This paper describes several additional analyses and results that go beyond the basic impact findings from the evaluation of the Working toward Wellness (WtW) program in Rhode Island. WtW was a one-year telephone care management intervention for depressed parents who were Medicaid recipients. To encourage individuals with depression to seek treatment from mental health professionals, the WtW program randomly assigned depressed Medicaid recipients to a program group, which had access to telephone care management for up to a year, or to a control group, which had access to the usual mental health services available to Medicaid recipients. Results from the study found that telephone care management modestly increased in-person treatment for depression during the year of the intervention but not after that point. No impacts on average depression severity were observed for the sample as a whole.

To understand which individuals showed reduced depression over time, the paper examines the relationship between participants’ characteristics and changes in depression scores from baseline to six months and to eighteen months. The results do not, however, suggest a clear means of targeting services like WtW to those who are least likely to improve on their own. Other than baseline depression severity, few participant characteristics were found to be associated with reduced depression over time. This suggests that most subgroups of participants could have benefited from a more effective intervention.

Also, because only about 40 percent of the study population participated in in-person mental health treatment, the paper examines which factors contributed to receiving treatment and the intensity of that treatment. The results suggest that a number of factors were associated with seeking mental health treatment. In particular, treatment occurred more frequently for those who were more severely depressed, those who were not working at baseline, white sample members, and those who had received treatment for depression prior to random assignment. This may suggest providing more resources and supports to encourage those groups to receive treatment who are least likely to participate, for whom the program might make a larger difference. It may also suggest excluding individuals with prior treatment for depression from future studies of similar interventions.

Lastly, because the eighteen-month results showed that there were significantly fewer program group members in the very severely depressed group, the paper investigates which baseline characteristics are associated with being very severely depressed at follow-up. It was found that although some characteristics are associated with having severe depression at follow-up, the impacts on depression severity for this high-risk subgroup are not statistically significant.