Electronic supplementary material
The online version of this article (doi:10.1186/s13063-016-1379-0) contains supplementary material, which is available to authorized users.
Process evaluation studies are recommended to improve our understanding of underlying mechanisms related to clinicians, patients, context and intervention delivery that may impact on trial or program results and on their potential transferability to practice. This paper aims to document the translation of a type-2 diabetes (T2D) prevention program into the routine context of several primary care centers, assessing process indicators related to clinician adoption, patient recruitment, exposure to the intervention components and baseline characteristics.
An observational descriptive process evaluation study was conducted of the 2.5-year implementation of the Prevention of Diabetes in Euskadi cluster randomized trial in 14 primary care centers of the Basque Health Service (Osakidetza). The clinical intervention consisted of three components: (1) risk screening, (2) an educational intervention promoting healthy lifestyles, and (3) remote support (follow-up). A passive dissemination strategy of providing training and materials was used to translate the intervention into practice. All non-diabetic patients aged 45 to 70 years who were identified as being at high risk of developing T2D were eligible for study inclusion. The RE-AIM framework guided the process evaluation.
Overall, 31.4 % of family physicians and 57.6 % of nurses participated in the study, while 4170 out of 67,293 (6.2 %) targeted patients who attended the centers during the implementation period were reached through the screening. Around half of the screened patients were identified as being at high risk of developing T2D (FINDRISC score ≥14). The rate of refusal to participate and the proportion of women were higher in the intervention group. Finally, 634 and 454 non-diabetic 45- to 70-year-old patients who were at high risk of T2D were included in the control and intervention group centers (intervention reach = 48 %). Significant variability in most process indicators was observed at center level.
The passive dissemination strategy has produced modest process indicators related to the adoption, reach and implementation of the intervention program, and reduced the possibility of its standardized application in heterogeneous contexts. The resulting different procedures and strategies used by the centers were associated with process outcomes. Context-specific variability and possible confounding will require rigorous procedures for analysis of the intervention effects.
The trial was registered in ClinicalTrials.gov (identifier: NCT01365013). Registered on June 2011.