Examining Resident Burnout Through the Lens of Self-Determination Theory: The Role of General Causality Orientations Open Access

Burnout—characterized by exhaustion and disengagement from job demands outweighing resources¹ —affects more than 50% of residents worldwide.^2,3  Job demands, such as long working hours, heavy workloads, and emotional stress, often exceed available resources like time for rest, supportive supervision, or opportunities for professional growth. This imbalance has serious consequences, including poorer conduct and patient care,^4-6  medical errors and safety incidents,^7,8  and mental health concerns such as depression, suicidal ideation, and substance abuse.^9,10  In response, the Accreditation Council for Graduate Medical Education (ACGME) has mandated that graduate medical education (GME) programs address resident burnout.^11,12  However, current approaches to combatting burnout have largely been ineffective.³  This is partly because they lack grounding in robust theoretical frameworks, limiting the design of targeted, impactful wellness interventions.

General causality orientations (GCOs), described in self-determination theory,¹³  represent a key personal resource that influences how people orient to their environment and regulate their behavior. Individuals vary in the extent to which they take interest and act with autonomous motivation (autonomy orientation); focus on rewards, punishments, or approval (controlled orientation); or perceive environments as uncontrollable and feel overwhelmed or disengaged (impersonal orientation).¹⁴  GCOs can thus affect how people experience and manage job demands and resources. For instance, autonomy-oriented individuals are more likely to stay engaged and motivated, even in demanding contexts.^15,16  In contrast, control- and impersonal-oriented individuals are more vulnerable to burnout when faced with stress or unsupportive environments.^15,17,18  Although these orientations have been shown to impact motivational, emotional, and well-being outcomes,^19-21  their role in resident burnout has not been investigated. This gap limits our understanding of how personal resources, such as GCOs, interact with the demanding GME environment to either exacerbate or mitigate burnout.

Examining how GCOs relate to burnout provides an opportunity to address this persistent and widespread issue. By understanding how individual differences in self-determination relate to burnout, this study aims to inform the development of systemic interventions that create learning/work environments where residents are supported and engaged. Such an approach moves beyond one-size-fits-all wellness strategies, offering a framework for addressing burnout that acknowledges personal resources and workplace demands.

All residents, across 3 Canadian medical schools, were invited to complete an anonymous online survey distributed via their medical program and resident newsletter. The survey (provided as online supplementary data) was sent in October 2023 and was open for 8 weeks with one reminder. It contained a brief demographic section and the Causality Orientations at Work Scale (COWS) and Oldenburg Burnout Inventory (OLBI) instruments. A random prize draw for a $50 Starbucks gift card was offered as an incentive (optional).

The COWS measures the strength of the 3 GCOs—autonomy, control, impersonal—at work.²²  The COWS has shown good reliability in medical education.²³  It contains 11 vignettes—each presenting 3 ways of responding on a scale from 1 (very unlikely) to 7 (very likely) based on each subscale—autonomy, control, and impersonal. Mean scores were computed for each subscale, with higher scores meaning a stronger workplace orientation.

The OLBI is a 16-item scale measuring occupational burnout. It has 2 subscales²⁴ —exhaustion and disengagement—and has been used with residents.²⁵  The OLBI uses a 4-point scale from 1 (strongly agree) to 4 (strongly disagree), where higher scores indicate more burnout.

We computed means, standard deviations, and Cronbach alpha reliability estimates. Variable relations were assessed using Pearson correlations or analysis of variance (ANOVA). As each GCO is said to coexist within an individual (ie, they are not considered mutually exclusive),¹⁴  a multiple linear regression was performed to assess the association between each GCO and burnout, controlling for significant demographic factors. Standardized regression coefficients and 95% confidence intervals (CI) were used.

Approval was obtained by the Research Ethics Board at the University of Saskatchewan (#3245), University of Calgary (#23-0469), and University of Alberta (#23-0469).

