Leadership behavior associations with domains of safety culture, engagement, and healthcare worker well-being

Abstract

Background:

Leadership is a key driver of healthcare worker well-being and engagement, and feedback is an essential leadership behavior. Methods for evaluating interaction norms of local leaders are not well developed. Moreover, associations between local leadership and related domains are poorly understood. This study sought to evaluate healthcare worker leadership behaviors in relation to burnout, safety culture, and engagement using the Local Leadership Scale of the Safety, Communication, Operational Reliability, and Engagement (SCORE) survey.

Methods:

The SCORE survey was administered to 31 midwestern hospitals as part of a broad effort to measure care context, with domains including Local Leadership, Emotional Exhaustion/Burnout, Safety Climate, and Engagement. We used mixed-effects hierarchical logistic regression to evaluate the relationships between local leadership scores and related domains, adjusted for role and work-setting characteristics.

Results:

Of the 23,853 distributed surveys, 16,797 (70.4%) were returned. Local leadership scores averaged 68.8 +/− 29.1, with 7338 (44.2%) reporting emotional exhaustion, 9147 (55.9%) reporting concerning safety climate, 10,974 (68.4%) reporting concerning teamwork climate, 7857 (47.5%) reporting high workload, and 3436 (20.7%) reporting intentions to leave. Each 10-point increase in local leadership score was associated with odds ratios of 0.72 (95% confidence interval [CI] 0.71–0.73) for burnout, 0.48 (95% CI 0.47–0.49) for concerning safety climate, 0.64 (95% CI 0.63–0.66) for concerning teamwork climate, 0.90 (95% CI 0.89–0.92) for high workload, and 0.80 (95% CI 0.78–0.81) for intentions to leave, after adjustment for unit and provider characteristics.

Conclusions:

Local leadership behaviors are readily measurable using a 5-item scale and strongly associate with established domains of health care worker well-being, safety culture, and engagement.

Modern health systems are complex, with hierarchical and overlapping configurations of provider teams and administrators with a shared goal of providing high quality patient care.^{1 2} The leaders responsible for overseeing and managing this complex system are often not chosen for or trained in effective leadership skills.^{3 4} Nevertheless, quality of patient care and health care worker (HCW) well-being are strongly influenced by leadership. Effective leadership drives better safety climate, better error reporting, fewer adverse events, and improved employee well-being.^{2 5–8} In addition, acute stressors such as the COVID-19 pandemic may exacerbate levels of stress on HCWs, such that the need for trust in leadership becomes particularly critical.^{9 10}

As the complexity of healthcare systems has accelerated, so has the perceived divide between front line workers and leadership, with a perceived misalignment between the priorities of healthcare leaders and those of HCWs.^11–13 This perceived values misalignment may be a key mediator that undergirds the relationship between leadership behavior and HCW well-being.^{11 14} Leadership may promote values alignment and thus improve HCW well-being with behaviors that serve to strengthen the relationship between HCWs and their leaders, such as being available, setting expectations and providing timely and effective feedback on performance.^14–16 The Local Leadership (LL) scale of the Safety, Communication, Operational Reliability, and Engagement (SCORE) survey is a brief set of items designed to evaluate interaction styles by leaders in health care settings related to feedback and availability. Identifying leaders doing particularly well could help identify strategies to more broadly improve leadership, with targeted interventions and coaching directed toward leaders with lower LL perceptions. The psychometric validity of this scale is described in an accompanying article. However, the convergent validity of LL with other established measures is not yet known, particularly in relation to key safety indicators, well-being markers, and employee engagement metrics.

This study sought to quantify the relations between LL perceptions and the established measures of well-being, safety culture, and engagement. We hypothesized that higher LL perceptions would correspond to lower emotional exhaustion (burnout), higher safety climate, higher teamwork climate, lower perceived workload, and lower intentions to leave.

METHODS

Design and Study Population

This cross-sectional observational study is a secondary analysis of the SCORE survey and demographic items distributed via e-mail through the Michigan Health and Hospital Association Keystone Center in 2015 as part of their routine patient safety and quality measurement.^{17 18} All employees working 0.5 full-time equivalent or higher for four consecutive weeks prior to survey administration were invited to participate. Respondents were assured of confidentiality, and participation at the individual and organizational level was voluntary. No tangible incentives were offered, but each organization was provided the opportunity for feedback and safety improvement based on their organizational data. The full SCORE survey consists of 80 short questions and takes an estimated 8 to 10 minutes to complete. The survey was administered by the Keystone Center, with each organization allowed to encourage participation locally to maximize response rates. This study was not considered human subjects research by the first author’s university and was approved by the Institutional Review Board at the second author’s university and by the ethics committee of the third author’s university. The study’s funder had no role in design, analysis, reporting, or the decision to publish findings.

