Table 3 Rationale for use of specific TDABC approaches in the case example pilot trials
Domain of implementation cost assessed [7] | Actual approach used to capture costs | Rationale for this approach | Alternative cost capture approaches (problems with this alternative) |
|---|---|---|---|
Case example 1: type 2 hybrid trial of physical activity coaching in primary care [26] | |||
Training (coach time) | Self-report: retrospective time diary | The coaches’ employer received time diary data to track the time spent on this project, in order to allocate a proper portion of their salary to the research grant | Self-report: uniform estimates of time for tasks (insufficient detail for the coaches’ employer’s goal to charge for the hours that each coach spent on the project) |
Delivering intervention coaching content (coach time) | Semi-automated: contemporaneous time diary embedded within the electronic health record (EHR, Epic Systems) | The PI captured the time each coach spent delivering the program, in order to compare the hours of coach time by certain patient demographic characteristics (e.g., age, gender, insurance type) | Automated: EHR estimate of time spent during this type of research encounter (grant funds were insufficient to pay the health informatics team to create a specific research encounter type for this project) |
“Hidden” costs of preparing for the program—including technical assistance with transmitting data between patients and clinics (research assistant time) | None | At the outset of the grant, it was not clear that the research assistant would need to support patients and serve in the role of a “technician” to support data-sharing between patients and clinics | Automated: EHR estimates of time spent during this type of research encounter (as noted above, grant funds were insufficient for this; in addition, it would not fit well with the research assistants’ work flow to document this technical assistance in study-specific documents rather than the EHR) |
Case example 2: Population-based cross-sectional study paired with observation of anticoagulation clinic processes [23] | |||
Delivering intervention (time for pharmacist, registered nurse, and clerical staff | Combination of automated and observation methods Observation: workflow process map Automated: a proprietary internal database (Workload Management Reporting system) estimated the time spent in each step of the process map | The proprietary database provided an accurate way to assess costs at each step of the process map | Observation alone: The authors also used direct observation to validate a subset of the automated cost capture timings derived from the system (sole use of observation methods would have resulted in far fewer data points than the combination of observation and automated approaches) |