Do Resident Coaching Programs Benefit their Coaches? Impact of a Professional Development Coaching Program on The Coaches

Abstract

Background

Coaching has been shown to decrease physician burnout; however, coachee outcomes have been the focus. We report the impact of coaching on women-identifying surgeons who participated as coaches in a 9-month virtual program.

Methods

A coaching program was implemented in the Association of Women Surgeons (AWS) to determine the effects of coaching on well-being and burnout from 2018 to 2020. AWS members volunteered and completed training in professional development coaching. Pre- and post-study measures were assessed, and bivariate analysis performed based on burnout and professional fulfillment score.

Results

Seventy-five coaches participated; 57 completed both pre- and post-study surveys. There were no significant changes in burnout or professional fulfillment including the Positive Emotion, Engagement, Relationship, Meaning, and Accomplishment scale, hardiness, self-valuation, coping, gratitude, or intolerance of uncertainty scores from baseline to post-survey. On bivariate analysis, hardiness was associated with lower burnout throughout the duration of the program. Coaches with lower burnout at the end of the program met with their coachee more frequently than coaches with higher burnout [mean (SD) 3.95(2.16) versus 2.35(2.13) p = 0.0099].

Discussion

Burnout and professional fulfillment demonstrated no change in women surgeons who participated as professional development coaches. Those with lower burnout and higher professional fulfillment at the end of the program were found to have higher hardiness, which may be worth future investigation.

Conclusions

Acquisition of coaching skills did not directly improve well-being in faculty who participated in a resident coaching program. Future studies would benefit from control groups and exploration of qualitative benefits of coaching.

Introduction

Burnout is highly prevalent among physicians, especially surgeons [1]. Recent studies have shown that half of the physicians in the US meet criteria for burnout [2] with rates ranging from 31 to 50% for surgery specialties, and even higher in women surgeons [1, 3–10]. Studies assessing burnout and well-being in women surgery residents and attendings have consistently found worse scores compared to their men counterparts [1, 5, 7–11].

Coaching has been proposed as a tool to decrease burnout among clinicians [12–15]. Recent data have demonstrated the successful implementation of coaching for residents with modest improvements in well-being parameters of professional fulfillment and burnout [13, 15–18]. However, these studies have analyzed the effects of coaching on the coachee, and there is a paucity of data on whether participating in a coaching program benefits the coaches [19, 20]. Additionally, little is known about how different levels of individual characteristics (e.g., hardiness, self-valuation, etc.) may influence well-being, burnout, and outcomes from coaching programs.

We report the impact of coaching on the well-being of practicing women surgeons (coaches) who underwent professional development coach training and participated in a 9-month virtual resident coaching program through the Association of Women Surgeons (AWS). We hypothesized that skills learned via this training would improve surgeon well-being and decreased burnout. Assessment used validated instruments that address well-being, including professional fulfillment, burnout, intolerance of uncertainty, and hardiness. We secondarily investigated whether there was any association between burnout and individual characteristics hypothesized to be correlated with burnout.

Methods

A coaching program was implemented through AWS to reduce burnout and improve well-being in women surgery residents from 2018 to 2020. AWS is a global professional surgical society whose mission is “to inspire, encourage, and enable women surgeons to realize their professional and personal goals.”[21] Concordant with its mission, AWS launched a program in which practicing surgeons volunteered for positive psychology coach training and subsequent pairing with women surgery trainees in a virtual coaching relationship. A secondary aim was to examine the effects on the coaches which is explored in this paper.

The research protocol, materials, and methods were approved by the Massachusetts General Brigham Institutional Review Board (Protocol #: 2017P00056) and funded through a grant to the Association of Women Surgeons Foundation by The Physicians Foundation. Methods and prior results of the clinical trial (ClinicalTrials.gov Identifier: NCT04883307) have been previously reported [17].

AWS members were recruited as unpaid volunteers via direct emails and social media to participate in the study as coaches (practicing surgeons) or coachees (residents). Sample size was based on the primary outcome in our previously published clinical trial in surgery residents [17]. Requirements for coaches included: AWS women-identifying member, currently in active surgical or surgical subspecialty practice in the United States (US) or Canada. Coachees were women-identifying surgery AWS resident members from the US and Canada.

