Table 5. Effectiveness of coaching on physician well-being, distress and burnout.
| First author, year | Study design, sample size | Coaching intervention (number, duration, frequency of sessions) vs. control | Outcome(s), measure, timing | Results |
|---|---|---|---|---|
| de Lasson, [30] 2016 | Qualitative study, 45 | 8, three whole-day sessions and five 2 h sessions, NR Control: NA |
Adoption to medical culture, career planning, work/life balance; thematic analysis; halfway point (2 months) and end of intervention (4 months) | “Participants typically stated that they had gained a new awareness of their patterns of thinking, feelings and reactions and found new ways of taking control of their professional lives. Participants were more at ease with themselves.” |
| Dyrbye, [19] 2019 | Randomized controlled trial, 88 | 5, initial: 1 hour; remainder: 30 minutes, every 2–3 weeks over 5 months Control: No coaching intervention (but given access to coaches after 5 months). |
Burnout: Maslach Burnout Inventory; Baseline and end of study (5 months) | Prevalence of symptoms decreased by 17.1% in the intervention and increased by 4.9% in the control group; absolute change -22.0 (95% CI, -25.2 to -18.7), p<0.001 |
| Quality of life; Single-item linear analog scale; Baseline and end of study (5 months) | Improved in intervention group: absolute change mean score (SD) = 1.2 (2.5) vs control group = 0.1 (1.7); absolute change, intervention vs. control = 1.1 (95% CI, 0.04 to 1.21), p = 0.005 | |||
| Resilience; Connor-Davidson Resilience Scale; Baseline and end of study (5 months) | Improved in intervention group: absolute change mean score (SD) = 1.3 (5.2) vs. control group = 0.6(4.0); absolute change intervention vs. control group = 0.7 (95% CI, 0.0 to 3.0), p = 0.04 | |||
| Job satisfaction; Global job satisfaction subscale of the Physician Job Satisfaction Scale; Baseline and end of study (5 months) | No difference (p = 0.79) | |||
| Work engagement; Utrecht Work Engagement Scale; Baseline and end of study (5 months) | No difference (Vigor: p = 0.16; Dedication: p = 0.73; Absorption: p = 0.77) | |||
| Empowerment at work; Empowerment at Work Scale; Baseline and end of study (5 months) | No difference (p = 0.95) | |||
| Dyrbye, 2022 [31] | Randomized controlled trial, 80 | 6, initial: 1 hour; remainder: 30 minutes, every 2–3 weeks over 5 months Control: No coaching intervention (but given access to coaches after 6 months). |
Burnout: Maslach Burnout Inventory (MBI); Baseline and end of study (6 months) | Decreased by 2.5% in intervention group; increased by 2.5% control group (delta -5.0%, 95% CI -8.6%, -1.4%, p = 0.007). |
| Quality of life; Single-item linear analog scale; Baseline and end of study (6 months) | No statistically significant difference. | |||
| Resilience; Connor-Davidson Resilience Scale; Baseline and end of study (6 months) | Increased by 1.9 points in intervention group; decreased by 0.2 points in control group (delta 2.2 points, 95% CI0.1, 4.3, p = 0.04). | |||
| Depersonalization; Subscale of MBI (6 months) | Greater reduction in intervention group vs. control group: mean (SD) = -1.3 (3.1) points vs. 0.4 (3.6) points, delta -1.7 points, 95% CI -3.2, -0.2; p = 0.03). Rates of high depersonalization decreased by 3.8% in intervention group; increased by 2.5% in control group (delta -6.3%, 95% CI -9.5%, -3.1%; p <0.001). |
|||
| Emotional exhaustion; Subscale of MBI (6 months) | Rates of high emotional exhaustion decreased by 5.4% in intervention group and by 2.5% in control group (delta -2.9%, 95% CI -6.4%, 0.7%; p = 0.11). | |||
| Fainstad, 2022 [32] | Randomized controlled trial, 101 | Intervention: Participants could participate in any or all of the following over a 6-month period: (1) two 1-hour group coaching calls per week; unlimited anonymous written coaching via online forum; unlimited access to weekly self-study modules Control: No coaching intervention (but offered coaching program after 6 months). |
