Wang 2017.

Study characteristics
Methods	Country: USA Type of study: observational Design: cohort study (retrospective) Time frame: January 2015 to December 2016 Duration of follow‐up: mean follow‐up was 60 months for HD patients and 46 months for PD patients
Participants	Number: treatment group (19); control group (28) Mean age ± SD (years): not reported Sex: not reported DM: not reported
Interventions	Treatment group PD Control group HD
Outcomes	Dialysis access‐related complications Total number of procedures required for dialysis access
Notes	Those who are unstable with critical illness were excluded The primary investigators were unable to contact for further information
Risk of bias
Bias	Authors' judgement	Support for judgement
Selection: representativeness of exposed cohort	Unclear risk	Insufficient data to permit judgement
Selection: non exposed cohort	Unclear risk	Insufficient data to permit judgement
Selection: ascertainment of exposure	Low risk	Medical records
Selection: demonstration that outcome of interest was not present at the start of the study	Low risk	Outcome of interest unlikely to be present before the study
Comparability of cohorts on basis of design or analysis	High risk	There was significant difference between PD and HD groups in baseline data including age, presence of heart failure, hypertension, and pre‐dialysis SCr
Outcome: assessment	Unclear risk	Insufficient data to permit judgement
Outcome: follow‐up length	Low risk	Adequate follow‐up period
Outcome: adequacy of follow‐up	Unclear risk	Insufficient data to permit judgement

CKD ‐ chronic kidney disease; DM ‐ diabetes mellitus; ESKD ‐ end‐stage kidney disease; HD ‐ haemodialysis; ITT ‐ intention‐to‐treat; KPNC ‐ Kaiser Permanente North Carolina; NYHA ‐ New York Heart Association; PD ‐ peritoneal dialysis; SCr ‐ serum creatinine; SD ‐ standard deviation; USHD ‐ urgent‐start HD; USPD ‐ urgent‐start PD