Prevention of Peritonitis in Peritoneal Dialysis

Time (days) after the date of randomization to the date of the first peritonitis episode was calculated for each participant. The cumulative time without peritonitis using the Cox regression model from which unadjusted hazard ratios were calculated. Actuarial survival curves showing proportions of peritonitis-free participants over time in the two groups were estimated by means of the Kaplan-Meier method. Log rank follow-up was used to compare the survival curves. Patients who stopped PD treatment due to causes other than peritonitis were managed according to the intention-to-treat principle.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis, i.e. to maximally 35 months after randomization. Time to first peritonitis was calculated for each participant.

Population: A number of 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. Participant who stopped PD treatment due to causes other than peritonitis were managed according to the intention-to-treat principle.

The incidence of a first peritonitis episode is expressed as the mean rate of a first peritonitis per patient year (event rate) with 95% confidence interval.

Time frame: The participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis, i.e. to maximally 35 months after randomization. The time in study was calculated for each participant

Population: 671 new patients with peritoneal dialysis, peritonitis-free, aged 18 year or more, and able to perform peritoneal dialysis without assistance were randomized to a control group or a follow-up group and followed up to 36 months after start of peritoneal dialysis. The time (patient-years) in the study to the first peritonitis episode was calculated for for each participant.

The incidence of peritonitis episodes is expressed as the mean rate of peritonitis per patient year (event rate) with 95% confidence interval.

Time frame: The participants were followed from 1 month after PD start (time of randomization) up to 36 months after start of peritoneal dialysis, The time (days) of participation in the study was calculated for each participant.

Population: 671 new patients with peritoneal dialysis, peritonitis-free, aged 18 year or more, and able to perform peritoneal dialysis without assistance were randomized to a control group or a follow-up group and followed up to 36 months after start of peritoneal dialysis.

The association of baseline variables (at PD start) was studied in the whole study group. Age was recorded at baseline. Data were collected at baseline except for medication, PD modality (CAPD/APD, continuous ambulatory PD/automatic PD), number of PD bags and volume of dialysis fluid used per day, and if help with exit-site care was needed. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for body mass index at PD start is presented. The association of baseline variables (at PD start) or time-updated variables with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis episode in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for body weight at PD start is presented. The association of baseline variables (at PD start) or time-updated variables with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Comorbidity in association with time to first peritonitis episode is presented. The Stoke grading system of cardiovascular comorbidity was used to reflect the comorbid burden. Comorbidity was graded 0 when absent, 1 when 1 or 2, and 2 when 3 or more conditions were present. The association of baseline variables (at PD start) or time-updated variables with time to first was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Analysis of Cytotoxic treatment (yes vs.no) in association with time to first peritonitis episode is presented. This variable may vary with time and is thus time-dependent. The variable was recorded at all study visits. The value recorded at the study visit which was closest to and prior to the first peritonitis episode was used in the time-dependent univariable and multivariable Cox regression analyses of the association of the variable with time to first peritonitis episode. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The count of participants on cytotoxic treatment at baseline is presented below.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Diabetes mellitus as comorbidity (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Diabetes mellitus as comorbidity or main cause of kidney failure (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Diabetic nephropathy as primary cause of kidney failure are presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Functional status (Karnofsky score) in association with time to first peritonitis episode in the whole study group is presented. Functional status was estimated using the Karnofsky Performance Scale with scores from 100 (Normal, no complains, no evidence of disease) to 0 (dead). 90 = Able to carry on normal activity, minor signs or symptoms of disease. 80 = Normal activity with effort, some signs or symptoms of disease. 70 = Cares for self, unable to carry on normal activity or do active work. 60 = Requires occasional assistance, but able to to care for most of her/his needs. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Gender is presented. The association of baseline variables (at PD start) or time-updated variables with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Glomerulonephritis as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Ischemic heart disease (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Ischemic kidney disease as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis episode in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Left ventricular dysfunction (yes vs. no) in association with time to first peritonitis episode is presented. Left ventricular dysfunction includes clinical evidence of pulmonary oedema not attributable to errors in fluid balance or moderate to severe left ventricular dysfunction on echocardiography, or clinical signs of symptomatic moderate to severe heart failure. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The exit site is the point where the PD catheter emerges from the skin of the abdomen. The number of participants who needed of help with exit-site care at PD start is reported below. The need of this help may vary with time. This variable is thus time-dependent. The time-dependent data were recorded at all study visits. The value recorded at the study visit, which was closest to and prior to the first peritonitis episode for each participant was used in the analyses. The data were included as time-varying covariates in a time-dependent Cox regression. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Nephrosclerosis/hypertension as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The association of the Number of PD bags used/24h and Time to First Peritonitis Episode was studied. The number of PD bags used/24h many vary with time, as this variable is thus time-dependent. Data from PD start is presented below. As described in Outcome measure 39, data was collected at each study visit. We recorded the number of bags connected to the PD catheter/24h in CAPD patients and in APD patients we recorded the number of bags connected to the tubing set/24h adding the daily number of bags connected to the PD catheter if additional manual daytime exchanges were used. The time-dependent data was included in a time-dependent univariable Cox regression to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Other diagnosis as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Other significant pathology (yes vs. no) in association with time to first peritonitis episode is presented. Other significant pathology includes diseases like severe chronic obstructive airway disease, liver cirrhosis, psychotic illness, and severe osteoarthritis. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

