Effects of low-dose dopamine on urine output in normotensive very low birth weight neonates

Abstract

Objective

Determine effects of low-dose dopamine on urine output in very low birth weight premature neonates.

Study Design

Retrospective cohort study of all low-dose (3-5 μg/kg/min) dopamine infusions >24 hours duration in neonates ≤1500 g and ≤32 weeks gestation from August 2009 through September 2011. Linear regression was used to estimate the impact of covariates on urine output.

Results

We identified 91 episodes of low-dose dopamine use in 65 neonates. Increased urine output occurred with 64% of episodes. Low-dose dopamine use was associated with a 0.6 mL/kg/hr increase in urine output (p<0.001) and a 1.3 mL/kg/hr increase when baseline urine output was <1.5 mL/kg/hr (p<0.001). The improvement remained statistically significant after controlling for medications (diuretics and hydrocortisone) and fluid intake.

Conclusions

Low-dose dopamine use was associated with increased urine output in very low birth weight neonates.

Introduction

Acute kidney failure is common in the neonatal intensive care unit (NICU), occurring in 3% to 24% of neonates (1-4). Oliguria is the predominant manifestation of acute kidney failure in neonates with an incidence of 46 to 93% (5). One suspected cause of oliguria is decreased renal perfusion. Dopamine is frequently used in the neonate to help preserve renal function (6). In one systematic review, 7 out of 19 NICU and pediatric intensive care units surveyed used low-dose dopamine to improve renal function; and 13 out of 19 used low-dose dopamine to increase urine output (UOP) (6). Dopamine does not carry an FDA approved indication for this condition (7), nor is its use well supported in the pediatric literature; the rationale for its use being extrapolated from the adult literature (8).

In our unit, low-dose dopamine is prescribed with the intent of increasing renal perfusion, and thus UOP, in extremely low birth weight (birth weight (BW) ≤1000g; ELBW) and very low birth weight (BW ≤1500g; VLBW) neonates. Low-dose dopamine has been shown in older premature populations to increase UOP (9) but has not been evaluated in the ELBW and VLBW populations or in neonates older than the first few days of life.

Methods

We conducted a retrospective cohort study of all neonates ≤1500 g and ≤32 weeks gestation at birth who received low-dose (3-5 μg/kg/min) dopamine infusion for ≥24 hours and who were subsequently admitted to the Duke Intensive Care Nursery from August 2009 through September 2011. Patients were started on low-dose dopamine at the discretion of the attending physician. Detailed patient information was extracted from patient electronic medical records. Daily progress notes on the day prior to dopamine initiation, the day dopamine was started, and the day after the dopamine infusion was started were reviewed for daily vital signs, including heart rate (HR) and mean arterial pressure (MAP) ranges, UOPs, total fluid volumes (TFV), and medications. Patients were excluded if the patient was on dopamine >5 μg/kg/min.

The data was analyzed using Stata 12 (College Station, TX). Demographic information was compared using summary statistics. Wilcoxon signed-rank test was used to analyze the changes in HRs and MAPs. Linear regression was used to estimate the impact of covariates on UOP. Covariates considered in this analysis included an increase in TFV during the dopamine infusion; an increase (dosage, frequency or change in diuretic class) or start of diuretics during the dopamine infusion; and increase or start of hydrocortisone during the dopamine infusion. Two sub-analyses focused on episodes where the patient’s starting UOP was <1.5 mL/kg/hr and another that included only ELBW patients ≤28 weeks gestational age (GA). Additionally, an analysis to assess the influence of chronological age on the association between low-dose dopamine and UOP was completed. This study was approved by the Duke University Institutional Review Board.

Results

During the 26 month period this study was conducted, 65 neonates met criteria for inclusion, accounting for 91 episodes of low-dose dopamine use. The median GA was 26 weeks (25^th%ile, 75^th%ile: 25, 28) and the median BW was 850 g (720, 980). In 93% of the episodes, the neonate was > 3 days of life at the start of the dopamine infusion. The median age at the start of dopamine was 15 days (7, 21). In 39% of the episodes in which dopamine was started, the patient was on the ventilator at the time dopamine was initiated (Table 1). Ten neonates were on non-steroidal anti-inflammatory drugs at the time dopamine was started. Of these ten neonates, half showed increased UOP when started on low-dose dopamine.

Table 1.