All participants were required to complete informed consent, a baseline questionnaire including demographics, surgery specialty, and several validated instruments related to physician well-being. Instruments included the Stanford Professional Fulfillment Index (PFI); Positive Emotion, Engagement, Relationship, Meaning, and Accomplishment scale (PERMA); Intolerance of Uncertainty Scale (IUS); Gratitude Questionnaire (GR-6); Maintenance of Current Status (MOCS); and Hardiness-Resilience Scale (HRS/DRS-15) [22–25]. Region was determined using self-reported cities and states determined by the US Census Divisions [26]. Coaches completed a 3-h in-person coach training program at the AWS national meeting led by author, Kerri Palamara (KP), who is a subject matter expert (SME) in professional development coaching and has trained physician coaches across multiple specialties including surgery. Training consisted of didactic teaching combined with breakout discussions to practice coaching techniques and reflection exercises. Skills taught focused on active listening, asking powerful questions, and using positive psychology tools to facilitate goal setting and maximizing potential. During training, coaches completed several exercises they would then use to coach their coachee; with each exercise, coaches practiced as both coach and coachee and therefore had the opportunity to experience coaching themselves. Exercises focused on IUS, hardiness and resilience, gratitude, coping with stressors, using energy in ways that serve you, and work–life integration. Exercises were designed to emphasize these skills in everyday scenarios in their personal and professional lives. Unique aspects of the curriculum tailored to the study population included dealing with negative feedback, microaggressions, imposter phenomenon, and self-doubt.

Coaches were invited to participate in two optional, virtual 90-min refresher training sessions months 3 and 6 of the program. These virtual trainings were similar to the in-person sessions with virtual breakout rooms used to practice skills one-on-one. After the 9-month coaching period, all coaches were administered a similar follow-up questionnaire. Additional details regarding the coach training and program are published separately [13, 14, 16, 17]. Coaching materials were created by the SME in conjunction with the Institute of Coaching and surgeons who were involved with this project (ALM and SLS) to tailor specifically to surgeons. Coaches received a training manual developed by the SME and AWS with step-by-step guides for three structured coaching sessions with their coachees. Further details of the training have been previously published [18].

Upon completion of training, coaches were paired with 1–2 residents. Coach-coachee pairings were constructed based on divergent surgical special and geographic location to limit potential hierarchical relationships between the coach and coachee to foster psychological safety and minimize the possibility of developing a more traditional mentoring relationship. All coaching sessions occurred virtually over voice or video calls. Participants were encouraged to meet for at least three coaching sessions over the 9-month period but could meet more frequently if desired. Each meeting was 45–60 min in length, and specific exercises were given to review at each meeting.

Two cohorts of coaching are included in this study, 2018–2019 and 2019–2020. Of note, the COVID-19 pandemic occurred following the training of the second coaching cohort. As the coaching interactions and subsequent trainings were all held remotely, there was no change in protocol during the pandemic.

Study outcome measures of physician well-being included two primary outcome measures, the PFI and the PERMA scales. The PFI is composed of three subscales, on a 5-point Likert scale which includes professional fulfillment (5-point Likert scale [from not at all (0) to completely true (4)], 6 items); PFI subscale for burnout as combination of work exhaustion and interpersonal disengagement (5-point Likert scale ranging from not at all (0) to extremely (4)], 10 items, range 0–4). The PERMA overall score ranges from 0 to 75 based on 15 items on a 5-point frequency scale (1–5). Other measures included self-valuation (5-point Likert scale [ranging from never (0) to always (4)], 4 items, range 0–4)); IUS (5-point frequency [1–5], 12 items, range 0–60); Hardiness-resilience score (HRS/DRS-15, 4-point frequency (0–3), 15 items, range 0–45); Gratitude score (7-point frequency [1–7], 2 items from the GQ-6 scale, range 0–14); and Measurement of Current Status score (MOCS, 5-point frequency [1–5], 3 items from MOCS part A, range 0–15). PFI scores are calculated as the average of the scoring items while all other scores are summed across items [22–25, 27–29]. Measures were selected based on previously published research demonstrating the impact of coaching on well-being and associated parameters [13, 14, 30].

