The Most Dangerous Procedure in the Hospital?
Communication is central to the human experience of illness, and therefore central to medical decision-making. Being an expert clinician now means being a skilled communicator. Fortunately, communication skills can be learnt, mastered and measured.
Communication is increasingly recognised as medicine’s most important non-technical skill. Perhaps this is self-evident: after all communication is how humans exchange meaning, reduce complexity, address uncertainty, manage emotions, inform, encourage, comfort and challenge. Communication is also central to the human experience of illness, and therefore central to medical decision-making. However, it is also complicated, nuanced and, consequently, error-prone. Therefore it should not be left to chance, or always left to junior team-members. Fortunately, communication can be learnt, mastered and measured (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014).
If communication is defined as “sharing, uniting, or making understanding common” (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014), then better communication is key to creating systems that are more trustworthy and patient-focused (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014). Whereas intensive care medicine previously focused on scientific discovery and technological advance, medicine can also be understood as a complex social system (Brindley 2010). Therefore intensive care medicine should now also make a “science of reducing complexity” and a “science of managing uncertainty” (St Pierre et al. 2008; Brindley 2010). Much of this will be achieved (or squandered) by how well we talk and listen.
Furthermore, patients often come to intensive care units (ICUs) following bad outcomes and bad decisions: not just bad pathology. Therefore, we are as much a Relationship Repair Unit (i.e. an RRU) as an ICU (personal communication, J Ronco). Overall, communication becomes one of our most potent ‘therapies’, and how we coordinate (or fragment) ongoing care, bolster (or impair) cooperation, and grow (or erode) trust (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014; Brindley 2010; Aron and Headrick 2002; Azoulay and Spring 2004; Azoulay et al. 2000).
This review cannot exhaustively cover (or comprehensively reference) a topic as capacious as communication. Therefore healthcare professionals should read widely. The goal should be to deliberately develop communication skills throughout our careers: such that ‘verbal dexterity’ matches manual dexterity and factual know-how (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014; Brindley 2010). This is because being an expert clinician now means being a skilled communicator (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014; Brindley 2010; Aron and Headrick 2002; Azoulay and Spring 2004; Azoulay et al. 2000). These skills will enhance difficult decisionmaking: whether during acute medical crises; during handover with colleagues, or, as is this article’s primary purpose, during discussions with patients and surrogates.
Medical Communication: The Basics
Communication skills are rarely innate, and do not necessarily improve through years of unstructured experience (eds. Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014). Similarly, communication is not onesize-fits-all, nor a panacea. However, communication training is associated with increased confidence, improved patient satisfaction, less anxiety, decreased depression and lower posttraumatic stress (eds Cyna et al. 2011; Dunn et al. 2007; Leonard et al. 2004). Communication can be a ‘placebo’ (i.e. good communication can reduce pain and anxiety) or ‘nocebo’ (i.e. bad communication can increase pain and anxiety) (eds Cyna et al. 2011). Better communication might also decrease litigation and maintain hospital reputation (eds Cyna et al. 2011). Accordingly, communication is everybody’s business: it should be taught to trainees, expected from practitioners and supported by administration (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014).
Communication is about listening as much as talking. When we do talk it is also about more than just what words are used (aka verbal communication) (St Pierre et al. 2008). We should also master good paraverbal communication: how words are said (pitch, volume, pacing and emphasis). Moreover, while this review focuses on verbal communication, non-verbal communication is just as important. This includes appropriate body language, suitable eye contact, response to emotions, the use of reflective silence and active listening (see below). We really cannot not communicate: failing to make the effort sends its own message (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014).
Why Communication (and Decision-Making) Is Often Difficult
Shannon and Weaver, working for Bell Laboratories, developed a model for verbal communication still relevant to medicine decades on (St Pierre et al. 2008). Simply put transmitters (i.e. speakers) encode messages, and receivers (i.e. listeners) decode them. However, both must be on the same channel (which in medicine could mean possessing similar situational awareness and emotional states), and there should be minimal interference (which in medicine could mean minimising chaos, stress, or cognitive bias). They also identified the danger of ‘channel-overload’ (which in medicine warns against communication that is unnecessarily complex). Overload, which often results in indecision, also occurs unless the receiver can filter data into usable information: You receive data (“his blood pressure is low”), but need to create usable information (“his body is failing”) (St Pierre et al. 2008).
