Journal of Cardiac Failure Vol. 13 No.

2 2007

Fluid Restriction in the Management of Decompensated Heart

Failure: No Impact on Time to Clinical Stability
BRONAGH TRAVERS, RN, CHRISTINA O’LOUGHLIN, PhD, NIAMH F. MURPHY, MB, BCh, MARY RYDER, RN,
CARMEL CONLON, RN, MARK LEDWIDGE, PhD, AND KENNETH MCDONALD, MD

Dublin, Ireland

ABSTRACT
Background: To examine the clinical effect of fluid restriction in patients admitted to the hospital with
class IV heart failure (HF).
Methods and Results: This is a single-blind randomized controlled study. Time to clinical stability was
compared between the fluid restricted (FR: n 5 34) and free fluid (FF: n 5 33) groups respectively show-
ing no significant difference (8.3 6 6.3 days versus 7.0 6 6.0 days, P 5 .17). There was no significant
difference between groups in time to discontinuation of intravenous diuretic therapy (FR: 2.7 6 4.5
days, FF: 3.2 6 5.6 days, P 5 .70). Changes from baseline to achievement of clinical stability in serum
urea (P 5 .23), serum creatinine (P 5 .14), BNP (P 5 .42), and sodium (P 5 .14) did not differ between
the FF and FR groups. Baseline serum sodium levels did not predict the time to clinical stability (b 5
0.11, 95% CI: 0.60, 0.23).
Conclusions: Fluid restriction is not an evidence-based therapy although it is occasionally applied in the
management of HF. These results suggest that FR is not of any clinical benefit in patients with acute de-
compensated HF and this hypothesis should be tested in a larger randomized controlled study. (J Cardiac
Fail 2007;13:128e132)
Key Words: Acute decompensated heart failure, Management, Fluid restriction.

The aim of treatment of acute decompensated heart fail- Methods

ure (HF) is to promptly achieve hemodynamic stability and
symptomatic improvement.1 In addition to a standard phar-
macologic approach, fluid restriction is frequently applied
to expedite recovery.2 However, to date there are no data
to support the use of this strategy. Indeed, recent guidelines
for the management of acute decompensated HF do not rec-
ommend the application of this intervention.1 Furthermore,
adherence to fluid restriction can be difficult for patients3
and should only be employed if benefit is likely.
The purpose of this study was to assess the impact of
fluid restriction on the clinical course of patients admitted
to hospital with New York Heart Association functional
class IV HF.
From the Heart Failure Unit, St. Vincent’s University Hospital, Dublin,
Ireland.
Manuscript received May 8, 2006; revised manuscript received Septem-
ber 21, 2006; revised manuscript accepted October 10, 2006.
Reprint requests: Dr. Kenneth McDonald, Department of Cardiology,
St. Vincent’s University Hospital, Dublin 4, Ireland.
1071-9164/$ - see front matter
Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.cardfail.2006.10.012
Study Design
This was a single-blind, randomized, controlled, proof of con-
cept study conducted at St. Vincent’s University Hospital. Ethical
approval was granted by the Medical Ethics and Research Board
and the principles of the study adhered to those outlined in the
Declaration of Helsinki.
Selection of Patients
Patients admitted age 18 years of age or older with a confirmed
diagnosis of class IV HF and giving informed consent were eligi-
ble for the study. The diagnosis of HF was based on presenting
symptoms, physical examination, radiologic evidence of HF on
chest x-ray and was confirmed by the HF team.
Exclusion criteria were patients with serum creatinine on ad-
mission greater than 2 times the upper limit of normal (ie, 230
mmol/L [mg/dL]), with life-threatening illness that could compro-
mise survival and patients considered unlikely to comply with
fluid restriction such as those with a history of cognitive impair-
ment or dementia.
Patients were consented within 24 hours of admission and ran-
domized by computer-generated protocol to either fluid restriction
(FR) of 1 L per day or free fluid (FF). Members of the HF team
with the exception of the HF nurse were blind to the

