Skip to main content

Table 3 Results of eligible studies. Results are presented per outcome category, i.e., attitude, knowledge / skills, and behavior

From: Empathy promoting interventions for health professionals: a systematic review of RCTs

First author, publication year

Rater/

Process

Outcome: measure

Participants analyzed (intervention / control

Difference within group

Difference between groups

Intervention

Estimatea (SD or 95 % CI)

Comparison group

Estimatea (SD or 95 % CI)

Estimatea (SD or 95 % CI)

P-value

Type of outcome: Attitude

Riess

2012b

Health professional/ questionnaire only

Attitude towards empathy: Jefferson Scale of Physician Empathy

99 (54 / 45)

1.2 (SD 9.3)

−1.1 (SD 6.7)

Effect size 0.3

NS

Type of outcome: Knowledge / Skills

Riess

2012b

Health professional/ questionnaire only

Physician’s knowledge of the neurobiology and physiology of empathy: Neurobiology and Physiology of Empathy Test

99 (54 / 45)

2.3 (SD 2.4)

0.4 (SD 2.3)

Effect size 0.8

<0.001

Riess

2012b

Health professional/ questionnaire only

Physician’s skill at decoding subtle facial expressions of emotion: Ekman test

99 (54 / 45)

2.1 (SD 2.5)

0.2 (SD 2.2)

Effect size 0.8

<0.001

Type of outcome: Behavior

Daeppen

2012c

External observer/

simulated interview

MITI 3.0 behavioral coding system Global Scores: Empathy

131 (66 / 65)

Before: NA

After: 4.0 (SD 0.6)

Before: NA

After: 3.4 (SD 0.7)

NA

NA

Butow

2008d

External observer/

simulated interview

Key doctor behaviors; category: creating an environment where emotion is likely to be expressed: Basic empathy subscale

30 (15 / 14)

Before: Median 5.0 (IQR 4.0, 5.0)

After: Median 5.0 (IQR 4.5, 5.0)

Before: Median 4.0 (IQR 3.8, 5.0)

After: Median 4.0 (IQR 3.0, 5.0)

ND

NS

Delvaux

2005e

External observer/

simulated interview

Empathy, educated guesses, alerting to reality, confronting, negotiating, summarizing: CRCWEM rated utterances directed to patients

56 (27 / 29)

ND

ND

RR 1.8 (95 % CI 1.0, 3.4)

NS

Delvaux

2005e

External observer/

simulated interview

Empathy, educated guesses, alerting to reality, confronting, negotiating, summarizing: CRCWEM rated utterances directed to relatives

56 (27 / 29)

ND

ND

RR 1.1 (95 % CI 0.4, 3.2)

NS

Razavi

2002

External observer/

simulated interview

Emotional depth of utterances: CRCWEM neutral (level 0) utterances

115 (57 / 58)

ND

ND

ND

NS

Razavi

2002f

External observer/

simulated interview

Use of emotional “distress” words: Frequency score

115 (57 / 58)

Before: 9.2 (SD 4.9)

After: 12.9 (SD 5.6)

Before 10.8 (SD 5.6)

After: 9.1 (SD 5.4)

ND

<0.001

Razavi

2002f

External observer/

simulated interview

Use of emotional “distress” words: Density score

115 (57 / 58)

Before: 3.6 (SD 2.1)

After: 4.4 (SD 2.2)

Before: 4.0 (SD 2.8)

After: 3.6 (SD 2.7)

ND

0.04

Tulsky

2011g

External observer/ actual interview

Emotion handling skills: Number of empathic statements per conversation

48 (24 / 24)

Before: 0.4 (SD 1.0)

After: 0.8 (SD 1.3)

Before: 0.3 (SD 0.7)

After: 0.4 (0.8)

RR 1.9 (95 % CI 1.1, 3.3)

0.024

Tulsky

2011g

External observer/ actual interview

Emotion handling skills: Number of continuer response to empathic opportunities

48 (24 / 24)

ND

ND

OR 2.1 (95 % CI 1.1, 4.2)

0.028

Bonvicini

2008

External observer/ actual interview

Physicians’ empathic expression: GRS, Empathy

155 (79 / 76)

