Demographic and occupational factors associated with work-life balance among physician assistants/associates: a cross-sectional study

Abstract

Objective

To examine demographic and occupational attributes associated with work-life balance (WLB) satisfaction among physician assistants/associates (PAs) using a national dataset.

Design

This is a cross-sectional study using 2023 national data.

Setting

USA.

Participants

The study included 149 909 board certified PAs who updated, confirmed or verified their profile questions.

Methods

Descriptive and bivariate statistics, followed by multivariate logistic regression, were conducted to identify factors associated with WLB satisfaction among PAs. The primary outcome was a binary variable derived from a 7-point scale assessing PAs’ satisfaction with WLB. Responses of ‘Somewhat’, ‘Mostly’ and ‘Completely’ satisfied were coded as ‘Satisfied’, while ‘Neither/Nor’, ‘Somewhat’, ‘Mostly’ and ‘Completely’ dissatisfied were coded as ‘Not satisfied’. Our analytical sample comprised 86,000 PAs who responded to a question inquiring about their satisfaction with WLB.

Results

Over two-thirds (71.7%) of PAs indicated satisfaction with WLB. The multivariate logistic regression revealed that the types of specialties that PAs practised were among the strongest factors associated with WLB satisfaction. Compared with PAs in primary care, those practising in dermatology (adjusted OR (aOR)=1.83; 95% CI 1.66 to 2.02), general surgery (aOR=1.64; 95% CI 1.48 to 1.83), pain medicine (aOR=1.63; 95% CI 1.41 to 1.89) and hospital medicine (aOR=1.52; 95% CI 1.37 to 1.68) had higher odds of being satisfied with WLB (all p<0.001). Moreover, compared with females, male PAs indicated nearly 25% higher odds of being satisfied with WLB (p<0.001). Lower odds of WLB satisfaction were observed among PAs with any education debt, those seeing more than 40 patients weekly, those working over 40 hours a week, and PAs in their mid- and late-career stages.

Conclusions

Our findings revealed that PAs practising in non-primary care specialties had the highest odds of reporting satisfaction with WLB. Identifying factors strongly associated with PA work-life balance can aid in developing targeted interventions. However, further research is needed to understand the intrinsic and extrinsic factors influencing PAs’ WLB.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The National Commission on Certification of Physician Assistants (NCCPA) provided data on all board certified physician assistants/associates (PAs) in the USA for 2023.

• The study utilises a comprehensive dataset that includes all board certified PAs in the USA and is the first national study to examine work-life balance among PAs.

• This study does not account for other possible confounding variables, such as social factors, family dynamics/norms, financial status and the number of children.

• The cross-sectional nature of the study does not allow for the establishment of a cause-and-effect relationship.

Introduction

The healthcare sector represents one of the fastest-growing occupations in the USA, comprising 17 million, or 10.8% of the workforce.¹ Physician assistants/associates (PAs) are part of this projected growth, and the profession is expected to increase by 28% over the next decade (2023–2033).² However, this rapid expansion comes at a time when the well-being of health workers is deteriorating, primarily attributed to increased burnout rates, persistent shortages of physicians and ongoing effects of the COVID-19 pandemic.3,6 According to the Centers for Disease Control and Prevention, the well-being of healthcare workers has declined since 2018, with nearly half (46%) reporting feelings of burnout in 2022.⁷ As a result, maintaining a positive work-life balance (WLB) is increasingly recognised as a crucial determinant of career satisfaction, personal well-being and overall professional longevity.⁸ The definition of WLB is elusive and varies depending on the framework of the research study; still, two key concepts include ‘engagement in work life and nonwork life and minimal conflict between social roles in work and nonwork life’.^{9 10}

For healthcare professionals, achieving WLB is essential for individual satisfaction, and it also plays a significant role in the quality of patient care.¹¹ Balancing professional duties with personal life is a challenge faced by many in the healthcare field, and PAs are no exception. The nature of the healthcare environment—characterised by long hours, high-pressure situations and emotional demands¹²—can make it challenging to maintain a sense of balance between career and personal/home life. This imbalance has been linked to increased levels of stress, depression and burnout among healthcare workers,^{13 14} especially during times of crisis. For example, studies conducted during the COVID-19 pandemic found that healthcare workers experienced increased interference between work and personal responsibilities, resulting in elevated stress levels.15,17 Even before the pandemic, studies showed that many physicians struggled with work-life integration. Factors such as excessive workloads, frequent on-call shifts and systemic gender disparities often contribute to dissatisfaction and fatigue.¹⁸ Moreover, a recent national survey has highlighted a significant decline in healthcare professionals’ satisfaction with their WLB during the pandemic, indicating a growing concern in the profession.¹⁷ These findings emphasise the importance of addressing organisational and individual factors that may affect WLB to protect the well-being and career longevity of healthcare providers. This is particularly true for PAs, who are integral to healthcare teams across a broad range of medical specialties and practice settings. PAs are employed in a variety of healthcare settings, ranging from hospitals and private practices to outpatient clinics and urgent care centres. They also practise in various specialties, including primary care, surgery, emergency medicine and more.¹⁹