128
 Fluid Restriction in Decompensated Heart Failure  Travers et al 129

randomization process. Patients, family members, nurses, and al- duration). Data were analyzed on an intention to treat basis; how-
lied health care professionals were asked not to reveal patients ever, the effects of patient withdrawal was assessed for all major out-
randomization status to the attending HF physician for the dura- comes and reported if effect was statistically significant.
tion of the study.
Results
Study Withdrawal
Demographics
Patients were withdrawn from the study in the following cir-
cumstances: an increase in serum creatinine of greater than 25% The baseline demographics for the FF (n 5 33) and FR
above baseline levels for 2 consecutive days; nonadherence to (n 5 34) groups are presented in Table 1. The patient pop-
fluid restriction as per protocol for 2 consecutive days; or with- ulation are representative of a typical community HF pop-
drawal at patients own request. ulation, being older (74 6 12 years) than most represented in
clinical trials.5,6 Statistical comparisons between the groups
Study End Points
baseline characteristics were made to ensure that randomiza-
The primary study end point was time in days to clinical stabil- tion eliminated any potential confounder variables. Signif-
ity defined by achieving all of the following parameters: symptom- icantly fewer patients were hypertensive in the FR group
atic improvement with no evidence of fluid overload; stable (P 5 .01). No statistical difference between the groups’
weight for 48 hours (no change greater than 1 kg); off intravenous baseline biochemical marker levels was observed (all
(IV) HF therapies for 48 hours; and no change in cardiac medica-
P O .05).
tion for 48 hours.4 The secondary end points for this study were as
follows: changes in renal parameters; changes in B-type natri- Time to Clinical Stability
uretic peptide (BNP) levels; duration of IV HF therapy; compli-
ance with fluid restriction protocol. There was no significant difference in time to clinical sta-
bility between the FF and FR groups, respectively (7.0 6 6.0
Assessment of Biochemical and Clinical Parameters days versus 8.3 6 6.3 days, P 5 .18) even when controlled
In addition to routine clinical assessments, the renal profile was for age and gender effects (Fig. 1). The time to clinical
measured daily for the duration of the intervention. BNP was as- stability in either arm was not influenced by baseline serum
sayed for the first 7 days and alternate days thereafter until clinical sodium levels (b 5 0.11, 95% CI: 0.601, 0.234), even
stability was achieved. Weight was recorded each morning (after when adjusted for treatment effects. No etiology factors
voiding, before breakfast, and before diuretic administration). were predictive of time to clinical stability nor did they ap-
Management of the patient’s HF status was reviewed on a daily pear to influence the model coefficients when introduced
basis by the HF physician and medications, including the contin- into the model.
uation of IV HF therapy, were adjusted accordingly. After the pri-
mary end point of clinical stability was achieved, patients reverted Duration of IV HF Therapy
to FF status.
There was no significant difference between FF and FR
Compliance with Fluid Restriction groups respectively in the number of days to discontinua-
tion of IV HF therapy (3.2 6 5.6 days versus 2.7 6 4.5
Patients who exceeded 200 mL above the recommended daily days, P 5 .70; Fig. 2). Furthermore, the average daily IV
fluid allowance for 2 consecutive days were deemed noncompli-
ant. Daily fluid intake was managed and recorded by the patient Table 1. Demographics of the Free-Fluid
with the aid of the primary nurse and monitored by the HF nurse and Fluid-Restricted Samples
for both groups. Patients reporting difficulty in adhering to the
Free Fluid Fluid Restricted
fluid restriction protocol were counseled by the HF nurse.
Baseline Characteristics n (%)/Mean 6 SD
Statistical Analysis n 33 34
Age (y) 73 6 13 75 6 11
Data are presented as the mean 6 standard deviation (SD) for Gender: Male 16 (49) 20 (59)
continuous variables, whereas the frequencies and percentages are Weight (kg) 72.1 6 20.5 76.2 6 16.2
used to summarize categorical (discrete) variables. Comparisons Etiology
between the FR and the FF groups were made using independent Ischemic 19 (59) 25 (76)
EF: !45% 19 (61) 20 (67)
sample t-tests for continuous normally distributed data and Mann- Mean EF 40.2 6 14.8 37.4 6 11.8
Whitney for nonnormal distributions (two-sided a 5 0.05). Treat- Diabetic 7 (21) 8 (23)
ment effects on outcome measures were assessed using repeated Raised cholesterol 10 (30) 16 (47)
measures analysis of variance. Linear regression was used to iden- Hypertensive 14 (42) 5 (15)
Pulmonary disease 6 (18) 4 (12)
tify the independent predictors of time to clinical stability and
Medications
time to discontinuation of IV HF treatments when adjusted for ACE inhibitor/ARB 23 (70) 23 (68)
age and gender. Because of the small sample size, it was not feasible b-Blocker 8 (24) 10 (29)
to adjust for all baseline demographic and etiology factors; however, Diuretic 6 (18) 2 (6)
each were introduced singularly into the final model to ensure that Nitrate 22 (67) 20 (59)
Statin 12 (36) 12 (35)
they did not effect the final parameter coefficients. Each patient’s
average compliance was calculated over the duration of the inter- EF, ejection fraction; ACE, angiotensin converting enzyme; ARB,
vention (ie, the proportion of compliant days to total intervention angiotensin receptor blocker.
130 Journal of Cardiac Failure Vol. 13 No. 2 March 2007