Demographics (n=65)

Gestational Age:
<28 weeks, n (%)	46 (71)
≥28 weeks, n (%)	19 (29)
Male, n (%)	35 (54)
Race/ethnicity:
White, n (%)	25 (38)
Black, n (%)	33 (51)
Hispanic, n (%)	5 (8)
Other, n (%)	2 (3)
Cesarean section, n (%)	52 (80)
APGAR, 1 minute*	4 (1, 5)
APGAR, 5 minute*	7 (6, 8)
Maternal prenatal steroids, n (%)	54 (83)

^*median (25^th, 75^th percentiles)

The mean UOP on the day prior to starting dopamine was 1.7 mL/kg/hr. This increased to 2.4 mL/kg/hr the day after dopamine was begun (p<0.001). Within our cohort, 58/91 (64%) episodes were associated with an increase in UOP from the day prior to starting dopamine to the day after starting dopamine. There were no significant increases in HR after starting dopamine (p=0.10). There was a significant increase in MAP from the day prior to the day after dopamine was started (MAP increased by 7 mm Hg, p<0.001).

Unadjusted analyses showed that low-dose dopamine was associated with a 0.6 mL/kg/hr increase in UOP (p<0.001). When episodes were restricted to those where the initial UOP was <1.5 mL/kg/hr, there was a 1.3 mL/kg/hr increase in UOP (p<0.001) (Table 2).

Table 2.

Change in Urine Output

	VLBW		ELBW
	Increase in UOP mL/kg/hr, (95% CI)	P	Increase in UOP mL/kg/hr, (95% CI)	P
Dopamine*	0.6 (0.3, 1.0)	<0.001	0.8 (0.4, 1.1)	<0.001
Dopamine* and starting UOP <1.5 mL/kg/hr	1.3 (0.8, 1.8)	<0.001	1.3 (0.9, 1.7)	<0.001
Dopamine controlling for:
Increase/start diuretics	0.6 (0.3, 1.0)	0.001	0.8 (0.4, 1.1)	<0.001
Increase total fluid volume	0.7 (0.3, 1.1)	<0.001	0.8 (0.4, 1.2)	<0.001
Increase/start hydrocortisone	0.4 (−0.2, 0.9)	0.06	0.7 (0.3, 1.1)	0.001

UOP=urine output

^*unadjusted

Controlling for an increase in total fluid volume from the day prior to the day after dopamine infusion, we observed an UOP increase of 0.7 mL/kg/hr (p<0.001). Controlling for an increase or start of corticosteroids in the same time period, there was an associated 0.4 mL/kg/hr increase in UOP that approached significance (p=0.06). Controlling for an increase or start of diuretics, there was an associated 0.6 mL/kg/hr increase in UOP (p=0.001).

An additional analysis was completed to assess the influence chronological age might have on the association between low-dose dopamine and UOP. When infants were older when dopamine was initiated, there was a slightly enhanced positive effect on urine output (p=0.04).

When the analysis was restricted to ELBW neonates with GA ≤28weeks (n=65), a similar increase in UOP was associated with use of low-dose dopamine (0.8 mL/kg/hr; p<0.001, unadjusted). Adjusted analyses revealed a similar increase in UOP in this population across all covariates (total fluid volume, hydrocortisone dosage, and diuretic use).

Discussion

In our cohort, low-dose dopamine was associated with a significant increase in UOP after adjustment for covariates. Furthermore, the increase in UOP associated with low-dose dopamine was amplified in neonates with UOP <1.5 mL/kg/hr and remained significant when the analyses were restricted to include only ELBW neonates. When restricted to neonates whose starting UOP was <1.5 mL/kg/hr, we found twice the increase in UOP compared to the entire cohort that received low-dose dopamine. Our findings are consistent with previous reports in which low-dose dopamine was associated with increased UOP, but not associated with improvements in serum creatinine (8, 10, 11).

Dopamine acts both centrally and peripherally (12). In the periphery, dopamine acts on dopamine receptors in the kidney causing increased renal blood flow, sodium excretion, and urine volume (6). Increased renal blood flow occurs through dopamine-induced smooth muscle relaxation and resulting vasodilation (12). In addition, dopamine causes efferent vasoconstriction via its endocrine effect through the renin-angiotension system, thus increasing the glomerular filtration rate and diuresis (12). Naturesis occurs through the hemodynamic and direct tubular actions of dopamine, as well as its endocrine effect (12). Theoretically, low-dose dopamine should result in increased UOP (13). Although there has been significant debate about its use for this purpose in pediatric and adult ICU populations (including suggestion that its use be abandoned) (6, 10, 14), there has been little study dedicated to the VLBW and ELBW neonate. There have been studies that show older animals have increased renal blood flow and naturesis secondary to dopamine compared to younger animals, suggesting a maturation of the dopamine receptor (15). This would suggest that very premature neonates might not have adequate adaptation of their dopamine receptors and may not respond to low-dose dopamine as much as older infants.