Measures

The SCORE survey measures common workplace issues and work setting norms,^{5 17 18} including scales of Local Leadership (LL, described below), Improvement Readiness, Burnout Climate, Emotional Exhaustion, Teamwork Climate, Safety Climate, and Work-Life Integration.^{17 19 20} SCORE also contains workforce engagement subscales (Growth Opportunities, Participation in Decision-Making, Advancement, Workload, Job Uncertainty, and Intentions to Leave), as well as questions regarding activity restriction due to illness and missed work for any reason, and demographic questions (i.e., number of years in specialty, job position, shift type and length). The complete questionnaire is published online, along with scoring instructions.²¹

Measurement of Local Leadership, Emotional Exhaustion, Safety Climate, Teamwork Climate, Workload, and Intentions to Leave

The LL scale (Cronbach’s alpha 0.96) of SCORE consists of five items assessing leadership behaviors regarding feedback and accessibility, designed to be both diagnostic and actionable. Providing effective feedback as a leadership behavior has been associated with well-being, and accessibility has been identified as one of three key leadership behaviors to promote psychological safety.^{4 15} LL psychometric validity is described in detail in an accompanying article. The LL scale begins with the prompt “In this work setting, local management…” (or “In this work setting, local physician leaders…” for physicians). Then individual items are as follows:

• Is available at predictable times.

• Regularly makes time to provide positive feedback to me about how I am doing.

• Provides frequent feedback about my performance.

• Provides useful feedback about my performance.

• Communicates their expectations to me about my performance.

Responses were scored on a five-point Likert scale, from “disagree strongly” to “agree strongly. At the individual respondent level, LL was calculated by transposing the mean score of the five scale items onto a 0 to 100 scale.

The emotional exhaustion scale (Cronbach’s alpha 0.92) consists of five items adapted from the nine-item Emotional Exhaustion sub-scale of the Maslach Burnout Inventory, (e.g., “Events in this work setting affect my life in an emotionally unhealthy way”), and has been demonstrated as valid for use on HCWs.^{5 17 18 22} The safety climate scale (Cronbach’s alpha 0.87) consists of seven items evaluating the safety attitudes and norms within the work setting (e.g., “The culture in this work setting makes it easy to learn from the errors of others”).^{5 17 18} The teamwork climate scale (Cronbach’s alpha 0.82) consists of seven items evaluating the ability of HCWs to work together effectively (e.g., “The people here from different disciplines/backgrounds work together as a well-coordinated team”).^{5 17 18} The workload scale (Cronbach’s alpha 0.84) consists of five items evaluating the demands of the workplace (e.g., “With respect to the workload in this work setting I have to attend to many things at the same time”).⁵ The intentions to leave scale (Cronbach’s alpha 0.90) consists of three items related to the desire and plans to leave one’s current position (e.g., “I often think about leaving this job”).⁵

Responses were scored on a five-point Likert scale from “disagree strongly” to “agree strongly.” We calculated each individual’s scale score responses by transposing the mean score of the scale items onto a 0 to 100 scale. For ease of interpretation we dichotomized the outcomes based on established thresholds. A score of 50 or higher on the emotional exhaustion scale was considered concerning for burnout, which reflects “not disagreeing” on average to the burnout items, in line with prior research.^{22 23} Scores of < 75 on the safety climate or teamwork climate scales, and scores ≥ 75 on the workload and intentions to leave scales were considered indicative of concerning responses to each of these items, also in line with prior research.^{5 18 19} Percent positive scores (scores ≥ 75) were calculated for the LL scale, aggregated by work setting and termed “LL climate.”

Work settings were classified as direct patient care (clinical) or indirect patient care (non-clinical, e.g., administrative or billing). Work settings providing direct patient care were further classified as either intensive/emergency care or acute care, as surgical or medical, and as inpatient, outpatient, or mixed inpatient/outpatient. To maintain confidentiality and reduce risk of response bias from small samples, respondents from work settings with fewer than five total respondents were excluded from regression analyses.