Burnout: Maslach Burnout Inventory subscales of Emotional Exhaustion (EE), Depersonalization (DP), Professional Accomplishment (PA); Baseline and end of study (6 months) | EE lower in intervention group, increased in control group: mean (SE) score = -3.26 (1.25) vs. 1.07 (1.12), p = 0.01 DP: No statistically significant effect for intervention or control group (−1.06 [0.64] vs −0.03 [0.58]; P = .23) PA: No statistically significant effect for intervention or control group (1.16[0.83] vs 0.25 [0.75]; P = .41) |
| Imposter syndrome: Young Imposter Syndrome Scale; Baseline and end of study (6 months) | Reduced in intervention group; increased in control group: mean (SE) score = −1.16(0.31) vs 0.11 (0.27), p = 0.003 | |||
| Self-compassion: Neff’s Self-Compassion Scale—Short Form; Baseline and end of study (6 months) | Improved in intervention group: mean (SE) score = 5.5(0.89) vs. -1.32 (0.80), p<0.001 | |||
| Moral Injury: Moral Injury Symptom Scale—Healthcare professionals; Baseline and end of study (6 months) | No statistically significant effect for intervention or control group: mean(SE) score = -5.39 (1.62) vs. -183 (1.47), p = 0.10 | |||
| Gardiner, [33] 2013 | Non-randomized experimental study, 69 | 8 coaching workshops and 6 weeks of email coaching, 9-hours total, Over 3 year period Control: No coaching intervention |
Distress; 10-item unnamed scale; before and 3 to 42 months after the intervention | Lower in intervention group: Mean score = 24.50 (95% CI, 21.71–27.29) vs. control group: mean score = 28.63 (95% CI, 27.08–30.17) |
| Intention to leave practice; 7-item unnamed scale; before and 3 to 42 months after the intervention | Before coaching, 81% of participants in the intervention group had considered leaving general practice; decreased to 40% after coaching (c2(2) = 16.31, P < .001). | |||
| Retention rate; calculated by comparing coaching participants with the total remaining population of rural physicians (n = 312); 2 time points, 3 years apart. | “Over a 3-year period, 94% of the coaching group remained in general practice compared with 80% of the control group (c2(1) = 4.89, P = 0.027).” | |||
| Kakarala, [34] 2018 | Pre-post study design, 12 | NR, 1 year program, NR Control: NA |
Emotional exhaustion; subscale of Maslach Burnout Inventory; Baseline and 1-year. | Emotional exhaustion was high or medium for 60% of participants at baseline, and 56% at 1-year. |
| McGonagle, [35] 2020 | Randomized controlled trial, 58 | 6, First session: 60min; remainder: 30 min, approximately every 2 weeks Control: No coaching intervention |
Burnout; Maslach Burnout Inventory; Pre- and post-intervention and 3- and 6-months post-intervention. | Intervention group vs. control group (p = 0.003): F = 9.82** Pre: M(SD) = 2.32 (0.68) vs. 2.37 (0.71) Post: M(SD) = 1.97 (0.72) vs. 2.45 (0.72) |
| Work stress; 15-item Stress in General Scale; Pre- and post-intervention and 3- and 6-months post-intervention. | Intervention group vs. control group (p = 0.077): Pre: M(SD) = 2.04 (0.77) vs. 2.13 (0.45) Post: M(SD) = 1.72 (0.76) vs. 2.09 (0.57) |
|||
| Turnover intentions; 3-item Turnover Intentions scale; Pre- and post-intervention and 3- and 6-months post-intervention. | Intervention group vs. control group (p = 0.062): Pre: M(SD) = 2.05 (0.91) vs. 1.75 (0.67) Post: M(SD) = 1.82 (0.85) vs. 1.76 (0.76) |
|||
| Engagement; 17-item Engagement Scale; Pre- and post-intervention and 3- and 6-months post-intervention. Psychological capital; 24-item Psychological Capital Questionnaire; Pre- and post-intervention and 3- and 6-months post-intervention. |
Intervention group vs. control group (p = 0.023): F = 5.49** Pre: M(SD) = 5.73 (0.78) vs. 5.87 (0.88) Post: M(SD) = 6.06 (0.68) vs. 5.92 (0.68) Intervention group vs. control group (p = 0.002): f = 10.39** Pre: M(SD) = 4.08 (0.69) vs. 4.23 (0.68) Post: M(SD) = 4.63 (0.68) vs. 4.39 (0.74) |
|||
| Compassion; 5-item Santa Clara | Intervention group vs. control group (p = 0.784): | |||