Shannon’s model has limitations. Complex communication also requires meaning, which is harder to encode, transmit and decode. This is one reason why we cannot assume that patients and surrogates have reached the same conclusion as medical practitioners (St Pierre et al. 2008). This in turn explains why doctors are commonly criticised for failing to say what they mean, or mean what they say (“did you ever actually say ‘he is dying’?”). This model also describes communication as unidirectional (transmitter to receiver), whilst medical decision-making is commonly multi-directional, across disciplines and across hierarchies (St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014). Location should not affect data transmission, but it affects communication quality, impact and efficiency. For example, when transmitter and receiver are no longer face to face communication loses important non-verbal cues (St Pierre et al. 2008). This is why the medical telephone call is important to practise, and why confirming understanding by routinely summarising and repeating back is an important fail-safe (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011; Brindley et al. 2014).
Newer communication models focus on relationships, not just tasks. The ‘four mouths and four ears model’ (St Pierre et al. 2008) has sender and listener separated by a message with four equal sides: i) content; ii) relationship; iii) self-revelation iv) appeal (St Pierre et al. 2008). Content refers to facts and words. The relationship aspect means that senders reveal (consciously and unconsciously) how they regard receivers through specific words, intonations and nonverbal signals. Senders also indicate how they feel about themselves, namely a ‘self-revelation’. Fourthly, there is an appeal (or request) where messages encourage the receiver to do (or not do) something. These four aspects apply to both talker and listener, namely we ‘speak with four mouths’ and ‘listen with four ears’. This is often unconscious, and depends upon mental state, expectation, and previous interactions (St Pierre et al. 2008). Notably the sender cannot fully force the listener’s mind (and vice-versa). A practical example follows:
When a doctor says to a patient or surrogate: “What do you want me to do?” the doctor may presume he asked a unambiguous question which respects autonomy. Perhaps he did, but intonation can suggest otherwise. He may also have revealed his inability to make difficult decisions or even frustration about the patient’s premorbid state (“it’s too late...what do you expect me to be able to do!”). Also, the doctor’s self-revelation could be one of either appropriate patient concern or resignation (“I don’t have the time/training/authority for these complex cases…just tell me what do you want”). His request or appeal (albeit unstated) might be to try to subtly persuade the family (“I’m not sure ICU would be best”). In contrast, senior doctors may wish to minimise the patient/ surrogate’s sense of responsibility/guilt. In this case, communication should unequivocally state what they believe is not appropriate but also what can be done (“life support doesn’t treat this so won’t be offered. Instead, we will treat reversible conditions and maintain comfort”).
The surrogate decision-maker also listens with four ears, and any one can be more or less open. For example, a content-based response would respond with objectivity (“I want everything”). If the family member hears the self-revelation he might reply: “she’s been through a lot, but she’s a fighter”. If the family member is attuned to relationship aspects, or has previous disappointments from the medical professions, then they may be more defensive (“You doctors give up too easily: I want everything”). Only rarely will the listener have the state of mind to decipher the appeal: “so what I think you’re telling me is..”. Regardless, this model shows how communication can create a virtuous cycle that builds cooperation, or a vicious cycle that destroys it (St Pierre et al. 2008; Brindley and Reynolds 2011).
Communication is affected (both in meaning and interpretation) by the paraverbal (volume, tone, pitch, pacing) and non-verbal (angry eyes; furrowed brow) (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011). These in turn are affected by subconscious emotions and attitudes. If verbal and non-verbal are incongruent (words say one thing; expressions another), then receivers typically deemphasise words, and amplify the importance of tone and body language. With incongruence receivers typically default to what they (or we) expected (“he said X, but I know what doctors mean”). Incongruence promotes misinterpretation, and can be interpreted as disingenuous (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011). Congruence is even more important when those involved are unfamiliar, or when the medical situation is novel, which is almost always for families! Investing the time to establish ‘rapport’ (usually defined as ‘common perspective’; ‘being in sync’ or ‘shared mental model’) can facilitate all future interactions (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley et al. 2014). The effort demonstrated can also reinforce the patient’s psychological reserves and resilience (eds Cyna et al. 2011; St Pierre et al. 2008; Brindley and Reynolds 2011).
