nephrologist Mujais 200678(40,869) USA ↑ ↑ ↑ McDonald 2003112 (9440) ANZ ↑ Indigenous ↑ ↔ ↑ Snyder 2003113 (41,197) USA ↑ Jager 1999114 (118) Netherlands ↑ ↓Urine volume ≥1000 mL/24hr ↓Peritoneal ultrafiltration Korbet 1999115 (233) USA ↑ ANZ = Australia/New Zealand; APD = ambulatory automated peritoneal dialysis; BMI = body mass index; BP = blood pressure; CAPD = continuous ambulatory peritoneal dialysis; CVD = cardiovascular disease; eGFR = estimated glomerular filtration rate; ESRD = end stage renal disease; PD = peritoneal dialysis ↑ = Significantly associated with higher rates of technique failure; ↔ = Not associated with higher rates of technique failure; ↓ = Significantly associated with lower rates of technique failure 51 Home-based versus In-center Dialysis Evidence-based Synthesis Program Technique Survival (Appendix C, Table 6) A prospective cohort study from the UK identified 142 HHD survivors and 24 who switched from HHD. 121 In a multivariate analysis, only comorbid diabetes was a significant predictor of technique failure (HR 3.96 [955CI 1.66, 9.49]). Patient-reported reasons for switching modalities (provided by 11 of 18 patients who were alive at the end of the study period) included family dynamics (20%), lack of carer support (17%), lack of confidence with procedure (15%), interference with home life (15%), and medical issues including access (12%). A retrospective study from Canada included data from 177 patients (145 successful, 32 failures) who initiated training for nocturnal HHD.122 The study site had a “home-first” policy whereby only patients with absolute contraindications were not invited to trial for HHD. In a multivariable analysis, ESRD due to diabetes and renting current residence were significantly associated with failure. The most common reasons for failure associated with training for HHD included inappropriate housing, deteriorating medical status, inability to cope with burden of HHD, nonadherence, and test failure. The most common reasons for technique failure included deteriorating medical status, inability to cope with burden of HHD, change in residence, inadequate family support, caregiver anxiety, and inability to perform cannulation. A second report from the same study site looked at differences in outcomes between patients characterized as dependent (partially or totally n=47) or independent (n=152) based on need for assistance with nocturnal HHD. 123 The adjusted hazard ratio for a composite outcome of time to all-cause hospitalization, technique failure, or death was not significant (HR 1.25 [95% CI 0.76, 2.04]). The need for back-up dialysis runs at an in-center or training facility did not differ between dependent and independent patients but dependent patients did require more home visits by nurses (RR 2.03 [95% CI 1.39, 2.97]). An analysis of data from the CAN-SLEEP Collaborative Group cohort study also included only nocturnal HHD patients.124 Most patients (74%) were able to perform HHD independently. Among 247 patients, there were 10 technique failures. The only significant predictor of failure was age with an HR of 1.09 (95% CI 1.03, 1.16) for each 1 year increase in age. Using a composite outcome of death or technique failure (36 events), age and diabetes were significant predictors. A prospective cohort study from Canada included all patients who began training for HHD.125 Patients had experienced a mean of 30 months of dialysis before entering the program. During follow-up of up to 3 years, 37 patients dropped out of the program including 13 who received transplants, 14 who died, 2 with inadequate social support, 2 with medical reasons, 2 with inadequate dialysis, 1 who moved, 1 who withdrew from dialysis, and 2 with unspecified reasons. No significant predictors of technique survival were identified. Risk of Bias for Key Question 3 We did not assess the risk of bias of individual registry studies. Registry studies are typically considered high risk of bias due to issues with selection bias and inability to assess and include all potential confounders in analyses. Other studies pertaining to Key Question 3 were one high risk of bias RCT and 2 CCTS (both moderate risk of bias), 5 reports of clinical cohort studies (4 rated high risk of bias and one moderate risk of bias), 24 cohort studies (2 low, 8 moderate, and 52 Home-based versus In-center Dialysis Evidence-based Synthesis Program 14 high risk of bias), and 9 cross-sectional studies (one low, one moderate, and 7 high risk of bias). 53 Home-based versus In-center Dialysis Evidence-based Synthesis Program KEY QUESTION 4. In the published literature, what are the costs of home hemodialysis or peritoneal dialysis compared to in-center hemodialysis? Summary of Findings Fifteen studies (2 from the US) reported cost outcomes. Cost analyses have typically reported lower costs for HHD and PD compared to in-center HD. However, what costs are considered in the analyses and factors that can influence costs (eg, failure rates, patient age, and comorbidity) vary across studies. Costs of Home Hemodialysis versus In-Center HD (Appendix C, Table 7) Cost-utility analysis of data from the randomized Alberta nocturnal HHD study found frequent nocturnal HHD led to incremental cost savings of $6700 Canadian dollars (US$5872 in 2014) and an additional 0.38 QALYs compared to