Statistical Analysis

Descriptive statistics are presented as means and standard deviations or frequencies and percentages as appropriate, and we compared group means via two-tailed T-tests. We excluded surveys that did not include LL responses and considered other missing data as informative missing. We evaluated the relationships between LL and the outcomes of emotional exhaustion (burnout), safety climate, teamwork climate, perceived workload, and intentions to leave using correlation coefficients and mixed effect logistic regressions, with work setting as random intercept, and job position, number of years in specialty, and work setting classifiers as fixed effects. Analyses were done using Stata 15.1 (StataCorp, College Station, TX). We used simple Bonferroni correction to account for multiple comparisons. With 5 comparisons and a desired familywise error rate of < 0.05, two-tailed p values < 0.01 were considered statistically significant.

RESULTS

Of 23,853 distributed surveys, 16,797 were returned for a 70.4% response rate. Descriptive statistics are shown in Table 1. Of the named positions represented, the most frequent were nurses (27.1%), administrative support personnel (11.7%), technologists (6.8%), and technicians (5.6%). The majority of respondents (67.3%) reported five or more years in their current specialty. Nearly half (47.7%) of respondents were from units not providing direct patient care, 9.2% were from units providing intensive or emergency care, and 9.9% were from units providing primarily surgical care. Of the 1140 work settings represented, 818 (71.8 %) had five or more unique respondents and were included in regression analyses. Characteristics of individuals from larger and smaller work settings are shown in Supplemental Table 1, with the 844 (5.0%) of respondents from smaller work settings less likely to be nurses, more experienced, more likely to be in indirect patient care, more likely to work day shift, and less likely to work 12-hour shifts.

Table 1.

Respondent Characteristics (N=16,797 Respondents from 1,140 Work Settings)

Position	Freq.	Percent
Nurse	4,555	27.1
Admin Support	1,961	11.7
Technologist	1,134	6.8
Technician	939	5.6
Admin/Manager	807	4.8
Therapist	742	4.4
Nurses’ Aide	672	4.0
Clinical Support	635	3.8
Other Manager	482	2.9
Physician Attending/Staff	387	2.3
Environmental Support	350	2.1
Pharmacist	240	1.4
Clinical Social Worker	129	0.8
Physician Assistant	110	0.7
Dietician	93	0.6
Mental Health Assistant	52	0.3
Physician Resident	42	0.3
Physician Non-employed	28	0.2
Other	3,439	20.5
Missing	562	3.4
Years in specialty
0 - 2 years	3329	20.16
3 - 4 years	2069	12.53
5 - 10 years	3650	22.10
11 - 20 years	3993	24.18
21 years or more	3472	21.03
Setting
Indirect patient care	8014	47.71
Direct patient care	8783	52.29
Acute care	7244	43.13
ICU	1539	9.16
Medical	7123	42.41
Surgical	1660	9.88
Inpatient	4695	27.95
Mixed	3207	19.09
Outpatient	881	5.24
Shift
Day	11753	69.97
Night	2458	14.63
Swing	885	5.27
Other	1371	8.16
Shift Length
8 hours	8515	50.69
10 hours	1359	8.09
12 hours	4334	25.80
Flex	978	5.82
Other	1372	8.17
Scale Percent Positive Responses
Emotional exhaustion	7338	44.22
Concerning safety climate	9147	55.89
Concerning teamwork climate	10,974	68.44
High workload	7857	47.49
Intentions to leave	3436	20.73

LL scores ranged from 0 to 100, with a mean of 68.8 ± 29.1, median of 75, and interquartile range of 50 to 95. LL scores were higher for respondents from indirect versus direct patient care units (mean 70.5 ± 28.5 vs. 67.2 ± 29.6, p < .001). A total of 8985 respondents (54.1%) reported positive LL. With regard to the primary outcome, 7338 (44.2%) reported emotional exhaustion symptoms concerning for burnout. In addition, 9147 respondents (55.9%) reported concerning safety climate, 10,974 (68.4%) reported concerning teamwork climate, 7857 (47.5%) reported high workload, and 3436 (20.7%) reported intentions to leave.