The demands of healthcare professionals are high. High expectations can lead to consequences, including burnout and decreased job satisfaction, which can, in turn, affect the quality of patient care.20,22 Similar to physicians, PAs often face challenges such as irregular working hours, high patient loads and the emotional toll of patient care.^{8 23} PAs may find that their practice environment, employer support and satisfaction with their specialty are contributing factors to their perception of WLB. When WLB is perceived as poor, a possible consequence is burnout.^{24 25} Burnout is a growing concern within the medical field. It can be characterised by emotional exhaustion, depersonalisation and a diminished sense of personal accomplishment.²⁶ For board-certified PAs, it has been reported that burnout rates were 26.6% in 2020 and increased to 34.1% by 2023.^{19 27 28}

Although burnout has received significant research attention since the COVID-19 pandemic, WLB among healthcare providers has remained underexplored, with only a few studies published. Shanafelt and colleagues examined burnout and WLB among US physicians relative to the general population and found that 48.2% of physicians indicated satisfaction with WLB.²⁸ Similarly, Dyrbye et al compared WLB of PAs to the US general population and found that nearly two-thirds (65.3%) of PAs reported being satisfied with WLB.²³ Keeton et al investigated demographic and practice characteristics related to WLB, career satisfaction and burnout among physicians.⁸ The authors found that age, gender and medical specialty did not predict WLB among physicians.⁸ Yet, more recent research revealed that age, race/ethnicity, gender, specialty and hours worked are associated with WLB among physicians and PAs.1823 28,31 The difference between these contradictory findings in the aforementioned studies was that most physician and PA literature research was conducted during or after the COVID-19 pandemic (2020 or later), which dramatically impacted the healthcare industry.^{32 33}

In summary, achieving WLB remains an ongoing challenge for all healthcare providers, including PAs. Despite the importance of WLB, there is a lack of comprehensive studies on WLB among PAs. This study’s objective was to investigate demographic and occupational factors associated with PAs’ satisfaction with WLB using a national workforce dataset, to inform more responsive workplace policies and support systems tailored to the unique needs of PAs. The primary research question guiding this study is: Which demographic and occupational factors are associated with WLB satisfaction among PAs?

Methods

Study design and data source

We conducted a quantitative, cross-sectional study of board certified PAs in the USA to examine demographic and occupational factors associated with WLB satisfaction. We utilised the 2023 PA Professional Profile dataset from the National Commission on Certification of PAs (NCCPA) and followed the STROBE reporting guidelines. The NCCPA certifies all PAs in the USA and has collected workforce data on PAs for over a decade through the PA Professional Profile.^{34 35} The primary role of the PA Professional Profile is to capture data on PAs’ demographics, practice attributes and other workforce-related characteristics. The PA Profile was designed based on the Minimum Data Set by the Health Resources and Services Administration to provide uniform and consistent health workforce information across various healthcare entities.³⁴

Setting and participants

The data for this study were from the 2023 PA Professional Profile and include board certified PAs in the USA who updated, reviewed or completed it (n=1 49 909) from February 2022 to 31 December 2023. Of those PAs who completed their profile, 86 000 (or 57.4%) responded to the WLB question, which comprised our final analytical sample. Participation in the PA Professional Profile is voluntary, and the responses are confidential. Administrative data—such as age, gender, US region and urban-rural settings are 99.9% complete; whereas the rest of the data used are self-reported.

