The Person-Centred Clinical Interview Free

The recent review by Giovanni Fava and colleagues [1] is an important reminder of the crucial role of the clinical interview in clinical practice. The review focuses predominantly on the clinical reasoning component of the clinical interview, as does the invited editorial by Dubovsky [2]. The other related editorials by Kroenke [3] and Starcevic [4] extend beyond clinical reasoning but none of these papers offers a direct answer to the questions: what do patients want from the clinical interview, and how can their priorities be best met? These questions are the focus of the present editorial.

In previous times, the choreography of the clinical encounter, and hence also the clinical interview, was relatively straightforward, with the clinician leading. Patients consulted doctors for their clinical expertise – to be asked about their symptoms, to receive a diagnosis, and to be given advice on treatment and outcome.

In the last two decades in particular, this process has changed with the emergence of person-centred and values-based medicine and also with the rise of consumerism in medicine. The theme common to all of these is the transformation of the clinical encounter, giving greater prominence to the role of the patient. The hallmark of person-centred health care is treating patients as individuals and as equal partners in the process of healing [5]. Values-based practice emphasises the centrality to decision-making of the patient’s values and preferences [6]. These offer two models of the contemporary clinical interview. These models have been developed independently of one another and therefore give the impression of perhaps being more distinct from one another than they probably are, even though they share similar underlying principles. In what follows, person-centred is intended to also include values-based. Narrative-based medicine is a further development of person-centred care [7]. The narrative interview is in some ways different from the person-centred interview [8] but will not be considered further in this editorial. If the relationship between clinicians and patients has changed to be more collaborative, what qualities and skills do patients expect of their clinicians to facilitate this?

In general, particularly in primary care, patients value their clinicians adopting a patient-centred approach [9]. However, this depends on the setting – emergency physicians are more directive than primary care doctors [10].

In the patient-centred interview, clinician behaviours which patients particularly favour include the use of open questions, summarising, and giving attention to what the patient brings to the consultation [11, 12]. Agenda setting is also valued (see below). Empathy shown by clinicians can have positive effects on patient outcomes [13]. Patients respond to clinicians’ non-verbal cues, particularly tone of voice, eye contact, and facial expression [14]. Where patients perceive contradictions between their clinician’s verbal and non-verbal behaviour, the non-verbal cues tend to override the verbal ones [15]. Many clinician behaviours endorsed by patients come under the umbrella of active listening. This involves attentive body language (including appropriate eye contact and behaviour showing engagement), giving the patient the opportunity to give their account in their own way, and reflecting skills like checking understanding and summarising. Particular clinician behaviours have been demonstrated to facilitate patient participation [16, 17] although, curiously, in one study, inhibiting behaviours by clinicians actually increased patient participation [16]. This relationship is often reciprocal – clinicians become more patient centred when with patients who participate more in the interview [18]. Similarly, if patients are involved in their care, they can be expected to gain more interest and understanding in appraising data concerning their health status [19], and they can then collaborate more effectively as producers of their own health [20].

As Kroenke and Dubovsky noted [2, 3], the context of the clinical interview may have a marked effect on these desired interactions, even when clinicians have the necessary skills and experience. Having limited time for the interview is an important factor. Where current practice in many countries and settings requires the clinician to make notes of the interview electronically, patients sometimes complain that during their consultation, their clinician spent time communicating with their computer and made no eye contact with them [21]. One longitudinal study found that primary care clinical interviews became less person centred over time, and the researchers queried whether this might have been due to the introduction of electronic note taking [22]. However, another study failed to find any difference in patient satisfaction before and after the introduction of computers into consultations – 75% of this sample considered that the clinician’s use of the computer had a positive effect on the consultation [23]. Positive views by patients regarding the clinician’s use of a computer in the consultation were associated with patients’ appraisals of the competence of their clinician in using the computer [23]. This is perhaps related to how intrusive patients find the clinician’s use of the computer during their consultations. Few clinicians have the luxury of their interviews with their patients not being compromised by time or resource constraints or the interposition of a computer between themselves and their patients. Acknowledging such restraints in the interview is likely to be helpful to the clinician-patient relationship, and some suggestions have been made about how this might be introduced to patients [21]. However, it appears that clinicians seldom consider this. This topic would be worthy of research.

