Umbilical Artery Catheters and Blood FlowVelocities in the Superior Mesenteric Artery:Effect of Insertion, Removal, Aspiration, andBolus Infusion

Jayesh B. Shah, MD,1 Luis A. Bracero, MD,2 Michael H. Gewitz, MD,1 Bernard G. Fish, MD,1

Harry S. Dweck, MD1

1 Department of Pediatrics, New York Medical College, Westchester County Medical Center, Valhalla, New York2 Department of Obstetrics and Gynecology, New York Medical College, Westchester County Medical Center,Valhalla, New York

Received 19 September 1996; accepted 18 July 1997

ABSTRACT: Purpose. We studied whether umbilicalartery catheters (UACs) affect blood flow in the supe-rior mesenteric artery (SMA) of neonates.

Methods. Noninvasive duplex pulsed Doppler so-nography was used to measure peak systolic velocity,end-diastolic velocity, and mean flow velocity in theSMA. The resistance index and pulsatility index werecalculated from these data. Thirty-two infants weigh-ing 450–2,520 g at birth were enrolled in the study.Gestational age at birth was 24–37 weeks. Eighteeninfants were studied before and after UAC insertion.Twenty infants were studied before and after UAC re-moval. Eleven infants with UACs were studied beforeand during aspiration of blood from the UAC and dur-ing bolus infusion of 5% dextrose solution into theUAC. Data were compared before and after UAC in-sertion; before and after UAC removal; and before andduring aspiration and during bolus infusion.

Results. Blood flow velocities and vascular resis-tance were similar in all comparisons except for in-creases in end-diastolic and mean velocities after UACinsertion.

Conclusions. Insertion and removal of UACs, aspi-ration of blood from UACs, and bolus infusion of flu-ids into UACs do not diminish blood flow velocity orincrease vascular resistance in the SMA. © 1998 JohnWiley & Sons, Inc. J Clin Ultrasound 26:73–77, 1998.

Keywords: newborns; umbilical artery catheters;

Doppler ultrasonography; blood flow velocity; supe-rior mesenteric artery

Umbilical artery catheterization is frequentlyperformed in critically ill preterm and full-

term neonates. Complications of umbilical arterycatheters (UACs) include thrombosis, embolism,vascular perforation, hemorrhage, bowel perfora-tion, and infection.1–3 UACs have also been impli-cated in the pathogenesis of necrotizing enteroco-litis.4 The most common sites of involvement innecrotizing enterocolitis are the terminal ileumand the ascending colon,4 whose blood supply isderived from the superior mesenteric artery(SMA).5 The SMA arises from the ventral wall ofthe aorta at an angle of approximately 90°, givingit a posteroanterior course proximally. This en-ables accurate measurement of blood flow velocitybecause the Doppler ultrasound probe can beplaced at an angle of less than 20°.

In critically ill newborns, aspiration of bloodsamples and bolus infusion of flush solutionsthrough UACs are performed frequently. Shrivas-tava et al6 observed an increase in the pulsatilityindex (PI) of the aorta during aspiration from andbolus infusion into UACs placed high in infants.The effect of these procedures on the SMA, how-ever, has not been reported.

The purpose of this study was to determinewhether blood flow velocity and vascular resis-

Correspondence to: L.A. Bracero, Department of OB/GYN,Catholic Medical Center, 88-25 153rd Street, Suite 4G, Ja-maica, NY 11432

© 1998 John Wiley & Sons, Inc. CCC 0091-2751/98/020073-05

73VOL. 26, NO. 2, FEBRUARY 1998

tance in the SMA of infants with catheters placedhigh within the umbilical artery are affected byinsertion or removal of the UAC or by aspirationof blood from and bolus infusion of fluid into theUAC.

PATIENTS AND METHODS

Thirty-two infants with birth weights of 450–2,520 g who required UACs were enrolled in thestudy. Gestational age at birth was 24–37 weeks.The infants were born at (n 4 26) or transferredto (n 4 6) the Regional Neonatal Intensive CareUnit at the Westchester County Medical Centerin Valhalla, NY. Infants with major congenitalanomalies or clinical evidence of patent ductusarteriosus and those receiving vasoactive drugswere excluded. The study protocol was approvedby the Institutional Review Board, and informedconsent was obtained from parents.

