the confidence of surrogates in the clinician's ability to provide an accurate prognosis, 88 percent of surrogate decision-makers doubted the clinician's ability, although all of the surrogates still found the information valuable for informing decisions and preparing for poor outcomes [122].

Even when surrogates rate the quality of communication about prognosis highly, they are often inaccurate in their expectations about prognosis [123]. This suggests that families and caregivers interpret the data clinicians provide through their own cultural and informational lens. Asking about what they are hearing from other individuals or what they are learning from the web may provide clinicians with important information. Other factors that may impact a family member's or caregiver’s estimation of prognosis include a perceived need to remain optimistic and their own estimation of the patient's strength of character and life history [120,124,125].

Family or caregiver optimism about prognosis impacts care at the end of life. In a study of 275 ICU patients, 45 percent of surrogates held more optimistic expectations than physicians and this was associated with longer time on life support at the end of life [126].

The uncertainty that exists in describing prognosis is best dealt with by being open [127]. Discussions of uncertainty need to accomplish three tasks:

●Explain that uncertainty exists (some clinicians do this by talking about best case and worst case [128] while others use language such as "I hope" and "I worry" [129])

●Explain the factors that will help decrease the uncertainty

●Give an estimate of when more will be known about prognosis

Estimating prognosis can be challenging since accurate predictors of functional outcomes and survival are lacking in critically ill patients. Limited data suggest that predictions are dependent upon the outcome and confidence level of the individual caring for the patient. A prospective cohort study of five ICUs investigated the ability of critical care nurses and physicians to predict outcomes in patients admitted to the ICU for at least three days who were mechanically ventilated and/or on vasopressors [130]. Physicians most accurately predicted six-month mortality (likelihood ratio [LR] 5.91) while nurses most accurately predicted in-hospital mortality (LR 4.71). Predictions were enhanced when physicians or nurses were confident in their predictions (LR 33). Compared with a predictive model that included objective clinical variables (Acute Physiologic and Chronic Health Evaluation [APACHE] III), a model that also included physician and nurses predictions, had greater discriminative accuracy in predicting mortality and return to original residence. (See "Predictive scoring systems in the intensive care unit", section on 'Acute Physiologic and Chronic Health Evaluation (APACHE)'.)

To deal with uncertainty early in the ICU course, one might say, "Your father is very sick. As we discussed, we are doing everything we can to help his body recover from the infection. While we hope he is going to recover, we worry that he might not. The next 48 hours are going to be really important regarding how he does. The kinds of things we are looking for are whether his blood pressure gets better, how his kidneys are doing, and whether he wakes up."

It may also be helpful to discuss both the best, worst, and most common situations with the family [131]. This gives the family both the range of outcomes and a rough view of their frequencies. A phrase that may help the family look at both sides is, "hoping for the best and preparing for the worst" [132]. This phrase emphasizes that the providers understand the family's hope that the patient will survive the hospitalization, while planting the seed that the patient may not survive. The importance of including negative predictions is supported by a study of 141 patients with advanced cancer who communicated with 51 oncologists [133]. While optimistic statements of prognosis were common, only pessimistic statements were associated with an increased likelihood that patients would agree with their oncologist's estimated chance of cure. To ascertain the family's fears and concerns, it may be helpful to ask if family members have thought about what might be done if the patient does not improve. This conversation may also lead to a discussion of when it is appropriate to change the treatment to less aggressive care or to palliative care.

Taking a minute at the end of the first meeting to foreshadow possible outcomes and normalize common reactions may help, especially since the rapidity with which a patient's status can change in the ICU is hard for families to comprehend (eg, "But he was doing so well just this morning"). An example might be, "I have tried to describe to you what is likely to happen in the ICU. Unfortunately, when people are really sick, like your parent, things may change very rapidly. Your time here may well feel like a roller coaster. That is normal. It is why we try not to get too caught up in each and every lab test but try to see how your parent is doing overall. We will try to point out to you when the changes are something to worry about or when they are just bumps in the road that do not change her prognosis. We will get hold of you if there is an important change in her status."