Before: 8.4 (95 % CI (7.8, 9.0)

After: 11.6 (95 % CI 11.0 12.2)

Before: 8.1 (95 % CI (7.6, 8.6)

After: 7.4 (95 % CI 7.0 7.8)

1.4 (ND)

<0.01

Bonvicini

2008

External observer/ actual interview

Physicians’ empathic expression: Hierarchical ECCS

155 (79 / 76)

Before: 2.7 (95 % CI (2.5, 2.8)

After: 4.0 (95 % CI 3.9 4.1)

Before: 2.6 (95 % CI 2.4, 2.8)

After: 2.5 (95 % CI 2.3 2.7)

3.9 (ND)

<0.01

Delvaux

2005e

External observer/

actual interview

Empathy, educated guesses, alerting to reality, confronting, negotiating, summarizing: CRCWEM rated utterances directed to patients

56 (27 / 29)

ND

ND

RR 1.4 (95 % CI 0.7, 2.8)

NS

Fallowfield

2002h

External observer/ actual interview

Empathy: Number of empathic expressions

160 (80 / 80)

ND

ND

RR 1.5 (ND)

0.005

Jenkins

2002i

External observer/ actual interview

Empathy: Presence of empathic expressions

93 (48 / 45)

186 (97/89)

Before: 57 59 % (ND)

After: 56 58 % (ND)

Before: 44 49 % (ND)

After: 38 42 % (ND)

ND

ND

Razavi

2002

External observer/

actual interview

Emotional depth of utterances: CRCWEM neutral (level 0) utterances

115 (57 / 58)

ND

ND

ND

NS

Razavi

2002f

External observer/

actual interview

Use of emotional “distress” words: Frequency score

115 (57 / 58)

Before: 4.4 (SD 3.8)

After: 7.0 (SD 5.8)

Before 4.6 (SD 4.9)

After: 4.3 (SD 4.2)

ND

NS

Razavi

2002f

External observer/

actual interview

Use of emotional “distress” words: Density score

115 (57 / 58)

Before: 3.7 (SD 3.4)

After: 2.7 (SD 4.8)

Before: 3.3 (SD 3.0)

After: 3.1 (SD 2.9)

ND

NS

Razavi

2002j

External observer/

actual interview

Use of emotional “anxiety” words: Density score

115 (57 / 58)

ND

ND

ND

0.028

Robbins

1979k

External observer/

actual interview

Empathy: Carkhuff scale - Empathy level

51 (26 / 25)

Before: 2.3 (ND)

After: 2.7 (ND)

Before: 2.3 (ND)

After: 2.3 (ND)

ND

ND

Robbins

1979k

External observer/

actual interview

Empathy: Carkhuff scale - Number of empathic responses

51 (26 / 25)

Before: 2.0 (ND)

After: 4.1 (ND)

Before: 2.3 (ND)

After: 2.3 (ND)

ND

ND

Tulsky

2011l

Patient/ actual interview

Empathy: Perceived Empathy Scale

48 (24 / 24)

Before: NA

After: 0.4 (95 % CI 0.3, 0.5)

Before: NA

After: 0.2 (95 % CI 0.1, 0.3)

NA

NA

Shapiro

2009

External observer/

questionnaire only

Expressed empathy: SPIR Scale

79 (38 / 41)

Before: 6.0 (SD 5.7)

After: 8.3 (SD 5.0)

Before: 7.3 (SD 6.7)

After: 6.6 (SD 5.1)

ND

0.04

Blair-Irvine

2012m

Health professional/ questionnaire only

Assessment of psychosocial construct: Empathy (4-item 7-point Likert scale)

172 (84 / 88)

Before: 5.1 (SD 1.0)

After: 5.5 (SD 0.9)

Before: 5.7 (SD 1.1)

After: 5.2 (SD 1.0)

ND

0.04

Riess

2012n

Health professional/ questionnaire only

Empathic responsiveness in personal life: BEES

99 (54 / 45)

0.9 (SD 14.5)

2.7 (SD 14.1)