There have been two randomized control trials examining the use of dopamine to increase UOP in premature neonates (16, 17). Both studies examined the effect in neonates that were younger (≤72 hours of life), more mature (≥30 weeks and a mean of 34 weeks, respectively), and exposed to different doses of dopamine (0.5-7.5 μg/kg/min) than the neonates in our cohort. Cuevas et al (16) found an insignificant increase in UOP in the groups receiving low-dose dopamine (1-2.5 μg/kg/min), and Lynch et al (17) reported significant increases in UOP in those receiving either 2.5 or 7.5 μg/kg/min of dopamine, but no significant increase in those receiving 0.5 μg/kg/min infusions. Seri et al demonstrated that premature neonates were sensitive to the dose of dopamine used and showed increased blood pressure and glomerular filtration rate at doses of 4 μg/kg/min but no changes in UOP or sodium excretion when compared to 2 μg/kg/min (9). Although our study addressed a more premature population that was >72 hours of age using a 2-5 μg/kg/min dose, the results of the four studies taken together suggest a dose-based threshold for dopamine treatment that is associated with increased UOP in premature neonates with response noted at doses as low as 2.5 μg/kg/min.

The clinical course following premature birth, particularly for the extremely premature neonate, is associated with several clinical variables that may augment UOP. To address this concern, we used linear regression to control for covariates that might influence UOP, including an increase in fluid, as well as the addition or increase in diuretics or corticosteroids. In addition to potentially increasing renal blood flow, dopamine has been suggested to inhibit the Na-KATPase in the proximal convoluted tubule, ascending loop of Henle, and the cortical collecting duct thereby increasing urinary sodium, and thus UOP (12). As loop diuretics also alter excretion of sodium, albeit through a different mechanism (Na+/K+/2Cl– co-transporter NKCC2), their use could theoretically limit the effect of concurrent dopamine on UOP. As a result, we suspected that neonates on loop diuretics would not experience as significant an increase in UOP as those neonates without concurrent loop diuretic exposure. However, in both VLBW and ELBW neonates, the increase in UOP associated with low-dose dopamine remained significant after adjustment for use of loop diuretics.

Another factor that can modify UOP is the daily fluid intake. When we controlled for any increase in TFV, the positive effect of low-dose dopamine on UOP remained significant. A paradoxical decrease in UOP was associated with an increase in the daily TFV in our population. One possible explanation for this finding is that many of these patients were extremely ill and were oliguric for reasons other than low fluid volume status where the increase in TFV would be less likely to result in any changes in UOP, and may even worsen the clinical status of the patient.

We also examined use of hydrocortisone and its effect on UOP. One would expect an increase in UOP when hydrocortisone was either started or increased in dosage due to its influence upon water reabsorption in the renal tubules and its resultant diuretic effect (18). Additionally, hydrocortisone is known to increase glomerular filtration rate (18). When we controlled for hydrocortisone in our linear regression analysis, there still seemed to be a trend towards an increase in UOP due to dopamine alone, although it did not reach significance.

The major strengths of this study are its sample size and the age (both gestational and post-natal) of the cohort studied as well as adjustment for potentially important clinical covariates. Previous studies on this topic, both prospective and retrospective, have reported much smaller cohorts ranging from 9 to 41 neonates that received dopamine (9, 16, 17, 19-21). Additionally, our study was limited to VLBW and ELBW neonates outside of the first three days of life, which has not been previously studied. Lastly, our study addressed the potential impact of important covariates known to affect UOP, including daily TFV and medication (hydrocortisone and diuretic) use. We acknowledge the limits associated with any retrospective cohort study performed without a control group. We also accept that we did not have the exact start time, a strict indication (e.g. UOP <1mL/kg/hr for previous 12 hours) for low-dose dopamine prescription or more frequent documentation of UOP beyond the average over a 24-hour period of time. Thus, we are unable to determine the exact time to effect between low-dose dopamine administration and an increase in UOP. These limitations are difficult to completely address without conducting a large randomized control trial but do not abrogate our findings supporting an increase in UOP associated with low-dose dopamine. Lastly, we found that low-dose dopamine had no overall effect on UOP in approximately one third of the episodes analyzed in this study. Future studies will be necessary to uncover the specific mechanism(s) behind the apparent lack of effect on UOP in these neonates and to identify potential side effects of its use.

Conclusion

Low-dose dopamine was associated with a significant increase in UOP for nearly two-thirds of the neonates in our cohort. The increase in UOP remained significant after adjustment for covariates and was more pronounced in ELBW neonates with low starting UOP (<1.5 mL/kg/hr). Even the most premature patients are likely to experience an increase in UOP in response to low-dose dopamine treatment.