PD modality (CAPD/APD) in Association With Time to First Peritonitis Episode in the whole study group was studied. A participant either used CAPD (Continuous Ambulatory Peritoneal Dialysis) or APD (Automated Peritoneal Dialysis) but may change the PD modality during the study. This variable may vary with time and is thus time-dependent. The time-dependent data was recorded at all study visits. The value recorded at the study visit which was closest to and prior to the first peritonitis episode for each participant was used in the analyses. The time-dependent data was included in a time-dependent Cox regression. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The c

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Peripheral vascular disease or stroke (yes vs. no) in association with time to first peritonitis episode is presented. Peripheral vascular disease or stroke includes distal aortic, renovascular, and cerebrovascular disease as well as symptomatic disease in these vascular territories e.g. cerebrovascular accident, claudication, amputation or significant stenosis \>50% on vascular imaging or Doppler ultrasound. The association of baseline variables with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Polycystic kidney disease as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Previous kidney replacement therapy (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable System collagen vascular disease (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Analysis of the association of the Study group (retraining group vs. control group) with time to first peritonitis episode is presented. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group.

The statistical analysis of the baseline biological characteristic Serum haemoglobin (mmol/L) in association with time to first peritonitis episode is presented. Serum haemoglobin was measured just before the start of peritoneal dialysis. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline biological characteristic Serum albumin (\< 35 g/L vs. 35 g/L or more) in association with time to first peritonitis episode is presented. Serum albumin was measured just before the start of peritoneal dialysis. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline biological characteristic Serum albumin (g/L) in association with time to first peritonitis episode is presented. Serum albumin was measured just before the start of peritoneal dialysis. The association of baseline variables (at PD start) with time to first peritonitis episode was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline biological characteristic Serum creatinine (micromol/L) in association with time to first peritonitis episode is presented. Serum creatinine was measured just before the start of peritoneal dialysis. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline biological characteristic Serum urea (mmol/L) in association with time to first peritonitis episode is presented. Serum urea was measured just before the start of peritoneal dialysis. The association of baseline variables (at PD start) with time to first peritonitis episode was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Current smoking (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Impaired hand function (yes vs. no) in association with time to first peritonitis episode is presented. The impairment of hand function was unspecified. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Living alone (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Working full or part time (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Retired (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline variable Visual impairment (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of the baseline social factor Working full or part time (yes vs. no) in association with time to first peritonitis episode is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

PD catheters could be either be straight or coiled. The statistical analysis of the type of PD catheter used (straight vs. coiled) in association with time to first peritonitis episode is presented. The association of the type of PD catheter used in association with time to first peritonitis episode was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The analysis of the use of a Corticosteroid drug (yes vs.no) in association with time to first peritonitis episode is presented. This variable may vary with time and is thus time-dependent. The variable was recorded at all study visits. The value recorded at the study visit which was closest to and prior to the first peritonitis episode was used in the univariable and multivariable Cox regression analyses of the association of the variable with time to first peritonitis episode. Time-dependent univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The count of participants on corticosteroid treatment at baseline is presented below.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of The use of antibiotics prior to PD catheter insertion (yes vs. no) in association with time to first peritonitis episode is presented. Antibiotics prior to PD catheter insertion may be given to prevent infection. The association of The use of antibiotics prior to PD catheter insertion in association with time to first peritonitis episode was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The count of participants who used antibiotics prior to PD catheter insertion is presented below.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis episode in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The association of The Volume of Dialysis Fluid used/24h with the time to first peritonitis episode is analysed in the whole study group. The volume of dialysis fluid used per 24h can vary with time and is thus time-dependent. The dialysis fluid volume/24h used at PD start is reported below. The time-dependent data were recorded at all study visits. The value recorded at the study visit, which was closest to and prior to the first peritonitis episode for each participant was used in the analyses. The time-dependent data were included in a time-dependent Cox regression. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Tubulointerstitial nephritis as primary cause of kidney failure is presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis episode in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

The statistical analysis of Type of start of peritoneal dialysis in association with time to first peritonitis episode is presented. The type of start of peritoneal dialysis could be either planned or acute. The association of Type of start of peritoneal dialysis in association with time to first peritonitis episode was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.

Outcome data for the baseline variable Unknown primary cause of kidney failure as primary cause of kidney failure are presented. The association of baseline variables (at PD start) with time to first peritonitis was studied in the whole study group. Univariable Cox regression was used to identify variables significantly associated with time to first peritonitis episode. Multivariable Cox regression was performed using a backward elimination procedure stopping when all remaining factors were significant at p \< 0.05. No collinearity among the factors affecting the backward elimination procedure was found. The risk (hazard ratio) associated with each studied variable is reported.

Time frame: Participants were followed from 1 month after PD start (time of randomization) up to a maximum of 36 months after start of peritoneal dialysis. The time (days) to the first peritonitis episode was calculated for each participant.

Population: 671 new patients with peritoneal dialysis (PD), peritonitis-free, aged 18 years or more, and able to perform PD without assistance were randomized to a control group and a retraining group. We decided to perform the analysis of factors associated with time to first peritonitis in the two groups together, since the possibility to find such associations is likely to be greater in a larger than in a smaller group.