UACs were inserted into 31 infants with respi-ratory distress syndrome that required an FiO2 of0.4 or more and in 1 infant born at 34 weeks ofgestation with unconjugated hyperbilirubinemia(peak bilirubin level of 19 mg/dl) that required anexchange transfusion. UACs were inserted amean of 5.3 ± 7.8 hours after birth. Polyvinyl chlo-ride UACs (Sherwood Medical, St. Louis, MO)were used. All UACs were placed in the high po-sition, with the catheter tip at T6–T10, and place-ment was confirmed by roentgenography.

Doppler sonographic studies were obtained us-ing a duplex Doppler ultrasound system (Inter-spec, Ambler, PA) with a 5-MHz transducer. TheSMA was visualized in real-time to allow properplacement of the Doppler sample volume. The in-sonation angle was between 0° and 20°, so anglecorrections were not necessary. All measure-ments were made and analyzed by 1 observer(JBS) to prevent interobserver differences. Dur-ing the measurements, infants lay comfortablysupine in an anti-Trendelenburg position of ap-proximately 15°. Warmed ultrasound gel was ap-plied to the abdomen. The sample volume lengthwas adjusted to include the lumen; vessel wallswere excluded to prevent disturbances in thewaveform. Waveform quality was assessed visu-ally and with an audible signal. All results areexpressed as means ± standard deviations.

Waveforms were recorded on videotape for sub-sequent analysis. Peak systolic, end-diastolic, andmean flow velocities were measured by tracing 5sequential cardiac cycles from the videotape withthe built-in analysis package and averaging thevalues. The mean values were used to calculatethe PI and the resistance index (RI).

Infants were put into 1 or more of 3 groups:insertion group, removal group, and aspirationand bolus group.

Insertion Group

Blood flow velocity was measured immediatelybefore and a mean of 46 ± 6.7 minutes after UACinsertion in 18 infants. The time lag reflects thetime it took to do the procedure and to confirm theUAC location by x-ray. None of these infants re-ceived enteral feedings at the time of study.

Removal Group

Twenty infants were studied immediately beforeand a mean of 2.6 ± 1.2 hours after removal of theUAC. In contrast to the insertion group, these in-fants had improved clinically to the point wherethe UAC was no longer necessary. Thirteen in-fants in this group were receiving a mean of 9.8 ±3.4 ml of bolus enteral feedings every 3 hours atthe time of study. Measurements taken just be-fore removal of the UAC were obtained a mean of112 ± 27 minutes after enteral feeding, and themeasurements taken after removal of the UACwere obtained a mean of 92 ± 18 minutes afterenteral feeding.

Aspiration and Bolus Group

Eleven infants with indwelling UACs were stud-ied to determine whether aspiration of blood fromthe UAC and bolus infusion of 5% dextrose solu-tion into the UAC would affect blood flow velocityin the SMA. The mean age of these infants was 69± 9 hours at the time of study.

Blood flow velocity was measured before andduring a 15–20-second aspiration of 1–2 ml ofblood from the UAC. (Although infusions and as-pirations may be performed faster, these weredone over 15–20 seconds to allow for Dopplermeasurements.) Blood flow velocity was thenmeasured during a 15–20-second bolus infusioninto the UAC of 1–2 ml of 5% dextrose solutiondone immediately after aspiration. Nine of theseinfants were receiving bolus enteral feedings at amean of 3.5 ± 1.0 ml every 3 hours at the time ofthe study. Blood flow velocity was measured amean of 178 ± 38 minutes after their last feeding.

Statistical Analysis

Statistical and data analyses were performedwith DataEase (Version 4.0, DataEase Interna-tional, Trumbull, CT) and SPSS PC+ (Version 3.1,

SHAH ET AL

74 JOURNAL OF CLINICAL ULTRASOUND

SPSS, Chicago, IL). Paired t-tests of the meanswere used for each infant to compare blood flowvelocities, RI, and PI within each group. A p valueof 0.05 or less was considered statistically signifi-cant.