Quantitative variables

Dependent variable

A new question asking PAs to rate their level of satisfaction with WLB was added to the PA Profile in 2022: ‘All things considered, how satisfied are you with WLB?’ Our primary outcome variable for this study was satisfaction with WLB. The response options were on a 7-point scale, ranging from ‘Completely Satisfied’ to ‘Completely Dissatisfied’. Participants who reported that they were ‘Somewhat’, ‘Mostly’ and ‘Completely’ satisfied were grouped into ‘Satisfied’ with WLB, whereas participants who reported ‘Neither/Nor’, ‘Somewhat’, ‘Mostly’ and ‘Completely’ dissatisfied were grouped as ‘Not satisfied’ with WLB. The factors that influence the decision to include ‘Neither/Nor’ responses into the ‘Not satisfied’ group were determined by the following: (1) most health services research36,38 interprets the ‘Neither/Nor’ option as neutral, ambivalence or indifference, which is not technically a positive response. We used ‘satisfied’ and ‘not satisfied’ as categories, since dissatisfied implies a distinct, strong negative sentiment. Neutral responses typically indicate a lack of positive endorsement and are therefore treated as more similar to mild dissatisfaction; (2) we conducted a sensitivity analysis and excluded the ‘Neither/Nor’ cases entirely and then re-ran the logistic regression. The results were consistent with the original model, which included ‘Neither/Nor’ into ‘Not satisfied’ group; (3) the design of the study includes a multivariate logistic regression model which requires a binary outcome variable, thus, based on the above two points ‘Neither/Nor’ would fall into the ‘Not satisfied’ category to interpret the OR for being ‘satisfied’ with WLB. For the reasons discussed above, we decided to include the ‘Neither/Nor’ option in the ‘Not satisfied’” group to increase the study’s sample size and statistical power.

Independent variables

Our independent variables include demographic characteristics (age, gender, race, ethnicity, geographic location (US region and urban-rural settings), and highest degree completed); occupational attributes (primary specialty where PAs practise (primary care, dermatology, general surgery, etc), primary practice setting (hospital, office-based private practice, etc), hours worked and patients seen weekly in the primary clinical position, career stages (early career (less than 10 years of practice), mid-career (11–20 years) and late career (over 20 years)),³⁹ income range (less than or equal to US$60 000 to over US$160 000 in a US$10 000 increments) and secondary position); and other occupational-related factors (education debt (none/not sure to over US$200 000 in a US$50 000 increments), satisfaction with career as a PA, satisfaction with number of hours worked, satisfaction with income, satisfaction with work benefits and satisfaction with employer (satisfied vs not satisfied). Online supplemental appendix A provides a complete description and coding of the independent variables.

Procedures

All responses were assessed for accuracy, and the results were independently validated by two researchers. Using comprehensive administrative data, we compared the characteristics of respondents and non-respondents. While gender distribution was similar (female: 71.1% vs 71.7%), the respondents were slightly older (median age: 38 vs 36 years), more likely to reside in rural/isolated areas (5.9% vs 4.5%) and more likely to indicate dwelling in the Midwest (19.5% vs 17.6%).

Missing data were handled using listwise deletion,⁴⁰ as the missingness patterns suggested the data were missing at random. This method allows complete data analysis without assumptions and maintains comparability across methods by analysing the same set of complete cases. The Cronbach’s alpha for the satisfaction items was 0.904, indicating good internal consistency. WLB was a predetermined primary outcome for this study. Correlation analyses were conducted using Spearman’s Rho⁴¹ (ρ) to determine whether satisfaction with WLB aligns with other satisfaction measures. Overall, WLB is moderately to strongly associated with other aspects of job satisfaction (Spearman’s ρ ranging from 0.417 to 0.837, all p<0.001). The strongest association was observed with satisfaction with work hours (ρ=0.837), and the weakest with job location (ρ=0.417).

Statistical methods

All analyses were performed using SPSS V.29 (IBM Corp, Armonk, NY, USA), with a two-sided statistical significance set at p≤0.05. Analyses included a summary of descriptive statistics, the Pearson’s χ² test for categorical variables and the Mann-Whitney U test for ordinal or continuous variables, comparing demographic and practice/occupational attributes of PAs who reported satisfaction with WLB versus those who did not. Continuous variables were expressed as mean with SD and median with IQR, and categorical variables were expressed as percentages.

Additionally, we conducted a multivariate logistic regression to identify factors associated with WLB satisfaction among PAs. The regression model included multiple covariates: age, gender, race, ethnicity, urban-rural setting, US region, highest degree completed, practice setting, practice area, years certified as a PA (a proxy for career stages), educational debt, secondary position, hours worked per week, patients seen per week, and other satisfaction metrics. The study was exploratory, aiming to explain the association between multiple independent variables (identified in prior literature) and the outcome without a predetermined hypothesis. A possible confounder identified in prior research was burnout. Burnout was considered but excluded from the final model to avoid overadjustment. Furthermore, burnout has been found to be strongly associated with job satisfaction and WLB,⁴² potentially obscuring some of the satisfaction variables included in the model. Covariate selection is based on prior research findings, and the model was focused on identifying patterns in this exploratory study. Lastly, we examined the variance inflation factors (VIFs) to assess multicollinearity before fitting the regression model.⁴³ A VIF greater than 10 suggests possible multicollinearity. In our study, the VIFs were less than 4, indicating that multicollinearity was not present.

