Late Breaking Abstract - Predicting time to next pleural procedure in patients with malignant pleural effusion following therapeutic aspiration

<bold>Background:</bold> There is variation in time to next pleural procedure in patients with malignant pleural effusion (MPE) having therapeutic aspiration (TA). <bold>Aim:</bold> To identify baseline variables associated with time to next procedure in patients with MPE undergoing TA. <bold>Method:</bold> A prospective cohort study recruiting in 10 UK hospitals. Patients with known or suspected MPE attending for TA were eligible. Assessments included chest X-ray, thoracic ultrasound, observations, blood tests, visual analogue score for dyspnoea (VASD) and EQ-5D-5L. Patients then underwent TA. They were followed up at 7 days, 30 days and 3 months with assessment of whether they had had further therapeutic procedures, EQ-5D-5L and survival. Multiple regression was performed to identify baseline variables which predicted time to next therapeutic procedure. <bold>Results:</bold> 241 patients were recruited. Of these, 211 had a final diagnosis of MPE and were included in analysis. Recruited patients were divided into a derivation cohort of 180 patients and a validation cohort of 31 patients. Logistic regression was performed comparing: 1. Patients who had their next therapeutic procedure within 1-13 days versus those who had a procedure 14 days or after or no procedure; 2. Patients who had their next therapeutic procedure within 6 weeks versus those who had a procedure after 6 weeks or no procedure. Independent predictive variables for repeat procedure were baseline 1) respiratory rate, 2) effusion depth on ultrasound and 3) VASD. ROC analysis showed the area under the curve was 0.73 for comparison 1 and 0.75 for comparison 2. <bold>Discussion:</bold> This study successfully identified variables predicting time to next procedure.