Ann Thorac Surg 2002;73:153-155
© 2002 The Society of Thoracic Surgeons
Original article: cardiovascular
Evaluating routine diuretics after coronary surgery: a prospective randomized controlled trial
Eric Lim, MRCSa,
Ziad A. Ali, MB, ChB (Hons)a,
Robert Attaran, MB, ChBa,
Graham Cooper, MD, FRCS (C/Th)*a
a Department of Cardiothoracic Surgery, Northern General Hospital, Sheffield, United Kingdom
Accepted for publication August 31, 2001.
* Address reprint requests to Dr Cooper, Department of Cardiothoracic Surgery, Northern General Hospital, Herries Rd, Sheffield S5 7AU, United Kingdom
e-mail: graham.cooper{at}sth.nhs.uk
 |
Abstract
|
|---|
Background. Routine diuretic administration after cardiopulmonary bypass is common despite the lack of evidence for its benefit. We performed a prospective study to evaluate if diuretics assist in weight loss or alter clinical outcome.
Methods. Seventy-nine patients undergoing primary elective coronary bypass surgery were randomized to either diuretic (fusemide and amiloride) or placebo (lactose) postoperatively until preoperative weight was achieved or for 5 days in total. Proportions were compared with 
2 or Fisher’s exact test.
Results. Forty patients were randomized to diuretics and 39 to placebo. By day 5, 97% of patients (37of 38) still in the diuretic arm, and 74% of patients (29 of 39) in the placebo arm, achieved preoperative weight (p = 0.02). There were no differences in clinical outcomes.
Conclusions. Routine diuretics promote an earlier diuresis but no clinical benefits are apparent in low risk patients with normal renal function. Clinicians should reconsider routine diuretic prescription in this setting.
|
Introduction
|
|---|
Routine diuretic administration is common after cardiopulmonary bypass, and it is recommended by standard texts for cardiac operations [1], [2]. The beneficial effects of diuretics are thought to be so obvious that trials have been unnecessary [3]; this is reflected by the lack of evidence for their routine use [4]. We have examined this in a prospective, double blind, placebo controlled trial.
|
Material and methods
|
|---|
Study population
Consecutive patients undergoing isolated, primary, elective coronary artery bypass operations on cardiopulmonary bypass were invited to participate. Exclusion criteria were preoperative diuretic use, contraindication to diuretics, possible drug interaction, elevated preoperative urea (more than 6.6 mmol/L) or creatinine (more than 109 mmol/L), and intolerance to lactose, which was the predominant constituent of the placebo. Patients were withdrawn if any of the following occurred: cardiac failure, renal failure, stroke, low output state, myocardial infarction, or cardiac arrest.
All patients gave written, informed consent. The North Sheffield Local Research Ethics Committee approved the protocol and permission to encapsulate treatment medication that was granted by the Department of Health Medicines Control Agency (United Kingdom).
Study design
The sample size estimation for our study was based on the assumption that 90% of patients on diuretics would achieve preoperative weight within 5 days. With a significance level (
) of 0.05 and statistical power (1-ß) of 0.9, the sample size required to detect an absolute difference of 15% between the two treatment arms would be 70 patients.
Enrolled patients were randomized into blocks of four, comprising two allocations to diuretics and two to placebos. Treatment allocation blocks were distributed among 5 operators according to caseload.
Study medication
Treatment was administered daily from the first day and continued until either preoperative weight was achieved or a total of 5 days was completed. Patients were randomized to receive either diuretics (frusemide, 40 mg and amiloride, 5 mg) or placebo (lactose tablets). Both medications were identically encapsulated. No restrictions were placed on administration of extra doses of frusemide for patients in either arm.
Study conduct
From the second postoperative day patients were weighed at 8 AM daily until discharge on standardized Avery 824/890 electronic scales (Walsall, West Midlands, UK) that were calibrated before commencement of the trial. When preoperative weight was achieved, study medication was discontinued. Upon completion of the study at 5 days, the decision to initiate further diuretic therapy was left to the discretion of the attending clinicians. Daily electrolyte, urea, and creatinine measurements were performed. Also recorded were frequency of atrial fibrillation, total supplementary diuretics use, renal dysfunction (defined as serum creatinine more than 177 µmol/L [5], low urine output (an output of less than 0.5 ml/kg quantified over the period of 1 hour), daily oxygen saturation, supplementary diuretic usage, patient withdrawals, patients requiring diuretics after 5 days, time to discharge, and mortality.
Statistical methods
Descriptive variables are presented as means with standard deviations, or medians with interquartile ranges, and compared with the t test or Mann-Whitney U test, respectively. Estimations of precision are presented with 95% confidence intervals. Categorical data and frequencies are expressed as percentage and compared with 2 or Fischer’s exact test where appropriate. Conventional levels of significance ( of 0.05) are applied throughout, except for the tests of significance applied to replicate readings when the Bonferroni correction was used for multiple testing and an adjusted significance level of 0.01 was used. Statistical analysis was undertaken using GraphPad Prism (GraphPad Software, Inc, San Diego, CA) and SPSS for Windows, Version 10.0 (SPSS Inc, Chicago, IL). Data were analyzed by both intention-to-treat and a case complete analysis. Data collectors were blind to the assignment of patients, and the results were compared and analyzed in two groups labeled A and B. Upon completion of all statistical analyses, the allocation of groups was revealed.
|
Results
|
|---|
Of 105 eligible patients, 79 were randomized (Fig 1).
There were no statistical significant differences in the preoperative or operative characteristics of the randomized patients (Table 1). All patients had urea and electrolytes within the normal range.
