Passing the torch: Evaluating exportability of a violence intervention program

Background A violence intervention program (VIP) developed at our trauma center resulted in a reduction of injury recidivism to 4% from a historical rate of 16%. Our aim was to investigate the feasibility of exporting our program to another trauma center by examining rates of and identifying potential barriers to recruitment, enrollment, and impact. We hypothesized that our VIP is feasible at another trauma center and successfully meets needs associated with risk reduction. Methods In January 2010, we introduced our VIP to another trauma center. To assess exportability of our program, we used a standard model of program evaluation for VIPs promoted by the Centers for Disease Control and Prevention. Specifically, the process and impact portions of the model evaluation were performed in this comparative analysis over a 1-year period. Recruitment, enrollment (process), and success at meeting risk reduction needs (impact) were our outcomes. This included patient and case manager characteristics in addition to rates at which eligible patients were approached and enrolled. These variables were compared using the Wilcoxon rank-sum and chi-square tests. Results During the study period, 155 patients were eligible for inclusion at the exported program compared with 119 at the original VIP. Rates at which eligible patients were approached at the exported program were significantly lower than the original program (44% vs 92%, P =.04). Rates at which approached patients were enrolled were also significantly lower (21% vs 55%, P =.002). The difference was associated with the time of injury and hospital length of stay because 40% of eligible patients were missed if injury occurred during a weekend and 70% were missed if the length of stay was less than or equal to 48 hours at the exported program. A cultural match between the client and case manager was assessed by race/ethnicity and language spoken; 2 of the 3 case managers at our site are Latino and bilingual and the other is black, whereas the 1 case manager at the exported program is black and monolingual. Cultural match was 91% versus 47%, respectively (P <.05). Impact: Both programs met more than 50% of identified client needs in several categories. Conclusions Program exportation is based on the replication of both the program model and the program infrastructure. The data in our study support success of the program model (case management process) at our export site, but the actual program infrastructure was not successfully exported to this hospital.