Item-level correlations between LL responses and each of the 5 established measures are shown in Table 2. After reverse scoring the positively phrased items, all safety climate, teamwork climate, emotional exhaustion, workload, and intention to leave items were negatively correlated with LL. Safety climate exhibited the strongest correlations with LL items (median −0.52, IQR −0.41 to −0.55), while workload exhibited the weakest correlations (median −0.17, IQR −0.19 to −0.08).

Table 2.

Item-level Correlations with Local Leadership Items

		LL scale	Available	Positive FB	Frequent FB	Useful FB	Sets Expectations
	Mean (SD)	68.77 (29.13)	74.20 (30.91)	66.91 (33.73)	64.29 (33.24)	66.97 (32.81)	71.41 (31.13)
Emotional exhaustion scale	41.70 (30.48)	−0.50	−0.41	−0.46	−0.45	−0.47	−0.44
Events in this work setting affect my life in an emotionally unhealthy way.	41.02 (34.82)	−0.47	−0.39	−0.44	−0.43	−0.45	−0.41
I feel burned out from my work.	40.61 (34.97)	−0.41	−0.34	−0.38	−0.37	−0.38	−0.36
I feel fatigued when I get up in the morning and have to face another day on the job.	40.84 (35.52)	−0.42	−0.34	−0.40	−0.39	−0.40	−0.37
I feel frustrated by my job.	43.98 (35.82)	−0.49	−0.40	−0.45	−0.45	−0.46	−0.43
I feel I am working too hard on my job.	42.38 (32.92)	−0.37	−0.31	−0.34	−0.34	−0.35	−0.33
Safety climate scale	67.66 (22.33)	−0.75	−0.58	−0.69	−0.69	−0.72	−0.68
My suggestions about quality would be acted upon if I expressed them to management.a	64.40 (30.95)	−0.61	−0.48	−0.58	−0.57	−0.58	−0.54
Errors are handled appropriately in this work setting.a	71.47 (29.25)	−0.58	−0.47	−0.53	−0.52	−0.55	−0.53
I receive appropriate feedback about my performance.a	67.87 (30.84)	−0.81	−0.55	−0.76	−0.78	−0.80	−0.73
The culture in this work setting makes it easy to learn from the errors of others.a	67.66 (27.84)	−0.57	−0.44	−0.52	−0.53	−0.54	−0.53
I would feel safe being treated here as a patient.a	78.24 (27.32)	−0.44	−0.35	−0.40	−0.39	−0.41	−0.40
In this work setting, it is difficult to discuss errors	64.48 (30.73)	−0.34	−0.27	−0.31	−0.30	−0.32	−0.30
The values of facility leadership are the same values that people in this work setting think are important.a	60.55 (31.68)	−0.55	−0.45	−0.50	−0.50	−0.52	−0.49
Teamwork climate scale	61.81 (22.04)	−0.56	−0.46	−0.51	−0.50	−0.52	−0.51
Disagreements in this work setting are appropriately resolved.a	66.02 (31.63)	−0.50	−0.41	−0.46	−0.46	−0.48	−0.46
In this work setting, it is difficult to speak up if I perceive a problem with patient care.	67.47 (31.94)	−0.35	−0.29	−0.32	−0.31	−0.33	−0.31
It is easy for personnel here to ask questions when there is something that they do not understand.a	77.26 (28.13)	−0.46	−0.39	−0.42	−0.40	−0.43	−0.43
The people here from different disciplines/backgrounds work together as a well-coordinated team.a	72.76 (28.77)	−0.40	−0.32	−0.36	−0.36	−0.37	−0.37
Dealing with difficult colleagues is consistently a challenging part of my job.	52.05 (34.92)	−0.27	−0.22	−0.25	−0.24	−0.25	−0.24
Communication breakdowns are common in this work setting.	49.77 (33.82)	−0.40	−0.34	−0.37	−0.37	−0.38	−0.36
Communication breakdowns are common when this work setting interacts with other work settings.	49.56 (32.18)	−0.34	−0.28	−0.30	−0.31	−0.32	−0.31
Workload scale	67.10 (22.92)	−0.20	−0.16	−0.19	−0.20	−0.19	−0.16
With respect to workload, I have too much work to do	53.17 (30.51)	−0.22	−0.19	−0.20	−0.19	−0.20	−0.19
With respect to workload, I have to work under time pressure	64.16 (30.45)	−0.21	−0.17	−0.20	−0.20	−0.20	−0.17
With respect to workload, I have to attend to many things at the same time	64.36 (32.32)	−0.18	−0.14	−0.17	−0.18	−0.18	−0.15
With respect to workload, I have to give continuous attention to work	75.16 (26.62)	−0.08	−0.05	−0.08	−0.09	−0.08	−0.05
With respect to workload, I have to remember many things	78.34 (26.53)	−0.08	−0.06	−0.09	−0.10	−0.08	−0.05
Intention to leave scale	34.46 (31.28)	−0.43	−0.35	−0.40	−0.39	−0.41	−0.38
I would like to find a better job.	39.66 (35.85)	−0.40	−0.31	−0.37	−0.36	−0.38	−0.35
I often think about leaving this job.	38.37 (35.99)	−0.43	−0.34	−0.39	−0.39	−0.41	−0.37
I have plans to leave this job within 1 year	25.94 (31.20)	−0.35	−0.28	−0.32	−0.30	−0.33	−0.31