Results

Demographic and occupational characteristics of participants

Our study found that 71.7% (n=61 642) of PAs reported being satisfied with WLB, and 28.3% (n=24 358) were not. Figure 1a,b illustrates the proportion of PAs who are satisfied with WLB on a 7-point scale, grouped into two categories (satisfied vs not satisfied). In bivariate analyses, statistically significant (all p<0.05) differences on likelihood of being satisfied with WLB were observed in terms of age, race, US region, years certified as a PA (career stages), medical specialty, practice setting, secondary position, hours worked and patient seen weekly, annual income, educational debt and various levels of job satisfaction (table 1 and; figures 2 and 3). PAs who were less than 30 years of age (77.6%), self-identified as Asian (74.0%) and White (72.1%) had among the highest satisfaction with WLB. Moreover, a slightly higher proportion of PAs who were satisfied with WLB reported residing in the Midwest (72.7%) and the South (72.4%) versus the Northeast (70.8%) and the West (70.6%). However, no statistically significant differences were observed between PAs who were satisfied with WLB and those who were not, regarding gender (p=0.069), ethnicity (p=0.424), and urban-rural setting (p=0.082).

Figure 1. (a) Full breakdown of satisfaction with work-life balance. (b) Proportion of PAs satisfied versus not satisfied with work-life balance. PA, physician assistant.

Table 1. Demographic attributes of PAs satisfied with work-life balance.

	Demographic characteristics
		Total number of respondents	Number of PAs satisfied with work-life balance*	% of PAs satisfied with work-life balance*	P value
Age group	Less than 30	11 872	9210	77.6	<0.001
	30–39	33 411	24 049	72.0
	40–49	21 477	15 048	70.1
	50–59	12 321	8393	68.1
	60+	6919	4942	71.4
Age	Mean (SD)		40.8 (11.1)		<0.001
	Median (IQR)		38 (32–48)
Gender	Female	61 263	44 020	71.9	0.069
	Male	24 715	17 606	71.2
Race	Asian	23 364	3975	74.0	<0.001
	White	51 464	50 065	72.1
	Multiple races	2308	1494	71.1
	Black/African American	2596	1989	71.0
	Other†	2721	1892	69.5
Ethnicity	Non-Hispanic/non-Latino(a/x)	57 193	55 507	72.1	0.424
	Hispanic/Latino(a/x)	25 726	4256	71.6
US region	Midwest	20 445	12 117	72.7	<0.001
	South	27 740	21 846	72.4
	Northeast	18 843	14 300	70.8
	West	18 645	13 164	70.6
Urban/rural setting	Urban	79 091	56 773	71.8	0.082
	Rural/Isolated	6410	4536	70.8
Highest degree	Doctorate degree	4759	1437	72.0	0.003
	Master’s degree	70 965	51 035	71.9
	Bachelor’s degree	8417	7859	70.3
	Other (certificate, associate, other)	1840	1296	70.4

^*PAs not satisfied with work-life balance are not shown.

^†Other race includes other, American Indian/Alaska Native and Native Hawaiian/Pacific Islander.

PA, physician assistant/associate.

Figure 2. Annual income comparing PAs satisfied with work-life balance with those not satisfied with work-life balance. PA, physician assistant/associate; WLB, work-life balance.

Statistical differences in WLB satisfaction were observed by medical specialty (p<0.001), with the highest proportion among PAs in dermatology (80.4%), pain medicine (77.2%), hospital medicine (74.1%), critical care medicine (74.0%) and psychiatry (73.0%). However, lower rates of WLB were observed among PAs practising in primary care (family medicine, general internal medicine and general paediatrics; 67.7%) and emergency medicine (67.4%) (table 2). There was also a significant association between WLB and practice setting. The highest proportion of WLB satisfaction was observed in PAs employed in urgent care (73.0%) and office-based private practice (72.3%). Moreover, a higher proportion of WLB was observed among PAs in the early stages of their career (less than 10 years of practice), 72.9%.

Table 2. Occupational attributes of PAs satisfied with work-life balance.

