cohort study from Spain enrolled 489 incident PD patients.76 Average follow-up was 13.4 months. Ninety-five (19%) had started dialysis on HD. The mortality rate was higher in patients that changed from HD to PD compared to those who initiated RRT with PD (11.5% vs 4.6%, P = .009). In a longitudinal study from Poland, 264 PD patients (67 of whom transferred to PD after a median of 18 months [range 3-268] on HD) were followed for a median of 21 months.77 No significant difference was observed in survival for the transferred patients versus the initial PD patients (RR 1.68 [95% CI 0.87, 3.22]). The result was similar for the combination of patient and technique survival (RR 1.45 [95% CI 0.89, 2.37]). A registry study from the US with 40,869 patients and follow-up of one to 4 years reported that survival was higher for patients who initially received PD compared to those who transferred from HD.78 At one year, the percentage of patients surviving was 86.7% in the initial PD group compared to 83.9 in the transfer to PD group. At 4 years, the values were 56.7% and 53.1%, respectively. The hazard ratio for patient survival for patients new to dialysis versus transfer from HD was 0.73 (P < .0001). It was noted that duration of ESRD was likely longer for the patients transferring from HD. Technique survival was longer for the initial PD patients. The hazard ratio for technique survival (new to dialysis versus transfer from HD) was 0.79 (P < .0001). Patients new to dialysis were more likely to undergo transplantation (HR 1.31, P < .0001). Details of these studies are presented in Appendix C, Tables 4-6. 39 Home-based versus In-center Dialysis Evidence-based Synthesis Program Risk of Bias for Key Question 2 As noted for Key Question 1, 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. There was one high risk of bias RCT that addressed KQ2. Of 8 clinical cohort reports, 4 were rated as high risk of bias and 4 as moderate risk of bias. There were 7 longitudinal studies – 2 high risk of bias and 5 moderate risk of bias. All of the cross-sectional studies (k=6) were rated high risk of bias. Additional information is presented in Appendix C, Table 2. 40 Home-based versus In-center Dialysis Evidence-based Synthesis Program KEY QUESTION 3. What are the a) health care system, b) provider, and c) patient factors associated with selection of and technique survival for home-based dialysis (including peritoneal dialysis)? Summary of Findings · Twenty-two articles (21 studies, 8 from the US) provided information on factors associated with selection of PD and 5 articles (none from the US) addressed factors associated with selection of HHD. · For PD selection, the following factors were reported: o Health Care System Factors: One US cross-sectional study reported that provision of home-based dialysis (including PD) was more likely in larger dialysis facilities (defined as 62 patients or more) with more years of facility Medicare certification and facilities with a higher population of employed 18 to 54 year old patients. Home-based dialysis was less likely at facilities in more rural areas, facilities offering evening care, and facilities with higher treatment capacity (based on number of patients, number of HD stations, and availability of a late shift). o Provider Factors: Several studies found that provision of patient education about dialysis modalities and a determination of medical (including comorbid conditions and decreased strength, manual dexterity, vision, or hearing) and psychosocial suitability (including fear of self-cannulation, anxiety, decreased cognition, psychiatric conditions, or history of non-compliance) for PD were associated with greater selection of PD. No studies reported on provider factors such as provider age, training, knowledge about PD, etcetera. o Patient Factors: Autonomy, ability to travel, and compatibility with employment were identified as positive features of PD. Conversely, lack of understanding, living alone, lack of space in the home, inability to perform PD in the place of residence, fear of social isolation, fear of inability to perform PD, and preference for medical supervision were patient barriers to selection of PD. · For HHD, the following factors were reported: o Health Care System Factors: As noted above, dialysis facility size, geographic location, and years of certification were all factors in provision of any homebased dialysis. o Provider Factors: From a provider perspective, patients with medical contraindications, psychosocial contraindications), unsuitable living conditions (including HHD not permitted, overcrowding, dampness/mold growth), lack of support in the home, and unplanned start or shorter pre-dialysis care by a nephrologist were less likely to be suitable for HHD. Providers with greater numbers of HHD patients reported having a dedicated education team. 41 Home-based versus In-center Dialysis Evidence-based Synthesis Program o Patient Factors: Patient-reported barriers to and advantages of HHD were similar to those noted above for PD. · Fifteen studies (8 from the US) reported factors associated with PD technique failure (the inverse of technique survival – a switch from PD to in-center HD): o Health Care System Factors: Patients from larger clinics had lower