Meta-Analysis: Low-dose dopamine Increases urine output but does not prevent renal dysfunction or death Annals of Internal Medicine 2005; 142: 510-524 Issaam Oozeerally Dopamine • Catecholamine • Dose dependent effects on systemic and renal vasculature • At low doses increases renal blood flow and promotes natriuresis [D1, D2, D4 receptors] • 1st clinical use in heart failure Purpose • Evaluate effects of low-dose dopamine c.f placebo /no therapy in patients with or at risk of ARF Methods: Search strategy • MEDLINE – 1966 –January 2005 • EMBASE – 1980- January 2005 • CINAHL – 1982 – January 2005 • CANCERLIT – 1975 – Oct 2002 • CENTRAL – 4th quarter 2004 • Renal Health library Search: MEDLINE 1. Dopamine/low dose dopamine/renal dose dopamine – Limited to clinical trial and meta-analysis MEDLINE Search no. 2 1 [Dopamine] 2 [Renal disease] 3 [physiology] Low Kidney Diseases Kidney Function Tests Renal Kidney Urine Kidney Renal circulation Dose Limited to RCTs [maximally sensitive strategy] Rest of the search • Modification of search from other databases – Strategy not specified – Authors happy to be contacted • Screened reference lists from articles against recent review articles to identify additional studies • No language restrictions Selection • • • • RCTs/quasi-randomization trials Any sample size Low dose dopamine vs. placebo/nil Outcomes: – – – – All cause mortality Requirement for RRT Renal physiological variables [UO, Creat, CrCl day 1, 2, 3] Adverse effects • Studies with pharmacologic co-interventions [e.g. mannitol, diuretics] • A priori adverse effects: – Ischaemia [myocardial, limb, cutaneous] or arrhythmias Data Abstraction • 2 independent reviewers • Study authors contacted if any disagreement – Consensus if persisted • Agreement for inclusion studies statistically tested [Cohen’s κ] Validity assessment • Individual studies methodology looked into – Randomization – Blinding – Outcome assessors – Reasons for withdrawals • Fluids and diuretic therapies – Standard or equally applied in both arms • Attempted to contact all authors of selected trials Exclusion Criteria – 70 studies • • • • • • • • • Small sample [3 patients] Dopamine dose >5mcg Not randomized cross-over study Combined intervention c.f control Duplicate Wrong topic Editorial 4 RCTs without group data available from authors No additional data provided Hence • 3359 patients identified in 61 trials Data Analysis • Pooled by outcome – E.g. mortality, RRT • If different doses of dopamine used data was combined • Random effects model used • Binary outcomes [RRT, mortality, adverse effects] reported as relative risk • Summary of relative risk : log scale – Used to calculate weight of study Data Analysis [cont] • If heterogeneity between studies; weight was adjusted • Clinical outcomes occurred infrequently and different statistical tests i.e. effect measures were undertaken – Similar results [not presented] Data Analysis [cont.] • Renal physiologic outcomes: – Relative change in dopamine gp c.f. control • Mean values were taken • Lack of standardized data e.g. weight Data Analysis [cont] • Between study homogeneity for each pool – [concept when there is more variation than chance alone] – Calculated statistically [I2 and Cochran Q-test] Patient profile • • • • • • • Cardiac surgery Vascular surgery Other surgery Iv Contrast dye Nephrotoxics Neonates Miscellaneous Data • Average numbers per trial 40 [12 to 347] • ANZICS trial included [328 patients; multicentre] • Only 6 trials used dopamine therapeutically: – Critical illness, contrast, malaria, CHF, preeclampsia Data • • • • Median dopamine dose 2.5 mcg 31 hours median [0.4 to 192 hours] 12 trials randomization not reported 11 trials fluids up to the clinician Clinical outcomes • No effect on mortality [0.96] or need for RRT [0.93] – No statistical evidence of heterogeneity • ANZICS trial and most heavily weighted trial removed and data analysed again – No change Mortality RRT Adverse effects • Arrhythmias & ischaemia [cutaneous, myocardial, limb] • 6 MI’s [4 on dopamine] • Statistically not significant Renal Physiologic Outcomes • Statistically significant increase in urine output on day 1 • Decrease in creat and increase in creat clearance on day 1 • Substantial heterogeneity Conclusion • 24% increase in urine output on day 1 – Probably explains continued popularity • ANZICS trial – same results • ANZICS trial – removal of data • Adverse effects – under reported Discussion • Methodology – Search strategies not specified – Age and sample size – Combining dopamine doses • Stats – If no events arbitrary figure of 0.5 given – Study weight adjusted in presence of heterogeneity – Use of relative risk