Mrs Maria Daraoay1, Dr Kamal Sud1,2, Dr Lukas Kairaitis1, Dr Eric Au3,4, Dr Katrina Chau1,5
1Regional Dialysis Centre, Blacktown Hospital, Blacktown, Australia, 2Nepean Kidney Research Centre, Department of Renal Medicine, Nepean Hospital and the Nepean Clinical School, The University of Sydney , Kingswood, Australia, 3The Alfred Hospital, Melbourne, Australia, 4Centre for Epidemiology and Biostatistics, University of Melbourne, Melbourne, Australia, 5Blacktown Clinical School, School of Medicine, Western Sydney University, Blacktown, Australia
Biography:
Maria Daraoay is a Clinical Nurse Consultant at the Western Renal Services at Blacktown Hospital with specialised experience in peritoneal dialysis. Her practice focuses on the clinical management of home-based PD therapies. She is actively involved in clinical governance, quality improvement, and staff education.
Abstract:
Infectious complications are a critical concern in patients undergoing peritoneal dialysis (PD). Identification of patients at higher risk may allow targeted interventions. “Surprise questions” have not been previously applied to predict PD-related outcomes.
Aim
To determine whether PD nurses can predict outcomes in incident PD patients using “surprise” questions.
Methods
Four PD nurses with varying experience (2–10 years) were enrolled in the study. Nurses independently answered three surprise questions for all patients initiating PD between June-December 2024. These were: Would you be surprised if this patient: 1) experienced PD related peritonitis (PDRP) or catheter related infection (CRI), 2) died, 3) discontinued PD within the next 12 months? These patient outcomes were recorded 12 months after initiation of PD.
Results
A total of 52 patients were assessed by each nurse. Mean age was 58 (+/-) years, 30 (57.7%) were male and 23 (44.2%) had diabetes. 21 episodes of infection (PDRP + CRI) occurred in n=16 (31 %) patients within 12 months. n=9 (17.3%) transferred to HD and n=2 (3.8%) died. Nurse prediction of infectious complications demonstrated positive predictive value of 64% (median, range 52.6-77.8%) and negative predictive value of 81% (median, range 78.6–82.9%). Median sensitivity and specificity were 53.1% (43.8 – 62.5%) and 87.5% (75.0 – 94.4%) respectively. Duration of nursing experience did not impact prediction accuracy.
Conclusions
PD nurses demonstrated moderate accuracy in predicting important PD related patient outcomes. A prospective study incorporating a validated health literacy tool and other known risk factors is underway to further improve predictive performance.