. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S526 Abstract Recovery From Operation Quality Assessment System: A Novel Technology for the Real-time Assessment of Recovery Following Cardiac Surgery David McCormack * , Damian Balmforth, Philipp Lohrmann, Sammra Ibrahim, Rakesh Uppal, Alex Shipolini, Adam El-Gamel Barts Heart Centre, London, UK Objective: Cardiac surgery in the UK has publicly reported mortality outcomes for >20 years. The audit tool provides detailed pre- and intraoperative information, but only a nar- row range of clinical outcomes are recorded. As clinicians, we created an efficient tool to provide a comprehensive, real-time analysis of postoperative outcome covering following cardiac surgery. Method: Between April 2011 and January 2015, all patients admitted for cardiac surgery were recruited to the Recov- ery from Operation Quality Assessment System (ROQAS). Patient recovery was prospectively recorded for inpatient days. Data were entered during the ward round (requiring an average of 8 seconds per patient). Data were collected on over 30 postoperative complications, functional recovery, and delayed discharge reasons. We present selected data on length of stay and contribution of social factors to delayed discharge. Results: In total, 1939 consecutive patients were recruited generating 25,422 continuous days of patient data. Group 1 (n = 660 [35.2%]) had no delay in discharge or complication. Patients in group 2 (n = 1215 [64.8%]) experienced at least one complication (mean 4.7). Mean lengths of stay for groups 1 and 2 were 5.0 and 11.8 days, respectively. In group 2, dis- charges were delayed an average of 6.4 days of which 5.0 days were for medical reasons and 1.4 days for social reasons. Conclusion: ROQAS provides unique insights into patient recovery following cardiac surgery. Its applications include quality enhancement and fiscal efficiency. It is being utilised to perform clinical outcome studies focusing on morbidity and functional recovery. Predictive artificial intelligence models for postoperative complications and to guide preoperative optimisation are in development. http://dx.doi.org/10.1016/j.hlc.2018.04.068 Sound Pressure Levels in Cardiac Recovery Units: An Invisible Harm David McCormack * , Eyal Ben-David, Gopal Soppa, Philippa Borra, Jonathan Anderson, Adam El-Gamel Hammersmith Hospital, London, UK Background: Excessive noise leads to sleep deprivation and distress. Prolonged exposure to loud noise increases plasma cortisol levels and is associated with hypertension. Nocturnal noise has been demonstrated to cause emotional instability and delirium. The World Health Organization (WHO) guidelines recommend noise levels should not exceed 35 dB during the day and 30 dB during the night in patient areas. Despite significant literature relating to acute wards and general critical care environments, noise levels in cardiac intensive care units (ICUs) have not previously been reported. Methods: The sound pressure levels in a central-London cardiac ICU were measured continuously for 13 day shifts and 12 night shifts. Measurements were taken both at the central nursing station and at patients’ bedsides. The central station is approximately 5 m from the bed spaces. Measurements were taken 1 m above the bed space. Results: Sound levels were found to exceed WHO guide- lines at all times of day and night. Peak noise levels also exceeded WHO guidelines. Specific review of night-time measurements revealed there were no continuous 15-minute periods compliant with the WHO criteria. Conclusions: Such levels of noise have been linked to delayed recovery and increased morbidity. Peak sound levels between 84 dB and 98 dB are in the range of a gas-powered lawnmower. While patient care is the primary focus of this audit, the sound-pressure levels demonstrated can be damaging to staff. Identification of contributing factors and reduction of excessive noise in the cardiac ICU is crucial. http://dx.doi.org/10.1016/j.hlc.2018.04.069 Readmission Following Cardiac Surgery: Does Inpatient Recovery Predict Readmission? David McCormack * , Damian Balmforth, Adam El-Gamel, Sammra Ibrahim, Philipp Lohrmann, Rakesh Uppal, Alex Shipolini Barts Heart Centre, London, UK Background: Readmission rates following cardiac surgery in the literature ranges between 5% and 20%. Medical prob- lems such as congestive cardiac failure and arrhythmias predominate. We sought to establish the readmission rate to our cardiac surgical unit and identify common causes. Method: Patients undergoing cardiac surgery were prospectively enrolled into the Recovery from Operation Quality Assessment System (ROQAS). Readmission rates were calculated for 30 days, 90 days, and 1 year. Univariate analysis was performed and variables with significant asso- ciations for readmission at 1 year were further analysed with logistical regression. Results: During a 42-month period, 1939 patients were recruited. In total, 60 patients died and were excluded from analysis. Readmission rates for 30 days, 90 days, and 1 year were 2%, 3.4%, and 3.9%, respectively. Causes for readmission at 1 year were pleural effusion (31%), surgical site infec- tion (27%), mediastinal collection (11%), sternal wire removal (5.4%), and other (25.6%). On univariate analysis significant predictors of readmission were age, female sex, body mass index, previous myocardial infarction, hypertension, renal dysfunction, pleural effusion, pericardial effusion, and total length of stay. On logistical regression analysis, female sex (odds ratio [OR] 2.18; 95% confidence interval [CI] 1.34–3.55), renal dysfunction (OR 2.86; 95% CI 1.15–7.13), and postopera- tive pericardial effusion (OR 2.49; 95% CI 1.09–5.71) remained independent risk factors for readmission. Conclusion: Readmission rates were less than reported elsewhere. The most common surgical reasons for readmis-