domain. Social support in the highest tertile was significantly associated with the chance of receiving PD (P = .02).63 Several reports with subsets of the NECOSAD cohort addressed quality of life outcomes. One (n=161) reported no difference between HD and PD in illness consequences or whether treatment controls the illness (both measured with the Brief Illness Perception Questionnaire). 64 Based on responses to the Treatment Effects Questionnaire, HD patients perceived more consequences of treatment than PD patients (P = .01).64 Another study (n=528) reported that the effect of social support on mortality was similar for HD and PD patients. 65 A third study (n=228) reported a significant adjusted mean difference over time in physical quality of life (SF-36) favoring HD (1.6 [95% CI 0.04, 3.20], P = .04) but no difference in mental quality of life.66 Several longitudinal or cross-sectional studies, not included in the existing reviews, also provided quality of life outcomes. A longitudinal study from the UK reported no significant differences in SF-36 Physical Component, SF-36 Mental Component, or KDQOL Symptom scores at 6 or 12 months follow-up between HD and PD patients who were 70 years of age or 34 Home-based versus In-center Dialysis Evidence-based Synthesis Program older at the start of dialysis. 58 One cross-sectional study from the UK reported scores for the Treatment Effects Questionnaire, Beck Depression Index, and Cognitive Depression Index.22 No differences were noted between in-center HD and PD. A 2002 study from the UK found mixed results for different quality of life instruments. 67 On the EuroQol EQ-5D, differences between incenter HD and PD patients were not significant. Using the Kidney Disease Quality of Life instrument, patients receiving PD scored significantly higher on effects of kidney disease, burden of kidney disease, and cognitive function but lower for sexual function. On the SF-36, PD patients had higher scores for the mental component summary but not the physical component summary. A 1999 study from the UK (cited above in the in-center HD vs HHD analysis) reported scores on components of the SF-36.23 Differences across groups (HHD, in-center HD, and PD) were noted for Physical Functioning, Role Physical, Social Functioning, and Role Emotional. The Canadian study (also cited above in the in-center HD vs HHD analysis) reported a non-significant difference between HD and PD in scores on the Self-Anchoring Striving Scale but significantly lower scores for HD vs PD on the Index of Well-Being and the Health State Utility/Time Trade-off assessment. 24 A study from the US reported that the risk of moderate to severe cognitive impairment for patients receiving either PD or in-center PD was significantly higher than that for patients age 55 and older without CKD.68 A second average-quality systematic review presented quality of life utilities. 69 Utilities represent the strength of a patient’s preference for specified health-related outcomes with values ranging from 0 (death) to one (full health). Some studies included in the review assessed utilities directly. For others, utilities were derived from SF-36 scores. The review included patients ranging from pre-treatment CKD to kidney transplant; 69% of the utilities evaluated in the review were from studies of dialysis. The mean utility estimate for HD (including both in-center HD and HHD) was 0.69 (95% CI 0.59, 0.80) while the estimate for PD was 0.72 (95% CI 0.62, 0.83). Although 0.03 is considered to be the minimum clinically important difference for utility scores, the test for interaction was not significant (P = .08). Studies of life participation activities were reported in a high-quality systematic review. 70 The activities of interest were physical function (eg, activities of daily living, self-reported physical functioning with the SF-36), travel abilities or restrictions, ability to engage in recreational or social activities, freedom (eg, perceived independence, ability to perform usual tasks), and work outcomes (eg, employment or working capacity). The review included English language cohort and cross-sectional studies published between 1980 and April 2012 and using a variety of outcome measures. For the comparison of HD and PD, there were 39 studies. Of 41 measures of physical function (some studies reporting more than one measure), only 10 showed a significant difference between HD and PD with 3 favoring HD and 7 favoring PD. Of 2 measures of travel, there was one significant difference favoring HD. There were 18 measures of recreation, 4 with significant differences favoring PD. Of 8 measures of freedom, one favored HD and one favored PD. Similarly, of 13 measures of work, 2 favored HD and 2 favored PD. The authors reported that the results were consistent across study designs, locations (US vs non-US), quality rating (appropriate adjustment for confounders vs no or minimal adjustment), and year of publication (1980-1990, 1991-2000, 2001-2012). Change in Dialysis Modality (Appendix C, Table 1) Seven registry studies reported changes in dialysis modality. A USRDS study reported that 6% of HD patients switched to PD and 57% of PD patients switched to HD during the 2 year followup period.26 A second USRDS reported similar findings; over a maximum follow-up of 5 years, 35 Home-based versus In-center Dialysis Evidence-based Synthesis Program 4% of