In total, 291 of 1200 residents (24.3%) participated in the survey. However, 48 of the 291 (16.5%) were excluded for being incomplete, leaving 243 (20.5% response rate; Table 1). Men and women were relatively under- and overrepresented in this study, based on local institutional data and the published national averages (approximately 45% and 55%, respectively).²⁶  The sample size in this study was considered sufficient, based on commonly used heuristics.²⁷  The Cronbach alphas for the GCO and burnout variables ranged from 0.79 to 0.91 (see Table 2), indicating good reliability. Two participants identified as nonbinary, and one did not indicate their gender. We thus excluded these cases from the ANOVA, due to low group size. Burnout scores did not differ by gender (F (1, 212)=0.01, P=.96) or program (F (8, 205)=1.63, P=.12), but differed by year, (F (4, 211)=3.56, P=.008). Tukey’s post hoc tests showed that first-year residents had the lowest burnout scores (M=2.41, SD=0.42) and differed from third-year residents (M=2.70, SD=0.44) who scored the highest (MD=-0.63; SE=0.20; P=.017).

Next, burnout was regressed onto the 3 GCOs, controlling for year. The overall model was significant (R²=0.315; F (4, 197)=22.66; P<.001). Autonomy was associated with lower burnout (B=-0.24; P<.001; 95% CI, -0.37 to -0.11); and control (B=0.20; P=.003; 95% CI, 0.07 to 0.33) and impersonal (B=0.28; P<.001; 95% CI, 0.13 to 0.42) were associated with higher burnout. Year was not associated with burnout overall (B=0.09; P=.12; 95% CI, -0.02 to 0.16).

In this study, GCOs accounted for 32% of the variance in resident burnout, with autonomy correlating negatively and control and impersonal correlating positively. This pattern aligns with other self-determination theory and education studies^28-31  and suggests that the autonomy GCO buffers burnout, while the control/impersonal GCOs facilitate it.

Autonomy-oriented individuals have an internal perceived locus of causality (I-PLOC: belief that one is the initiator/sustainer of one’s own behavior³² ), which promotes engagement and resilience.³⁰  Conversely, control-oriented individuals have a more external PLOC (E-PLOC: belief that one “must” or “should” do something, due to external/internal pressure). They respond to stress in more reactive and defensive ways,³⁰  undermining well-being.³²  Being impersonally oriented invokes a more inactive PLOC (belief that one cannot control outcomes), which promotes anxiety and helpless ways of coping.³²  While scores for autonomy and control were similar and higher than impersonal (Table 2), all GCOs can be primed and affect behavior and well-being, even if that GCO is, itself, relatively weak.³³  Thus, a lower impersonal score should not be overlooked.

We further observed that burnout did not differ by gender, but did by year, with first years scoring lowest and third years scoring highest. This could be due to increasing demands and prolonged exposure to challenging work environments. The third year—at least in Canada—also represents a stressful transition when residents are working hard while competing for limited spots in their desired subspecialty. Burnout scores not differing by specialty suggests that it is prevalent across all GME programs. These findings align with prior studies,^34-36  reinforcing the urgent need to address burnout in a systematic way.

This study is cross-sectional, correlational, and employs self-report scales. Causal conclusions are thus not possible, and there is potential for response bias. The research also occurred in only 2 Canadian provinces, with a relatively low sample size and response rate, and men were underrepresented relatively to the study population. Finally, it is not uncommon for GCOs to intercorrelate, as in this study: a resident could feel pressured by the environment (controlled) and also powerless to influence desired outcomes (impersonal), or see opportunities for choices and engagement regardless of the situation at hand (autonomy).¹⁴  Nonetheless, both larger-scale and longitudinal studies are suggested to help confirm the generalizability, representativeness, and stability of our findings.

This study found that when residents’ self-determination (autonomy causality orientation) was higher at work, it was associated with lower burnout. Conversely, when residents’ self-determination was lower or missing altogether (control and impersonal causality orientations), it was associated with higher burnout.

The authors would like to thank Professor Richard M. Ryan, PhD, for his contributions to this manuscript.

The online supplementary data contains the survey used in the study.

Funding: The authors report no external funding source for this study.

Conflict of interest: The authors declare they have no competing interests.