MO702DIFFERENCES IN HOSPITALISATION BETWEEN PERITONEAL DIALYSIS AND IN-CENTRE HAEMODIALYSIS PATIENTS

Abstract Background and Aims End stage kidney disease (ESKD) and dialysis treatment are associated with high morbidity, frequently resulting in hospitalisation. However, studies comparing hospitalisation between different dialysis modalities report conflicting results. Some studies report an equal number and length of hospital admissions, while others conclude that peritoneal dialysis (PD) patients are more likely to be hospitalised. In addition, most studies only analyse data of patients that remain on their initial dialysis modality. However, a transition from one dialysis modality to another, e.g. from PD to in-centre haemodialysis (ICHD), certainly occurs in current dialysis practice. Therefore, the aim of this study was to compare hospitalisations between PD and ICHD patients, taking into account transfers between dialysis modalities. Method The retrospective Dutch nOcturnal and hoME dialysis Study To Improve Clinical Outcomes (DOMESTICO) collected hospitalisation data of ESKD patients who started dialysis treatment between 2012 and 2017. Eligible patients had a minimum dialysis duration of 3 months. For baseline comparison, groups were defined based on the dialysis modality (i.e. PD or ICHD) 3 months after dialysis initiation. Primary outcome was hospitalisation rate, which was analysed with a multi-state model that attributed each hospitalisation to the dialysis modality the patient was treated with at that the time. Secondary outcomes were time to first hospitalisation, number of hospitalisations and length of hospitalisation. Time to first hospitalisation was analysed with Cox regression analysis, with dialysis modality as a time-varying covariate. Number of hospitalisations was analysed with negative binomial regression, and length of stay with Poisson regression. All analyses were adjusted for potential confounders. Results In total, 252 PD and 443 ICHD patients from 31 Dutch dialysis centres were included. Baseline characteristics of the groups were comparable, apart from a lower dialysis vintage and a slightly lower comorbidity score in the PD group. Patients transferred more often from PD to ICHD (33%), than from ICHD to PD (11%) during a median follow-up period of 22.0 months [IQR 11.1-36.4]. The crude hospitalisation rate for PD was 2.3 (±5.0) and for ICHD 1.4 (±3.2) hospitalisations per patient-year. Using a multistate model, the adjusted hazard ratio (HR) for hospitalisation rate was 1.1 (95%CI 1.02-1.3) for PD compared to ICHD patients. Cox regression analysis showed a significant difference in time to first hospitalisation with an adjusted HR of 1.3 (95%CI 1.1 - 1.6) for PD compared to ICHD patients in the first year after dialysis initiation. After the first year, the time to first hospitalisation had an adjusted HR of 1.9 (95%CI 1.4–2.5) for PD compared to ICHD patients. The number of hospitalisations was significantly higher, while the length of stay was non-significantly higher for PD patients. In PD patients, the most common cause of all hospitalisations was peritonitis (23%). In ICHD patients, the most common cause was access-related (33%). Conclusion PD was associated with a higher hospitalisation rate, a shorter time to first hospitalisation and more hospitalisations compared to ICHD. One explanation might be that the threshold for admission to hospital is lower for PD patients compared to ICHD patients.