Factors That Affect Quality of Care Among Mental Health Providers

Background and Purpose: High quality, person-centered care is a priority for mental health services. Stressors (e.g. intensive workload) could negatively impact the quality of care, while resources (e.g., organizational support) could have positive impacts. However, research on quality of care and predictive factors is still scarce in mental health. The current study examined factors which affect quality of care among mental health providers focusing on stressors and resources.      

Methods: We conducted a secondary data analysis from an organizational change study of 145 direct care providers at a community mental health center. They were mostly female (80%) and white (67%). The average age was 37.8 (±14) years old; the average work experience within the agency was 3.6 (±7) years; and 63% held less than graduate degree. Structural equation modeling was used to test a hierarchical regression model – examining the impact of providers’ stressors and resources on their perceived quality of care. Quality of Care consisted of two subscales; discordant care (e.g., delayed response to client requests, missed appointments, making mistakes) and person-centered care (Luther et al, 2018). Stressors included working with a high-risk population (assertive community treatment vs outpatient), the number of clients seen weekly, and emotional exhaustion, a subscale of Maslach Burnout Inventory (Maslach et al, 1996). Resources included perceived organizational support (Rhoades & Eisenberger, 2002), support for work-family balance (Allen, 2001), work autonomy (Spector & Fox, 1998), and role clarity (Lyons, 1971). Mplus ver. 8 was used for the analysis.

Results: Providers who were less burned out [β=-.3] and older [β=.3] reported increased person-centered care [R²=.17]. After adding resources into the model, increased role clarity [β=.3], organizational support [β=.2], and more client cases [β=.2] were significant predictors for improved person-centered care; however burnout and age were no longer significant [R²=.25]. For discordant care, increased burnout [β=-.4], working with a high-risk population [β=-.2], and more client cases [β=-.2] negatively affected discordant care [R²=.25]. After adding resources into the model, increased burnout [β=-.2], more client cases [β=-.2], and working with a high-risk population [β=-.2] still remained as significant factors for increased discordance in care, but the positive relationship with more role clarity [β=.3] was also found [R²=.33]. The resource factors significantly improved the prediction in both models (p<.001).

Conclusions and implications: Stressors including burnout in the form of emotional exhaustion, more client cases, and a high-risk population may lead to discordance in care. As the strongest predictor in both models, improved role clarity may be particularly important for improving quality of care in the context of high stress. Interestingly, more client cases were associated with increased perception of person-centered care. Because providers have multiple tasks (e.g., direct care, paper work, administration), increased opportunity for direct client interactions might be needed to improve perceived person-centered care. Yet, this should be interpreted with caution. Longitudinal mediation designs may be required to more rigorously evaluate the variable associations. However, our study clearly indicates the need to increase resources available to providers, particularly role clarity. The implications in social work practices will be discussed.