To follow up or not? A new model of supportive care for early breast cancer.

9098 Background: Routine follow-up after curative treatment for early breast cancer exists to monitor for local recurrence and provide support for patients. Hospital visits can be stressful for patients and evidence indicates most recurrences are first identified by the patient. The value of resource-intense clinical follow-up is constantly being questioned. Many believe that time spent seeing essentially well-women is not clinically beneficial or an efficient use of time. METHODS This pilot study tested the feasibility and acceptability of a new supportive model of follow-up using quality-of-life (QOL) questionnaires plus qualitative diary evaluations. All patients attended four half day patient education workshops with course evaluations, followed by randomised to open access (OA) or hospital follow up (HFU). QOL including Hospital Anxiety and Depression Score (HADS), EORTC QLQ-C30 and BR23 were performed at baseline and 6 months, with further results awaited for 12, 18 and 24 months. RESULTS 106 women were recruited to the pilot study. 53 were randomised to HFU and 53 to OA. Multivariate analyses of covariance (MANCOVA) tests were conducted on all QOL data. Age was included as a covariate. Follow-up ANCOVAs on individual function and symptom outcomes were also conducted. The MANCOVA analyses indicated no statistically significant differences in change scores between HFU and OA groups, or between patients of different ages, on any of the three questionnaires. In all cases the effect of group had a greater effect on change (baseline-6 months) scores than the effect of age. Univariate ANCOVA tests and descriptive analyses showed performance improvement in many of the individual function and symptom scales in the OA group. CONCLUSIONS While open access patients showed greater mean improvements in more scales than patients seen in clinic, follow-up method does not appear to significantly affect overall outcomes on any of the three questionnaires. As an influencing factor, the patients' age is less significant than the assigned group. This new model is feasible and acceptable. It is being adopted as standard across the Yorkshire Cancer Network.