Tracking implementation strategies: a description of a practical approach and early findings

Published descriptions of implementation strategies often lack precision and consistency, limiting replicability and slowing accumulation of knowledge. Recent publication guidelines for implementation strategies call for improved description of the activities, dose, rationale and expected outcome(s) of strategies. However, capturing implementation strategies with this level of detail can be challenging, as responsibility for implementation is often diffuse and strategies may be flexibly applied as barriers and challenges emerge. We describe and demonstrate the development and application of a practical approach to identifying implementation strategies used in research and practice that could be used to guide their description and specification.

An approach to tracking implementation strategies using activity logs completed by project personnel was developed to facilitate identification of discrete strategies. This approach was piloted in the context of a multi-component project to improve children's access to behavioural health services in a county-based child welfare agency. Key project personnel completed monthly activity logs that gathered data on strategies used over 17 months. Logs collected information about implementation activities, intent, duration and individuals involved. Using a consensus approach, two sets of coders categorised each activity based upon Powell et al.'s (Med Care Res Rev 69:123-57, 2012) taxonomy of implementation strategies.

Participants reported on 473 activities, which represent 45 unique strategies. Initial implementation was characterised by planning strategies followed by educational strategies. After project launch, quality management strategies predominated, suggesting a progression of implementation over time. Together, these strategies accounted for 1594 person-hours, many of which were reported by the leadership team that was responsible for project design, implementation and oversight.

This approach allows for identifying discrete implementation strategies used over time, estimating dose, describing temporal ordering of implementation strategies, and pinpointing the major implementation actors. This detail could facilitate clear reporting of a full range of implementation strategies, including those that may be less observable. This approach could lead to a more nuanced understanding of what it takes to implement different innovations, the types of strategies that are most useful during specific phases of implementation, and how implementation strategies need to be adaptively applied throughout the course of a given initiative.