Specific components of a complex depression care program can affect staff outcomes differently: post-hoc analyses of a stepped-wedge cluster-randomized trial in nursing homes

Abstract

ABSTRACTObjective:The Act in case of Depression program showed effects on the quality of life and depression in nursing home (NH) residents. We aimed to explore the effects of this complex multidisciplinary program on job satisfaction, job demands, and autonomy in nursing home staff.Design:Four data points from a stepped-wedge cluster-randomized trial on patient outcomes were used for secondary analyses on staff outcomes.Setting:Sixteen dementia special care and 17 somatic care units in Dutch NHs.Participants were 717 (90.1%) care staff or trainees, 34 (4.3%) paramedical staff, and 45 (5.7%) other staff members.Intervention describes procedures for nursing staff, activity therapists, psychologists, and physicians. It contains evidence-based pathways for depression assessment, treatment, and monitoring treatment results.Results:Mixed models for intention-to-treat analyses showed no significant changes in job demands, job satisfaction, or autonomy. Models corrected for the ratio of unit residents who received, when indicated, a specific program component revealed reduced job demands and improved job satisfaction and autonomy when treatment procedures were used. A better use of assessment procedures was associated with increased job demands, while conducting monitoring procedures was associated with increased job demands and decreased autonomy.Conclusions:Components of complex care programs may affect the staff outcomes in opposite directions and, taken together, produce a zero-sum or a statistically insignificant effect. While implementing treatment protocols affecting patients directly can also improve job outcomes such as satisfaction and autonomy and decrease job demands, it is possible that other procedures of complex programs may have unfavorable effects on job outcomes. It is important to account for specific components of complex interventions when evaluating intervention effects.