Factorial Validity of the Maslach Burnout Inventory in Peer Support Providers

Background:

In the mental health field, there are persistent assumptions regarding peer support providers’ susceptibility to job burnout, however, this phenomenon has not been sufficiently examined and current conceptualization of work-related stress is unclear.  As part of an in-depth analysis of stress in peer support providers nationwide, this study examined the factor structure of the Maslach Burnout Inventory-Human Services Survey (MBI-HSS).   The MBI-HSS is a widely-used 22-item measure of occupational burnout that is specifically adapted to helping professionals, and assesses according to three hypothesized domains: emotional exhaustion (EE), personal accomplishment (PA), and depersonalization (DP).

Method:

In a cross-sectional study design, the MBI-HSS was administered through an online survey platform.  The study was promoted to national networks, mental health organizations, and social media.  Subjects were included if they were trained to use their own recovery experiences to assist others with mental health challenges, and provided direct services (paid or unpaid) in a mental health organization.

The sample was first assessed on their performance on the MBI-HSS as presented in the 22-item form with three proposed subscales.  Then, a probable factor structure was identified with principal components analysis, using a polychoric correlation matrix to account for ordinal response data. Eigenvalues were assessed to determine which factors should be retained, and these were rotated using the Varimax procedure.  Item loadings were evaluated, and those loading at less than .4 were dropped from the model.  

Results:

Responses were collected from 46 states over six months.  Preliminary analyses in this sample (N=927) reveal a mean age of 48.4, with 64.4% identifying as female.  The MBI-HSS was completed by 726 participants.  According to Maslach’s proposed scoring of subscales, most respondents (72.1%) reported a “high” level of PA; 79.8% reported “low” levels of DP.  Performance on the EE subscale was mixed, with over half of respondents reporting a “moderate” or “high” level. 

Initial exploratory analyses revealed three likely components, however, items did not load as expected onto the three subscales.   One item, “I feel service recipients blame me for their problems,” was expected to load on the DP subscale but failed to load on any.  Another item, “I feel very energetic,” fell short of the cutoff for the PA subscale.  Two additional other items loaded as expected on the DP subscale, but crossloaded onto the EE subscale.   

Implications:

Due to the highly personal nature of their work, peer support specialists may have a unique experience of job burnout that warrants additional examination of the construct.  Broadening understanding of work stress has implications for training programs and worker retention. These preliminary results suggest that in the assessment of burnout, a reduced item form of the MBI may be more appropriate to use with peer support providers.  Further, there is precedence for exclusion of the ill-fitting items discovered here.  This analysis is part of a larger investigation on work-related stress in this population, and alternative explanations of stress and burnout will be discussed.