RESULTS

Patient characteristics such as gestational ageand birth weight were similar in the 3 groups (p >0.05 by chi-squared analysis) (Table 1). The de-velopment of 30 infants was appropriate for theirgestational ages; 2 infants in the removal groupwere small for their gestational ages. There wereno statistically significant differences in bloodflow velocities or vascular resistance between the3 groups when compared by mode of delivery, gen-der, or hematocrit. There was a trend of increasedflow velocity with higher birth weight and oldergestational age, but the differences were not sta-tistically significant.

UACs of 3.5 Fr were used in infants with abirth weight of 1,000 g or less (n 4 9). A 3.5-Fr (n4 13) or 5-Fr (n 4 10) UAC was used in infantswith a birth weight over 1,000 g. Catheter size inthe latter group had no effect on blood flow veloc-ity. Necrotizing enterocolitis did not develop inany infant included in this study.

Insertion Group

Values of blood flow velocity, RI, and PI beforeand after UAC insertion are shown in Table 2.

The peak systolic velocity was similar before andafter insertion of the UAC, whereas the end-diastolic and mean velocities increased signifi-cantly after UAC insertion. Both RI and PItended to be lower after UAC insertion, but thedifferences did not reach significance.

Removal Group

No statistically significant differences in bloodflow velocity, RI, or PI before and after UAC re-moval were found (Table 3), nor were differencesobserved when the 18 infants of appropriate sizefor gestational age were compared with the 2 in-fants who were small for gestational age.

Aspiration and Bolus Group

Values of blood flow velocity, RI, and PI obtainedduring aspiration and during bolus infusion were

TABLE 1

Patient Characteristics

CharacteristicInsertion Group

(n = 18)Removal Group

(n = 20)

Aspiration andBolus Group

(n = 11)

Male/female 10/8 12/8 5/6Birth weight, g (range) 1,397 ± 706* 1,615 ± 644* 1,601 ± 584*

(590–2,520) (450–2,520) (770–2,370)Gestational age, weeks (range) 29.0 ± 4.5* 31.0 ± 3.8* 30.0 ± 3.3*

(24–37) (24–37) (25–35)Mode of delivery

Vaginal 10 5 3Cesarean 8 15 8

Apgar score <7 at 1 minute 13 15 9Apgar score <7 at 5 minutes 6 7 4AGA/SGA 18/0 18/2 11/0Age at study, hours 4.0 ± 6.4* 134.0 ± 123.0* 69.0 ± 28.8*

HCT at study, % 48.0 ± 7.3* 43.0 ± 5.4* 44.0 ± 7.3*

Catheter size3.5 Fr 13 14 75 Fr 5 6 4

Abbreviations: AGA, appropriate for gestational age; SGA, small for gestational age; HCT, hemato-crit.

*Mean ± standard deviation.

TABLE 2

Insertion Group (n = 18): Blood Flow Velocities, RI, and PI

Before UACInsertion

(mean ± SD)

After UACInsertion

(mean ± SD) p

Velocity, cm/secPeak systolic 50.8 ± 14.3 53.8 ± 16.1 0.375End-diastolic 5.7 ± 4.6 8.3 ± 6.5 0.026Mean 19.6 ± 5.8 23.1 ± 8.8 0.024

RI 0.89 ± 0.09 0.85 ± 0.12 0.058PI 2.41 ± 0.74 2.16 ± 0.79 0.086

Abbreviations: RI, resistance index; PI, pulsatility index; UAC, um-bilical artery catheter; SD, standard deviation.

UMBILICAL ARTERY CATHETERS AND BLOOD FLOW VELOCITY

75VOL. 26, NO. 2, FEBRUARY 1998

not significantly different from each other or frommeasurements obtained before aspiration (Ta-ble 4).