The literature stresses the importance of maintaining hope while remaining truthful [134]. Surrogates may try to maintain hope in the face of prognostic information by focusing on small details, relying on gut feelings that the patient may be different, seeking more positive information from other sources or, rarely, avoiding prognostic information entirely. In the end, this is a difficult task and surrogates want physicians to help them hope for the best and prepare for the worst [135]. Clinicians can help with this balance in several ways [136]. Most importantly, providers should share the family's or caregiver's hope and describe what is being done to attempt to achieve the desired outcome. If one believes the family's goals are unrealistic, providers may acknowledge them and express their sorrow that they are unlikely to be fulfilled. Wish statements (eg, "I wish I could tell you that your parent is going to pull through") are a particularly powerful way to do this. Finally, providers can gently inquire about other goals. For most families, the initial hope is that the patient recovers and returns to his or her previous baseline. If this is not possible, families must be given time to understand and grieve the loss. Chaplains or other religious counselors may be particularly helpful in this process. The family can then turn to discussing what the patient's new goals might be. This might mean staying alive until all family can get into town, living until a fiftieth wedding anniversary, or dying in a dignified, peaceful manner.

Therapeutic plan — It is critical to describe what is going to be done to improve the patient's situation. This information helps them cope with their emotions by being able to plan for the future. In addition, the medical plan may be important to assure the family or caregiver that the clinicians are competent and trustworthy. Precise medical details are only important if they convince the family of the competence of the ICU team. It is more important that the family understands what the therapy means for the patient (eg, "a tube needs to be placed in a large blood vessel to give your father blood" or "he is going on the dialysis machine today to help out his kidneys, which are not working"). Common complications can be mentioned with an offer to answer any other questions the family might have.

The discussion should focus on what the family believes the patient's goals are, rather than asking the family to choose specific treatments. The family should be seen as the experts on the patient's goals and values and a careful, comprehensive values history should be obtained [137]. Unfortunately, audiotapes of family meetings show that clinicians do not spend very much time talking about the patient as a person and inquiries into the patient's goals and values are superficial [138].

Clinicians, on the other hand, are the medical experts and have the technical expertise to determine the most efficacious way to achieve the patient's goals and should offer recommendations about specific treatments that can achieve the patient's goals. As an example, there should be an explicit discussion about what will be done to keep the patient comfortable and what tradeoffs may exist between interventions that improve comfort and interventions that improve prognosis.

How success or failure of the therapeutic plan will be measured should also be discussed. This is particularly important when there is conflict and a therapeutic trial is tried. Often the only way to determine if a trial worked is to be clear about what counts as success and when it will be evaluated. As an example, for a patient who is septic and requiring vasopressors, the following might be said to the family: "The blood pressure is low because of the infection in his blood. We're giving him antibiotics to fight the infection and medicines to raise the blood pressure. If one more day passes and we are still unable to stop the medicines which are artificially raising the blood pressure, that will tell us that the antibiotics are not working. That would mean that this infection might not be treatable. Let's meet again tomorrow to see if we have been able to stop the blood pressure medicines." This is referred to as a time-limited trial and the various measures that are examined are termed clinical milestones [1,139].

Families are generally reassured by knowing that a large team is caring for the patient and that the patient is receiving multiple therapies in the hopes of improvement. Explaining all of the therapies being undertaken is a subtle way of reinforcing the seriousness of the situation. Paradoxically, it is even more important to stress the team's efforts when you think the discussion should focus on end-of-life topics. Families of critically ill patients often worry about giving up too early. Emphasizing the efforts that were taken may reassure them that the health care team wants the patient to improve and is trying all reasonable interventions.

End-of-life issues — Family meetings are an opportunity to foreshadow end-of-life discussions, allowing family members to psychologically prepare themselves for bad news. We and many experts believe that the possibility of death should be mentioned early in the ICU stay for appropriate patients and evolve as the clinical situation unfolds [104]. End-of-life conversations tend to be less jarring for families when they occur during a family meeting because they appear to be part of the natural progression of dialogue about the course of an illness. In contrast, waiting to raise the notion of palliative care until there is nothing more to be done may seem cold or disheartening to families.

