Organizational Correlates of Staff Turnover Perceptions in Substance Abuse Treatment Programs

Objective 

An estimated 1 in 3 clinicians and 1 in 5 supervisors turnover in substance abuse treatment (SAT) organizations each year, leading to costs associated with hiring and training new staff, increased caseloads, reduced organizational functioning, and disrupted client care (Garner & Hunter, 2014).  Previous research focuses on individual level predictors with few studies explaining variation in turnover among SAT organizations.  The present study treats turnover as a collective, organizational level attribute and explores four organizational level predictors of turnover: organizational functioning, directorial leadership, attitudes toward evidence-based practice (EBP), and severity of client population served. 

Method

The study sample consists of 88 supervisors of SAT programs in Los Angeles County and 12,392 clients treated by these programs.  To measure organizational turnover, supervisors rated whether or not frequent turnover is a problem in their agencies on a five-point Likert scale.  We measured organizational functioning, attitudes toward EBP, and directorial leadership using well-established scales, and measured client severity at the program level by creating a composite score based on client reports of ‘harder’ drug-of-choice, homelessness, mental illness, and criminal justice involvement.  Based on the trimodal distribution of the organizational turnover variable, we divided programs into three turnover categories: 1) problematic; 2) neutral; and 3) not problematic. We used ANOVAs with Tukey’s tests to determine significant differences in organizational variables between the three program categories.  Next, we used ordinal logistic regression to examine a model with the four organizational variables as predictors of turnover.

Results

There were significant differences between turnover categories with regard to organizational functioning (F[ 2, 85] = 8.50, p < .001).  Programs with problematic turnover had lower organizational functioning than programs with non-problematic turnover (p < .05). The climate subscale was the only significant subscale of the measure of organizational functioning; programs with problematic turnover had more negative climates that programs with non-problematic turnover (p< .05).  Findings from logistic regression showed that programs with problematic turnover had significantly: 1) lower organizational functioning (OR = 0.63; CI = 0.48-0.82); 2) better supervisor attitudes toward EBP (OR = 1.13; CI = 1.02-1.25); and 3) less severe client populations (OR = 0.36; CI = 0.14-0.92) than programs with neutral or non-problematic turnover.   

Discussion

Results provide preliminary evidence for understanding turnover as an organizational level phenomenon.  Organizational functioning, particularly organizational climate, emerged as a robust predictor of turnover with implications for managers seeking to improve retention.  This study also had several surprising findings that warrant further investigation.  Problematic turnover was associated with better supervisor attitudes toward EBP, perhaps because staff in these programs experience more stress as a result of pressure to use EBPs.  In addition, problematic turnover was associated with less severe client populations, possibly because programs with less severe clients do not have access to as many funding sources.  These findings can inform future research on turnover and guide SAT management practices. 

Garner, B.R. & Hunter, B.D. (2014). Predictors of staff turnover and turnover intentions within addiction treatment settings: Change over time matters. Substance Abuse: Research and Treatment, 8, 63-71.