Interpersonal (mis)perceptions and (mis)predictions in patient-clinician interactions

Abstract

Patient-clinician interactions require an interpersonal exchange of information, preferences, expectations, values, and priorities. Given the brief interaction patients and clinicians are allowed, many barriers to effective communication exist, resulting in patients and clinicians leaving an interaction with discordant perceptions of what has occurred and what is to come. We review literature on concordance, and lack thereof, between patient and clinician perceptions, reasons why discordance may occur, as well as how it impacts patient care and outcomes.

When people think about medical encounters, they often imagine the simplest scenario in which a patient approaches their doctor with a symptom, the doctor recommends a treatment, and the patient takes the treatment and gets better. In reality, patient-clinician encounters are usually far more complicated. Even in simple cases where there is a clear diagnosis and a single best treatment, clinicians must still identify the multiple motivations and concerns that triggered the medical appointment (e.g., is this just about pain, or is the patient also concerned he might have cancer?). In more complex cases there may be multiple treatment options to choose from, each with a different amount of treatment burden, chance of benefits, and possible harms. In these cases, especially, the best treatment decisions depend on the patients’ preferences[1], and therefore it is particularly important that clinicians accurately perceive their patients’ motivations, expectations, and preferences. Although “concordance” has been operationalized in many ways[2], herein we use this term to refer to alignment between patient and clinician perspectives. Literature shows a consistent lack of concordance between patient and clinician perceptions. This can arise due to many factors but is deeply rooted in the ways that patients and clinicians communicate.

Clinicians often mispredict patients’ preferences, expectations, and priorities

Clinicians often misperceive patients’ preferences and priorities, and one of the more surprising reasons why is that clinicians and patients often systematically disagree about decision-making fundamentals, such as what qualifies as a treatment harm or benefit. For example, in cancer screening, clinicians usually characterize false positive results and overdiagnoses as harms. However, when forced to choose whether false positives and overdiagnosis were a harm or benefit, a large proportion of patients viewed them as benefits of screening[3].

Additionally, patients and clinicians disagree about which facts are most important in guiding their decisions. Patients considering breast reconstruction following mastectomy thought differences in immediate v. delayed reconstruction and avoiding use of a prosthesis were important, while clinicians focused on the impact of radiation and women’s satisfaction with reconstruction[4]. A systematic review examining agreement between patients’ and clinicians’ ranking of important treatment attributes (e.g., morbidity, mortality, health related quality of life) found that in most cases patients and clinicians did not agree[5]. Moreover, even if patients and clinicians rank the importance of treatment attributes similarly, patients may evaluate the likelihood of risks and benefits differently. Patients at high risk of stroke generally agree with doctors that preventing stroke is very important, but they are willing to accept medication with lower benefit and higher risk of side effects, relative to what physicians believe is appropriate[6].

Clinicians also frequently mispredict what patients want out of a clinical encounter. For example, one study found clinicians misinterpreted their patients’ desire for antibiotics about half the time, believing patients expected antibiotics when they did not[7]. These clinicians likely felt pressure to offer a prescription, when in reality this was based on a misperception.

Moreover, patient expectations for treatments are often also incorrect. In one study, patient and caregiver expectations of a chance of a cancer cure were only in agreement with the oncologist about 40% of the time[8]. Further, one acute myeloid leukemia study suggested that patients and clinicians often did not even agree on what symptoms a patient was suffering from[9].

Often patients experience multiple symptoms or comorbidities, requiring patients and clinicians to agree on priorities and make a plan together. Here, too, we find disagreements; patients with cancer who had chosen to undergo chemotherapy had complete concordance with their physicians in terms of their complex treatment plans less than 14% of the time[10]. Another study found that clinicians caring for patients with advanced chronic kidney disease were correct only 35% of the time when asked to identify the patient’s top health outcome priorities[11]. In one study only 44% of clinicians and patients agreed on the most important condition to manage[8].

In summary, clinicians and patients are frequently discordant in a variety of ways, including 1) their perceptions of harms or benefits of medical interventions, 2) how to weigh those harms and benefits in decision-making, 3) expectations for the clinical encounter and for treatment outcomes, and 4) healthcare priorities. These misperceptions have been observed in relatively simple clinician situations where the clinician makes a straightforward recommendation, as well as situations where there are multiple treatment options and the best choice depends on the patients’ preferences and goals.

Why do clinicians often mispredict patients’ preferences and motives?

Perhaps the most obvious reason why clinicians and patients have discordant perceptions is poor communication within a clinical encounter. Clinicians are not routinely trained to elicit their patients’ preferences, or to have the kinds of conversations that lead to more accurate understanding[12, 13]. Moreover, there is an inherent imbalance of power and knowledge between clinicians and patients. Clinicians often speak with jargon, frequently interrupt patients, and infrequently elicit their patients’ preferences[14]. Patients often assume clinicians want them to follow recommendations without asking questions or expressing their preferences. Patients may feel uncomfortable correcting their clinicians’ misperceptions about their preferences or understanding[15]. While communication barriers may be lessened when patients have more information to draw on (e.g. greater education or experience)[10] or when patients actively participate in their own care[16], pressures on clinicians to keep visits brief limit their ability to identify and overcome these barriers.

Another reason clinicians may systematically mispredict what their patients want is that people tend to weigh risks and benefits differently when making a recommendation for another person versus deciding for themselves[17]–[20]. When people make decisions for others they may focus on only one decision attribute (e.g., minimizing the chance of death) while ignoring other attributes that are important to the person affected by the decision (e.g., quality of life;[17], [20]). Clinicians may mispredict patients’ preferences because clinicians and patients view each decision from a different vantage point, with clinicians always making recommendations for another person, and patients weighing the harms and benefits from the perspective of the one personally affected.