Effect size 0.12

NS

Rask

2009o

Health professional/

questionnaire only

Nurse’s perception of patient’s experiencing empathy during their communication: NPRI, Empathy subscale

23 (12 / 11)

Before: 21.4 (SD 2.4)

After: 24.0 (SD 2.9)

Before: 22.1 (SD 3.6)

After: 23.2 (SD 3.2)

Effect size 0.42

NS

Shapiro

2004p

Health professional/ questionnaire only

Empathy: ECRS

16 (10 / 6)

ND

ND

ND

ND

Shapiro

2004p

Health professional/ questionnaire only

Empathy: BEES

16 (10 / 6)

ND

ND

ND

ND

Daniels

1988q

Health professional/ questionnaire only

Empathy: Carkhuff index of communication

53 (24 / 29)

Before: 1.7 (SD 0.3)

After: 2.4 (SD 0.2)

Before: 1.8 (SD 0.4)

After: 1.9 (SD 0.3)

ND

<0.05

Daniels

1988q

Health professional/ questionnaire only

Empathy: ECRS

53 (24 / 29)

Before: NA

After: 188.8 (42.1)

Before: NA

After: 151.9 (40.9)

ND

NA

Wolf

1987

Health professional/ questionnaire only

Ability to respond to patients’ emotional concerns in hypothetical scenarios, MCI

134 (65 / 69)

Before: 0.7

(SD 0.7)

After: 2.0

(SD 1.0)

Before: 0.9

(SD 0.9)

After: 1.1

(SD 0.8)

ND

<0.001

Wolf

1987

Health professional/ questionnaire only

Exhibit greater preferences for responses that addressed patients’ emotions: HRI

134 (65 / 69)

Before: 33.0 (SD 7.9)

After: 17.9 (SD 6.7)

Before33.1 (SD 9.5)

After: 22.3 (SD 8.8)

ND

<0.001

Robbins

1979k

Health professional/ questionnaire only

Affective sensitivity - empathy: Affect Sensitivity Scale

51 (26 / 25)

Before: 26.2 (ND)

After: 28.8 (ND)

Before: 26.5 (ND)

After: 28.0 (ND)

ND

NS

Riess

2012n

Patient/ questionnaire only

Physician’s empathy: CARE

99 (54 / 45)

0.7 (SD 7.9)

−1.5 (SD 6.0)

Effect size 0.31

0.04

Rask

2009o

Patient/

questionnaire only

Patient perception of nurse empathy: PPRI, Empathy subscale

23 (12 / 10)

Before: 22.8 (SD 1.9)

After: 22.8 (SD 1.8)

Before: 22.4 (SD 1.7)

After: 23.5 (SD 2.9)

Effect size 0.05

NS

Smith

1995r

Patient/

questionnaire only

Patient Satisfaction Questionnaire, Patient perception of physician’s empathy

26 (14 / 12)