When ICU interventions are unlikely to accomplish the patient's goals, it is appropriate to raise the issue of transitioning to a solely palliative approach [11]. Family members of patients who die in the ICU report a higher quality of dying and death when the ICU team held a family conference and discussed the patient's wish to withdraw life support during that conference [140]. Factors to be considered in making a decision to withdraw support are the patient's premorbid quality of life, the likelihood of survival, the likelihood of returning the patient to an acceptable post-ICU quality of life, and the pain and suffering that may result from ICU interventions [141-143]. Such conversations frequently occur later in the patient's illness than may be appropriate, a belief that is supported by the observation that do-not-resuscitate (DNR) orders are typically written only three days before death.

Discussion — Talking to surrogates about end-of-life issues raises unique concerns. Clinicians first need to operationalize the ethical principle underlying surrogate decision-making: family members should be guided by what they believe the patient would think if they could speak. This may be helpful to surrogates who feel as if they are being asked to make life and death decisions for their loved one. Rather than asking, "What do you think we should do next?" the clinician might say, "If your father was sitting here and could hear what we have said about his condition, what would he say?" or "What do you think your mother’s opinion on this would be?" Specific questions such as, "Have you ever talked to your mother about what her wishes would be if she became critically ill?" may also be helpful. Empirical evidence suggests that wording the question this way results in the family member being better able to predict what the patient would say [144]. Advance directives may help inform these discussions and should be reviewed if available. Recent data suggest that advance directives impact decision-making at the end of life, especially if created as part of an advance care plan that involves the patient, surrogate decision maker, and involved clinicians [145-147]. A systematic review showed that advance care planning is associated with decreased ICU admissions and reduced ICU length of stay [148]. Advance care planning has also been associated with surrogates reporting increased confidence in their role as decision-maker [48] and lower levels of decisional conflict [149].

Physicians do not often talk to surrogates about the patient's values or preferences. In a study of family meetings, there was no discussion of preferences or values in about one-third of cases [44]. It is hard to make decisions based on the patient's view if these views are never discussed. A theoretical article tried to delineate the various domains that one might talk to a surrogate about to better understand the patient's values: topics ranging from function to social values to longevity [137].

Even if an advance directive is available to aid in this discussion, family members may find these questions difficult. They often struggle balancing a desire to follow their loved one's wishes and not wanting to feel responsible for their death [150]. A common response is, "My father would never have wanted to be like this, but I can't tell you to stop." This answer gives the providers two important pieces of information. First, one knows that the ethically appropriate treatment plan is the one that is consistent with the patient's goals. Second, the answer emphasizes that the family members feel like they are being asked to make a life and death decision, which makes them feel uncomfortable. In end-of-life decision-making, we believe it is often important to take the decision-making burden away from the family by presenting a recommendation based upon a synthesis of the medical facts and the patient's values. Such value-centered recommendations have been incorporated into most guidelines for making decisions with families [151-153], although there is conflicting evidence about whether such recommendations are preferred by families [36,154-156]. In particular, families with low trust in their physician prefer more control over value-laden life support decisions [157]. Physicians also vary in what role they play in the decision-making process with some only giving information, some eliciting preferences/values, some offering recommendations, and some making independent decisions [158]. Given the variability of decision-making preferences, negotiation of roles can be helpful [159]. We suggest the phrasing: "Would it be helpful if I gave you my recommendation?"

A possible response to the above son who "can’t tell you to stop" using a value-centered recommendation would be, "I would never ask you to make the decision to stop. That is a horrible thing to ask a child to think about doing for their parent. What I needed you to do was to tell me what your parent would have said so that I could tell you the best way to respect their goals and values. And you have done that." After determining that the patient would never want to be kept alive if he could not hunt and fish, one might say, "Given these values, I think we should focus our goals on keeping your father comfortable." At this point, keeping the patient comfortable should be described. Finally, one might say, "If he gets sicker, I do not think we should put him on machines to prolong his dying process."

