Representativeness of the PDOPPS cohort compared to the Australian PD population

Background: The Peritoneal Dialysis Outcomes and Practice Patterns Study (PDOPPS) is an international, prospective study following persons treated by peritoneal dialysis (PD) to identify modifiable practices associated with improvements in PD technique and person survival. The aim of this study was to assess the representativeness of the Australian cohort included in PDOPPS compared to the complete Australian PD population, as reported to the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry. Methods: Adults with at least one PD treatment reported to ANZDATA Registry during the census period of PDOPPS Phase I (November 2014 to April 2018) were compared to the Australian PDOPPS cohort. The primary outcomes were the representativeness of centres and persons. Secondary outcomes explored the association of person characteristics with consent to study participation. Results: After data linkage, 511 PDOPPS participants were compared to 5616 Australians treated with PD. Within centres eligible for PDOPPS, selected centres were similar to other Australian centres. The PDOPPS participants’ cohort tended to include older persons, more males, a higher proportion of Caucasians and more persons with higher socioeconomic advantage compared to the Australian PD population. Differences in distribution across sex and ethnicities between the PDOPPS cohort and the overall PD population were in part due to the selection and consent processes, during which females and non-Caucasians were more likely to not consent to PDOPPS participation. Conclusion: Sampling methods used in PDOPPS allowed for good national representativeness of the included centres. However, representativeness of the unweighted PDOPPS sample was suboptimal in regard to some participant characteristics.

MO037SYMPTOM MONITORING WITH FEEDBACK TRIAL (SWIFT) PILOT TO EXPLORE THE FEASIBILITY AND ACCEPTABILITY OF ELECTRONIC PATIENT REPORTED OUTCOME MEASURES (E-PROMS) DATA CAPTURE AND FEEDBACK

Background and Aims People on haemodialysis (HD) often experience overwhelming and debilitating symptoms of fatigue, pain, nausea, itching, sleeping problems and depression that contribute to poor quality of life. Trials in oncology have shown that active symptom monitoring can improve quality of life and survival, but it remains unknown if this intervention is effective in people on HD. The use of PROMs is increasing in nephrology care; but they are not routinely collected in kidney registries. The SWIFT pilot aims to assess the feasibility and acceptability of tablet-based symptom monitoring with feedback to clinicians in preparation of a large-scale, registry-based cluster-randomised trial to assess the clinical- and cost-effectiveness of this intervention in HD (SWIFT; ACTRN12618001976279). Method This pragmatic, multicenter cluster-randomised controlled pilot study tests the hypothesis that 3-monthly tablet-based PROMs monitoring using the Integrated Palliative Outcome Scale-Renal Symptom Tool with electronic feedback to clinicians using the Australia and New Zealand Dialysis and Transplant Registry (ANZDATA) compared with usual care will improve health-related quality of life measured by EQ-5D-5L and is technically feasible and clinically acceptable to patients, nurses and nephrologists in different HD settings. Acceptability and feasibility outcomes are assessed through a process evaluation using the Medical Research Council framework: (1) patient acceptability of ePROMS (5-point Likert scale; focus groups with 20 purposively sampled participants), (2) feasibility of embedding ePROMS in routine care (response rates, time from data capture to feedback, ease of generating PROMs reports by the ANZDATA registry), (3) acceptability and usefulness of ePROMs data for nephrologists and nurses (qualitative semi-structured interviews with thematic analysis). Results 163 adult patients from 4 HD units across Australia were enrolled (83 in the intervention units and 80 to the control units) and completed baseline PROMs, representing 77% of eligible patients (range 44-90%). At 3-months, 61% of enrolled patients in the intervention units completed the PROMs (range 48-73%). Reasons for non-completion included transfer to another unit and lack of ability to read English. Severe or overwhelming symptoms were reported by 30 of 83 enrolled patients (36%) at baseline and by 32 of 62 (52%) at 3 months. Technical feasibility was demonstrated by successful development of a Qualtrics survey platform presented on tablet computers, use of QR reader codes for linkage with the ANZDATA registry which captures clinical outcomes, and linkage to the relevant survey for the patient’s allocation and trial timepoint. Emails containing a short report of any symptoms identified as severe or overwhelming along with a summary of evidence-based guidelines for management of those symptoms were sent to and opened by treating clinicians. Assessment of patient and clinician acceptability of the feedback mechanism is ongoing. Conclusion Electronic symptom monitoring in adults on HD with feedback to clinicians is feasible and provides a template for ongoing registry-based monitoring of PROMs to facilitate patient-centred care and the infrastructure for future evaluations of interventions to improve patients’ quality of life and survival.

Effect of patient- and center-level characteristics on uptake of home dialysis in Australia and New Zealand: a multicenter registry analysis

Background Home-based dialysis therapies, home hemodialysis (HHD) and peritoneal dialysis (PD) are underutilized in many countries and significant variation in the uptake of home dialysis exists across dialysis centers. This study aimed to evaluate the patient- and center-level characteristics associated with uptake of home dialysis. Methods The Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry was used to include incident dialysis patients in Australia and New Zealand from 1997 to 2017. Uptake of home dialysis was defined as any HHD or PD treatment reported to ANZDATA within 6 months of dialysis initiation. Characteristics associated with home dialysis uptake were evaluated using mixed effects logistic regression models with patient- and center-level covariates, era as a fixed effect and dialysis center as a random effect. Results Overall, 54 773 patients were included. Uptake of home-based dialysis was reported in 24 399 (45%) patients but varied between 0 and 87% across the 76 centers. Patient-level factors associated with lower uptake included male sex, ethnicity (particularly indigenous peoples), older age, presence of comorbidities, late referral to a nephrology service, remote residence and obesity. Center-level predictors of lower uptake included small center size, smaller proportion of patients with permanent access at dialysis initiation and lower weekly facility hemodialysis hours. The variation in odds of home dialysis uptake across centers increased by 3% after adjusting for the era and patient-level characteristics but decreased by 24% after adjusting for center-level characteristics. Conclusion Center-specific factors are associated with the variation in uptake of home dialysis across centers in Australia and New Zealand.

Representativeness of Honeypot Trial Participants to Australasian PD Patients

BackgroundThe HONEYPOT trial failed to establish the superiority of exit-site application of Medihoney compared with nasal mupirocin prophylaxis for the prevention of peritonitis in peritoneal dialysis (PD) patients. This study aimed to assess the representativeness of the patients in the HONEYPOT trial to the Australian and New Zealand PD population.MethodsThis study compared baseline characteristics of the 371 PD patients in the HONEYPOT trial with those of 6,085 PD patients recorded on the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry.ResultsCompared with the PD population, the HONEYPOT sample was older (standardized difference [ d] = 0.19, p = 0.003), more likely to be treated with automated PD ( d = 0.58, p < 0.001), had higher residual renal function ( d = 0.26, p < 0.001) and a higher proportion of participants with end-stage kidney disease due to polycystic kidney disease ( d = 0.17) and lower proportion due to diabetes ( d = -0.17) and glomerulonephritis ( d = -0.18) ( p < 0.001), and lower proportions of indigenous people ( d = -0.17, p < 0.001), current smokers ( d = -0.10, p < 0.001), and people with prior histories of hemodialysis ( d = -0.16, p < 0.001), diabetes mellitus ( d = -0.18, p < 0.001), and coronary artery disease ( d = -0.15, p < 0.001).ConclusionsHONEYPOT trial participants tended to be healthier than the Australian and New Zealand PD patient population. Although the differences between the groups were generally modest, it is possible that their cumulative effect may have had some impact on external generalizability, which is not an uncommon occurrence in clinical trials.