INTRODUCTION

Caring for complex patients may be an underreported source of stress for primary care physicians (PCPs). Prior research on reducing physician burnout has largely focused on communication, work-life balance, and self-management techniques.1 Perceived patient complexity has been linked to decreased PCP work satisfaction, but little is known about how specific components of patient complexity such as socioeconomic or personal circumstances, burden of medical conditions, or behavioral health impact PCP stress.2, 3 We examined the association between PCP self-reported stress and the factors that define their most complex patients.

METHODS

PCPs from a large integrated health care system in Northern California completed a survey as part of a quality improvement project related to complex patient care management. Printed surveys were distributed to 156 PCPs from 6 community primary care practices across 3 medical facilities with similar access to case management services. The survey prompted PCPs to consider their 10–15 most complex, challenging patients and asked them to report levels of stress and ease of care coordination. PCPs were also asked to rate the importance of pre-defined patient factors that contribute to making complex patients challenging to manage (Table 1). For our analysis, we dichotomized responses as very high or high importance vs. moderate or low importance and compared PCPs reporting “extreme” stress vs. lower stress levels (moderate, somewhat, or low). We used multivariate logistic regression to control for years’ experience with their employer to adjust for familiarity with practice setting. This study was reviewed by the Research Determination Office of Kaiser Permanente of Northern California and deemed to be quality improvement exempt from full review by a KP Institutional Review Board.

Table 1 PCP Rating of Factors That Contribute to the Challenge of Managing Complex Patients, by Self-reported Physician Stress Level
Full size table

RESULTS

The survey response rate was 87% (135/156). PCP respondents were 65% women; 77% were internal medicine and 23% were family medicine trained. The mean years from medical school graduation was 20.9 (SD 8.2) years. Over half (56%) of physicians had > 10 years’ experience in their current practice setting. Social or financial issues, poor patient adherence, and pain management were among the most commonly cited patient complexity factors (Table 1).

Overall, 80% of PCPs reported their work with complex patients to be either moderately stressful (47%) or extremely stressful (32%). In multivariate analyses, PCPs who reported extreme stress were significantly more likely to rate social or financial issues (aOR 4.43; 1.24–15.28) as the most relevant contributors to defining their complex patients but not other patient factors like number of diagnoses or pain management (Table 1). PCPs reporting extreme stress were also more likely to report greater difficulty with care coordination (2.20; 1.01–4.79).

DISCUSSION

Most PCPs reported that primary care of complex patients was moderately or extremely stressful, and PCPs reporting the greatest stress were more likely to identify the contribution of patients’ social or financial issues rather than medical or behavioral problems as major management challenges.

Even though social adversity and poor health are linked, programs that meet patients’ social and financial needs are uncommon.4 Without such resources, PCPs can become de facto responsible for these issues. PCPs have the training, experience, and resources to treat patients’ medical needs, but management of socioeconomic difficulties may lie outside of physicians’ expertise or workflow.5 Without the necessary resources, physicians can feel ill-equipped to provide the interventions patients truly need, contributing to increased stress.

We used cross-sectional self-report survey data, so our results must be interpreted in the context of the study design. While we were unable to use objective measures of patient complexity, PCP-defined patient complexity is well established in the literature.3 We found that several non-medical patient factors had an association with extreme stress but did not reach statistical significance due to wide confidence intervals. Despite a robust response rate, this was likely due to sample size constraints. Nevertheless, this correlation supports the notion that PCP stress is not driven by medical complexity alone and could be further investigated in a larger study.

Our results suggest that the existing systems of care delivery for socially complex patients contribute to extreme PCP stress and that improving care coordination may be protective. Efforts to reduce PCP burnout should therefore recognize the impact of caring for complex patients with limited social and financial resources on PCP stress by implementing interventions to address patients’ unmet needs.