ND

ND

ND

NS

  1. SD Standard Deviation, CI Confidence Interval, ND No Data, NS Non Significant, IQR Interquartile Range, RR Risk Ratio, NA Non Applicable, MITI Motivational Interviewing Treatment Integrity, SPIR Staff-Patient Interaction Rating Scale, ECRS Empathy Construct Rating Scale, BEES Balanced Emotional Empathy Scale, GRS Global Rating Score, ECCS Empathy Communication Coding System, CARE Consultation and Relational Empathy Measure, CRCWEM Cancer Research Campaign Workshop Evaluation Manual, NPRI, Nurse—Patient Relationship Inventory, PPRI Physician—Patient Relationship Inventory, MCI Medical Communication Index, HRI Helping Relationship Inventory
  2. a Estimate: mean unless otherwise indicated
  3. b Numbers for estimate represent pre-post change
  4. c There was not pre-intervention assessment of evaluation; authors reported that higher post-intervention scores on empathy were found in the experimental group (effect size 0.7, 95 % CI 0.6–0.8, P < 0.001)
  5. d Results correspond at baseline and 6 month follow up values; difference between groups was also not significant after 12 month follow up
  6. e Estimated relative risks were based on a multivariable Poisson regression model was adjusted for physician age, oncology practice, and gender in both types of interviews; for patients’ and their relatives’ age, gender, and educational level; for patients’ Karnofsky performance status; for relatives’ ties with the patients; for the number of months since diagnosis; for the presence of a previous and/or current cancer treatment; for the type of information (diagnosis or prognosis-focused and treatment-focused) and of the news given (good, bad, or neutral); and for the fact that it was or not the first interview of the
  7. patient with the physician. There was no convergence of the model for relatives in actual interviews
  8. f Frequency score and density score were calculated by PROTocol ANalyzer (PROTAN); comparisons with multivariate analyses of variances (MANOVA) included pre-intervention scores, post-intervention scores, and scores at 3 month follow up (non-significant difference); “distress” words as tagged by the French version of the Harvard Third Psychosociological Dictionary
  9. g Analysis for the number of empathic statements was based on total conversations (n = 264) while for the number of continuer response to empathic opportunities on conversations with at least one empathic opportunity (n = 135) including 275 empathic opportunities (range: 1 to 11 opportunities per conversation). Results for the number of empathic statements were from a mixed-effect Poisson regression model adjusted for study site; physician sex; and mean number of empathic statements, defined as NURSE (name, understand, respect, support, explore) statements used by the physician in the pre-intervention phase. Results for the number of continuer response to empathic opportunities were from a logistic mixed-effect regression model adjusted for study site and physician sex
  10. h Number of empathic expressions was assessed after videotapes including consultations were rated by Medical Interaction Process System (MIPS)
  11. i Presence or absence of empathic expressions was based on Medical Interaction Process System (MIPS) summary data from the videotaped consultations between the physicians and patients; authors presented the results of comparison between the groups only after intervention stating that the intervention group were more likely to exhibit empathy (56 [58 %] vs. 38 [42 %]; χ 2 = 4.823, df =1, P = 0.02)
  12. j “Anxiety” words as tagged by the French version of the Regressive Imagery Dictionary (MRID); authors stated
  13. that in actual interviews, trained health care professionals used different words (density scores) tagged by ‘anxiety’ MRID subcategory compared to untrained health care professionals (group by time effect MANOVA F value: 3.66, P = 0.028). Authors do not report the results on frequency or density scores of any other MRID emotion subcategories assessed in the study for health care professionals
  14. k Authors presented only comparisons within but not between groups stating that the mean empathy level scores increased significantly in the experimental group (P < 0.05 by F tests of group means) while control scores did not. In addition, the number of responses dealing with patient feelings increased in the experimental group (P < 0.05 by F tests of group means) but not in the control group
  15. l Perceived empathy was not evaluated before intervention. Two hundred two patients (109 in the intervention group and 93 in the control group) were included in the analysis for perceived empathy. Mixed-effect models are adjusted for site and physician sex. Marginal standardized estimates are predicted proportion for binary outcomes, and the 95 % CIs for the standardized estimates and relative risks are from 1000 bootstrap samples. There was no statistically significant difference between the two groups (P = 0.058)
  16. m Comparison of the posttest scores, adjusted for pretest scores, showed greater gains by the intervention condition compared to the control condition for empathy. Results from the ANCOVA analysis adjusted for pretest scores did not show significant group differences at 8-week follow-up
  17. n Numbers represent pre-post change
  18. o Group x time interaction effects were investigated by within-between-subjects ANOVAs. For nurses, Group x time interaction effects were also investigated at 3 month follow up without statistically significant between group differences (P = 0.94)
  19. p Authors presented analyses after combining the two randomized groups to one; results per group and on the comparison between groups were not provided
  20. q Significant difference between groups was not confirmed after 9 month follow up; authors reported that after the intervention experimental trainees performed better in ECRS (P < 0.01); however, pre-post intervention comparison between groups is not applicable
  21. r Mean post-intervention scores were given separately for male and female residents while pre-intervention scores were not provided. Analysis of covariance with groups (trained vs. untrained) and pre-training satisfaction score as a covariate did not yield statistically significant results for patients’ perceived physician empathy (P = 0.65)
Back to article page

Journal of Compassionate Health Care

ISSN: 2053-2393

Contact us