Participant demographics and characteristics were summarized for all who responded to the baseline survey and those who responded to both the baseline and follow-up surveys. To assess the difference in outcomes, our analysis was limited to participants who answered both baseline and follow-up surveys. Continuous variables are presented as mean (standard deviation (SD)), and categorical variables are presented as frequency (%). A paired t-test was used to compare the outcomes between pre- and post-period.

A bivariate analysis of burnout and PFI categories was performed to assess how certain characteristics (IUS, HRS/DRS-15, gratitude, self-valuation, coping, and PERMA) changed over time during the program and whether certain characteristics were associated with lower rates of burnout and higher professional fulfillment. To compare the outcomes between burnout and PFI categories, a two independent sample t-test was used. A two-sided with p-value < 0.05 was considered as statistically significant. All analyses were conducted using SAS 9.4 (SAS Institute, Cary, NC).

Results

From 2018 to 2019, we trained 75 attending surgeon coaches from 65 different surgery programs throughout the US and Canada. About 76% of coaches were from an academic affiliated program, and 42% were junior attendings in their first 5 years of practice. Demographics are detailed in Table 1.

Table 1.

Demographics of coaches

	N (%)
N	75
Race, N (%)
Asian	10 (13.3%)
Black or African American	3 (4.0%)
White	30 (40.0%)
More than one race	1 (1.3%)
Other	3 (4.0%)
Ethnicity, N (%)
Hispanic	3 (4.0%)
Non-Hispanic	45 (60.0%)
Not answered	27 (36.0%)
Region, N (%)
Northeast	21 (28%)
Midwest	18 (24%)
South	25 (33.3%)
West	11 (14.7%)
Outside of the US (Canada)	0 (0%)
Specialty, N (%)
ACS/Trauma	13 (17.3%)
Non-general specialty	7 (9.3%)
Surgery subspecialty	47 (62.7%)
Cardiothoracic/vascular	5 (6.7%)
General	3 (4.0%)
Number of years in practice
0–5	33 (44.0%)
6–10	22 (29.3%)
11–15	11–15 (14.7%)
16 +	9 (12%)
Academic affiliated institution
Yes	57 (76%)
No	18 (24%)

Of 75 coaches, 46 were matched with two coachees, and 29 were matched with one coachee. Fifty-seven coaches completed both the pre-survey and post-survey and are included in the pre/post analysis. Response rate was lower in year 2 (22/30, 73.3%) than in year 1 (35/45, 77.8%). Data on meeting frequency were determined from coachee responses. About 12% of all respondents reported no coaching interaction. About 79.6% of coaching pairs (66/83) met at least once, 62.7% (52/83) met at least twice, and 49.4% (41/83) had three or more meetings.

About 24.5% (14/57) of coaches met criteria for burnout prior to the coaching program, and 35% (20/57) met criteria for burnout after the coaching program (p = 0.1573). There were no significant changes in our primary outcomes of well-being (PFI and PERMA) from baseline to post-survey. There were also no significant changes from baseline to post-study in measures of hardiness, self-valuation, coping, gratitude, or IUS (see Table 2).

Table 2.

Study outcomes in coaches before and after the study period

Scores		Pre			Post			p-value
		N	Mean	SD	N	Mean	SD
Hardiness	Communication	57	11.175	2.450	53	11.151	2.389	0.2224
	Control	57	11.053	2.166	52	11.154	2.043	0.8423
	Challenge	57	9.386	3.075	51	9.373	2.891	0.6677
	Total	57	31.614	5.599	50	31.580	5.470	0.3216
PERMA	Positive Emotion	57	11.070	1.761	54	11.111	1.766	0.5686
	Engagement	57	11.702	1.463	55	11.691	1.914	0.6819
	Relationships	56	12.000	1.945	55	12.200	2.004	0.8059
	Meaning	57	12.614	1.868	55	12.782	2.323	0.9395
	Accomplishment	57	11.702	1.658	55	12.127	1.963	0.2218
	Total	56	59.018	6.751	54	59.778	8.004	0.9794
Intolerance of uncertainty	Prospective	57	18.509	5.619	53	17.943	4.975	0.2596
	Inhibitory	57	8.386	2.975	53	8.283	3.472	0.4278
	Total	57	26.895	7.734	52	26.212	7.929	0.6892
WES		57	1.320	0.673	57	1.469	0.762	0.1457
ID		57	0.848	0.656	57	0.833	0.604	0.8672
SC		57	2.175	0.778	57	2.075	0.939	0.3598
Personal fulfillment		55	6.705	2.021	55	6.530	2.250	0.4138
Burnout		57	2.592	1.467	57	2.719	1.548	0.5325
Self-valuation		57	1.825	0.778	57	1.925	0.939	0.3598
Coping		57	8.316	2.037	55	7.909	2.390	0.0766
Gratitude		57	12.035	2.275	56	11.821	2.552	0.2538