Two elements of the person-centred interview in particular demand collaboration between the clinician and their patient – agenda setting and shared decision-making. Even when clinicians claim that these are integral to their clinical practice, research indicates that these activities are neither comprehensively nor consistently collaborative.

Important features of agenda setting include eliciting a full list of agenda items before moving on, postponing premature elaboration of any details, prioritising items, and confirmation of priorities [24]. Establishing a conversational focus is also important [25]. The start of the consultation offers a good opportunity to set the interview off in the right direction. Echoing comments by Kroenke [3], in a study of family physicians, patients’ opening statements were interrupted on average after 23 s but when patients were allowed to complete their opening statements, this took only 6 s longer on average and patients were then also less likely to offer further queries later in the interview [26]. In another study involving clinicians from different specialties, the clinician interrupted after a median of 11 s while uninterrupted patients took a median of 6 s longer to state their concerns [27]. When patients attending an internal medicine clinic were allowed to talk uninterrupted until they themselves declared that they had finished their opening statement, the median time taken was 59 s and the clinicians judged that the patients were giving valuable information and should not be interrupted [28]. Compared with untrained physicians, physicians who received training in agenda setting elicited patient concerns more comprehensively and promptly and did not have additional concerns emerge later in the interview, with no significant differences in interview length [29]. In this study, physicians had 16 h of training and coaching over four consecutive weeks.

The other crucial collaborative aspect of the person-centred interview is shared decision-making. This has attracted increasing interest, for example, as an essential component in lifestyle medicine [20]. Shared decision-making involves those participating in a clinical decision (usually patient and clinician but sometimes also family and/or carers) collaborating to make the decision, taking into account the patient’s preferences and priorities, as well as the best available clinical evidence. Decision aids may be useful before or during these discussions [30]. The underlying principle is that shared decision-making is a meeting of experts – the clinician as the expert in clinical judgement and evidence-based care and the patient as the expert in their experience of their illness and their views regarding risks, preferences, and priorities. The clinician’s tasks include facilitating the patient’s involvement in their collaboration and actively contributing to making the discussion one between equals rather than reflecting the imbalance of power between clinician and patient in “traditional” clinical interviews. Some have argued that shared decision-making can be effective in engaging patients who may be disadvantaged [31, 32], and there is some evidence for this [33]. Another skill required of the clinician is to be able to negotiate which decisions particularly benefit from shared decision-making. These are particularly decisions involving equipoise [31], but not exclusively so. Given that the key purpose of decision-making is respecting and incorporating the patient’s views, it is also important to note that some patients under some circumstances prefer to let the clinician make decisions without collaboration [34].

An early systematic review concluded that shared decision-making could be effective, especially when patients with chronic illnesses needed to make long-term decisions [35]. Despite much subsequent research, there is still relatively little evidence that shared decision-making improves the course of the patient’s illness although there is consistent evidence for improved patient satisfaction and reduced decisional regret. A Cochrane review concluded that no definite conclusions could be reached about the benefits of shared decision-making because of the low certainty of the available evidence [36]. However, there are numerous isolated studies reporting more encouraging results. Examples include a systematic review involving elderly patients in primary care, reporting the benefits of shared decision-making in improving symptoms and reducing hospital visits as well as patient quality of life [37]. Also in primary care, shared decision-making reduced short-term use of antibiotics [38]. In a surgical clinic, patient-focused interviews increased patient knowledge and reduced decisional regret [39]. A systematic review involving patients with lung cancer reported that, in some studies at least, participating in shared decision-making reduced anxiety and depression and also led to reductions in healthcare utilisation [40]. As noted above, decision aids are sometimes used to facilitate shared decision-making and there is evidence that these can be effective [41, 42].