The specific language used to discuss end-of-life issues is particularly critical. Many clinicians, in an attempt to operationalize the principles of surrogate decision-making, ask what the patient would "want." We find this language evokes fantasies that may be divorced from the real options available and focuses too much on treatment options rather than goals. This then results in clinicians trying to convince the family why their choice may not be best. Instead, we recommend asking what the patient would "say" or "think" about the situation which then often leads to a discussion of goals and values [160]. Another example of the importance of precise language is that surrogates are more likely to choose a do not resuscitate (DNR) status if it is framed as the norm and if it called "allow natural death" [161]. It is also important to ensure that decisions to forgo a particular treatment are not viewed as "withdrawing care." Abandonment is a critical issue for dying patients and their families [162]. Families need to be assured that the patient will receive appropriate, competent care regardless of the stage of his or her illness.

Providers should pay attention to family dynamics throughout the process, since families typically make end-of-life decisions based on consensus even if there is a durable power of attorney. The legally appointed decision-maker still needs to get along with other family members after the patient dies. As a result, providers need to ensure that all opinions can be voiced and discussed in family meetings. It also may be helpful to leave the room and let the family members talk among themselves, since they may not wish to disagree in front of a stranger.

Family satisfaction with the care provided at the end of life is associated with the degree to which the decisions made were patient-centered and collaborative [41]. While shared decision-making is generally favored and associated with increased satisfaction, it is uncommon. In one study, only 2 percent of decisions met all of the criteria for shared decision-making [163].

It can also be helpful to ask about spiritual or religious concerns [34]. Unfortunately, even among highly religious families, one study found that discussion of spiritual or religious issues happened in less than 20 percent of ICU goals of care meetings [164].

Types of statements from clinicians that are associated with family satisfaction include assurances that the patient will not be abandoned before death, assurances that the patient will be comfortable and will not suffer, and support for the family's decisions about end-of-life care, whether the decision was to withdraw or not withdraw life support [165].

The following are some phrases that might be helpful when discussing end-of-life issues:

●"If your father were sitting here at the table, what would he say?"

●"Sometimes what you would say is different from what your mother would say. I really respect your ability to focus on her perspective."

●"I do not want you to feel like you are making medical decisions; that is my job. You can help greatly by telling me what your father would have said if he were here."

●"I want you to be able to look back on this six months from now and know you did all the right things."

●"Your father went to some trouble to make sure we knew his views regarding this form of treatment. We need to respect this decision like we would anything else he told us to do."

●"There are no more life and death decisions to be made, just type of death decisions."

●"How are you doing? What can we do for you?"

●"Are there any spiritual, religious, or cultural issues that we should know about?"

Estimating the value of fulfilling a dying patient’s wishes has been poorly studied. However, one study reported that implementation of an end-of-life program that fulfilled three wishes for the dying patient was valuable to family members and healthcare staff [166]. This study is discussed separately. (See "Palliative care: Issues in the intensive care unit in adults", section on 'Models for palliative care provision'.)

Family emotions — Families may react to these difficult conversations in a number of ways. Psychophysiologic responses include the fight or flight response (leading some family members to need to get up and walk around) and the parasympathetic response (leading some family members to become quiet, numb, and withdrawn). One may also see a variety of emotional reactions to impending death, including sadness, anger, grief, guilt, fear, and anxiety [167].

All of these reactions are normal. Launching into a biomedical discussion or providing false reassurance will convey to family members that providers are uncomfortable with their emotions. Supportive reactions include active listening and letting family members know that health care providers are present and can deal with their strong emotions. For the family members, knowing that someone understands what they are going through may be psychotherapeutic [168].

Two mnemonics have been developed that can guide clinicians during their interaction with surrogates and help manage the emotional impact of the interaction:

●NURSE describes how a clinician can be empathic when surrogates express strong emotions [169]. Its components include naming (N), understanding (U), respecting (R), supporting (S), and exploring (E), as described in the table (table 1). The use of more empathic statements has been associated with higher family satisfaction with communication [29].

