Psychometric properties of the St. Elizabeth Youngstown hospital wellbeing inventory and non-burnout inventory for physicians and nurses

Abstract

Background: Physicians and nurses have substantial problems with wellbeing and burnout. We examined the reliability and construct validity of a wellbeing inventory (WBI) administered to some physicians and nurses working in St. Elizabeth Youngstown Hospital (SEYH). Methods: The SEYH-WBI, consisting of 4 positive affect (PA) items and 7 negative affect (NA) items developed from 5 validated surveys, was administered (n = 419). A non-burnout inventory (SEYH-NBI) consisting of 2 PA items and 3 NA items was derived from the SEYH-WBI. The Positive and Negative Affect Schedule (PANAS), a validated survey consisting of 10 PA items and 10 NA items, was conducted (n = 191). The Maslach Burnout Inventory (MBI), a validated survey consisting of 3 domains (3 items each), was completed (n = 150). Results: For the SEYH-WBI, Cronbach coefficients were 0.76 for PA items and 0.83 for NA items. The NA item loading on factor 1 was 0.55-0.84 and the PA item loading on factor 2 was 0.47-0.89. Confirmatory indices were as follows: root mean square residual, 0.07 and Bentler Comparative Fit Index, 0.92. For the SEYH-NBI, Cronbach coefficients were 0.76 for PA items and 0.79 for NA items. The NA item loading on factor 1 was 0.80-0.87 and the PA item loading on factor 2 was 0.89-0.90. Confirmatory indices were as follows: root mean square residual, 0.02; and Bentler Comparative Fit Index, 0.99. PANAS correlations were as follows: SEYH-WBI PA and PANAS PA scores, r = 0.9; p < 0.0001; SEYH-WBI NA and PANAS NA scores, r = 0.9; p < 0.0001; SEYH-NBI PA and PANAS PA scores, r = 0.8; p < 0.0001; and SEYH-NBI NA and PANAS NA scores, r = 0.7; p < 0.0001. Correlations for SEYH-NBI and MBI were as follows: total NBI and total MBI, r = - 0.6, p < 0.0001; NA and emotional exhaustion, r = 0.6, p < 0.0001; PA and personal accomplishment, r = 0.3, p = 0.0003; and NA and depersonalization, r = 0.3, p = 0.0008. Conclusions: Validation assessments indicate that the SEYH-WBI and SEYH-NBI have acceptable psychometric performance. Similar findings in a larger cohort would be more compelling.