Problems in demonstrating the effectiveness of shared decision-making include the multiple models of shared decision-making, each emphasising different core features and procedures [43]. Intervention studies commonly neglect to identify the model used, sometimes include little information about how clinicians were trained, and often give little if any information about the success of that training. Adverse outcomes are seldom reported. What is clear is that simply aspiring to practise shared decision-making is not enough and being able to incorporate this effectively into the clinical interview requires training [44, 45], probably including real patient scenarios [46, 47]. Clinicians trying to implement shared decision-making sometimes revert back to their previous patterns of communication with patients [48] or fail to implement shared decision-making even when they intend to [49]. Despite the widespread endorsement of shared decision-making as a key component of the clinical interview, its successful implementation in practice has been disappointing.

One reason is the assumption that shared decision-making will inevitably lead to longer interviews [50]. In some settings at least, shared decision-making does not appear to lead to longer consultations [51‒53] although others have reported the opposite [54]. Even when a single decision-making interview is longer than its “biomedical” counterpart might have been, it might save time and resources in the longer term by reducing overdiagnosis and overtreatment [51]. Nevertheless, successful implementation probably requires organisational endorsement [31, 36]. In the United States, at least one state offers reduced malpractice insurance premiums to clinicians trained in and implementing shared decision-making and the Centers for Medicare and Medicaid Services have mandated shared decision-making for particular cardiac procedures [32]. A similar approach has been taken in the Netherlands [55].

Another potential collaboration in person-centred care is the involvement of the multidisciplinary team. While multidisciplinary team meetings are the norm in many clinical specialties, their involvement with the patient in shared decision-making is less common and the common practices of team decision-making might even discourage direct involvement by patients [56].

Self-reflection is an essential career-long skill for all clinicians, but it is particularly important in the clinical interview. All clinicians are aware of clinical interviews that have not gone as expected and reflecting on the reasons why this might have happened is important for both patients and clinicians. This is one of the functions of Balint groups and Schwartz rounds, but there are also other methods of fostering self-reflection. For clinicians, knowing why some patients arouse irritation or other negative emotions is particularly important if interviews are to be person-centred. Physicians show different levels of respect towards individual patients, and this is manifested in their behaviour towards their patients [57]. Consultations improve when physicians show behavioural adaptability, the ability to modify their behaviour in response to their patients, leading to a better match between clinician and patient [58]. It has been assumed that clinicians’ emotional intelligence enhances their practice [59] although not all evidence to date has not supported this [60]. Patients’ satisfaction is negatively related to their clinicians’ stress levels [61]. As already noted, patients respond to their doctor’s non-verbal cues, particularly tone of voice, eye contact, and facial expression [14].

Many of the potential difficulties in the interview present in subtle ways. Before it became routine for doctors to receive training in psychological medicine, it was observed that when terminally ill patients mentioned their emotions during consultations, their doctors tended to change the discussion to pain control, which they considered themselves more competent to manage [62]. In an oncology clinic, interviews rated as less satisfactory by patients were also rated as such by their physicians but the physicians rated interviews later in their clinics as less satisfactory than earlier appointments, even though their patients’ ratings showed no distinctions between early and late appointments [63].

It is widely recognised that people from minority groups are commonly disadvantaged in receiving health care. One commonly implemented solution is training in cultural competency – learning about the values and practices of different cultural groups. However, this approach is necessarily restricted (clinicians cannot learn about all the cultures they encounter among their patients) and has been criticised for perpetuating cultural stereotypes. A preferred alternative is to train clinicians in cultural humility – being open to the aspects of cultural identity which are most important to their patient [64]. This fits well into the person-centred interview.

The COVID pandemic led to consultations moving from face-to-face to video based, and this has persisted to some extent following the end of the pandemic. Regardless of the overall relative merits of video versus face-to-face interviews, many of the tasks described above can be accomplished in a video interview. Paradoxically, during video interviews, patients might not notice that they are competing for attention with the clinician’s computer because the patient does not see the clinician interacting with the computer. A recent systematic review reported that in psychiatry, face-to-face interviews were superior to video interviews for people with eating disorders but otherwise video interviews had no significant disadvantage and were superior to face-to-face interviews for people with some diagnoses [65].

