Ietje Perfors

157 is in line with earlier reports of the Dutch Cancer patients organisation, 10 which demonstrated a strong wish for more GP involvement. Unfortunately, high uptake was followed by suboptimal implementation to the intended intervention and, in its current form, it did not result in improved satisfaction. Therefore, adjustment of the design and/or implementation of the intervention is required. Several findings provide clues for improving the intervention, including its integration in the daily workflow. First, the majority of the TOCs were scheduled after the treatment decision was already made, which makes active participation in the treatment decision impossible. The time between diagnosis and therapy choice was too short to enable active TOC planning by the patient. Hence, to enable adequate timing of TOCs, its planning should preferably be embedded in the hospital care pathway. Second, almost half of the patients did not want the suggested contacts with the HON after treatment completion (46%). Obviously, the design and content of the HON intervention did not match the needs of almost half of the patients. Possibly, the patients expected their GP to be personally involved, and not to delegate it to the primary care nurse. This needs further exploration. In contrast to our hypothesis, patients in the intervention group were less satisfied with their GP and slightly more with their nurse. This may be explained by several reasons. First, the intervention itself may have raised expectations about GP involvement in the intervention group, that were not met in practice. Patients receiving the intervention were notified that they would receive extra care from the primary care team: both their GP and a HON. They might have expected more contact with their GP, but met the HON instead. The significantly lower scores on “GP-Availability” in the intervention group support this hypothesis. Another possible explanation may be found in the GP involvement in the control group, as a result of an independent proactive approach by the GP. More proactive GP contacts might have led to higher patient satisfaction in the control group. Finally, the lack of difference in satisfaction with caremay be the result of a ceiling effect, which is supported by the high overall satisfaction scores in both study groups. In the Netherlands, patients usually have a nurse as case manager in the hospital, which might contribute to the high satisfaction. 6