DISCUSSION

Necrotizing enterocolitis is seen in 1–5% of in-fants admitted to neonatal intensive care units.7

Its multifactorial etiology includes poor perfusionof the gut, enteral feeding, and infection. UACshave been implicated as contributing to ischemiaof the gut, which leads to necrotizing enterocoli-tis.4

To assess the effect of UACs on blood flow inthe SMA, we measured blood flow velocity in theSMA with the noninvasive method of duplexpulsed Doppler sonography. This technique hasbeen used previously in adults8,9 and in new-borns.10–15 In our study, blood flow velocities andvascular resistance were similar in all compari-sons except for an increase in end-diastolic andmean velocities after UAC insertion. This sug-gests that insertion of the UAC, its removal, andaspiration of blood from or bolus fluid infusioninto a UAC do not affect blood flow velocity orvascular resistance in the SMA. The increaseseen in end-diastolic and mean velocities afterUAC insertion may be due to hemodynamicchanges occurring at the level of the ductuswithin the first few hours after birth. Low end-diastolic velocity, and therefore low mean veloc-ity, before insertion of the UAC could be due tosome left-to-right shunting at the level of the duc-tus. Serial Doppler observations of blood flow ve-locity made during the first few hours after birthin infants of comparable birth weight and gesta-tional age are required to confirm this observa-tion. Kempley and Gamsu15 randomized 69 in-fants to high or low UAC placement andperformed Doppler velocimetry on days 1, 3, 7,and 14 on various vessels, including the SMA.They found no difference between the 2 groups in

mean blood flow velocity in any studied vesselsuntil day 14. A third group of infants withoutUACs (UACs had been removed or had failed af-ter placement) had a significantly higher meanblood flow velocity in the SMA compared with in-fants in the high UAC placement group on day 14.Their data for all groups before day 14 are com-patible with our findings. UACs were removed be-fore 14 days (within a mean of 5.7 ± 4.9 days) inall our infants.

Enteral feeding has been reported to increaseblood flow velocity in the SMA, with a peak effectoccurring 45 minutes after the feeding.12 In ourremoval group, the observations before UAC re-moval were made a mean of 112 ± 27 minutesafter the last feeding and those after UAC re-moval were made a mean of 92 ± 18 minutes afterthe last feeding. These infants received a mean ofonly 9.8 ± 3.4 ml per enteral feeding. In the 9infants receiving enteral feeding in the aspirationand bolus group, there was a mean interval of 178± 38 minutes between their last feeding and ve-locity measurements, and they received a mean ofonly 3.5 ± 1.0 ml per feeding. The small amount offeeding they received and the long interval afterfeeding would be expected to have minimal or noeffects on blood flow velocity.

An earlier study using continuous-wave Dopp-ler sonography reported a significant increase inthe PI in the descending aorta during aspirationand during bolus infusion through a UAC withhigh placement.6 The speculation was that theincreased aortic resistance with a high UAC mayaffect the blood flow in the splanchnic artery andcause significant vascular and organ compromise.The current study using pulsed-wave Doppler so-nography shows no change in the blood flow ve-locity or vascular resistance in the SMA duringaspiration and bolus infusion. The discrepancy

TABLE 4

Aspiration and Bolus Group (n = 11): Blood Flow

Velocities, RI, and PI*

BeforeAspiration

(mean ± SD)

DuringAspiration

(mean ± SD)

DuringBolus Infusion(mean ± SD)

Velocity, cm/secPeak systolic 54.2 ± 20.9 55.6 ± 19.5 50.1 ± 18.2End-diastolic 11.1 ± 5.9 11.2 ± 5.9 9.1 ± 3.0Mean 23.4 ± 8.9 23.4 ± 8.9 20.5 ± 6.6

RI 0.79 ± 0.10 0.80 ± 0.10 0.81 ± 0.07PI 1.9 ± 0.66 2.0 ± 0.56 2.0 ± 0.43

Abbreviations: RI, resistance index; PI, pulsatility index; SD, stan-dard deviation.

*Comparing values obtained before aspiration with values obtainedduring aspiration and during bolus infusion showed no significantdifferences.

TABLE 3

Removal Group (n = 20): Blood Flow Velocities, RI, and PI

Before UACRemoval

(mean ± SD)

After UACRemoval

(mean ± SD) p

Velocity, cm/secPeak systolic 72.7 ± 22.0 71.1 ± 26.1 0.677End-diastolic 15.8 ± 10.0 17.0 ± 10.5 0.515Mean 33.0 ± 13.0 33.4 ± 15.2 0.870

RI 0.79 ± 0.10 0.76 ± 0.09 0.224PI 1.86 ± 0.60 1.77 ± 0.90 0.553

Abbreviations: RI, resistance index; PI, pulsatility index; UAC, um-bilical artery catheter; SD, standard deviation.

