Treatment Costs of Kidney Injury in Patients Having Heart Surgery in the UK

BACKGROUND Cardiac surgery associated acute kidney injury (CSA-AKI) is reported in 25-40% of patients. This variability is due to a spectrum of definitions used to classify AKI. AKI definition and staging are based on the Risk, Injury, Failure, Loss, End-stage kidney disease (RIFLE) criteria and the Acute Kidney Injury Network (AKIN) criteria, which have been further refined in the Kidney Disease: Improving Global Outcomes (KDIGO) AKI working group guidelines. It is now customary to use the KDIGO definition or its modified version - the AKIN consensus criteria.

The development of AKI in the post-op cardiac surgery setting is associated with significant morbidity, mortality, prolonged intensive care (ICU) \& in-hospital stays, increased treatment costs and poorer long-term survival and quality of life. Despite the ability to identify patients at high risk preoperatively, no effective strategy has yet been established to prevent CSA-AKI.

RATIONALE FOR CURRENT STUDY Acute kidney injury (AKI) is associated with substantial increases in healthcare costs among patients undergoing cardiac surgery. One study estimated that AKI increases costs by $19,212 (£14,437) per patient (non-UK data). In a more recent US study published in 2018, the additional cost attributable to postoperative AKI was estimated at $38,358 (£28,824). This was accompanied by a nearly 10-fold increase in mortality (13.9% vs 1.3%) and a doubling of the mean in-hospital length of stay (18.2 vs 8.6 days) compared with patients without AKI. Similarly, ICU length of stay is significantly prolonged in patients who develop AKI after coronary artery bypass grafting (CABG), with a mean of 5.5 days versus 2.2 days in those without AKI.

STUDY OBJECTIVES Primary Objective: To compare the costs of treatment for patients who develop different stages of cardiac surgery-associated acute kidney injury (CSA-AKI), with the costs for patients who do not develop CSA-AKI, in a UK cardiac surgery population.

Secondary objectives: To determine the incidence of cardiac surgery-associated acute kidney injury (CSA-AKI) using the modified KDIGO criteria (overall and stratified by stage). To compare ICU and in-hospital length of stay. To evaluate the costs of ICU treatment and overall in-hospital stay. To assess survival rates during the index admission and at the time of study follow-up. To calculate the overall hospital costs for the cardiac surgery episode.

STUDY DESIGN This is a single centre, propensity-matched, retrospective study involving patients who developed AKI after cardiac surgery (CSA-AKI) compared to propensity-matched (1:1) non-AKI patients. It involves retrieving patients' clinical data which was prospectively recorded during their stay when they underwent cardiac surgery at The Royal Wolverhampton NHS Trust in the UK. The data are currently within the PATS Dendrite cardiac surgery database, Electronic Medical Records, TD-web (results Portal), MedTrack (ICU database), iSite (Radiology), SPY (Long-term status) databases, or their equivalents. These are clinical databases used during routine care.

The PATS dendrite database records pre-op, intra-op and post-op data pertaining to patient characteristics, operative details, ICU and ward stay including complications. Data recorded include date of operation, participant's age, body mass index \& gender, risks factors (diabetes, left ventricular function, urgency of procedure), type of procedure, Log Euroscore, cardiopulmonary bypass time, date of extubation, need for inotropes, any complications (respiratory including additional respiratory support, re-exploration, arrhythmias, neurological dysfunction, renal dysfunction, blood transfusion, gastro-intestinal problems), date of ICU discharge, ICU re-admission, date of hospital discharge, participant status at time of discharge and discharge destination. This data is independently verified by the cardiac audit personnel and then the core data is submitted to NICOR from where surgeon-specific and Unit-specific mortality data is confirmed and published on the DoH/SCTS website.

The MedTrack Database records the patients' complications and length of stay on the ICU. The iSite details all the radiological investigations while the TD-Web includes all the investigations results (e.g. blood analysis). Data on all databases are recorded contemporaneously. The Clinical Web Portal (Electronic Medical Records) will be used to cross check this information. Some data may need to be retrieved from the medical records. Survival data will be retrieved from the SPY database while hospital cost data will be requested from the hospital finance department.

This study (data collection) will run over a period of three months. Thereafter, data analysis \& interpretation, manuscript writing \& publication may take up to another three months.

The site research team will extract data from the clinical databases directly into Microsoft Excel spreadsheets. Initial data extraction will involve extracting the pre-propensity matching data for all eligible patients, which will be shared with the statistician for propensity matching. Following propensity matching, further data extraction will take place to collect the post-propensity matching data from the AKI and non-AKI propensity-matched groups. Any missing data will be cross-checked using the Clinical Web Portal or original medical records where necessary.