	Practice characteristics
		Total number of respondents	Number of PAs satisfied with work-life balance*	% of PAs satisfied with work-life balance*	P value
Years certified as a PA/career stage	Less than 10 (early career)	45 169	32 912	72.9	<0.001
	11 to 20 (mid-career)	24 853	17 487	70.4
	21 and over (later career)	15 978	11 243	70.4
Current practice area (top 10)	Dermatology	9306	2989	80.4	<0.001
	Pain medicine	5977	984	77.2
	Hospital medicine	4609	2238	74.1
	Critical care medicine	4129	1266	74.0
	Psychiatry	2333	1605	73.0
	Surgery—general	922	138	72.3
	Internal medicine—subspecialties	6722	6128	72.1
	Surgery—subspecialties	12 800	12 355	71.2
	Primary care†	13 543	13 253	67.7
	Emergency medicine	5976	5923	67.4
	Other	19 661	14 749	75.0
Primary practice setting/location	Urgent care	13 837	3858	73.0	<0.001
	Office-based private practice	31 870	23 029	72.3
	Hospital	26 332	24 902	71.4
	Federal government facilities	3662	2821	70.9
	Rural health clinic (federally certified)	1995	836	67.0
	FQHC	2021	1610	65.7
	Other	6251	4563	73.0
Hours worked weekly at the principal clinical position	Mean (SD)		38.5 (10.1)		<0.001
	Median (IQR)		40 (36–40)
Hours worked weekly at the principal clinical position (in categories)	Less than 30	11 689	9483	81.1	<0.001
	31–40	49 977	38 016	76.1
	41–50	19 213	11 800	61.4
	51 or more hours	5099	2328	45.7
Patients seen weekly at the principal position	Mean (SD)		64.0 (41.0)		<0.001
	Median (IQR)		60 (40–80)
Patients seen weekly at the principal position (in categories)	40 or fewer patients	26 281	20 149	76.7	<0.001
	41–60 patients	21 813	15 590	71.5
	61–80 patients	16 008	10 947	68.4
	81–100 patients	12 288	8378	68.2
	More than 100 patients	9528	6520	68.4
Secondary PA position	No, I work only in one clinical position	73 163	52 617	71.9	<0.001
	Yes, I also work in a position where I do not provide direct patient care (ie, education, research, administration)	3085	2201	71.3
	Yes, I work in two or more clinical PA positions	9668	6768	70.0

^*PAs not satisfied with work-life balance are not shown.

^†Primary care includes family medicine/general practice, internal medicine-general and paediatrics–general.

FQHC, Federally Qualified Healthcare Center; PA, physician assistant.

Furthermore, PAs who reported holding only one clinical job (71.9%), seeing fewer than 40 patients per week (76.7%) and working less than 30 hours weekly (81.1%) reported the highest satisfaction with WLB. Lastly, when examining various aspects of job satisfaction (figure 3), PAs who reported satisfaction with WLB versus those who did not were more likely to indicate satisfaction with their present job, career as a PA, number of hours worked, income, benefits at work and their employer (all p<0.001).

Figure 3. Satisfaction with job position attributes and career of PAs who indicated satisfaction with work-life balance. *Satisfied includes ‘completely satisfied’, ‘mostly satisfied’ and ‘somewhat satisfied’. **Not satisfied includes ‘neither/nor satisfied’, ‘somewhat dissatisfied’ and ‘mostly dissatisfied’. PA, physician assistant/associate.

Figure 2 shows the income ranges of the two groups. PAs who were satisfied with WLB indicated a lower income range than those who were not satisfied with WLB. PAs who were satisfied with WLB reported earning US$3158 less in mean yearly income than their PA colleagues who indicated not being satisfied with WLB (mean (SD), US$123 767 (US$35 848) versus US$126 925 (US$34 366); p<0.001).

Factors associated with WLB among PAs

In multivariate logistic regression (table 3), the types of medical specialties that PAs reported practising were among the strongest factors associated with WLB satisfaction. Compared with PAs in primary care, PAs practising in dermatology (adjusted OR (aOR)=1.83; 95% CI 1.66 to 2.02), general surgery (aOR=1.64; 95% CI 1.48 to 1.83), pain medicine (aOR=1.64; 95% CI 1.41 to 1.89), hospital medicine (aOR=1.52; 95% CI 1.37 to 1.68), surgical subspecialties (aOR=1.40; 95% CI 1.33 to 1.48), critical care medicine (aOR=1.34; 95% CI 1.18 to 1.53), internal medicine subspecialties (aOR=1.30; 95% CI 1.22 to 1.39), and psychiatry (aOR=1.16; 95% CI 1.04 to 1.29)—all had higher odds of being satisfied with WLB.