●VALUE describes a communication system that can promote harmony between providers and surrogates. Its components include valuing and appreciating what surrogates communicate (V), acknowledging their emotions with reflective summary statements (A), listening carefully (L), understanding who the patient is as a person by asking open-ended questions (U), and eliciting questions (E) [2,170].

The purpose of attending to emotions is not just to allow them to pass, but to use them as a point of connection from which the clinician can infer what's truly important then develop a plan to address these needs [171]. Unfortunately, health care providers often miss opportunities to address family emotions [172]. In one study, clinicians did not make any empathic statements in one-third of family meetings [29]. This appears to be especially problematic when clinicians are meeting with families with limited English proficiency. In meetings requiring a translator, clinicians made fewer supportive statements such as valuing the family's input, easing emotional burdens, and active listening [173]. Some providers hesitate to address emotions because they fear that it will take too much time, but this concern may be unfounded. Studies of clinician-patient interactions show that addressing emotions shortens outpatient visits, reduces patient anxiety, elicits clinically important information, improves the clinician-patient relationships, and promotes satisfaction with the meeting [29,174-177].

Conflict — Conflicts between families and health care providers regarding end-of-life decisions are common. In one study, over one-third of clinicians reported conflicts with families, of which 63 percent involved decision-making at the end of life [178]. In another study, 46 percent of families reported conflict with providers, most commonly regarding communication [25]. Interestingly, there is little correlation between whether the physician and the surrogate even believe conflict has occurred, with one study showing that the physicians identified conflict in 28 percent of cases while the surrogates identified conflict in 43 percent of cases. Both physicians and surrogates identified conflict in only 13 percent of cases [179]. Conflict is also common between members of the clinical team [180,181].

When conflicts occur, clinicians may feel that their competence or judgment is not trusted. Distrust may be particularly high in the ICU because patients have often had extensive exposure to the health system prior to admission, then are cared for by ICU physicians with whom they have no long-term relationship [182]. The conflict may lead clinicians to redouble their effort to convince the family that they are wrong, or they may begin to question the family's competence or intentions. Families may feel isolated, misunderstood, or abandoned. In addition, they may feel pressured to give up or they may doubt the health care team's commitment to the patient's well-being. What should be a cooperative effort to achieve treatment goals can quickly turn into frustration and distress.

A strategy for dealing with conflict is to attempt to understand why the family is choosing therapies that the providers believe are unlikely to work [183-185]. Approaching the problem as a differential diagnosis helps ensure that all of the possible explanations for the disagreement are carefully considered. A list of possible reasons why families may disagree with providers' recommendations and potential responses are listed in the table (table 2). It is also important to acknowledge the emotions that underlie the conflict. The underlying issue is often sadness or frustration over the patient's dying [139]. Social work interventions to help address such issues improve decision-making in families that are at high risk for conflict [186]. Palliative care consultations can also be helpful in navigating this differential diagnosis. A study showed that palliative care specialists communicate differently than intensivists, using fewer task-focused statements and more relationship-building statements [187].

Another strategy is to use time-limited trials [188] and clinical milestones, as discussed above (see 'Therapeutic plan' above). It is often easier to withdraw an intervention when certain criteria are not met if the family understands the criteria that the team uses to determine the success of an intervention [139]. A recent study compared an academic medical center with low end-of-life treatment intensity with one with high end-of-life treatment intensity and found that the low intensity center was more likely to use time-limited trials of life sustaining therapy [189]. A quality improvement initiative in the medical ICUs of three academic public hospitals taught clinicians to use time-limited trials as their default approach with families or caregivers. This was associated with improved quality of meetings as well as shorter ICU length of stay and less intensive care with no difference in mortality compared to the pre-intervention period [46]. Time-limited trials are most likely to be helpful when initiated by the clinician in charge of the patient's care, in a setting that ensures continuity, and in medical conditions where evidence exists to formulate a clear timeframe in which the patient could be expected to improve or deteriorate [190].