12 changes in these markers (Table 2). No significant change

P = 0.176
was observed in creatinine (P 5 .35) and sodium (P 5 .60).
At clinical stability there was no significant difference
between groups in levels of serum sodium (FF: 137.6 6
Time to Clinical Stability (days)

9 3.2 mmol/L versus FR: 137.8 6 3.8 mmol/L, P 5 .84), se-

rum urea (FF: 8.4 6 3.1 mmol/L vs FR: 10.8 6 5.5 mmol/
L, P 5 .10), serum creatinine (FF: 111.7 6 30.5 mmol/L
[mg/dL] versus FR: 128.2 6 37.5 mmol/L [mg/dL], P 5
6
.06), and BNP (FF: 560.9 6 439.2 pg/mL versus FR:
602.0 6 644.8 pg/mL, P 5 .94).

Compliance With Fluid Restriction

3 The total daily volume of fluid consumed was signifi-

0 tween the groups (FF: 1625.4 6 748.4 mL, FR: 1971.5 6
FF
Error Bars show Mean +/- 1.0 SE
Fig. 1. Time to clinical stability (days) in free-fluid and fluid-
restricted groups.
diuretic dose in the FF and FR groups did not significantly
differ (73.6 6 30.4 mg versus 75.9 6 18.0 mg, P 5 .22). In
addition the average daily dose of oral diuretics was not sig-
nificantly different between the FF and FR groups, respec-
tively (52.7 6 23.2 mg versus 61.8 6 27.1 mg, P 5 .24).
Biochemical Markers
cantly different between both groups (1466.6 6 607.3 mL
versus 1074.3 6 319.2 mL; P 5 .001; Fig. 3). In addition,
the total daily fluid output did not differ significantly be-
FR 721.3 mL, P 5 .25). There was no significant difference
in the average weight loss in both groups from baseline
to clinical stability (FF: 3.0 6 3.6 kg, FR: 2.6 6 3.0 kg,
P 5 .65). No patients were withdrawn from the FR because
of failure to comply. The average number of compliant
days in the FR arm was 73% (ranging from 40% to
100%). Fourteen patients (21%) were withdrawn because
of increased creatinine levels, 4 (12%) from FF and 10
(29%) from FR group.
Discussion

Using repeated measures analysis of variance, an in-
crease in urea (P ! .01) and a reduction in BNP (P !
.01) was observed in the sample. There was no significant
interaction effect between the randomization arm and
5
P = 0.70
4 situations.7
Duration of IV Therapy (days)
The results of this study question the routine application
of fluid restriction in the management of acute decompen-
sated HF. Commencing within 24 hours of admission this
intervention did not reduce the time to clinical stability.
Furthermore, fluid restriction did not reduce the duration
of IV HF therapy, the average dose of diuretic used or alter
the reduction of BNP. Finally, there were more withdrawals
due to elevated creatinine in the FR group, which may
suggest an adverse impact of this intervention in certain

Fluid restriction may seem a logical intervention in the

management of HF, although there are no scientific data
3 to support its application. Its uncertain status as a therapeu-
tic strategy in HF is reflected in the recently published
guidelines for HF care, where its use is recommended in
2
chronic HF as a nonpharmacologic approach,8 but not dis-
cussed in new guidelines for the management of acute de-
compensated HF.1 Indeed, it could be argued that fluid
restriction in the setting of IVs diuretic use could in certain
1
circumstances predispose to an adverse outcome. This
could be explained by accentuating the impact of fluid
loss with the resultant heightened activation of the renin-
0 angiotensin-aldosterone system predisposing to renal
FF FR dysfunction.9 Therefore, as a practice that prevails in med-
Error Bars show Mean +/- 1.0 SE
icine it is clear that proof of impact is needed to support or
Fig. 2. Duration of intravenous therapy (days) in free-fluid and refute its continued use. Holst et al have described the de-
fluid-restricted groups. sign of a prospective randomized study examining the
 Fluid Restriction in Decompensated Heart Failure  Travers et al 131

