Ietje Perfors

159 the individual elements of the intervention and to optimize the definition and assessments of outcomes. Another limitation was that patients and healthcare providers could not be blinded, due to the nature of the intervention. This may have affected outcomes, e.g., by the previously mentioned consequences of raising expectations among patients. Furthermore, several patients stopped study participation and we found a higher drop-out in the intervention group compared to the control group, which might have caused selection bias. However, patients’ characteristics of drop-outs did not differ. Conclusion The GRIP intervention, which aimed to structure involvement of primary care during and after cancer treatment with curative intent, was well accepted but sub-optimally implemented and adhered to. It slightly increased primary healthcare contacts, did not improve patient satisfaction with care and slightly increased use of the ED. As the high uptake of the intervention suggests that it addresses patients’ needs, future research should focus on optimizing the design and implementation of primary care involvement. This future effort may benefit from an integrated and collaborative approach with patients and healthcare professionals. 6