Often, clinicians are confronted with families who say "do everything." Instead of seeing this as a conversation closer, it can be helpful to ask what "do everything" means to the patient. This can lead to a deeper exploration of the multitude of factors that may underlie the request. The physician can then offer a recommendation for a treatment plan that aligns with the patient's goals and values within the context of what may be achievable with the medical condition [191]. A variation of this request is the family that is "hoping for a miracle." Although there are no empirical data directly addressing the best way to address this question, we find a similar exploratory approach to be helpful [192]. We recommend sharing the family's hope for a miracle, addressing the emotion that often underlies such statements, exploring what a miracle would look like, and seeing if there are other things for which they are hoping. Often a "hope for the best, prepare for the worst" approach is helpful [132]. A chaplain or other religious figure can also be of assistance.

There are two issues that frequently underlie the conflict. First, families may need time to come to grips with end-of-life decisions. Family members are not familiar with the course of life-limiting illnesses and, early in the course of the illness, they may have been told to remain hopeful or that the ICU was where the patient was most likely to get better. Allowing a family a day or two to think about new information that they received may be enough for them to acknowledge a patient's terminal condition. A family may sometimes need to have small changes made over time, as illustrated by a study that found that withdrawal of all life-sustaining interventions took more than a day in over one-half of patients and that such stuttering withdrawal was associated with increased satisfaction [193]. Second, having a family member in the ICU is stressful and this may interfere with decision-making. One way to build a relationship with family members is to acknowledge the stress that they are under and to offer to help. Many families live far from the hospital and it may be helpful if the social worker can arrange for closer accommodations or less expensive parking.

Wrap up — Prior to finishing up the family meeting, the major issues discussed should be summarized. If the discussion focused on giving news, summarize the major issues. Giving family members a written summary may increase their knowledge and understanding of what was emphasized within the meeting. Another option is to ask the family to summarize the conversation. This allows the providers to make sure that the family has heard the major points.

The summary should focus both on consensus decisions and areas of disagreement. There should be a discussion about what will happen next. Always ask the family if they have questions, especially since it is likely that the family will forget many of the topics discussed.

Finally, clinicians must ensure that there is an adequate follow-up plan. This includes a family spokesperson for less formal conversations who can then communicate with other family members (particularly for large families), a way for the family to contact the providers if questions or issues arise, and a clear plan regarding when the next family meeting will take place. The plan can be a time or the criteria that will guide the occurrence of a next meeting. Given the importance of continuity, family members should be informed if a follow-up meeting will not occur and should be given an explanation so that they do not feel abandoned.

After leaving the room, it is often helpful to debrief as a team, especially if it was a particularly difficult family meeting due to emotions, conflict, or ethical issues. This can be an opportunity for teaching and to address any moral distress felt by team members [194].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Intensive care unit ethics".)

SUMMARY AND RECOMMENDATIONS

●Family meetings facilitate communication between health care providers and families of patients in the intensive care unit (ICU). Effective communication improves family satisfaction, and clinical decision-making. The effect on psychological burden is unclear. (See 'Importance of communication' above and 'Purpose of a meeting' above.)

●It is generally worthwhile to schedule an initial family meeting within the first 72 hours after admission. Scheduling family meetings should not be dictated solely by the need to make an important health care decision. (See 'When is a meeting appropriate?' above.)

●If a previously competent patient expressed prior wishes that certain people not be included, these wishes should be honored. In the absence of such restrictions, the meeting generally should be open to all family members. Family members should also be asked if there are any friends or religious counselors they would like to attend the meeting. (See 'Who should be invited?' above.)

●Family conferences may bring together a number of different providers. It may be useful for the providers to meet before the family conference so that conflicting information is not presented. (See 'Meeting before the family meeting' above.)

●All providers and family members should be introduced at the beginning of a family meeting. Family members should be asked what they already know about the patient's condition and then the clinical situation should be described. The types of information to be discussed generally include diagnosis, prognosis, and therapy. (See 'Starting a meeting' above and 'Sharing clinical information' above.)

●Family meetings are an opportunity to foreshadow end-of-life discussions, allowing family members to psychologically prepare themselves for bad news. When ICU interventions are unlikely to accomplish the patient's goals, it is appropriate to discuss other goals that are achievable. These may include comfort, dignity, and family member peace with the decision-making process. (See 'End-of-life issues' above.)