Table 3. Multivariate logistic regression of work-life balance factors among PAs.

Independent variables		aOR	95% CI		P Value
			Lower limit (LL)	Upper limit (UL)
Gender	Male vs Female	1.25	1.20	1.30	<0.001
Race	Asian vs White	1.04	0.97	1.11	0.294
	Black/African American vs White	1.01	0.92	1.11	0.840
	Multiple races vs White	0.96	0.86	1.07	0.450
	Other race vs White	0.94	0.85	1.03	0.204
Ethnicity	Hispanic/Latino(a/x) vs non-Hispanic	1.04	0.97	1.12	0.233
Urban/rural setting	Rural/isolated vs urban	1.05	0.99	1.12	0.131
Practice location	Office-based private practice vs hospital	1.08	1.03	1.13	0.001
	Urgent care vs hospital	1.02	0.94	1.10	0.642
	FQHC vs hospital	0.85	0.77	0.94	0.001
	Federal government vs hospital	1.13	1.03	1.23	0.007
	Rural health clinic (federally certified) vs hospital	0.99	0.86	1.13	0.836
	Other practice settings vs hospital	1.05	0.98	1.13	0.141
Patients seen weekly at the principal clinical position	Number of patients seen weekly: 41–60 vs ≤40	0.85	0.81	0.89	<0.001
	Number of patients seen weekly: 61–80 vs ≤40	0.77	0.73	0.81	<0.001
	Number of patients seen weekly: 81–100 vs ≤40	0.76	0.72	0.81	<0.001
	Number of patients seen weekly: 101+ vs ≤ 40	0.75	0.70	0.80	<0.001
Practice area	Emergency medicine vs primary care	0.95	0.88	1.01	0.104
	Dermatology vs primary care	1.83	1.66	2.02	<0.001
	Hospital medicine vs primary care	1.52	1.37	1.68	<0.001
	Surgery-general vs primary care	1.64	1.48	1.83	<0.001
	Psychiatry vs primary care	1.16	1.04	1.29	0.008
	Critical care medicine vs primary care	1.34	1.18	1.53	<0.001
	Pain medicine vs primary care	1.63	1.41	1.89	<0.001
	Surgery-subspecialties vs primary care	1.40	1.33	1.48	<0.001
	Internal medicine- subspecialties vs primary care	1.30	1.22	1.39	<0.001
	Other specialties vs primary care	1.42	1.35	1.50	<0.001
Hours worked weekly at the principal clinical location	Hours worked weekly: 31–40 vs ≤30	0.72	0.68	0.77	<0.001
	Hours worked weekly: 41–50 vs ≤30	0.34	0.32	0.36	<0.001
	Hours worked weekly: 51+ vs ≤ 30	0.18	0.17	0.20	<0.001
Years certified (career stages)	Years certified: 21 and over (late-career) vs ≤10 (early career)	0.81	0.77	0.84	<0.001
	Years certified: 11 to 20 (mid-career) vs ≤10 (early career)	0.81	0.77	0.86	<0.001
Education debt	Education debt ≤US$49 999 vs no debt	0.94	0.89	0.99	0.021
	Education debt US$50 000 - US$99 999 vs no debt	0.92	0.87	0.97	0.003
	Education debt US$100 000 - US$149 999 vs no debt	0.91	0.86	0.96	<0.001
	Education debt US$150 000 - US$199 999 vs no debt	0.85	0.80	0.90	<0.001
	Education debt US$200 000 or more vs no debt	0.75	0.70	0.80	<0.001
Highest degree completed	Bachelor’s vs master’s	0.99	0.93	1.04	0.617
	Doctorate vs master’s	1.09	0.98	1.22	0.118
	Other vs master’s	1.04	0.92	1.17	0.522
USA region	Northeast vs South	0.85	0.82	0.89	<0.001
	Midwest vs South	0.97	0.93	1.02	0.229
	West vs South	0.89	0.86	0.94	<0.001
Secondary position	Yes, also work in a non-direct patient care job vs only one clinical job	0.87	0.79	0.95	0.002
	Yes, work in two or more clinical PA jobs vs only one clinical job	0.98	0.93	1.03	0.379

Summary of the overall model fit: −2 log likelihood=86 9920.195, Cox & Snell R2=0.054 and Nagelkerke R2=0.078.

