Unintended consequences: a qualitative study exploring the impact of collecting implementation process data with phone interviews on implementation activities

BACKGROUND: Qualitative data are crucial for capturing implementation processes, and thus necessary for understanding implementation trial outcomes. Typical methods for capturing such data include observations, focus groups, and interviews. Yet little consideration has been given to how such methods create interactions between researchers and study participants, which may affect participants’ engagement, and thus implementation activities and study outcomes. In the context of a clinical trial, we assessed whether and how ongoing telephone check-ins to collect data about implementation activities impacted the quality of collected data, and participants’ engagement in study activities. METHODS: Researchers conducted regular phone check-ins with clinic staff serving as implementers in an implementation study. Approximately 1 year into this trial, 19 of these study implementers were queried about the impact of these calls on study engagement and implementation activities. The two researchers who collected implementation process data through phone check-ins with the study implementers were also interviewed about their perceptions of the impact of the check-ins. RESULTS: Study implementers’ assessment of the check-ins’ impact fell into three categories: (1) the check-ins had no effect on implementation activities, (2) the check-ins served as a reminder about study participation (without relating a clear impact on implementation activities), and (3) the check-ins caused changes in implementation activities. The researchers similarly perceived that the phone check-ins served as reminders and encouraged some implementers’ engagement in implementation activities; their ongoing nature also created personal connections with study implementers that may have impacted implementation activities. Among some study implementers, anticipation of the check-in calls also improved their ability to recount implementation activities and positively affected quality of the data collected. CONCLUSION: These results illustrate the potential impact of qualitative data collection on implementation activities during implementation science trials. Mitigating such effects may prove challenging, but acknowledging these consequences—or even embracing them, perhaps by designing data collection methods as implementation strategies—could enhance scientific rigor. This work is presented to stimulate debate about the complexities involved in capturing data on implementation processes using common qualitative data collection methods. TRIAL REGISTRATION: ClinicalTrials.gov, NCT02325531. Registered 15 December 2014.