Implementation timelines across conditions: a universal Stages of Implementation Completion^® (SIC) comparison of...

Despite ongoing efforts, achieving consistent success in implementing evidence-based practices remains challenging due to diverse, context-specific implementation needs. This analysis examined the time required to implement practices currently represented in the Universal Stages of Implementation Completion^® (SIC) data repository under research-led conditions (supported by researchers) versus real-world conditions (supported by purveyors or other technical support providers). The SIC measures the implementation process across three well-established phases of implementation: Pre-implementation (Phase 1), Implementation (Phase 2), and Sustainment (Phase 3).

Implementation outcomes were compared between research-led and real-world implementation conditions, representing 547 and 818 sites, respectively, across 43 program practices (i.e., interventions) in the Universal SIC repository. Descriptive analyses were conducted using data aggregated at the practice level to contrast implementation conditions. Intraclass Correlation Coefficients (ICC) from multilevel models (MLM) were calculated to compare outcome variation within and between practices. MLMs examined whether implementation condition was associated with SIC scores (proportion, duration, and Final Stage reached) and key implementation process outcomes (i.e., achieving practice Start-Up and Competency in practice delivery).

In aggregated descriptive analyses, research-led sites demonstrated slightly higher rates of practice Start-Up (48% vs. 42%) and Competency (41% vs. 36%), and faster pace in Phase 2 (mean difference = 322 days) than real-world sites. ICCs from unconditional multilevel models ranged from 0.31 to 0.62 across outcomes, confirming substantial variance attributable to between-practice differences. MLMs found few significant differences in outcomes between conditions. MLMs confirmed that research-led sites showed significantly shorter durations during Phase 2 (461 vs. 740 days, p = 0.02), and they additionally showed lower activity completion proportions in research sites (61% vs. 83%, p = 0.02) compared to real-world sites. MLMs found no significant differences between research-led and real-world sites in achieving program Start-Up or Competency in practice delivery.

While real-world sites had longer Phase 2 durations than research sites, there were no differences in achieving Start-Up or Competency between the two implementation conditions. Low success rates (less than 50% of sites achieving practice Start-Up and Competency in both conditions) highlight shared implementation challenges across contexts and support unified SIC benchmarks to inform future implementations.