FQHC, Federally Qualified Healthcare Center; PA, physician assistant/associate.

Compared with female PAs, male PAs had 25% higher odds of being satisfied with WLB (aOR=1.25; 95% CI 1.20 to 1.30). Additionally, PAs who reported working in federal government offices and office-based private practices had higher odds of indicating satisfaction with WLB than PAs working in hospital settings (aOR=1.13; 95% CI 1.03 to 1.23, and aOR=1.08; 95% CI 1.03 to 1.13, respectively). On the other hand, lower odds of being satisfied with WLB were observed for PAs with any education debt, those seeing more than 40 patients weekly, those working over 40 hours a week, and those in the late- or mid-career stages (all p<0.001).

The overall model was assessed using –2 log Llkelihood statistics and Pseudo R² measures (Cox & Snell R²=0.054 and Nagelkerke R²=0.078), suggesting that the independent variables included in the model account for only 5.4% to 7.8% of the variance.

Discussion

In this study, using national data, we investigated demographic and occupational factors associated with PAs’ satisfaction with WLB. Our analysis found that the strongest factors associated with satisfaction with WLB were for PAs practising in non-primary care specialties compared with primary care, those working in federal government facilities and private offices, versus those working in hospitals, and male versus female gender. Decreased odds of satisfaction with WLB were observed among PAs at the mid- and late-stages of their careers, those working over 40 hours per week, those seeing more than 40 patients per week, and those with any educational debt. In the following sections, we outline the key factors associated with WLB among PAs identified in our study and compare these findings with existing literature.

WLB and gender

In terms of gender, our study found that male PAs had 25% higher odds of being satisfied with WLB compared with female PAs. This finding is similar to previous studies on gender disparities regarding WLB. Dyrbye and colleagues found that, compared with men, women PAs had 48% lower odds of being satisfied with their WLB.²³ Concurrently, female physicians reported lower WLB satisfaction than male physicians.⁴⁴ These findings are not surprising, as research has shown that working women report more disparities with their WLB than men.⁴⁵ Generally, due to social norms and expectations, women tend to be primary caregivers for children and household tasks.46,48 A study examining burnout and gender differences in the general population found that women who worked full-time also spent 8.5 hours a week on home activities compared with 40 minutes by men who worked full-time.⁴⁹ These findings underline that women PAs may have more family responsibilities at home, in addition to working full-time as healthcare providers, leading to increased stressors and possibly lower levels of satisfaction with WLB; however, more research is needed to test our assumption for women PAs.

WLB and career stages

Our study found that PAs in the mid-career (certified for 11–20 years) and late-career (certified for 21+years) stages had lower odds of being satisfied with WLB than those in the early-career stage (less than 10 years). Similarly, Dyrbye et al noted that mid-career physicians were more likely to report lower satisfaction with their WLB,³⁹ which may be related to working more hours and taking more calls than early-career physicians, and therefore being more susceptible to burnout. Goldman and Barnett also echoed our results, noting that physicians aged 35–44 years reported working more hours than physicians over 65.⁵⁰ On the contrary, another study found that older age was linked to better WLB and lower burnout rates among physicians, although age was not a strong indicator of WLB.⁸ In a later study, Dyrbye and colleagues examined burnout and WLB among PAs and found no statistical differences by age.²³ Considering that the PA profession is relatively young (median age 38),³⁵ it could be that younger PAs may have fewer family responsibilities, are more enthusiastic about starting their new career, and therefore can maintain a better WLB. Although this interpretation is speculative, further studies are needed to examine the relationship between WLB and age or career stages among PAs.

WLB and practice characteristics (specialty, practice setting)

Our study discovered that the top three specialties with the strongest association with increased odds of WLB satisfaction were dermatology, general surgery and pain medicine. Compared with PAs working in primary care, PAs in medical subspecialties had higher odds of being satisfied with WLB. Prior studies have reported that PAs working in non-primary care specialties typically see fewer patients, may work fewer hours, report lower burnout levels and report the highest satisfaction rate with WLB,²⁸ which aligns with our study’s findings. Likewise, Dyrbye and colleagues found that PAs working in paediatric or internal medicine subspecialties had nearly threefold higher odds of being satisfied with WLB than those working in primary care settings.²³ Physician literature also highlighted similar findings that physicians in dermatology, general paediatrics and preventative medicine specialties had the highest proportion of WLB satisfaction.²⁸ Yet, another study did not find any association between WLB and medical specialty among physicians in South Dakota.⁵¹ The above findings emphasise the importance of WLB among clinicians working in medical specialties, which may be another potential explanation for the declining number of PAs working in primary care and the increasing shift toward medical subspecialties.^{52 53} However, further research is needed to explore the relationship between WLB and specialty selection among PAs.

