Compensatory cognitive training for people with severe mental illnesses in supported employment: A randomized controlled trial (Twamley et al., 2019)

This study was a randomized controlled trial in which individuals were randomized to the intervention group who received CCT in the context of supported employment (77 participants) or to the comparison group who received Enhanced Supported Employment only (76 participants). To be eligible for the study, individuals had to be adult outpatients with severe mental illness who had been unemployed for at least one month but had a goal of working, who were literate and fluent in English, and who did not have dementia or an intellectual disability.  

On average, participants were 44 years old. Forty-three percent of participants were female, and 38 percent identified as being from a racial/ethnic minority. Participants had experienced mental illness for an average of 24 years and were recruited via clinician referrals from the University of California, San Diego.  

CCT was delivered within the context of an employment intervention, Individual Placement and Support (IPS). IPS focuses on rapid, individualized help searching for competitive employment paired with follow-along mental health and employment services for an unlimited amount of time. In this case, supported employment services were available to both groups as needed for two years (the duration of the study). A different master’s-level employment specialist provided services to the comparison group. During the first 12 weeks of the study when the intervention group was receiving CCT, the comparison group was assigned to receive one extra supported employment session per week.  

Employment and earnings outcomes were collected weekly by the employment specialist or by a study rater. Other outcomes come from assessments conducted at baseline and quarterly for the first year; the symptom severity outcomes were also collected at 18- and 24-month follow-ups. Competitive work was defined as work that paid at least minimum wage and was not set aside for individuals with disabilities. Individuals who dropped out of the study were coded as not working for the duration of the study. Two standardized assessments were used to measure the severity of positive and negative psychosis symptoms (PANSS measures) and of depressive symptoms (HAM-D).  The MATRICS Consensus Cognitive Battery, along with supplemental standardized assessments, was used for the cognitive functioning measures; using these assessments, the authors created composite scores in the domains of processing speed, working memory, learning/memory, and executive functioning. The Continuous Performance Test - Identical Pairs (CPT-IP) was used to measure sustained attention and the Memory for Intentions Screening Test (MIST) was used to measure prospective memory.  

For the employment and earnings outcomes, the authors used t-tests to compare the two groups. For the remaining outcomes, the authors used analyses of covariance (ANCOVA) to compare changes from baseline to three months for the two groups and used hierarchical linear models to examine changes over time from baseline to 12- or 24-months depending on the outcomes. These analytic models controlled for diagnosis, age, education, gender, minority status, and the interaction of intervention group by diagnosis.