Pax Good Behavior Game: Robust Prevention Engineering for Population-Level Impact

Introduction: In 2009, the Institute of Medicine issued its report, Prevention of Mental, Emotional, and Behavioral Disorders in Young People (O'Connell, Boat, & Warner, 2009). One of the most potent and scalable prevention strategies discussed for schools and communities was the good behavior game. In 2010, SAMSHA issued a call for applications to fund 20 diverse school districts, and 18 of the funded sites used the PAX Good Behavior Game. PAX GBG was the first commercialized version of the Hopkins research (see http://bit.ly/NREPP) and began in 1999 as an effort to scale-up GBG for population-level benefits in an easy-to-use, simple, robust and reliable format (Domitrovich et al., in press; Embry, 2002).

Method: Our team scaled-up PAX GBG within 14 schools and 247 classrooms (1^st grade thru 7^th grade, involving 6,500 students) in three South Chicago school districts. The Chicago replication involved more than 10 times the number of students than the original Hopkins cohort (Dolan, Kellam, Brown, Werthamer-Larsson, & et al., 1993). Additional replications of PAX GBG implementation and scale-up have included a longitudinal randomized trial within a Canadian province, replications within hundreds of communities in 28 U.S. states, and now multiple replications within other countries. In 2014 alone, some 8,000 teachers received training affecting some 200,000 students.

Results: In the Chicago replication trial, we were able to replicate the proximal results demonstrated in the original Hopkins study. While there are contextual variables that have influenced implementation and sustainability amongst the several thousand PAX GBG implementation sites, the principles and components of “robust prevention engineering” have enabled PAX GBG to work in very diverse settings—despite contextual issues such as school characteristics.

Discussion and Conclusion: The PAX GBG history tells the story about how a simple behavior analysis procedure (Barrish, Saunders, & Wolf, 1969), which had a training protocol and manual (Turkkan, 1988) and powerful short and long-term results, can be scaled up as a population-level prevention strategy. Such a strategy might enable society to protect against the epidemic of mental, emotional, or behavioral disorders in whole communities, states/provinces, and even nations. Our team has been able to transition from prevention efficacy and effectiveness trials to “robust prevention engineering” principles and policies and real-world monitoring for continuous improvement—all in service of population-level, public health benefits. Those “robust” principles can be applied widely to other prevention programs, practices and strategies to mitigate the variation that local context factors can cause on the success of implementation and scale-up.