Shifting to practice settings, we found that PAs working in federal government facilities and office-based private practices had 13% and 8% higher odds of being satisfied with WLB than those working in hospitals. On the contrary, Dyrbye and colleagues found no statistically significant differences based on practice settings in a national sample of PAs.²³ Meanwhile, Hussenoeder et al found that physicians working in outpatient settings had higher satisfaction with WLB than those working in hospitals.²⁹ Potential reasons could be that clinicians working in hospital settings are obligated to accommodate shift hours, including nights, weekends and holidays, have less control over their patient load and schedule, and see more complex cases, which may contribute to an imbalanced WLB and increased burnout.

WLB and workload

We found that PAs who reported seeing more than 40 patients weekly and working more than 40 hours weekly were more likely to report lower satisfaction with WLB. Our results align with prior studies in the physician literature, which have highlighted that longer work shifts and high patient volumes are indicative of lower WLB,^{8 18} as increased workload could impact clinicians’ well-being,^{54 55} increase stress and increase the risk for adverse health outcomes.⁵⁶ A study examining the WLB among various groups of healthcare professionals found that those working on average more than 9 hours daily have a lower WLB than those working less than 9 hours.⁵⁷ Lin et al found that healthcare workers working more than 40 hours per week were at higher risk for burnout.⁵⁸

Furthermore, increasing patient load does not necessarily lead to increased productivity. Fairchild and colleagues⁵⁹ examined the productivity of part-time and full-time primary care physicians. The authors observed that physicians’ hourly productivity was higher in part-time than in full-time physicians.⁵⁹ As patient complexity rises, clinicians may have less time for in-depth discussions with their patients and thus less time to document. This time constraint may lead to hurried patient visits and inadequate coding and billing, which could ultimately result in insufficient compensation. In conclusion, an imbalanced WLB could contribute to burnout among healthcare providers.⁶⁰ Since burnout may be closely linked to WLB, similar factors such as workload and time pressures may also impact WLB among PAs.

Implications for policy and practice

 Interventions focusing on WLB should be directed not only to the individual PA but also to the institution that may have control over specific job tasks and roles. For instance, institutional policies and interventions should foster components of employee engagement, work satisfaction and practical organisational commitment that are focused on mitigating stressors.^{45 61} Other factors to consider include greater flexibility and more predictability with the work schedule, shorter work week schedule, job sharing and teamwork.⁴⁵

Strengths and limitations of the study

Despite using a national comprehensive dataset, this study has some limitations that should be considered when interpreting the findings. Due to the study’s cross-sectional nature, we cannot establish a cause-and-effect relationship between the included variables. Moreover, the self-report measure of WLB is based on the perceptions of PAs, which are subjective and susceptible to recall bias. Selection bias may have occurred if early-career PAs with lower WLB were more likely to leave the profession, while mid-career or late-career PAs may stay despite not being satisfied with WLB—overrepresenting the sample and potentially skewing the results. The lower response rate (57.4%) should also be considered when interpreting the results, as the WLB question was new to the PA Professional Profile. Additionally, Pseudo R² values (Cox & Snell R²=0.054 and Nagelkerke R²=0.078) indicate that the model accounts for only a small proportion of the variance (5.4% to 7.8%). We also did not account for the unmeasured variables, such as social factors, family dynamics/norms, financial status and number of children (not available in the dataset), which have been shown to affect WLB in previous research.⁶² Future studies should incorporate theory-guided research, longitudinal design and models to improve the exploratory nature of this study and address these limitations.

Conclusions and future directions

Our findings revealed that most (71.7%) PAs are satisfied with WLB. Non-primary care specialties and male gender were the strongest factors associated with increased odds of WLB satisfaction among PAs. However, PAs in the mid- and late stages of their careers, those working over 40 hours per week, those seeing more than 40 patients per week, and those with any educational debt were less likely to report satisfaction with WLB. Thus, further research is needed to understand the intrinsic and extrinsic factors associated with WLB among PAs and how WLB compares among physicians and NPs. Additionally, qualitative studies should be conducted to understand the intersectionality of WLB with gender and family norms among PAs.

Data availability statement

Data are available upon reasonable request.