| Brief Compassion Scale; Pre- and post-intervention and 3- and 6-months post-intervention. | Pre: M(SD) = 5.47 (0.93) vs. 5.48 (0.99) Post: M(SD) = 5.63 (0.83) vs. 5.58(1.01) |
|||
| Job self-efficacy; 7-item Job Self-Efficacy Scale; Pre- and post-intervention and 3- and 6-months post-intervention. | Intervention group vs. control group (p = 0.62): Pre: M(SD) = 3.78 (0.66) vs. 3.46 (0.72) Post: M(SD) = 4.01 (0.60) vs. 3.61 (0.70) |
|||
| Job satisfaction; three items from Cammann, Fichman, Jenkins, and Klesh Scale; Pre- and post-intervention and 3- and 6-months post-intervention. | Intervention group vs. control group (p = 0.021): F = 5.74** Pre: M(SD) = 3.59 (0.95) vs. 4.11 (0.54) Post: M(SD) = 3.91 (0.80) vs. 4.04 (0.53) |
|||
| McKimm, [36] 2018 | Mixed-methods study (pre/post design + qualitative interviews), 52 | 4, NR, a period of up to 18 months. Control: NA |
Likelihood of leaving profession; unnamed 10-point scale; baseline and immediately after intervention | “Pre-coaching, 75% of respondents said they were likely (score of 7+) to leave general practice; this fell to 21% (nine GPs) of post-coaching questionnaire respondents. Of the latter group, half of them were aged between 50 and –60 and 4 planned to continue working as a doctor in different roles.” |
| Performance under pressure; Human Function Curve; first and last coaching sessions. | “At the start of coaching, 4 GPs rated themselves at breakdown point, 27 at the point of exhaustion and a further 4 considered themselves fatigued. GPs’ self-rated performance was considerably improved at the end of coaching. The final ratings of all except 4 GPs moved from ‘distress’, ‘boredom’ and excess pressure nearer to the ‘safe zone’, with 27 moving to the ‘safe zone’. Only one GP moved further up the ‘pressure scale’; however, they explained that coaching had revealed they were in denial about the pressure they were working under at the start of coaching.” |
|||
| Palamara, [37] 2015 | Pre-post study design, 72 | Up to 4, 40 minutes, every 3 months Control: NA |
Personal accomplishment; subscale of Maslach Burnout Inventory; Baseline, 3-months, 12-months (post) | High personal accomplishment: pre-coaching 49/59 participants; post-coaching: 48/59 participants |
| Emotional exhaustion; subscale of Maslach Burnout Inventory; Baseline, 3-months, 12-months (post) | High emotional exhaustion: pre-coaching: 44/59 participants; post-coaching: 33/59 participants | |||
| Palamara, [38] 2018 | Pre-post study design, 104 | 3 or more per year, NR, every 3 months Control: NA |
Coping skills; unnamed survey; 3 years | 70% of those who participated fully in the coaching program indicated improved coping skills |
| Perception of professional relationships; unnamed survey; 3 years | 70% of those who participated fully in the coaching program indicated improved relationships | |||
| Emotional exhaustion; subscale of Maslach Burnout Inventory; 3 years | 62.8% of participants who reported excellent opportunities for reflection with their coach also reported lower levels of emotional exhaustion | |||
| Palamara, [37] 2021 | Pre-post study design, 129 | 2–3 per year, 45–60 minutes, every 3 months Control: NA |
Burnout; Depersonalization and Emotional exhaustion subscales of Maslach Burnout Inventory; baseline and 8 months | Participants with higher burnout (measured by depersonalization and emotional exhaustion scores) at baseline more likely to improve and experience lower burnout at 8 months (-0.393, p<0.001; -0.476, p<0.001). |
| Emotional exhaustion; subscale of Maslach Burnout Inventory; baseline and 8 months | Black/Asian/Hispanic participants more likely to experience increase compared to non-Hispanic white participants (2.608, 0.46). Decrease for non-Hispanic white participants (mean difference = -1.86, p = 0.02) |
|||
| Well-being; PERMA Well-being Scale; baseline and 8 months | Well-being changed from baseline to follow-up in all participants; females showed a decline while males showed an increase (−1.41 vs. 0.83, p = 0.04). Less improvement observed if participant had higher well-being at baseline (-0.407, p<0.001). |