Pearson product-moment correlation coefficients

All correlations significant at the P < .01 level

The 20% largest correlations for each subscale are highlighted in green

^aItem is reverse-coded for scoring purposes.

Overall LL score was associated with better scores on all domains evaluated via multivariable logistic regressions, as shown in Table 3. Each 10-point increase in LL was associated with a 28.3% reduction in the odds of emotional exhaustion (burnout), 51.9% reduction in the odds of concerning safety climate, 35.7% reduction in the odds of concerning teamwork climate, 9.7% reduction in odds of high perceived workload, and 20.1% reduction in the odds of intention to leave.

Table 3.

Local Leadership Scale as a Predictor of Burnout, Safety Climate, Teamwork Climate, Workload, and Intentions to Leave

	Emotional exhaustion		Concerning Safety Climate		Concerning Teamwork Climate		High Workload		Intentions to Leave
	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI
Leadership Score (each 10 pts)	0.72*	0.71-0.73	0.48*	0.47-0.49	0.64*	0.63-0.66	0.90*	0.89-0.92	0.80*	0.78-0.81

N = 14,924 respondents in 818 work settings. Estimates via mixed model with unit as random intercept, adjusted for job type, years of experience, patient care type (ICU vs. not, surgical vs. not, inpatient vs. not), and direct patient care vs. not.

^*Statistically significant at the P < .01 level.

CI, confidence interval.

Emotional Exhaustion (Burnout).

Four of the 5 LL items were independently related to burnout, each associating with a 4.6% to 10.6% reduction in odds of emotional exhaustion for each 1-point improvement in leadership item, as shown in Table 4. The strongest independent associations were for the LL items regarding useful feedback and availability at predictable times. Frequent feedback was not independently associated with exhaustion. Results were similar when stratified by direct patient care versus indirect patient care. Interaction terms for direct patient care versus not in relation to local leadership items were all non-significant (p values 0.41 to 0.96).

Table 4.

Local Leadership Items as Independent Predictors of Emotional Exhaustion

	All respondentsa		Direct patient careb		Indirect patient carec
In this work setting, local management…	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI
..is available at predictable times	0.90*	0.89-0.92	0.90*	0.88-0.93	0.90*	0.88-0.93
..regularly makes time to provide positive feedback to me about how I am doing	0.93*	0.91-0.95	0.93*	0.90-0.96	0.92*	0.89-0.95
..provides frequent feedback about my performance	0.99	0.97–1.02	1.01	0.97–1.05	0.98	0.94–1.02
..provides useful feedback to me about my performance	0.89*	0.87-0.92	0.90*	0.86-0.93	0.89*	0.85-0.93
..communicates their expectations to me about my performance	0.95*	0.93–0.98	0.95*	0.92–0.98	0.96	0.93–0.99

Estimates via mixed model with unit as random intercept. Odds ratios reflect the odds of emotional exhaustion for each 1-point increase in local leadership item (5-point scale).

^aN= 14,924 respondents in 818 work settings. Adjusted for job type, years of experience, patient care type (ICU vs. not, surgical vs. not, inpatient vs. not), and direct patient care vs. not.

^bN = 7971 respondents in 395 work settings. Adjusted for job type, years of experience, and patient care type (ICU vs. not, surgical vs. not, inpatient vs. not)

^cN = 6952 respondents in 423 work settings. Adjusted for job type and years of experience

^*statistically significant at the P < .01 level.