To explore the potential relationships between individual characteristics and well-being outcomes, we performed bivariate analyses of study outcomes, with PERMA, hardiness, self-valuation, coping, gratitude, and intolerance of uncertainty. We stratified coaches by burnout status and by PFI score. Coaches with lower burnout at baseline (< 3.3) were more likely to have higher HRS/DRS-15 scores [mean (SD) 32.91(5.18) vs. 27.64(5.09) p = 0.0016] than those who had higher burnout (> 3.3) at baseline. This association between low levels of burnout and higher HRS/DRS-15 scores [mean (SD) 33.36(5.04) vs. 28.11(4.65) p = 0.0008] remained true after completion of the coaching program. There were no differences in PERMA, IUS, coping, self-valuation, and gratitude scores for those with high burnout versus low burnout prior to the program but after the program, coaches with lower burnout scores had significantly higher scores across all domains. Coaches at the end of the program with low burnout scores met with their coachee more frequently than the coaches with higher levels of burnout [mean (SD) 3.95(2.16) vs. 2.35(2.13) p = 0.0099] (see Table 3).

Table 3.

Bivariate analysis of study outcomes by burnout score in coaches

Scores		Pre					Post
		Burnout < 3.3 (n = 43)		Burnout >  = 3.3 (n = 14)		p-value	Burnout < 3.3 (n = 37)		Burnout >  = 3.3 (n = 20)		p-value
		Mean	SD	Mean	SD		Mean	SD	Mean	SD
Hardiness	Communication	11.698	2.231	9.571	2.472	0.0039	12.000	2.128	9.500	2.007	0.0001
	Control	11.302	2.099	10.286	2.268	0.1284	11.735	1.763	10.056	2.128	0.0037
	Challenge	9.907	3.100	7.786	2.455	0.0236	9.706	2.866	8.706	2.91	0.2482
	Total	32.907	5.177	27.643	5.093	0.0016	33.364	5.036	28.118	4.649	0.0008
PERMA	Positive Emotion	11.233	1.757	10.571	1.742	0.2257	11.722	1.632	9.889	1.367	0.0001
	Engagement	11.767	1.360	11.500	1.787	0.5573	11.972	1.764	11.158	2.115	0.1348
	Relationships	12.209	1.910	11.308	1.974	0.1445	12.667	1.773	11.316	2.162	0.016
	Meaning	12.791	1.780	12.071	2.093	0.2139	13.472	1.828	11.474	2.633	0.0017
	Accomplishment	11.907	1.571	11.071	1.817	0.1019	12.667	1.852	11.105	1.792	0.004
	Total	59.907	6.298	56.077	7.610	0.0727	62.500	6.893	54.333	7.404	0.0002
Intolerance of uncertainty	Prospective	17.884	5.328	20.429	6.248	0.1426	17.167	4.748	19.588	5.185	0.0985
	Inhibitory	8.000	2.645	9.571	3.674	0.086	7.343	2.828	10.111	3.939	0.0049
	Total	25.884	7.145	30.000	8.884	0.0836	24.400	7.022	29.941	8.584	0.0165
Self-valuation		1.890	0.739	1.625	0.887	0.2728	2.095	1.033	1.613	0.646	0.0351
Coping		8.488	2.142	7.786	1.626	0.266	8.694	2.227	6.421	1.981	0.0005
Gratitude		11.884	2.710	11.615	2.022	0.7032	12.730	1.661	10.053	3.064	0.0017
Total times met with trainee				3.946	2.160	2.350	0.0099

