Study design: Longitudinal cohort design.

Objectives: First, to explore the longitudinal outcomes for people who received early intervention vocational rehabilitation (EIVR); second, to examine the nature and extent of relationships between contextual factors and employment outcomes over time.

Setting: Both inpatient and community-based clients of a Spinal Community Integration Service (SCIS).

Methods: People of workforce age undergoing inpatient rehabilitation for traumatic spinal cord injury were invited to participate in EIVR as part of SCIS. Data were collected at the following three time points: Discharge and at 1 year and 2+ years post discharge. Measures included the spinal cord independence measure, hospital anxiety and depression scale, impact on participation and autonomy scale, numerical pain-rating scale and personal wellbeing index. A range of chi square, correlation and regression tests were undertaken to look for relationships between employment outcomes and demographic, emotional and physical characteristics.

Results: Ninety-seven participants were recruited and 60 were available at the final time point where 33% (95% confidence interval (CI): 24-42%) had achieved an employment outcome. Greater social participation was strongly correlated with wellbeing (Ï =0.692), and reduced anxiety (Ï =â '0.522), depression (Ï =â '0.643) and pain (Ï =â '0.427) at the final time point. In a generalised linear mixed effect model, education status, relationship status and subjective wellbeing increased significantly the odds of being employed at the final time point. Tertiary education prior to injury was associated with eight times increased odds of being in employment at the final time point; being in a relationship at the time of injury was associated with increased odds of being in employment of more than 3.5; subjective wellbeing, while being the least powerful predictor was still associated with increased odds (1.8 times) of being employed at the final time point.

Conclusions: EIVR shows promise in delivering similar return-to-work rates as those traditionally reported, but sooner. The dynamics around relationships, subjective wellbeing, social participation and employment outcomes require further exploration.