A total of 42 studies (n=4,064, as reported in the text) were included: 21 studies of endovascular revascularisation (n=1,750) and 21 studies of surgical procedures (n=2,314). All studies were either prospective or retrospective observational studies. The number of patients ranged from 12 to 500 for the surgical studies and 18 to 300 for the endovascular studies. Mean length of follow-up was 15 months in the endovascular studies and 42 months in the surgical studies.
Hypertension: There was a 21% higher rate of improvement in hypertension in the surgical procedures group compared to the endovascular procedures group (95% CI 9% to 33%). The difference in post-procedural mean arterial pressure was 3.1mmHg (95% CI -1.0 to 7.3) lower in the surgical group than the endovascular procedures group.
Renal function: There was a 34% higher rate of improvement in renal function in the surgical procedures group compared to the endovascular procedures group (95% CI 1.8% to 54%) and a significant 8.8% lower rate of deterioration in renal function in surgical procedures (95% CI 2.6 to 14.9).
Mortality: Surgical treatments had a 3.1% excess 30-day procedure-related mortality compared with endovascular procedures (95% CI 1.8 to 4.4) for all revascularisations. When the type of revascularisation was analysed further, excess mortality with surgical procedures was found to be related to combined procedures (6.5% excess mortality, 95% CI 3.8 to 9.3) and was not apparent with isolated renal artery surgical procedures.
Technical success: Technical success was not statistically different between surgical and endovascular procedures.
There were insufficient data to undertake patency and survival analyses. Other outcomes were presented in the paper.