Roslyn Prichard, Louise Kershaw, Stephen Goodall, Patricia Davidson, Phillip Newton, Frederick McNeil, Timothy Homer, Christopher Hayward

Widespread application of left ventricular assist devices (LVADs) in advanced heart failure, is limited by costs, and access to technical expertise. Hospitalisation drives both cost and inversely, quality of life - but cross institutional and pre-surgical inpatient length of stay data is missing in the Australian literature. We describe changes in hospitalisation rates, in the year before and after bridge to transplant LVAD therapy and preceding heart transplant (HTX). Hospitalisation, Australian refined diagnosis group (ArDRG), and clinical data were assessed for 77/100 consecutive patients listed for heart transplant between July of 2009 and June of 2012. Twenty-five of the patients required ventricular assist device (VAD) therapy whilst waitlisted. Hospitalisation was defined as the proportion of "days at risk" that were spent in hospital and included all public and private admissions identified in the year before and after VAD implant, or before HTX, in a linked administrative dataset of admissions across New South Wales. Patients requiring VADs were clinically more unstable and spent proportionally more time in hospital than pre-HTX patients, (13% (IQR 10-20%) vs 4% (IQR1-10%), p<0.01). During the index admission, they spent 22days (IQR 10-33) in hospital before implantation, including 13 days in non-transplant centres (IQR 7-20). Following implantation, median inpatient stay was 31(IQR 26-70) - including rehabilitation in 8 of the 25 patients. The number of admissions per patient reduced in the year after VAD-implant to two (IQR1-3), from five pre-implant (IQR 3-7) p=0.002. This was similar to the pre-HTX group's three admissions (IQR1-6), p=0.33. Overall hospitalisation decreased in VAD patients beyond the first year - from 14% (IQR 10-20%) at one-year to 0.5% (IQR 0-10%) at two-years (p=0.002). A high percentage of hospitalisation prior to VAD (41%) and HTX (66%) occurred outside the transplant centre. A high proportion of activity in the pre-implant and transplant year occurs outside the implanting hospital with higher pre-implant hospitalisation in VAD patients reflecting clinical severity. Ventricular assist device implantation is significantly associated with reduced admissions, and hospitalisation once reconditioning has occurred.