reported no difference in hospitalizations.19 Follow-up periods ranged from 8 weeks16 to 20 months.17 Risk of Bias 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 were 2 RCTs that addressed Key Question 1, one 25 Home-based versus In-center Dialysis Evidence-based Synthesis Program moderate risk of bias and one high risk of bias. The 3 CCTs were rated as moderate (k=1) and high (k=2) risk of bias. Three cross-sectional studies were all rated as high risk of bias. Additional information is presented in Appendix C, Table 2. 26 Home-based versus In-center Dialysis Evidence-based Synthesis Program KEY QUESTION 2. What are the benefits and harms (ie, all-cause mortality, cardiovascular events, hospitalizations, depression, cognitive impairment, quality of life, conversion to a different type of dialysis, complications related to vascular access, complications of dialysis) of peritoneal dialysis compared to in-home hemodialysis or in-center hemodialysis? Summary of Findings · Evidence is inconsistent whether mortality differs between patients treated with PD compared to in-center HD. · Strength of evidence for mortality was low based on high risk of bias associated with the registry studies. Results from registry studies should be interpreted with caution due to likely residual confounding. · Twenty-seven registry studies, one RCT, and 2 clinical cohort studies provided evidence for the comparison of PD to in-center HD. Of 22 registry studies reporting mortality for the total sample, 12 (2 from the US, 3 from Canada, 1 from Australia/New Zealand, and 6 from Europe/UK) found no difference in mortality between PD and in-center HD. Four studies (2 from the US, one from Canada, and one from Europe/UK) found a mortality benefit for PD while 6 studies (3 from the US, 2 from Australia/New Zealand, and one from Europe/UK) found a mortality benefit for in-center HD. It is difficult to assess if results vary by time of publication due to differences in study populations, length of follow-up reported, and methods of data analysis, but publication dates suggest that a trend may exist. Studies showing increased mortality with PD compared to in-center HD were generally published before 2003 while studies showing no difference or reduced mortality with PD were generally published after 2003. · A small RCT from the Netherlands found no difference in mortality between PD and incenter HD. This study was terminated due to low enrollment. A prospective, clinical cohort study from the United States with 1,041 patients and a follow-up of up to 7 years found no difference in mortality. Data from a prospective cohort study from the Netherlands showed no difference in 2-year mortality. · Analyses of interactions between dialysis modality and age (10 studies), gender (4 studies), race (5 studies), BMI (5 studies), diabetes (12 studies), cardiovascular disease (6 studies), and duration of ESRD (3 studies) yielded mixed results. · Of 5 registry studies reporting cardiovascular disease risk, one reported a significantly higher percentage of deaths due to cardiovascular disease in the PD group. In 3 of 5 studies reporting, hospitalizations were higher in the HD groups. Mixed results were reported for quality of life outcomes including mental and physical health components, quality of life utilities, and life participation activities. Changes in treatment modality and kidney transplantation were generally more likely for patients receiving PD. · Few studies reported adverse events. 27 Home-based versus In-center Dialysis Evidence-based Synthesis Program · There is limited evidence for the comparison of PD and HHD. In 2 registry studies, results were mixed with a study from the United States finding no difference in mortality and a study from the United Kingdom finding a mortality benefit for HHD. Other outcomes were not reported. · Two studies reported higher mortality among patients who initiated ESRD treatment with HD and then switched to PD compared to patients who initiated PD as their first modality. Overall duration of ESRD was likely longer in the patients who initiated with HD. Peritoneal Dialysis (PD) Compared to In-Center Hemodialysis (HD) Study Characteristics Twenty-seven registry studies reported mortality outcomes for patients receiving HD or PD. There were 11 reports of Centers for Medicare and Medicaid (CMS) data26-36 representing patient data from 1987 to 2006. Maximum follow-up ranged from one to 6 years. Sample sizes ranged from 3,337 to 684,426 and all but one35 reported data from incident patients. In 7 of the 11 studies, the PD patients were younger and in all of the studies, PD patients were less likely to be African-American. Two studies reported that PD patients were more likely male. 30,33 All studies used an intent-to-treat approach. Three used a Poisson regression model,30,32,36 5 used Cox proportional hazards models,29,31,33,34 one used a MSM approach,28 2 used both Cox and MSM models,26,35 and one did not specify.27 Two studies included matched-pair data. 27,29 Additional patient characteristics and details about the analyses are presented in Appendix C, Table 1. One additional study reported US data. 37 This analysis