When stratifying by PFI score, those surgeon coaches with a higher PFI score (> 7) before the program had significantly higher hardiness [mean (SD) 34.192(5.426) vs. 29.452(4.836) p = 0.001] and PERMA [mean (SD) 62.720(4.668) vs. 56.032(6.745) p < 0.0001] scores. After the program, those with higher PFI scores had greater HRS/DRS-15 scores [mean (SD) 34.440(4.691) vs. 28.720(4.695) p < 0.0001], PERMA [mean (SD) 64.8983(4.717) vs. 54.269(7.136) p < 0.0001], IUS, coping [mean (SD) 9.250(1.602) vs. 6.519(2.293) p =  < 0.0001], self-valuation [mean (SD) 2.317(1.011) vs. 1.491(0.622) p = 0.0005], and gratitude scores [mean (SD) 13.103(1.319) vs. 10.444(2.847) p =  < 0.0001] than those with lower PFI scores (see Table 4).

Table 4.

Bivariate analysis of study outcomes by PFI score in coaches

Scores		Pre					Post
		PF < 7 (n = 31)		PF >  = 7 (n = 26)		p-value	PF < 7 (n = 27)		PF >  = 7 (n = 30)		p-value
		Mean	SD	Mean	SD		Mean	SD	Mean	SD
Hardiness	Communication	9.936	2.22	12.654	1.832	< .0001	9.692	2.205	12.556	1.601	< .0001
	Control	10.581	1.996	11.615	2.264	0.0721	10.462	2.213	11.846	1.617	0.013
	Challenge	8.936	2.670	9.923	3.475	0.2304	8.640	3.040	10.077	2.607	0.0757
	Total	29.452	4.836	34.192	5.426	0.001	28.720	4.695	34.440	4.691	< .0001
PERMA	Positive Emotion	10.323	1.869	11.962	1.113	0.0002	10.077	1.672	12.071	1.245	< .0001
	Engagement	11.258	1.549	12.231	1.177	0.0111	11.000	1.819	12.357	1.789	0.0073
	Relationships	11.613	1.978	12.480	1.829	0.0975	11.185	2.076	13.179	1.362	0.0001
	Meaning	11.71	1.755	13.692	1.379	< .0001	11.333	2.270	14.177	1.307	< .0001
	Accomplishment	11.129	1.450	12.385	1.602	0.0035	11.111	2.026	13.107	1.315	< .0001
	Total	56.032	6.745	62.720	4.668	< .0001	54.269	7.136	64.893	4.717	< .0001
Intolerance of uncertainty	Prospective	18.355	5.851	18.692	5.439	0.8237	19.000	4.983	17.000	4.861	0.1457
	Inhibitory	9.226	3.222	7.385	2.334	0.0185	9.885	3.830	6.741	2.229	0.0008
	Total	27.581	8.398	26.080	6.934	0.4697	29.040	8.438	23.593	6.541	0.0118
Self-valuation		1.758	0.666	1.904	0.900	0.4858	1.491	0.622	2.317	1.011	0.0005
Coping		7.871	2.125	8.846	1.826	0.0714	6.519	2.293	9.250	1.602	< .0001
Gratitude		11.400	2.848	12.308	2.112	0.1779	10.444	2.847	13.103	1.319	< .0001
Total times met with trainee				3.482	2.119	3.300	0.7657

Discussion

This paper analyzes the effects of positive psychology coaching on the faculty physician coaches with respect to burnout and well-being. We found no significant improvements in PFI, burnout, or PERMA scores after the completion of the program for coaches. There may be several reasons for this including low sample size, lack of control group, survey response rate, as well as other organizational/individual factors. It is notable that we did not see any significant worsening of these scores during the study, as our program spanned the start of the COVID-19 pandemic. There are much data on increased physician burnout rates during the COVID-19 pandemic, particularly female physicians [31–33]. The fact that well-being measures were stable may demonstrate the nurturing effects of the program on coaches. In comparing cohorts from year 1 (pre-COVID-19) and year 2 (which spanned the pandemic), there were no significant differences in mean burnout score before or after the program. The coaching program potentially may have supported these female surgeons during this uniquely stressful time by training them in personal reflection, goal setting, coping, communication, tolerating uncertainty, and self-valuation.

