Investigating Psychosocial Predictors & Employment Outcomes of Enhanced IPS Intervention

Overview

The purpose of this cohort study, using an existing database, is to seek to describe the relationship between psychosocial factors (independent variables) and employment outcomes (dependent variables) in a population of people with complex mental health problems receiving evidence-based supported employment (EBSE) augmented with theory-driven occupational therapy interventions. The intention is to understand the predictive nature of psychosocial factors on the positive outcome of success in achieving paid employment. The Worker role interview (WRI) is routinely used in the research site; an National health Service (NHS) occupational therapy led vocational rehabilitation delivering EBSE service for people with complex mental health problems. An existing database of initial assessments and employment outcomes is available for investigation.

The cohort study will be undertaken utilising a secondary dataset from an established vocational rehabilitation service (VRS). This service is experienced in providing intervention based on individual placement and support (IPS) with established high fidelity to the model. The team's interventions are also guided by the model of human occupation, (MOHO) and the Worker Role Interview is routinely completed with all patients accessing the service. The Worker Role Interview is a valid reliable measure of psychosocial items measuring of readiness for employment. The consistency of intervention will be ensured as the service is routinely appraised for fidelity to the IPS by two experts, using a standardised assessment tool the IPS fidelity scale and the team has received ongoing training and regular caseload supervision to integrate evidence-based occupational therapy into practice. The experts appraising the service are independent of both the clinical service and the proposed research. The established VRS provides an assessment and intervention for people with complex mental health problems. A dataset has been routinely collected over several years which is currently used to evaluate the service. To address potential limitations of secondary data analysis the researcher has an established relationship with clinicians in the VRS and was involved in the selection of routine measures now used to evaluate the service. It is, therefore, possible to be confident that a consistent and standard dataset is available for analysis. Ongoing training and consultation have been provided by the researcher to the clinical service to ensure the data quality. Data Collection The data will be managed and transferred by a trained researcher on site at both services.Data will be robustly anonymised prior to transfer. Data will be transferred from NHS secure network to secure network within Queen Margaret University using a secure email service approved by the Department of Health. The dataset comprises; Research identification (ID) code Gender Age on Admission Ethnicity Marital Status Living situation (alone/shared/supported accommodation) Quintile range of Scottish Index of Multiple Deprivation Educational attainment Previous employment ICD10 Worker role interview items Date of assessment WRI Assesses Abilities \& Limitations WRI Expectation of Success in Work WRI Takes Responsibility WRI Commitment to Work WRI Work-Related Goals WRI Enjoys Work WRI Pursues Interests WRI Appraises Work Expectations WRI Influence of Other Roles WRI Work Habits WRI Daily Routines WRI Adapts Routine to Minimize Difficulties WRI Perception of Physical Work Setting WRI Perception of Family and Peers WRI Perception of Boss and/or Company WRI Perception of Co-workers Employment goal Date of final goal review Outcome of goal (successful/ unsuccessful in attaining employment) Sample; Based on the guideline of a minimum of 10 events per predictor variable, a sample size of 200 would allow for the robust estimation of 8 variables in the model. Analysis Plan The collected data will be subject to analysis with the use of multiple logistic regression using the R data analysis software. Multiple logistic regression will allow the construction of models which examine the effects of psychosocial predictors on the probability of a person with complex mental health problems attaining paid employment. Possible confounds which may influence outcomes will be controlled for (e.g. age, gender, educational achievement, history of employment). Interpretation of the analysis will be shared and discussed with clinicians who are expert in their field to ensure that findings are considered from both a statistical significance and clinical significance position.