Safety Climate.

All 5 LL items were independently related to safety climate (Supplemental Table 1), with an 8.99% to 19.4% decrease in odds of reporting concerning safety climate for each 1-point improvement in LL item. As shown in Table 2, all safety climate items were significantly correlated with all LL items, with the strongest correlations observed for the safety climate item about adequate feedback.

Teamwork Climate.

Three of 5 LL items were independently related to teamwork climate (Supplemental Table 2), with a 3.4% to 14.7% decrease in odds of reporting concerning teamwork climate for each 1-point improvement in LL item. The teamwork climate items most strongly correlated with LL were about appropriate conflict resolution and asking questions to clarify uncertainty, both of which are rooted in psychological safety.

Workload.

Three of 5 LL items were independently related to workload (Supplemental Table 2). The relationships between workload and LL were less consistent than for the other safety culture and engagement domains, with one item (clear expectations) associated with 3.2% higher odds of high perceived workload and two items (predictable times and useful feedback) associated with 2.7% to 3.8% lower odds of high perceived workload.

Intentions to Leave.

Four of 5 LL items were independently related to intentions to leave (Supplemental Table 2), with a 3.7% to 10.1% reduction in odds of intentions to leave for each 1-point improvement in LL item. In particular, intentions to leave were most highly associated with the LL item about getting useful feedback, followed by getting feedback that is positive, predictable availability, and setting clear expectations.

Figure 1 demonstrates the unit-level correlations between LL climate and the 5 employee engagement scales, showing positive associations for safety climate and teamwork climate, with negative associations for emotional exhaustion, workload, and intention to leave. The quartiles of LL climate were associated with significant differences in all domains assessed.

Figure 1.

The bar graph shows the mean of the percent positive response for each of the employee engagement scales, stratified by quartile of local leadership score. T-test results are shown for illustration purposes.

DISCUSSION

This study found that positive assessments of local leadership behaviors were strongly associated with lower emotional exhaustion (burnout), better safety climate, better teamwork climate, lower perceptions of workload, and reduced intentions to leave. Respondents reporting positive local leadership behaviors also reported: less frustration with their jobs; getting appropriate feedback about their performance; learning from errors; better handling of errors, resolving conflicts, better ability to ask questions, and less frequent intentions to leave the job. These relationships were most prominent for the domains of emotional exhaustion, safety climate, and teamwork climate. Surprisingly, similar effects were apparent for respondents in direct clinical care versus those in indirect clinical care.

These results highlight and build on a growing body of literature documenting of the importance of leadership in healthcare settings, with a recognition that leaders are central to team communication, establishment of clear expectations, career development, and goal setting.^{4 24–26} Much of the reported relationships focus on leadership of physicians or nurses, but our findings of nearly identical relationships among indirect care providers suggest that similar phenomena may be occurring across diverse work settings.^{3 27–29} Despite the common tendency to stratify or isolate research findings to reflect one particular population or work setting, our results suggest that leadership findings (using this scale) from one type of work setting may carry relevance to other settings, despite the differences in clinical or provider roles among settings.

Local leadership behaviors were found to be strongly negatively correlated with the primary outcome of health care worker emotional exhaustion, with each 10-point increase in LL score corresponding to a 28% decrease in the odds of exhaustion in multivariate analysis. A recent analysis of 2813 physicians by Shanafelt et al. found a similar widespread pattern of correlations between 12 leadership items and burnout domains, focused largely on empowerment and satisfaction with leadership.⁴ Although all LL items in the present study were significantly correlated with emotional exhaustion, the item related to receiving useful performance feedback exhibited the strongest relationship in univariate correlations and in multivariate regressions. Although directionality of the relationship requires further evaluation, the observed pattern of item-level correlations (Table 2) may suggest that leaders who are able to provide effective feedback may be driving improved outcomes via psychological safety. Psychological safety refers to a shared belief that risk-taking (i.e., speaking up or proposing new ideas) among team members is acceptable, promoting team learning and continuous improvement—advancements that may promote professional satisfaction and quality of care by extension.^30–32 As leaders provide feedback, they may also be providing mechanisms for reducing effort (e.g., by proposing more efficient ways of performing tasks) and providing rewards (e.g., by commending for a job well done), both of which could serve to mitigate effort-reward imbalance and promote well-being.^33–37

Local leadership also strongly correlates with better safety climate and better teamwork climate, consistent with the expected degree of overlap among these domains.^{2 6} For example, perceptions of leadership are largely predicated on the communication and feedback received at the individual level, while safety climate may be influenced by both individual and group feedback regarding performance, errors, and near-misses.³¹ Similarly, teamwork climate is primarily related to the degree of openness and communication between individual team members and their leaders as well as among team members themselves.³⁸ It is possible that leaders influence this climate directly through their interactions with their team members as well as indirectly through the example they set.

