Factors Predicting Entry into Medicaid Funded Wraparound Care

Purpose:  In Massachusetts, a class action lawsuit (Rosie D. v. Romney) identified the need for care coordination services for children with serious emotional disturbances. In response, Wraparound was added to Medicaid funded mental health services as a care coordination mechanism; access to these services relied on natural referral patterns rather than delineating strict utilization criteria.  Yet, little is known about the predictors of a child and family entry into care coordination services in cases of open access.  This study explored whether access to Wraparound services could be predicted by clinical characteristics, case complexity and/or whether access was driven by other factors, including sociodemographics and caregivers’ needs. 

Methods:  A total of 45,991 electronic health records were extracted from a statewide database of children accessing specialty behavioral health services funded by Medicaid for years 2011-2012 (once the services were fully in place). A multivariate regression model was utilized to examine the odds for accessing Wraparound services with a binary dependent variable.  The independent variables were informed by the Anderson (1968) model, including individuals’ clinical characteristics, case complexity, sociodemographic variables and/or caregiver needs. Clinical characteristics and case complexity were derived from the Child and Adolescence Needs and Strengths Assessment. 

Results:  Among Medicaid youth receiving behavioral health treatment in in Massachusetts in years 2011 and 2012, age, (“p”<.005) and speaking English (“p<”0.001) decreased the likelihood of accessing Wraparound in addition to or in place of other services while being male (“p”<0.001) and/or white (“p<”0.001) significantly increased the odds of accessing Wraparound. Geography appeared highly predictive of a child’s odds of receiving Wraparound (“p”<.001).  The clinical picture was a bit more complex.  Higher acuity on three CANS domains (life functioning, risk behaviors and caregiver needs, p<.001)) increased the likelihood of Wraparound while a higher score on the behavioral/emotional domain reduced the likelihood. The CANS complexity score increased the likelihood of access (p<.001), though less so (p<.05) when caregiver needs were in the model, suggesting that an important component of case complexity is caregiver needs.

Implications: Several factors were predictive of entry into Wraparound services.  While case complexity and clinical characteristics were predictive of access to Wraparound (which would suggest meeting the targeted need of care coordination), sociodemographic and caregiver characteristics were also an important part of the model. These findings suggest that Wraparound may be filling gaps in the service delivery system not related to care coordination.  It is possible that Wraparound is meeting a need not only for care coordination, but also for foreign language fluency and geographic availability of services for children with serious emotional disturbance.  These trends can inform the service delivery system in Massachusetts and in other states facing similar mandates and reiterate the need for a person-in-environment perspective when examining trends in service use.