Clinician Response to Patient Emotion: Impact on Subsequent Communication and Visit Length

INTRODUCTION

Clinician empathy in health care settings has been defined as a communication behavior “that demonstrates an accurate understanding and acceptance of the patient’s feelings or concerns,”¹ and it is generally believed that providing empathy in response to patient emotions is an important part of the clinician’s therapeutic response.^2,3 Several communication coding systems have facilitated the study of emotional communication; these systems typically involve procedures for identifying and coding emotional expressions made by the patient (an empathic opportunity) and then coding different types of clinician response.^4-6 Using these methods, studies have repeatedly demonstrated that clinicians miss most opportunities to express empathy in response to patient emotional expressions.^7,8

One hypothesis for why clinicians do not make explicit empathy statements is that they may not feel they have enough time to address their patient’s emotions in addition to all the other competing concerns. To alleviate clinicians’ concerns about the time involved in responding to emotions, it is argued that providing empathic responses to patients saves time. The evidence for this claim is based on a 1997 study of 116 patient-clinician encounters, which discovered that visits with missed opportunities to acknowledge patients’ feelings averaged 3 minutes longer than visits with a more positive response.⁸ The same study also found that in 53% of visits with missed opportunities, the patient brought up the same concern more than once. This finding suggests that patients who have not had their emotional state adequately acknowledged and do not perceive being heard by their clinician may feel compelled to repeat themselves.

The conclusions from this landmark study are appealing for anyone who believes that expressions of empathy are important, particularly for those who teach communication, yet it has never been replicated. Knowing the impact of clinician communication behaviors on visit length is highly relevant for clinicians, who have to make choices at each moment in each encounter about how to spend a fixed amount of time in the best interests of their patients. We therefore designed this study to evaluate the relationship between the clinician’s response to a patient’s emotion and subsequent communication, including the patient’s repetition of emotional expressions and length of visit, using a comprehensive, detailed emotional communication coding system.

METHODS

RESULTS

DISCUSSION

Emotional communication may have a more complicated relationship with visit length than previously thought. We found that clinicians vary in how they respond to patient emotions as time passes during visits, and that empathic statements are rare and not associated with visit length. Our results add nuance to the interpretation of an earlier landmark study by Levinson et al⁸ demonstrating the timesaving nature of empathic emotional responses. We found a trend suggesting that providing space for patients to elaborate could be associated with longer visit length. We also found, however, that responses explicitly focusing on patient affect were associated with shorter visit length, which perhaps provides a patient-centered solution for acknowledging emotion without necessarily taking more time and is of practical utility when teaching about communication skills.

The comparisons between our results and those of the previous landmark study are worth a deeper analysis. Levinson et al⁸ included in their study 65 emotional expressions in both surgical and primary care encounters, and used a different methodology to characterize clinician responses to patient emotional expressions as either positive or missed opportunities. These categories were therefore evaluative (ie, good vs bad), whereas the VR-CoDES is descriptive. It has been the subsequent interpretations of the study by Levinson and colleagues that have emphasized the benefits of “empathy” specifically, but this interpretation of empathy may differ from the more limited definition provided by the VR-CoDES. This discordance is evident from the differences in rates of responses categorized as “positive” in the that study (28%) vs responses categorized as “empathy” in ours (5%). The low rate of empathy seen in our study also makes it more difficult to observe significant associations between empathy and outcomes. It is possible that our finding about the time-saving nature of explicit responses is reflective of what Levinson and colleagues contributed to their categorization of positive responses.

When clinician responses were explicit (vs nonexplicit), we found that patients were less likely to repeat their emotional expression, and there was a trend toward shorter length of visit. Although the association with visit length was reduced after accounting for clinicians’ tendency to be more explicit toward the end of the encounter, we still estimate a 1.37-minute shorter visit from being explicit vs nonexplicit. The corresponding finding that a patient was less likely to repeat their emotion supports the theory that this additional time is saved because the patient feels heard. It is worth considering, however, that the time may be saved because of the active listening required to make an explicit response, rather than or in addition to the fact that the patient feels heard. In either case, the advice to clinicians would be to respond explicitly, which requires mindful attention and focus, which is of practical use in teaching about communication skills.

It is interesting to note that clinician responses varied over the course of the encounter. As time passed in medical visits, clinicians were less likely to respond to patient emotional expressions by providing space and more likely to make their response explicit. These findings are perhaps not surprising, although to our knowledge they have not been previously demonstrated. In a related line of reasoning, however, one study found oncologists gave more attention to the first emotional expression raised by the patient and less attention to later concerns.¹⁴ The typical arc of a routine patient encounter allows for a greater amount of patient talk in the beginning when the clinician is gathering information and may feel less hurried, rather than toward the end when the clinician may be providing information and trying to wrap up loose ends. By reducing space and making it known to the patient that the clinician has heard the emotional expression, the clinician may be attempting to limit the conversation further. This phenomenon is important to consider when evaluating the true effect of clinician behavior on visit length because failing to account for it will distort the association between clinician response and visit length.

The results of this study should be interpreted in light of several limitations. We collected data from HIV specialty centers at 2 urban academic medical centers located in the United States, which may limit the generalizability of our results to other health care settings and populations. Because HIV has become a chronic condition and many of the patients in these clinics rely on their HIV clinicians for primary care, we believe this setting is similar to many primary encounters with medically and socially complex patients. Although all patients had HIV, we did not collect data on other comorbidities such as mood disorders (depression or anxiety), substance use disorders, or other physical health issues. These comorbidities could impact visit length as well as emotional expressions, and would be important to explore in future studies.

Our study sample comprises primarily White, female clinicians and African American, male patients. Although these demographics are reflective of the HIV epidemic overall, the sample’s skewed racial and sex composition, the diversity of health professionals represented, or race-sex discordance may influence clinician communication around patient emotions.^12,14,15 Finally, all patients and clinicians in this study spoke English in the encounter, and all had (as specified by inclusion criteria) been in a relationship with each other for at least 6 months. Whether emotional communication changes over time in patient-clinician relationships is not known.

Because we assessed verbal behavior through audio-recorded encounters, we were not able to evaluate nonverbal behaviors that could provide clues to patients’ negative emotions and contribute to empathic responses by the clinician. The definition of an empathic response was limited to verbal behavior, as it is in many studies using the VR-CoDES.^16,17 Similarly, the limitation of our analysis to audio-recorded interactions means that we could not know the intention of the clinician responses (e.g., whether the responses intended to be helpful or perhaps were an automatic pattern of response), nor could we ascertain the impact that the responses had on the patient’s experience. Finally, our data were coded by a single experienced coder, and we did not assess intercoder reliability. Previous studies that have done so, however, have found discrepancies to be small.¹⁰

In contrast to common understanding, our study found that there was no association between the specific response of empathy and visit length. Our findings suggest that some clinician responses—such as providing space for patients to talk about their emotions—can be considered beneficial but may add time to the visit. As medical educators think about how to teach communication skills to students and as clinicians make decisions about how to allocate a fixed amount of time for their patients, it makes sense to consider that coherent explicit responses to patient emotion may save time and are considered at least as good as—if not better than—nonexplicit responses. Providing space for patients to talk about their emotions in the HIV ambulatory care setting does not save time, but may be important for other reasons.