Local leadership behaviors also strongly correlate with perceived values alignment within the safety climate scale, a finding supported by a recent study involving physicians.¹² This relationship may be generalizable beyond physicians to all healthcare workers, an important finding considering the role of values alignment in both well-being and engagement.^{11 14}

Perceived workload and intentions to leave may be less directly related to the communication between leaders and team members, but still demonstrated strong relationships.³⁹ Availability of leaders may result in reduced workload if the leaders recognize that workload is excessive and institute changes to remedy this, or it may improve the perception of workload merely by the process of having concerns heard and validated.⁴⁰ Similarly, an individual may be less likely to leave a work setting if their concerns are addressed or validated, which is directly related to the ability of leadership to recognize pain points and address them satisfactorily.³⁹

This study also builds on our anecdotal findings from experience debriefing healthcare leaders on their LL scores. For one, high scores on leadership availability and low scores on leadership feedback suggest an opportunity to improve an essential leadership behavior—that of coaching and generating positive relationships with individuals who report to those leaders and managers. Conversely, the opposite pattern of scores may identify an organization’s structural defects, such as when a Nurse Manager has 150 individuals who report to them and no assistants helping to create regular monthly feedback loops. Lower scores across the entire domain may highlight leaders whose behaviors do not support cultural attributes needed for highly reliable care. Delving into these details of LL influences how managers and directors apportion time in their calendars. In our experience, LL domain analysis done well can improve a leader’s daily personal satisfaction, improve engagement by frontline workers, and enhance the structural foundations on which highly reliable care is built. Thus, incorporating such a LL domain analysis into annual assessments of HCW well-being and safety culture—when paired to timely and actionable feedback to leaders—could serve to improve HCW perceptions of leadership and leaders’ professional satisfaction.

This study must be interpreted in the context of its design. As a cross-sectional, observational study it cannot determine causality or directionality of the observed correlations (i.e., whether better leadership reduces emotional exhaustion, emotional exhaustion reduces perceptions of leaders, an external factor influences both, or a combination of these effects is present). Although the response rate of over 70% compares favorably with other studies of this magnitude and exceeds commonly accepted thresholds for survey-based research, it is possible that some sampling bias remains.^{41 42} We do not have data from non-responders and thus are unable to directly compare respondent and non-respondent characteristics. Similarly, although the survey was confidential, some responses may be susceptible to recall bias or social desirability bias. Although this survey was administered within a single US state, the respondents represented a diverse array of settings, ranging from independent community hospitals to regional referral centers for 9 of the largest health systems in the state of Michigan, making our results likely to be generalizable to a wide variety of community and academic hospitals across the United States.

It remains unknown whether improvements to leadership through improved training or selection of local leaders will be effective in reducing emotional exhaustion or improving safety and teamwork climates. There is some evidence that LL is responsive to interventions. Cross-sectional results from this same dataset show that work settings have much higher local leadership scores when they use patient safety leadership walk-rounds and directly communicate to HCWs about actions taken as a result of the rounds.⁵ In another independent study, LL was responsive to positive leader walkrounds.⁴³ Longitudinal observational studies may enhance our understanding of the directionality of these relationships, but prospective trials will be needed to fully evaluate the effect of interventions to improve health care worker well-being and engagement.

CONCLUSION

Local leadership behaviors related to feedback and accessibility are readily measurable using a five-item scale and strongly associate with lower health care worker emotional exhaustion (burnout), better safety climate, better teamwork climate, and lower perceived workload and intentions to leave. This construct of leadership behaviors is essentially being predictably available, setting expectations, and providing feedback that is frequent, useful, and positive. Consistently robust associations across a diverse set of metrics illustrate that these core leadership norms are essential across roles, work-setting types, acuity levels and even clinical vs. non-clinical work settings, suggesting a widely generalized response to the local leadership behaviors reported here.