Primary outcome measures
Two patients in the diuretic arm were too unwell to be weighed; 1 patient developed renal failure and the other cardiac failure. Both were excluded from analysis. After 5 days, 97% of patients (37of 38) originally assigned to the diuretic arm and 74% of patients (29 of 39) originally assigned to the placebo arm achieved preoperative weight (p = 0.02), a difference of 23% (95% Confidence Limits, 7.5, 38.6).
On complete case analysis, 97% of patients (34 of 35) in the diuretic arm and 71% of patients (25 of 35) in the placebo arm who completed the assigned treatment achieved preoperative weight by 5 days (p = 0.02). A difference of 26% (95% Confidence Limits, 8.8, 42.4) between the two treatment arms. For the 69 patients that achieved preoperative weight before discharge, the median time to achieve preoperative weight was 2.5 days (interquartile ranges, 2, 4) in the diuretic arm and 3 days (interquartile ranges, 2, 4) in the placebo arm (p = 0.44). This included all patients in the diuretic group and 79% of patients (31of 39) in the placebo group (p = 0.008).
Secondary outcome measures
There were no deaths in either group and no significant difference in other outcome measures (Table 2).
There were no differences in daily oxygen saturations between the groups. The median amount of total diuretic use per patient in the diuretic arm was 240 mg of frusemide and 30 mg of amiloride. The median amount of supplementary diuretic use per patient was 20 mg of frusemide in both arms. A higher proportion of patients in the diuretic arm required supplementary diuretics for oliguria (95% [38 of 40] compared with 79% [31 of 39], p = 0.05).
Daily sodium, potassium and urea measurements remained within the normal range for both groups. Two patients (5%) in the diuretic group had a rise in creatinine to above 177 mmol/L (p = not significant compared to placebo). Upon completion of the trial, 3 patients in the diuretic arm and 6 patients in the placebo arm were commenced on further diuretics (p = 0.31).
|
Comment
|
|---|
Diuretics are usually prescribed to augment fluid losses after cardiac operations [2]; this may have originated from early reports of renal dysfunction with progressive hypervolemia resulting in cardiac failure and pulmonary edema after cardiopulmonary bypass [6, 7]. Strategies of salt and fluid restriction accompanied by obligatory diuresis have been recommended since [8]. A combination of frusemide and amiloride are most commonly used.
Previous studies report significant improvements in both diuresis and creatinine clearance with diuretics [9, 10]. Our study confirms the earlier diuresis in patients given diuretics, but failed to demonstrate any clinical benefit associated with this. Indeed more patients in the diuretic group developed oliguria, and the 2 patients whose creatinine rose above 177 mmol/L were in the diuretic group. This suggests that the forced diuresis obtained with diuretics nay have a detrimental effect on renal function. Reports of renal protection against ischemic injury conferred by frusemide in animals have not been confirmed in humans [11]. In fact there is increasing evidence for the deleterious effect of frusemide on renal function after cardiac operations [12, 13], and a higher rate of frusemide administration has been shown in patients who develop moderate to severe renal failure after cardiac operations [14].
In conclusion, low-risk patients undergoing coronary artery bypass are given a routine prescription of diuretics that promote an earlier diuresis, which is not associated with clinical benefit and may promote renal dysfunction. Clinicians should reconsider routine diuretic prescriptions in this setting.
|
Acknowledgments
|
|---|
Grant support was provided by the Northern General Heart Research Fund. We would like to gratefully acknowledge the help of Mr Casula, Mr Locke, Mr Sharpe, and Mr Wilkinson for allowing us to randomize their patients.
|
References
|
|---|
-
Sabiston D.C., Spencer F. Surgery of the chest. Philadelphia: WB Saunders, 1995.
-
Bojar R.M. Manual of perioperative care in cardiac surgery. Boston: Blackwell Science, 1999.
-
Maseri A. Evidence-based medicine: progress but not a final solution. Clin Cardiol 1998;21:463-464.[Medline]
-
Vanky F., Broquist M., Svedjeholm R. Addition of a thiazide: an effective remedy for Furosemide resistance after cardiac operations. Ann Thorac Surg 1997;63:993-997.[Abstract/Free Full Text]
-
Mangano C.M., Diamondstone L.S., Ramsay J.G., Aggarwal A., Herskowitz A., Mangano D.T. Renal dysfunction after myocardial revascularization: risk factors, adverse outcomes and hospital resource utilization. Ann Intern Med 1998;128:194-203.[Abstract/Free Full Text]
-
Myers B.D., Moran S.M. Hemodynamically mediated acute renal failure. New Engl J Med 1986;314:97-105.[Medline]
-
Neville W.E., Thomason R.D., Hirsch D.M. Post perfusion hypervolemia after hemodilution cardiopulmonary bypass. Ann Thorac Surg 1966;93:715-723.
-
Clarke A.D., Jackson P.W. Postoperative care of patients undergoing cardiopulmonary bypass. Br J Anaesth 1971;43:248-260.[Free Full Text]
-
Copeland J.G., Campbell D.W., Plachetka J.R., Salomon N.W., Larson D.F. Diuresis with continuous infusion of furosemide after cardiac surgery. Am J Surg 1983;146:796-799.[Medline]
-
Nuutien L.S., Kairaluoma M., Tuononen S., Larini T.K. The effect of furosemide on renal function in open-heart surgery. J Cardiovasc Surg 1978;19:471-479.[Medline]
This article has been cited by other articles:
|
|
|
|
 
B. E Gulbis and A. P Spencer
Efficacy and Safety of a Furosemide Continuous Infusion Following Cardiac Surgery
Ann. Pharmacother.,
October 1, 2006;
40(10):
1797 - 1803.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Abu-Omar and C. Ratnatunga
Cardiopulmonary Bypass and Renal Injury
Perfusion,
July 1, 2006;
21(4):
209 - 213.
[Abstract]
[PDF]
|
|
|
Top
Abstract
Introduction
