Physician, practice, and patient characteristics related to primary care physician physical and mental health: Results from the physician worklife study

Health Services Research, Feb, 2002 by Eric S. Williams, Thomas R. Konrad, Mark Linzer, Julia McMurray, Donald E. Pathman, Martha Gerrity, Mark D. Schwartz, William E. Scheckler, Jeff Douglas

Twenty years ago it would have been unusual to have a session on physician job satisfaction at a conference on health care research. Everyone knew that doctors had good jobs, even if they were difficult and challenging. The recent past has seen numerous and well-documented changes (Scott 1993). Financial, technological, and delivery system changes have been important, but more directly affecting physician job satisfaction and stress levels are changes in the actual organization of the medical workplace. Many observers have suggested that the autonomy of physicians is being constrained (Navarro 1988), as purchasers, employers (McKinlay and Stoeckle 1988), and consumers (Haug 1988) exercise countervailing power (Light 1993).

The reactions to these changes have been documented in various sources. Newspapers chronicle the woes of a medical career (Hall 1995), linking surging disability claims to job dissatisfaction (Hilzenrath 1998). Similarly, research journals have linked poor physician satisfaction to higher rates of patient noncompliance (DiMatteo, Sherbourne, Hays, et al. 1993) and patient dissatisfaction (Linn et al. 1985) and go further to suggest that dissatisfied physicians may have riskier prescribing profiles (Melville 1980). Associated with this decrease in satisfaction is a corresponding increase in perceived levels of stress, which may lead to such outcomes as burnout, mental health problems, or even suicide (Arnetz, Horte, Hedberg, et al. 1987). More ominous is the linkage of stress with disruption of work performance, including absenteeism, turnover, decline in job performance, accidents and errors, and alcohol and drug use (Kahn and Byosiere 1992). Taken together, these findings suggest that distress and dissatisf action have significant costs not only to the individual physician, but also to the patient and health care organization. This is even more important as increasing numbers of physicians practice in organized settings.

In looking for insight on how to address these issues, we must draw on the physician job satisfaction and job stress literature. However, the literature is subject to two limitations. The first is that they are chiefly devoted to description and prescription. Many recommendations are made, often based on common sense or intuition, which are not supported by empirical findings. The second limitation lies in the observation that most of the empirical work has focused on the causes of job satisfaction and job stress rather than examining their impact on physicians, patients, and health care organizations. Our purpose in this study is to begin to redress some of these limitations and to provide "good science" upon which recommendations can be made to physicians, managers, and policy makers. The following section features a conceptual model of physician satisfaction and stress that explores both their causes as well as their consequences.

CONCEPTUAL MODEL

Theoretical Basis

Our model (Figure 1) draws on the theoretical models presented by Lazarus and Folkman (1984) and Ivancevich and Matteson (1980). The strength of the Lazarus model lies in the specification of the cognitive processing that attends stressful events. Specifically, stress is defined as a troubled relationship between the person and environment in which environmental demands tax or exceed a person's resources. The cognitive processing that attends the stressful experiences tries to address questions such as "Am I in trouble?" and "What can I do about this situation?" Subsequent to processing, Lazarus theorizes that some immediate effects like physiological (hormonal or blood pressure changes) or emotional (positive or negative feelings) reactions occur. In the longer term, chronic stress can lead to somatic complaints or illness, lower morale, and impaired social functioning. Ivancevich and Matteson's model features four stages: antecedents (stressors), stress, outcomes, and consequences. Both intraorganizational and extraorganizational factors are considered as antecedents. Furthermore, intraorganizational antecedents are offered at the individual, group, and organizational levels. These antecedents produce job, career, or life stress, which results in several outcomes, including physiological (serum cholesterol and blood pressure), behavioral (satisfaction, performance, absenteeism, and turnover), and, in the longer term, diseases of adaptation (coronary heart disease, anxiety, and depression). The outcomes and consequences of Ivancevich and Matteson map cleanly on the immediate and long-term effects of Lazarus.

Conceptual Model Elements

Our conceptual model (Figure 1) draws specifically on Ivancevich and Matteson by specifying three specific sets of characteristics (physician, practice, and patient) that may act as stressors. Drawing on Lazarus, we theorize that these stressors are subject to some cognitive processing, which results in an appraisal of stress we term perceived stress. Additionally, both of these antecedents and stress affect physician job satisfaction, and together, perceived stress and job satisfaction influence physician perceptions of physical and mental health, Job satisfaction here is the immediate effect of Lazarus or the outcome of Ivancevich and Matteson, and mental and physical health are the longer term effects of Lazarus or the consequences of Ivancevich and Matteson. The remainder of this section will detail three sets of characteristics that are theorized to effect both stress and satisfaction.