Two generations ago, almost all the knowledge physicians required for their practice could be contained in the textbooks on the shelves of their offices. This has changed dramatically with the rapid advances in medical knowledge and the growth of evidence-based medicine. It is no longer likely that a clinician will be able to answer from their existing knowledge all questions arising in the clinical interview, whether from the patient or otherwise. Some clinicians may be reluctant to express uncertainty in their interviews with patients, but there is evidence that doing this is associated with greater patient satisfaction [66]. The clinician telling a patient that they do not know the answer to a particular question and will search for the answer before the next appointment is seldom construed by the patient as a reflection of the clinician’s incompetence, but rather is appreciated as an indication that the clinician is making extra efforts to understand the patient’s illness and situation.

More problematic, as Fava and colleagues acknowledged [1], is dealing effectively with the substantial information resources now available to patients. The editorial by Vladan Starcevic [4] has already reviewed some of the challenges of online information seeking by patients. As Starcevic noted, access to information online has the advantage that this can contribute to patients’ empowerment and should in principle therefore be encouraged. However, a major disadvantage is that patients cannot be expected to be able to discriminate between reliable online sources and others. This is particularly problematic with social media and with the development of artificial intelligence resources, specifically Large Language Models (LLM) like Chat-GPT and CoPilot, with chatbot interfaces. The huge rise in conspiracy theories during the COVID pandemic was in part due to groups on social media espousing particular views. Someone joining such a group is likely to find its membership reinforcing the group’s views. This is also reinforced by the social media providers’ algorithms – having reached a particular site, the algorithms direct users to similar sites. This can cause what has been described as an “echo chamber” for disinformation, with consequent confirmation bias. Social media providers have made some attempts to manage such disinformation but with limited success since those employed to moderate sites disseminating disinformation are hardly ever clinically qualified [67]. A recent study showed that LLM platforms were inconsistent in their safeguards against generating and recognising deliberate misinformation [68, 69]. LLM chatbots are not optimised to give accurate answers but rather plausible ones. Chatbots sometimes give incorrect information (termed “hallucinations”) but because the answers aim to be plausible, such incorrect information can be difficult to detect for the clinician, let alone for patients. Curiously, if an answer from Chat-GPT is fed back to Chat-GPT with the instruction to identify any errors, it might identify any “hallucinations” in the previous answer [70].

It is difficult to envisage how information which the patient brings from social media or LLM chatbots can be discussed or otherwise accommodated in the clinical interview. As Starcevic noted [4], clinicians can recommend reputable internet sources, such as the websites of large healthcare organisations, medical charities, and specialist societies. However, this does not deal with the potential problems just outlined of large amounts of information, potentially including disinformation, intruding in the clinical interview. One option that to my knowledge has not yet been tried is to offer patients who wish to gather their own information from the internet a session with a medical librarian prior to the clinical interview. This would be worth testing in a randomised controlled trial. However, there are already other innovative potential solutions, such as User-Driven Health Care, in which patients, professionals, and even medical students can collaborate online to learn about health and illness [71].

An early publication on shared decision-making was subtitled “it takes at least two to tango” [72]. This creates an appropriate image of the person-centred interview as an active collaboration between clinician and patient, sometimes led by one party and sometimes by the other. Person-centred interviews are consistently associated with increased patient satisfaction, but the evidence for benefits to wider patient outcomes remains inconsistent. This appears largely attributable to the variable quality of the relevant research. Given that psychosomatic medicine is based on the biopsychosocial approach to ill health, and key biopsychosocial factors will vary from person to person [19], it is impossible to envisage a truly psychosomatic approach without the person-centred interview. It is clear that for clinicians, aspiring to be person centred is not enough – where benefits have been demonstrated from person-centred interviews, the clinicians involved have received specific training. There is no doubt that patients prefer person-centred interviews. The benefits of these, in terms of patient empowerment and potentially improving the experience of health care by disadvantaged groups, make person-centred interviews worth pursuing.

The author has no conflict of interest to declare.

This study did not receive any funding.

Tom Sensky is the sole author of this editorial.