AK VIECELLI1, E O’LONE2, B SAUTENET3, JC CRAIG2, A TONG2, E CHEMLA4, LS HOOI5, T LEE6, C LOK7, KR POLKINGHORNE8, RR QUINN9, T VACHHARAJANI10, R VANHOLDER11, L ZUO12, AB IRISH13, TA MORI13, EM PASCOE1, DW JOHNSON1, CM HAWLEY1
1School of Medicine, University of Queensland, Brisbane, Queensland; 2Centre for Kidney Research, The Children’s Hospital at Westmead, Sydney, New South Wales; 3University Francois Rabelais, Tours, France; 4St George’s University NHS Foundation Trust, London, United Kingdom; 5Department of Medicine and Haemodialysis Unit, Hospital Sultanah Aminah, Johor Bahru, Malaysia; 6Department of Medicine and Division of Nephrology, Veterans Affairs Medical Center, Birmingham, Alabama, United States; 7Department of Medicine, University of Toronto, Toronto, Ontario, Canada; 8School of Public Health and Preventive Medicine, Monash University, Melbourne, Victoria; 9Departments of Medicine & Community Health Sciences, University of Calgary, Calgary, Canada; 10Division of Nephrology, W.G. (Bill) Hefner Veterans Affairs Medical Center, Salisbury, North Carolina, United States; 11Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium; 12Peking University People’s Hospital, Beijing, China; 13School of Medicine and Pharmacology, University of Western Australia, Perth, Western Australia
Aim: To assess the scope and consistency of vascular access (VA) outcomes reported in haemodialysis trials.
Background: Many randomised controlled trials have been performed to improve VA outcomes. These trials can inform decision-making, provided the reported outcomes are relevant and measured consistently to allow interventions to be compared across trials.
Methods: All randomised controlled trials and trial protocols reporting any VA outcome in adult haemodialysis patients from January 2011 to June 2016 were identified from clinicaltrials.gov, Cochrane CENTRAL, Embase, and MEDLINE. The frequency and characteristics of VA outcomes were analysed and classified.
Results: From 168 relevant trials, 1426 access-related outcome measures were extracted and classified into 23 different outcomes. The three most commonly reported outcomes were function (136 [81%] trials), infection (63 [38%]), and maturation (31 [18%]). Function was measured in 489 different ways and at 46 different time points, but most frequently reported as “mean access blood flow (mL/min)” (37 [27%] trials) and “number of thromboses” (30 [22%]). Infection was assessed in 136 different ways with “number of access-related infections” being the most frequently used measure. Maturation was assessed in 44 different ways, at 15 different time points, and most commonly characterised by vein diameter and blood flow. Patient-reported outcomes, including pain (19 [11%]) and quality of life (5 [3%]), were reported infrequently. Only a minority of trials used previously standardised outcome definitions.
Conclusions: The reporting of access outcomes in haemodialysis trials is highly heterogeneous with limited patient-reported outcomes and infrequent use of standardised outcome measures. Efforts to standardise outcome reporting for VA are critical to optimising the comparability, reliability and value of trial evidence to improve outcomes for patients requiring haemodialysis.