Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Use of dopamine in acute renal failure: a meta-analysis

Kellum JA, M Decker J.

Crit Care Med. 2001;29(8):1526-31. [abstract]

Reviewed by Caroline Davision, MBBch, FRCA, Boston Children's Hospital

Review posted December 17, 2001

---

I. Are the results of the study valid?

A. Primary questions:

1. Did the overview address a focused clinical question?

No. The authors attempted to determine if the use of low dose dopamine has an impact on the prevention, development, and course of acute renal failure (ARF), mortality, and hemodialysis requirements in critically ill patients. However, this was too broad a question as there was no definition of critical illness, no distinction between prophylactic use of dopamine in patients with normal renal function on admission and patients with established renal failure. The analysis included a heterogeneous group of patients including non-septic patients, post-transplant patients, cardiac patients and patients exposed to contrast dye.

2. Were the criteria used to select articles for inclusion appropriate?

No. The inclusion criteria:

• The initial search was not restricted to randomized control trials
• There was no definition of critical illness; were they even patients in an ICU?
• They did define low dose dopamine as < 5 mcg/kg/min.
• There was no standard definition for development of ARF.
• Hemodialysis was taken to mean any form of renal support.

The outcome data:
To be included in the meta-analysis studies had to include outcome data on at least one of the following outcomes; mortality, requirement for hemodialysis, development of ARF, worsening of existing ARF.

B. Secondary questions:

3. Is it unlikely that important, relevant studies were missed?

Maybe. The authors did conduct a MEDLINE search, however, we are not told what search terms were used. They did hand search bibliographies of these and other relevant articles. They did not restrict themselves to English language. They did not attempt to make personal contacts with authors or experts in the field to search for unpublished studies, and they did not search the Cochrane database. Thus there is some concern that the studies selected may lead to bias.

Only 58 studies were identified, of these outcome data was reported in 24.

4. Was the validity of the included studies appraised?

Unclear. They did grade the study designs from I to V, however, the only thing we are told about the study design is that level V is a case series with no control group. Thus we are unable to tell if the studies included for meta-analysis have good methodology or not. They identified 17 "randomized trials" and restricted their analysis to these studies (total n=854).

5. Were assessments of studies reproducible?

Unclear We are told that the 2 authors reviewed each study independently. We are not given any information as to whether there was good agreement between the reviewers, and how any disagreements were resolved.

6. Were the results similar from study to study?

We are told there were no statistically significant differences between groups for any of the outcomes analyzed, but they do not specifically refer to a test for between-study heterogeneity. However, we are not provided with the results and confidence intervals for the individual studies so it is difficult to interpret.

The authors did determine 3 subgroups presumably to try and avoid heterogeneity. Two groups were defined a priori, the first group included patients receiving contrast dye, the second group was limited to those studies enrolling patients with heart disease (either medical or surgical patients). The third subgroup was defined post hoc and excluded studies in which the control group event rate or the effect size for each outcome was significantly different from the mean (i.e., significant outliers) as determined by analysis of variance. The authors used a fixed effect model for their meta-analysis; as there may have been a lot of heterogeneity between the study population, perhaps a random effect model would have been more appropriate.

II. What are the results?

1. What are the overall results of the review?

Mortality was reported for 508 patients enrolled in 11 trials and was similar between treatment and control groups (4.7% vs. 5.6%; RR, 0.83, 95% confidence intervals 0.39, 1.77).

Development of ARF was reported to be 15.3% in the dopamine arms and 19.5% in the control arms (RR, 0.79 95% CI 0.54, 1.13) for 511 patients in 11 trials.

Requirement for hemodialysis was reported to be 13.9% in the dopamine arms vs. 16.5% in control arms (RR, 0.89 95% CI 0.66, 1.21) for 618 patients in 10 studies.

Worsening of existing ARF was only reported in one trial and was not analyzed further.

Overall the results suggest there was no differences between dopamine and control with respect to any outcome for any of the subgroups analyzed.

2. How precise were the results?

The confidence intervals for all outcomes and all groups analyzed crossed 1, thus there was no statistically significant benefit or harm from treatment with dopamine.

The authors comment that their analysis had 85% power to detect a RR of 0.5 in prevention of ARF. So it is possible that a type II error was committed in concluding that a RR of 0.79 was not significant. This could represent a 21% relative risk reduction, which most practitioners might well find clinically significant.

III. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

The study population was presumably adult patients and my patients are pediatric intensive care patients. One must also realize these were a heterogeneous group of patients and that an analysis specifically looking at patients with sepsis (or other specific subgroups), either as prophylaxis or for established renal impairment, would be more informative.

However, I believe that the results of this meta-analysis are likely to be relevant to my population, and thus the use of renal dose dopamine on pediatric critical care patients, with the associated cost and potential for adverse effects is not justified. Further studies on the pediatric population may be justified.

2. Were all clinically important outcomes considered?

Mortality, development of acute renal failure and need for renal replacement therapy were all included. However, there was no discussion on long term follow up and persistence to chronic renal failure.

3. Are the benefits worth the harms and costs?

No. This meta-analysis agrees with the findings of a previous meta-analysis (1) and a recent large multi-centered RCT (2), specifically looking at patients with SIRS, that there is no evidence to support the use of dopamine in critically ill patients to try to prevent the onset of ARF. There are many potential side affects from the use of dopamine, including arrhythmias, gut ischemia, gastrointestinal ileus, potential interaction with the endocrine system and decreased respiratory drive. Thus with no evidence to support its use, renal dose dopamine should no longer be used.

References

1. Kellum JA. The use of diuretics and dopamine in acute renal failure: A systematic review of the evidence. Critical Care 1997; 1:53-59. [abstract]
2. Bellomo R, Chapman M, Finfer S, Hickling K, Myburgh J. Low-dose dopamine in patients with early renal dysfunction: a placebo-controlled randomised trial. Australian and New Zealand Intensive Care Society (ANZICS) Clinical Trials Group. Lancet. 2000;356(9248):2139-43. [abstract]

Comments

Back to the EB Journal Club Index