|||
| Palamara, [40] 2022 | Randomized controlled trial, 150 | Intervention: Minimum 3 coaching sessions, 45–60 minutes, 9-month period Control: three emails over study period containing well-being resources |
Professional fulfillment; Professional fulfillment index (PFI) subscale; baseline and post-intervention (9 months) | Increase: mean (SD) pre 2.33(0.67) vs post 2.52(0.77) p = 0.021; Cohen’s d = 0.26 |
| Burnout (work exhaustion + interpersonal disengagement); PFI subscale; baseline and post-intervention (9 months) | Decrease in burnout: mean (SD) pre 1.39(0.68) vs post 1.19(0.58) p = 0.026; Cohen’s d = -0.26], Decrease in work exhaustion: mean (SD) pre 1.72(0.78) vs post 1.48(0.67) p = 0.017; Cohen’s d = -0.27 Decrease in interpersonal disengagement: mean (SD) pre 1.17(0.71) vs post 1.00(0.62) p = 0.071; Cohen’s d = -0.21 |
|||
| Self-evaluation; unnamed 5-point Likert scale; baseline and post-intervention (9 months) | Increase: mean (SD) pre 1.42(0.71) vs post 1.75(0.82) p = 0.0003; Cohen’s d = 0.35 | |||
| Positive emotions, engagement, relationship, meaning and accomplishment (PERMA); PERMA scale; baseline and post-intervention (9 months) | Increase: mean (SD) pre 55.47(7.60) vs post 58.05(7.90) p = 0.002; Cohen’s d = 0.37 | |||
| Intolerance of Uncertainty (IUS); IUS scale; baseline and post-intervention (9 months) | Improvement (reduction in IUS score) in intervention group (53.0% to 40.8%), but not statistically significant (p = 0.07) | |||
| Resilience: Hardiness Resilience Score scale; baseline and post-intervention (9 months) | No significant change | |||
| Solms, [41] 2021 | Non-randomized experimental study, 114 | 6, 60–90 minutes, varied (time in between sessions determined by participants) Control: No coaching intervention |
Emotional exhaustion; subscale of Maslach Burnout Inventory; baseline and 10 months | Pre: mean(SD) = 2.75 (1.08) Post: mean(SD) = 2.25 (0.79), p = 0.000 |
| Cynicism; subscale of Maslach Burnout Inventory; baseline and 10 months | Pre: mean(SD) = 2.11(1.08) Post: mean(SD) = 1.90 (SD 0.75), p = 0.151 |
|||
| Work engagement; Utrecht Work Engagement Scale; baseline and 10 months | Pre: mean(SD) = 5.08(0.78) Post: mean(SD) = 5.28 (0.59), p = 0.033 |
|||
| Psychological capital; PsyCap questionnaire; baseline and 10 months | Pre: mean (SD) = 4.83(0.69) Post: mean (SD) = 5.16(0.65), p = 0.000 |
|||
| Self-compassion; Self-Compassion Scale; baseline and 10 months | Pre: mean (SD) = 3.07(0.60) Post: mean (SD) = 3.27(0.52), p = 0.009 |
|||
| Psychological flexibility; Work Acceptance and Action Questionnaire; baseline and 10 months | Pre: mean (SD) = 3.43(0.63) Post: mean (SD) = 3.47(0.65), p = 0.6 |
|||
| Song, [42] 2020 | Mixed-methods study (pre/post design + qualitative interviews), 25 | 1 group workshop (2 hours) + 8 individual sessions (1 hour), spaced throughout academic year Control: NA |
Burnout; Abbreviated Maslach Burnout Inventory; first and last coaching sessions | No statistically significant changes in burnout scores or the proportion of participants at risk of burnout before (60%) and after (52%) coaching (p = 0.78). |
| Resilience; Brief Resilience Scale; first and last coaching sessions | Mean resilience score improved: mean (SD) = 3.8 (0.8) to 4.2 (0.7) (p = 0.002). Improvement was observed primarily among men (mean [SD], 4.1 [0.7] vs. 4.5 [0.5], p = 0.007) rather than women (mean [SD], 3.4 [1.0] vs 3.6 [0.8], p = 0.16). |
|||
| Positive and negative feelings; Scale of Positive and Negative Experience; first and last coaching sessions | No statistically significant improvement (mean [SD] = 6.7 [8.2] vs 8.4 [8.3], p = 0.14) | |||
| Perceptions of coaching experience, burnout and wellness; grounded theory analysis; end of intervention | Most participants reported that coaching provided useful skills and was helpful for improve well-being, but felt it should be offered for a longer duration. |