Table 2. Differences Between Free-Fluid and Fluid-Restricted Groups in Biochemical Marker Levels
From Baseline to Clinical Stability
Median (Range)
Free Fluid Fluid Restriction
Baseline Stability Baseline Stability P*
Sodium (mmol/L) 138 (133e144) 138 (130e143) 138 (123e145) 139 (128e146) .14
Urea (mmol/L) 6.5 (3e17) 8 (3e17) 7 (3e16) 9 (4e24) .23
Creatinine (mmol/L)(mg/dL) 105.5 (62e177) 104.0 (66e186) 111 (63e205) 125 (65e213) .14
Brain natriuretic peptide (pg/mL) 608 (148e3240) 479.5 (24e1520) 853 (217e2310) 346 (37e2960) .42

*The P value relates to the interaction effect between randomization arm and changes in the biochemical marker from baseline to clinical stability using
repeated measures analysis of variance. The removal of withdrawn individuals from the analysis did not significantly change the P values for any marker.

benefits of fluid restriction in the outpatient management of
HF,10 although the results have not been published. Another
study examining the impact of a nutritional intervention on
chronic HF could not attribute the observed benefit specif-
ically to fluid restriction.11 The results presented here are
the first published data addressing the role of fluid
restriction in acute decompensated HF.
Though arguing against routine application of fluid re-
striction, the presented data do not necessarily exclude
a role for this intervention in certain subsets of patients. Hy-
ponatremia is a poor prognostic indicator in HF12 and has
been routinely suggested as an appropriate indication for
fluid restriction.12,13 In this study, admission serum sodium
did not influence the time to clinical stability or response to
fluid restriction. However, this study did not have sufficient
patients with low serum sodium levels to adequately ad-
dress this issue, because the mean sodium level on admis-
sion was 138 6 4.0 mmol/L with only 3% having
a sodium level lower than 130 mmol/L. Furthermore, this
1800
P < 0.01
1500
Average Fluid Intake (ml)
study did not address the value of fluid restriction in pa-
tients with severe renal dysfunction. Because this group
may not respond as well to diuretics,14 it is possible that re-
stricting fluid intake may be of benefit in this subgroup. Fi-
nally, the majority of the patient population studied in this
protocol responded promptly to IV diuretic and vasodilator
therapy. It is possible that those failing to respond to initial
strategies may benefit from fluid restriction.
Previously reported findings in other clinical settings3,15
and among chronic HF patients 16 have reported difficulties
in adhering to a fluid restriction regimen. Difficulty with
fluid restriction in HF may in part reflect the heightened
stimulation of the thirst center in acute decompensated
HF as a result of the activation of the renin-aldosterone-
angiotensin system.17 Van der Wal et al16 reported that
73% of stable community HF patients were compliant
with fluid restriction of 1500 mL using a self-reported mea-
surement of compliance. Although the restriction of fluid
intake to 1 L/day may appear severe, our data indicate
that fluid restriction can be adhered to in this population
with an average of 73% compliant days over the study
period. However, we did not assess tolerability of this
intervention by questionnaire.
It is an interesting observation that the mean difference
between the 2 arms was only 400 mL/day. This may in
part explain the lack of clinical benefit of fluid restriction.

