in diverse settings (as discussed in Chapter 8) but they are often stymied by policy, institutional, and practice barriers (Dausey et al., 2007). Rather than providing incentives for treatment strategies that recognize comorbidities, the policy environment often sets up barriers. Thus, for the most part, even the limited empirical knowledge about how best to treat adults with comorbid conditions remains largely unexploited (Drake et al., 1998).
One striking example involves the absence of strategies that can respond effectively to depressed parents who are also substance abusers. While it is known that 24 percent of adults with depression have a co-occurring substance use disorder (Kessler et al., 2003), less than 19 percent of individuals with co-occurring depression and substance use disorders receive treatment (Mark, 2003). This is fewer than the 30 percent of individuals with either disorder alone. For those individuals with co-occurring disorders (depression and alcohol addiction) that are treated, treatment is more costly, 68–80 percent higher than treatment for either disorder alone (Mark, 2003). Many parents, particularly low-income parents, experience not just depression, but depression and trauma, substance abuse, domestic violence, and other conditions that impair effective parenting (Cooper et al., 2007; Flynn and Chermack, 2008). Furthermore, evidence suggests that when parents experience multiple risk factors, this is reflected in negative cognitive and behavioral outcomes for children, particularly very young children (Whitaker, Orzol, and Kahn, 2006).
Many model programs tested in controlled trials and found to be effective for parental depression prevention and care involve complex strategies such as integrated services, coordinated programs, and comprehensive care, and strive to align the complexity of an intervention or organization with the complexity of the environment in which it functions (Dietrich et al., 2004; Katon et al., 2004; Wells et al., 2000). This approach, an example of the concept of requisite variety in systems theory, involves more of the variables that can affect a problem, but at the cost of increased difficulties in replicating complex solutions in new settings (Ashby, 1958; Bodenheimer, Wagner, and Grumbach, 2002; Pfeffer and Sutton, 2006). The limited evidence available on the implementation of depression care interventions, either after clinical trials or in nontrial settings, has shown increased variation in performance (Fisher, Goodman, and Chandra, 2008; Pearson et al., 2005; Pincus et al., 2005), which could be owing to many factors. But certainly the multiple working parts of these interventions—that is, their complexity—may be among the relevant influences.
The limited available evidence casts collaboration in depression care in a promising light (Bluthenthal et al., 2006; Wells et al., 2006). But the