%0 Journal Article %A J Davies %A K Hartshorn %A L Honeyman %A S Kolhe %A A Phillips %T 453 External validation of the Norwich DEFAB score for up-front prediction of endometrial cancer %D 2021 %R 10.1136/ijgc-2021-ESGO.149 %J International Journal of Gynecologic Cancer %P A96-A96 %V 31 %N Suppl 3 %X Introduction/Background*The DEFAB score combines clinical data and patient characteristics such as Age, endometrial thickness, diabetes, BMI and bleeding characteristics to calculate the risk of endometrial cancer causing PMB. It was based on 3047 patients cared for between 2006 and 2009. Inclusion was all women presenting with PMB. The original authors proposed a cut off score of 3 – whereby a score of ≤2 is managed based on ultrasound assessment of endometrial thickness; whereas a score of ≥3 was additionally investigated with endometrial biopsy regardless of any ultrasound findings, and proceeded to hysteroscopy if this biopsy is negative.Methodology We compared the baseline characteristics of our cohort to the original internal validation study and retrospectively applied the DEFAB score to our patient group to determine how it performed.Result(s)*Mean age was 61 (50-95) years, BMI was 30.7 (25-36) and 30.2 (25-36) in cancer and non-cancer patients. Diabetes was present in 20 (20.6%) and 93 (9.7%) cancer and non-cancer patients. The bleeding pattern was recurrent in 39 (40%), and 309 (32.2%) in cancer and non-cancer patients.42 cancer patients had a DEFAB score of ≤2 , of whom 41 met the endometrial thickness criteria and went on to have a cancer diagnosed. There was one cancer found in a patient who the DEFAB protocol would have discharged without diagnosis, a false negative rate of 2.4%.The incidence of cancer in patients with a DEFAB score of >3 (n=457), was 58 (12.7%.). The sensitivity of the DEFAB & ET thresholding protocol was 98.7% if the cut-off DEFAB score of 3 is used, with a specificity of 60.1%.Conclusion*The DEFAB score performs well at the proposed cutoff in our cohort. %U https://ijgc.bmj.com/content/ijgc/31/Suppl_3/A96.2.full.pdf