Quantifying the process of empathy during crisis interventions

NATIONAL CENTER FOR VETERANS STUDIES

Quantifying the process of empathy during crisis interventions: synchrony in voice pitch among clinicians and

patients

Craig J. Bryan, PsyD, ABPPNational Center for Veterans Studies &

The University of Utah



Co-Authors

Brian Baucom, PhD National Center for Veterans Studies & The University of UtahAlex O. Crenshaw, BS National Center for Veterans Studies & The University of Utah



Disclosure

This work was in part supported by the Military Suicide Research Consortium (MSRC), an effort supported by the Office of the Assistant Secretary of Defense for Health Affairs under Award #W81XWH-10-2-0181.

Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Government, the Department of Defense, the Department of the Army, or the MSRC.



• The ability to maintain a “collaborative, non-adversarial stance” with a patient in crisis is considered to be a core competency for the clinical care and management of suicide risk (Suicide Prevention Resource Center, n.d.)– Assumes an ability to establish and maintain a good working alliance with the patient (Rudd, Cukrowicz, & Bryan, 2008)

• Working alliance (a.k.a., therapeutic alliance or therapeutic relationship) refers to the extent of agreement and/or harmony between the patient and clinician during psychotherapy

• Meta-analyses indicate positive correlations among various measures of the therapeutic relationship with symptom reduction during psychotherapy (Elliott, Bohart, Watson, & Greenberg, 2011; Horvath, Del Re, Flückiger, & Symonds, 2011)– Correlation is largely accounted for by variability among clinicians more than patients (Baldwin, Wampold, & Imel, 2007; Del Re,

Fluckiger, Horvath, Symonds, & Wampold, 2012)

• Few studies have examined the correlation of working alliance with suicide-related outcomes

The ability to develop a strong working alliance with suicidal patients is viewed as an essential skill for clinicians and is consistent with meta-analyses indicating positive correlations among various measures of the therapeutic relationship with symptom reduction during psychotherapy ( Elliott, Bohart, Watson, & Greenberg, 2011; Horvath, Del Re, Flückiger, & Symonds, 2011)

(Baldwin, Wampold, & Imel, 2007; Del Re, Fluckiger, Horvath, Symonds, & Wampold, 2012)



• Bedics et al. (2012): RCT of female patients with borderline personality disorder, DBT vs. expert therapists– DBT: stronger therapeutic relationship fewer instances of NSSI– Control: no relation between therapeutic relationship and NSSI

• Bryan et al. (2012): primary care patients– Therapeutic bond after first appointment not related to subsequent change in suicide ideation

• Conflicting results could be due to limitations in measuring working alliance, notably empathy



• Empathy: the ability for a person to experience an emotional or psychological state that is similar to another person’s state as a result of observing and/or understanding the others’ situation (Preston & De Waal, 2002)

• Empathy entails perceptual and behavioral processes that are functionally intertwined (Decety & Jackson, 2004)

• Two dyadic processes may be involved in therapeutic relationship: synchrony and regulation– Synchrony: degree of “match” or ability to mirror another person’s experience– Regulation: decreased variability in affective expression



• Synchrony: – Greater match in patient/clinician voice frequency during psychotherapy correlated with empathy ratings (Imel et al., 2014)– Individuals who score high on empathy measures tend to mimic the actions of others (Chartrand & Bargh, 1999)– Patient ratings of relationship quality associated with coordinated movements in patient/clinician (Ramseyer & Tschacher, 2011)

• Empathy:– Trust and safety reflected by extent to which one person promotes affective regulation in the other (Geller & Porges, 2015)

• Although relationship dynamics and empathy are considered key to effective care of suicidal individuals, few studies have investigated moment-to-moment dynamics within the context of therapeutic relationships among clinicians and suicidal patients



Hypotheses

1. Emotional bond will be positively associated with synchrony in f0

2. Emotional bond will be positively associated with patient stability in f0

3. Differences in associations between emotional bond, synchrony in f0, and regulation of f0 during the suicide risk assessment interview relative to during the intervention


Method



Participants

• 54 active duty Soldiers– Gender: 45 male, 9 female– Years of service: M=4.89 (SD=3.91)– Deployments: M=1.20 (SD=1.12)– Race: 76% white, 9% black, 6% Native American, 2% other– Ethnicity: 7% Hispanic

• Soldiers voluntarily presenting to emergency departments and mental health clinics for emergency behavioral health evaluation referred to researchers for evaluation

• Inclusion criteria: active suicide ideation during the past week and/or lifetime suicide attempt

• Exclusion criteria: inability to provide informed consent



Procedures

1. Eligibility assessment2. Informed consent3. Turn on digital recorder (TASCAM DR-40 linear PCM)4. Conduct structured interview focused on suicide attempt history5. Turn off recorder6. Patient completes self-report scales7. Turn on recorder8. Crisis intervention completed by clinician and patient9. Turn of recorder 10. Patient completes self-report scales (including therapeutic alliance)



Measures

• Mean fundamental frequency (f0)• Working Alliance Inventory Short Form (bond subscale)



Data Analysis


Results



High emotional bond Low emotional bond



Suicide risk interview Intervention

Low

emotio

nal

bond

High

emotio

nal

bond

+ p<.10, *p<.05, **p<.01, ***p<.001


Conclusions



• Patient and clinician mean f0 were positively correlated with each other (r=0.52) during both the risk assessment and the intervention, suggesting patients and clinicians have a moderate degree of “match” in vocally expressed emotional arousal

• Patient/clinician match was correlated with therapeutic alliance ratings during the intervention but not risk assessment interview phase– Information-gathering function of the risk assessment may lend itself to different relationship dynamics than the collaborative

intervention phase of a crisis encounter

• Affect regulation/dysregulation differed across interview and intervention– During interview, dysregulating effect of clinician on patient associated with higher alliance ratings– During intervention, regulating effect of clinician on patient associated with higher alliance ratings



• Overall (mean) levels of emotional arousal were less informative than the modeling of moment-to-moment transactions between clinician and patient

• Emotional bonding during crisis encounters related to the patient’s overall level of emotional arousal during the intervention phase (but not the interview) as well as the extent of synchrony that exists between clinicians and patients during the intervention phase (but not the interview).

• Considering how patients and clinicians mutually influence each other reveals more nuanced details of the interpersonal process – Empathy and alliance are strongly related to the influence of the clinician on the patient – Relationship is also related to the patient’s influence on the clinician


Questions

Craig J. Bryan, PsyD, [email protected]

mailto:[email protected]

mailto:[email protected]

Quantifying the process of empathy during crisis interventions

Science