1200
It is somewhat surprising that fluid intake was not greater
in the FF arm given the heightened thirst drive in heart
failure.
900
Limitations

600
The sample size was small and these results should be
tested in a larger population. Moreover, the small sample
size did not allow for effective analysis of fluid restriction
300 in certain subgroups, such as those with hyponatremia, as
well as a more complete analysis on the impact of param-
eters such as BNP. The difference in fluid intake between
0 the 2 groups was small (400 mL/day). Although more re-
FF FR strictive fluid management may be of benefit, compliance
Error Bars show Mean +/- 1.0 SE
would likely become an issue. Patients with severe renal
Fig. 3. Average fluid intake in free-fluid and fluid-restricted dysfunction were excluded, and it is possible that fluid
groups. restriction may be of benefit in this population.
132 Journal of Cardiac Failure Vol. 13 No. 2 March 2007
Summary
These are the first published data addressing the role of
fluid restriction in acute decompensated HF. The results
of this preliminary study indicate that this intervention is
not of clinical benefit in this setting and this hypothesis
needs to be tested in a larger population. Furthermore,
more evidence is required to see if subgroups of heart fail-
ure patients, such as those with hyponatremia, raised creat-
inine, or those who fail to respond to IV diuretics would
benefit from fluid restriction.
References
1. Executive summary of the guidelines on the diagnosis and treatment
of acute heart failure: the Task Force on Acute Heart Failure of the
European Society of Cardiology. Eur Heart J 2005;26:384e416.
2. Moser D, Lennie TA, Doering LV. Non pharmacologic management of
heart failure. In: Stewart S, Moser DK, Thompson DR, editors. Caring
for the heart failure patient: a textbook for the healthcare professional.
London: Martin Dunitz; 2004.
3. Welch JL, Thomas-Hawkins C. Psycho-educational strategies to pro-
mote fluid adherence in adult haemodialysis patients; a review of
intervention studies. Int J Nurs Studies 2005;42:597e608.
4. Mc Donald K, Ledwidge M, Cahill J, Kelly J, Quigley P, Maurer B,
et al. Elimination of early rehospitalisation in a randomised, con-
trolled trial of multisciplinary care in a high-risk, elderly heart failure
population: the potential contributions of specialist care, clinical sta-
bility and optimal angiotensin-converting enzyme inhibitor dose at
discharge. Eur J Heart Fail 2001;3:209e15.
5. Mc Murray J. Heart failure: we need more trials in typical patients.
Eur Heart J 2000;21:699e700.
6. Lloyd-Williams F, Mair F, Shiels C, Hanratty B, Goldstein P,
Beaton S, et al. Why are patients in clinical trials of heart failure
not like those we see in everyday practice? J Clin Epidemiol 2003;
56:1157e62.
7. Krumholz HM, Chen YT, Vaccarino V, Wang Y, Radford MJ,
Bradford WD, et al. Correlates and impact on outcomes of worsening
renal function in patients $65 years of age with heart failure. Am J
Cardiol 2000;85:1110e3.
8. Remme WJ, Swedberg K. Guidelines for the diagnosis and treatment
of chronic heart failure. Eur Heart J 2001;22:527e60.
9. Shirreffs SM, Merson SJ, Fraser SM, Archer DT. The effects of fluid
restriction on hydration status and subjective feelings in man. Br J
Nutr 2004;91:951e8.
10. Holst M, Stromberg A, Lindholm M, Uden G, Willenheimer R.
Fluid restriction in heart failure patients: is it useful? The design
of a prospective, randomised study. Eur J Cardiovasc Nurs 2003;
2:237e42.
11. Ramirez EC, Martinez LC, Tejeda AO, Gonzalez VR, David RN,
Lafuente EA. Effects of a nutritional intervention on body composi-
tion, clinical status and quality of life in patients with heart failure.
Nutrition 2004;20:890e5.
12. Goldsmith SR. Current treatments and novel pharmacologic treat-
ments for hyponatremia in congestive cardiac failure. Am J Cardiol
2005;95:14Be23B.
13. Goh KP. Management of hyponatremia. Am Fam Physician 2004;69:
2387e94.
14. Krämer BK, Schweda F, Riegger GAJ. Diuretic treatment and diuretic
resistance in heart failure. Am J Med 1999;106:90e6.
15. Welch JL, Perkins SM, Evans JD, Bajpai S. Differences in perceptions
by stage of fluid adherence. J Ren Nutr 2003;13:275e81.
16. Van der Wal MH, Jaarsma T, Moser DK, Veeger NJ, van Gilst WH,
van Veldhuisen DJ. Compliance in heart failure patients: the impor-
tance of knowledge and beliefs. Eur Heart J 2006;27:434e40.
17. Braunwald E, Zipes DP, Libby P. Heart disease: a textbook of cardio-
vascular medicine. London: W.B. Saunders; 2001.