Overall, we observed low levels of burnout throughout the study, in contrast with prior studies of women surgeons [17, 34, 35]. Mete et al. reported that 46% of female surgeons in their study met criteria for burnout compared to 33% in male surgeons [10] while Dyrbye et al. reported a 43.3% burnout rate compared to 39.0% in men [8]. In our study, only 22.8% met criteria for burnout at baseline and 35% at the end of the program. This could be explained by self-selection by coaches, as it might be hypothesized that those with higher burnout might be less likely to volunteer as a coach. In the future, it would be valuable to ask surgeons about their motivation in volunteering for the program. Interestingly, we did find that lower burnout was associated with more frequent coaching sessions, suggesting perhaps that those with lower burnout were more available and engaged in coaching than those with higher burnout, or perhaps that coaching (helping others) was protective from burnout. Additionally, being part of this project created new connections among female surgeons which could have contributed to decreased burnout.

It is worth noting that certain individual characteristics were different in those with or without burnout and that these varied from baseline to post-study survey. Specifically, those with higher hardiness scores were less likely to be burned out, suggesting a possible protective effect of hardiness. Hardiness has been defined as the ability to adapt under stressful conditions while maintaining emotional stability [36] whereas grit has been defined as an individual’s passion and dedication to achieve a long-term goal [37]. A prior study by Betchen et al. found no association between grit and gender [11]. Additionally, based on the definition of grit, one might not expect that to change with coaching, whereas one might expect hardiness to improve with coaching (which was found in our trainee study) [17]. Therefore, we chose to use the measure of hardiness by the HRS/DRS-15 score for this study.

After completion of the coaching program, we did see significantly higher levels of all individual characteristics hypothesized to be correlated with burnout (e.g., hardiness, IUS, coping, and gratitude) in those who were not burned out. This may suggest that some of these characteristics may be modifiable through training and acquiring new skills, such as coaching skills. Research has suggested that certain characteristics are innate, present in us from birth, while others are acquired throughout our lives through experience[38]. Hypothesized innate characteristics include tolerance of uncertainty, resilience, and gratitude [39–41]. In contrast, professional fulfillment, burnout, and self-valuation are viewed as acquired skills. All of these skills, whether innate or acquired, can be mutable and liable to change with certain skill development [27, 42, 43]. Our results demonstrating the correlation of hardiness and burnout suggest that we might target improving these skills and personality characteristics in future coaching programs and research.

There are several limitations to our study. Our small sample size can limit results and fail to detect key differences, particularly without a control group. As the study was powered to identify changes in coachees, this may suggest a type 2 error. The response rate, although good comparatively, limits the ability to evaluate whether those who did not complete the surveys had different responses and potentially benefited less from the program, further altering results. The low burnout rates in our study may limit generalizability of the study. More importantly, this coaching program did not occur in a vacuum. Other life factors, including the initial stages of the COVID pandemic, could have influenced the burnout and well-being scores beyond just this program. Additionally, coaching is just one component of a burnout prevention toolkit. Lastly, we only studied women surgeons in the present study. It is possible that we would have found different outcomes in more inclusive populations. KP has been working to expand this program to other societies including the American Pediatric Surgical Association and the American College of Physicians, as well as multispecialty implementation within graduate medical education programs nationally [44, 45].

Conclusion

In this study which trained women surgeons in positive psychology coaching techniques for a 9-month resident coaching program, burnout and well-being scores were stable over time among coaches. While burnout and professional fulfillment outcomes did not change as a result of the program, we did find an association in certain characteristics associated with well-being in those with low burnout scores at the end of the program.

This paper suggests several possible avenues for future study. Most of our analysis was quantitative using previously validated tools to measure burnout and well-being. A qualitative study may reveal more insight over a purely quantitative analysis. Additionally, perhaps, eliminating burnout should not be the ultimate outcome and goal of these studies. Rather, we should focus on targeting individual characteristics that can improve coping skills and experiences in domains of physicians’ personal and professional lives. Expanding this program to all surgeons in the future may give more insight into the impact of this program on coaches.

Funding

This study was supported by a Grant from the Physicians Foundation in the amount of $100,000. Grant monies paid for the statistical analysis (Sarode and Donelan), training (Palamara), and curriculum development (Palamara).

Declarations

Conflict of interest

Dr. Palamara was paid a consulting fee for coach training and curriculum development. No other disclosures are reported.

Ethical approval

The study was approved by the Massachusetts General Hospital IRB.

Informed consent

Informed consent was obtained from all individual participants included in the study.