Finally, a broad factor that may contribute to clinicians’ misperceptions of their patients are stereotypes and implicit biases[21]–[24]. Research has shown correlations between clinicians’ implicit bias and clinical care[25]–[29]. Implicit bias can negatively affect communication: When interacting with a White clinician who has high implicit race bias (but not necessarily high explicit bias), Black patients talk less and report poorer clinician communication and quality of care, and less of a feeling of being on the same team with their clinician[29]–[32]. Implicit biases could cause clinicians to make inaccurate assumptions about what their patient is like, what they want or are willing to accept, which could in turn influence what options are discussed or offered[33].

Together, a lack of preference-elicitation training, the power and knowledge imbalance, differences in decision making perspectives, and stereotypes and implicit bias likely contribute to the lack of concordance between patients’ and clinicians’ perspectives.

Improving concordance

Broadly, a patient-centered approach to communication can help improve concordance[34]. A patient-centered approach to care focuses on eliciting a patient’s concerns, their perspective on their illness, and their preferences and understanding, which can dramatically increase the concordance between patients and clinicians. One specific approach to patient-centered communication in clinical practice is shared decision making (SDM). SDM upends an older, paternalistic model of medical care, and is a style of communication in which clinicians actively elicit their patients’ care preferences, and collaboratively develop a treatment plan with the patient’s input[35]. While SDM is used in situations where there is no one best treatment option (and therefore the patient’s preferences should guide the decision), the tenants of patient-centered communication can improve concordance between the clinician’s and patient’s perspectives even when there is a single best option. Patient-centered communication encourages the clinician to assess the patient’s understanding of the disease, the treatment, and how the treatment fits with their lifestyle and values.

Training clinicians in SDM is key to addressing patient-clinician discordance, because SDM skills emphasize bidirectional communication and understanding the patient’s perspective. SDM could also increase clinicians’ awareness of the effect of the power imbalance, encouraging them to communicate in more accessible ways for patients (e.g. plain language). Well over 100 SDM clinician training programs have been developed, but only a fraction have been formally evaluated and implemented[14], [36]–[39].

SDM also encourages patient activation[40]--that is, patients’ ability and willingness to express themselves--which is necessary for clinicians to obtain an accurate view of patients’ perspectives. Decision aids (DAs) are a common tool to support SDM that provides information about a medical decision (e.g., options, benefits, and harms) and attempts to activate patients by showing them they can be involved in decision-making. Most DAs also help patients to clarify their values related to the decision. A Cochrane Systematic review showed that DAs increase patient knowledge and participation in decision-making[41]. Other methods for increasing patient activation, particularly among patients with chronic conditions, are currently being explored [42]–[44][45].

Another strategy for increasing patient-clinician concordance could be to work to decrease implicit and explicit biases. Recommended strategies for bias reduction include recognizing that you are fallible, cultivating motivations to be fair, displaying counterstereotypic examples, and diversifying talent[48]. Tested bias reduction interventions vary widely in approach and intensity[47]. A recent meta-analysis of diversity training programs found that training improved participant knowledge, whereas effects on attitudes and behavior were smaller and decayed relatively quickly[49]. While many bias reduction trainings exist, there is a need for evidence that these trainings change behaviors of interest, either within or outside of healthcare settings. More high-quality studies are needed.

In sum, concordance can potentially be improved by training clinicians in patient-centered communication and encouraging patient activation through SDM, DAs, and self-management. Additional work is still needed to identify an effective way of curbing explicit and implicit bias.

Impact of Concordance on Health Outcomes

When patient-clinician concordance is reached, patients benefit from improved health outcomes. Extensive literature on medication adherence has shown that physicians directly contribute to lack of adherence by failing to explain benefits and side effects of a medication adequately, not considering how treatment plans fit with patients’ lifestyles, and having poor relationships with patients[50]. In one study of patients with poorly controlled asthma, clinicians who were trained in patient-centered communication had patients who later showed not only better adherence to their asthma medications, but also better asthma-related quality of life and fewer asthma-related medical visits even a year later[51]. Similar SDM-focused interventions have been promoted in surgical decision making. One scoping review found that engaging in SDM increased patient knowledge, decreased decisional conflict, increased physician trust, and decreased surgical intervention rates[52]. Although improved SDM generally results in decreased use of major elective surgery[41], studies of SDM in populations that generally have low surgical rates have resulted in increases in surgical rates[53]. Taken together, this suggests that when patient-clinician concordance occurs we might reach the “right” rate of surgical procedures.

While patient-clinician concordance can result in better downstream health outcomes, discordance can negatively impact outcomes. For instance, a systematic review and meta-analysis focusing on adult psychosocial mental health interventions found that patients who received their preferred treatment had lower dropout rates from their treatment, and had greater partnerships with their mental health clinicians[54]. However, one study of symptom reports for patients with acute myeloid leukemia found that discordance in symptom reporting between patients and physicians lead to lower patient reported health-related quality of life[9].

Critically, patients benefit from patient-clinician concordance. When patients and clinicians have concordant perceptions, patients are more adherent to medication regimens, experience greater quality of life and improved treatment outcomes, and are less likely to drop out of mental health interventions.

Closing remarks

Patient-clinician communication has an increasingly important role in the process of medical decision making, because patients are increasingly asked to participate in care decisions. Currently, clinicians are not well-poised to consistently, accurately understand patients’ perspectives. Discordance may arise from power dynamics, poor communication, and biases; therefore, interventions increasing concordance, activating patients, and improving communication may help. However, additional work is needed to understand how best to empower patients to actively participate in decision making, as well as to train clinicians to become aware of how biases may impact their practice. If done well, patient-clinician concordance can lead to increased medication adherence, improved quality of life, and better treatment outcomes.