BS06 PATIENT‐RATED OUTCOME MEASURES WERE MORE SENSITIVE THAN CLINICIAN‐RATED MEASURES AT DISTINGUISHING THE EFFECTS OF SENTINEL NODE BIOPSY AND AXILLARY CLEARANCE IN THE SNAC TRIAL

Abstract

Purpose We sought to determine which outcome measures were most sensitive at detecting the benefits of Sentinel node based management (SNBM) over routine axillary clearance (RAC) in the SNAC trial.Patients and Methods 1088 women with early breast cancer were randomised to either SNBM or RAC. The primary endpoint was the percent increase in arm volume based on clinicians’ measurements of arm circumference at 10cm intervals. Secondary endpoints included patients self‐ratings of arm swelling and other aspects of quality of life, assessed using the SNAC Study Specific Scales (SSSS: 15 questions asking about symptoms, dysfunction and disabilities). We report a comparison of the relative sensitivity of these endpoints in detecting differences between the treatment groups.Results Patients’ ratings on the SSSS were 3.2 times as sensitive as clinicians’ ratings of arm swelling, requiring 68% fewer patients to detect a given treatment effect. The 7 questions asking patients about symptoms were most sensitive. Questions asking about dysfunctions and disabilities were less sensitive.Conclusion Patient‐rated measures were more sensitive in this trial than clinician‐rated measures at distinguishing the effects of SNBM and RAC. Similar trials would require only a third as many patients if the primary endpoint was rated by patients rather than clinicians.