Helping Patients and Surrogates Feel Safe to Communicate
As popularised by Pincince (2013) and others, decisions are best made when people are not in hot emotional states (anger, stress). To ‘cool’ emotions, patients and surrogates need familiarity, which can be undermined by changing staff too frequently; predictability, which can be threatened by keeping families waiting; intact support systems, which can be destabilised if families feel excluded and a sense of control, which can be weakened by illness and unfamiliarity. Hot emotional states lead to anger or intransigence (aka ‘neural hijacking of the rational brain’). The psychologist Stephen Porges explains this idea using the polyvagal theory: when calm our vagus nerve can facilitate engagement; when angry it stimulates conflict ( Porges 2011). These ideas apply to those that cannot talk (endotracheal tube; stroke), and those that do not talk (fear, confusion, deference) (eds Cyna et al. 2011; Brindley et al. 2014). For patients already burdened with illness, not being able to verbalise, and not being understood, can accelerate a downward spiral into frustration and disengagement. Also when healthcare workers do speak we may speak differently than patients (eds. Cyna et al. 2011). Physicians often use technical language and focus upon gathering information and delivering news. Patient language and surrogate language often focuses on beliefs, fears and hopes, which explains why they may cling onto any positive news (“so it’s not the worst you’ve ever seen!”). Patient and surrogate coping strategies may include denial or aggression, whereas caregivers intellectualise to protect emotions. Communication that is sensitive, but objective, can bridge the caregiver’s ‘scientific world’ and the care-receiver’s ‘natural world’(eds. Cyna et al. 2011; Brindley et al. 2014).
Communication Tools That Can Aid Decision-Making
Especially when communicating bad news, doctors and nurses should understand that, while routine for us, for families these are sentinel moments, unlikely to be forgotten (eds. Cyna et al. 2011; Brindley et al. 2014). Combined with ‘active listening’ (where listeners pay close attention and use feedback or rephrasing to demonstrate engagement and understanding), the effort put into communication is a way to demonstrate non-abandonment (eds. Cyna et al. 2011; Brindley et al. 2014). Communication tools and bundles (see below) can also provide structure and reliability to complex communication. However, they should never make interactionsrobotic and devoid of personal connection.
We also have tools to audit communication with patients
and surrogates (Black et al. 2013; Davidson et al. 2007). Black et al. (2013)
promoted a communication bundle with six requirements within 24 hours:
identification of i) the surrogate-decision maker; ii) code status; iii)
advance directive; iv) pain v) dyspnoea, and also vi) distribution of a brochure.
Four additional goals should be met within 72 hours: i) family meeting, ii)
discuss prognosis, iii) assess patient-specific goals, and iv) offer spiritual
care. This approach emphasises that decision-making requires more than just
data transmission. Patients are validated as people with values (not just
diseases) and part of a larger ‘lifesupport system’ that includes family,
friends and community (eds. Cyna et al. 2011; Brindley et al. 2014 Black et al.
2013; Davidson et al. 2007). All of the above should also explain why patients
and surrogates value clinicians’ communication skills at least as much as their
technical skills (Heyland et al. 2002). For difficult decision-making (and even
for simple decision-making!) communication is likely our greatest clinical
asset, or keenest liability.
This article is adapted from a more comprehensive review of high-stakes medical communication in the Handbook of Intensive Care Organization and Management (Editors A. Webb and G. Ramsay; publication pending).
- Communication is central to the human experience
of illness, and therefore central to medical decision-making.
- Being an expert clinician now means being a
- Fortunately, communication skills can be learnt,
mastered and measured.