SHAH ET AL

76 JOURNAL OF CLINICAL ULTRASOUND

between the previous study and this one may bedue to the increase in vascular resistance in ves-sels other than the SMA branching from the aortaor to a type I error in the previous study.

Our sample size for the aspiration and bolusgroup is small, and we cannot exclude the occur-rence of a type II error in this group. To detect adrop of 10 cm/sec in mean velocity after aspirationand bolus infusion, we estimate that 17 subjectsare needed for an adequate sample size, assuminga standard deviation of 6.6 (for p < 0.05).

Our study indicates that indwelling UACs (tipat T6–T10) in the abdominal aorta do not dimin-ish blood flow velocity or increase vascular resis-tance in the SMA. Inexplicably, instead of the ex-pected decreases in flow velocities, the end-diastolic and mean velocities increased afterinsertion of the UACs. We also found that aspira-tion of blood and bolus infusion of fluids throughthe UAC do not affect flow velocity or vascularresistance in the SMA.

REFERENCES

1. Mokrohisky ST, Levine RL, Blumhagen JD, et al:Low positioning of umbilical artery catheters in-creases associated complications in newborn in-fants. N Engl J Med 1978;299:561.

2. Neal WA, Reynolds JW, Jarvis CW, et al: Umbilicalartery catheterization: demonstration of arterialthrombosis by aortography. Pediatrics 1972;50:6.

3. Harris MS, Little GA: Umbilical artery catheters:high, low or no. J Perinat Med 1978;6:15.

4. Kliegman RM, Fanaroff AA: Necrotizing enteroco-litis. N Engl J Med 1984;310:1093.

5. Clemente CD, ed: Gray’s Anatomy of the HumanBody, 30th Am. edn. Philadelphia, Lea & Febiger,1985, pp. 739–744.

6. Shrivastava B, Dweck HS, Bracero LA, et al: Effectof umbilical artery catheters on aortic velocitywaveforms. J Matern Fetal Invest 1991;1:87.

7. Fanaroff AA, Martin RJ: Neonatal-Perinatal Medi-cine: Diseases of the Fetus and Infant, 5th edn. St.Louis, Mosby–Year Book, 1992, pp. 1068–1072.

8. Sato S, Ohnishi K, Sugita S, et al: Splenic arteryand superior mesenteric artery blood flow: nonsur-gical Doppler US measurement in healthy subjectsand patients with chronic liver disease. Radiology1987;164:347.

9. Qamar MI, Read AE, Skidmore R, et al: Transcu-taneous Doppler ultrasound measurement of supe-rior mesenteric artery blood flow in man. Gut 1986;27:100.

10. Van Bel F, Van Zwieten PHT, Guit GL, et al: Supe-rior mesenteric artery blood flow velocity and esti-mated volume flow: duplex Doppler US study of pre-term and term neonates. Radiology 1990;174:165.

11. Van Bel F, Zoeren DV, Schipper J, et al: Effect ofindomethacin on superior mesenteric artery bloodflow velocity in preterm infants. J Pediatr 1990;116:965.

12. Leidig E: Doppler analysis of superior mesentericartery blood flow in preterm infants. Arch DisChild 1989;64:476.

13. Coombs RC, Morgan MEI, Durbin GM, et al: Dopp-ler assessment of human neonatal gut blood flowvelocities: postnatal adaptation and response tofeeds. J Pediatr Gastroenterol Nutr 1992;15:6.

14. Coombs RC, Morgan MEI, Durbin GM, et al: Ab-normal gut blood flow velocities in neonates at riskof necrotizing enterocolitis. J Pediatr GastroenterolNutr 1992;15:13.

15. Kempley ST, Gamsu HR: Randomised trial of um-bilical artery catheter position: Doppler ultrasoundfindings. Arch Dis Child 1992;67:855.

UMBILICAL ARTERY CATHETERS AND BLOOD FLOW VELOCITY

77VOL. 26, NO. 2, FEBRUARY 1998