DOI: 10.1542/peds.2007-1962; originally published online April 7, 2008; 2008;121;e1152Pediatrics

Houchang D. ModanlouViet Hoang, Jack Sills, Michelle Chandler, Erin Busalani, Robin Clifton-Koeppel and

InsertionNeonates: Complications Rates Related to Upper Versus Lower Extremity Percutaneously Inserted Central Catheter for Total Parenteral Nutrition in

  

  http://pediatrics.aappublications.org/content/121/5/e1152.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

ARTICLE

Percutaneously Inserted Central Catheter for TotalParenteral Nutrition in Neonates: ComplicationsRates Related to Upper Versus LowerExtremity InsertionViet Hoang, MD, PharmDa, Jack Sills, MDa, Michelle Chandler, MDb, Erin Busalani, BSa, Robin Clifton-Koeppel, MS, RNC, CPNPa,

Houchang D. Modanlou, MDa

aDivision of Neonatal-Perinatal Medicine, Department of Pediatrics, University of California, School of Medicine, Irvine, California; bDivision of Pediatric Radiology,Department of Radiology, University of California, School of Medicine, Irvine, California

The authors have indicated they have no financial relationships relevant to this article to disclose.

What’s Known on This Subject

Uses of peripherally inserted central venous catheters in ill neonates, to administer totalparenteral nutrition, are frequently associated with catheter-related bloodstream infec-tion and other complications.

What This Study Adds

Lower extremity percutaneously inserted central venous catheters have lower rates ofcatheter-related bloodstream infection, longer time to first complication, and lowercholestasis despite longer duration of total parenteral nutrition.

ABSTRACT

OBJECTIVE. The objective of this study was to compare the complication rates of upperversus lower extremity percutaneously inserted central catheters used for totalparenteral nutrition in neonates.

METHODS.During a 48-month study period, 396 neonates were identified as havinghad percutaneously inserted central venous catheters. A total of 370 catheters wereinserted from the upper and 107 from the lower extremity. Data retrieved andanalyzed were birth weight, gestational age, age at placement, duration in place,duration of total parenteral nutrition, type of infusates, catheter-related bloodstreaminfection, phlebitis, leakage, occlusion, necrotizing enterocolitis, intraventricularhemorrhage, serum creatinine, liver function tests, and length of hospitalization.

RESULTS. The median birth weight and gestational age were 940 g and 28 weeks. Therate of catheter-related bloodstream infection was 11.6% for the upper and 9.3% inthe lower extremity catheters. The most common organism was coagulase-negativeStaphylococcus for both upper and lower extremity catheters and significantly higherwith catheters from the upper extremity. Lower extremity catheters were in placelonger, and the time from insertion to complication was also longer. The rate ofcholestasis was higher for the upper extremity catheters. Multiple regression analysisshowed that the most significant contributor to cholestasis was duration of time thecatheters were in place and the duration of total parenteral nutrition administration.Receiver operating characteristics curve demonstrated higher sensitivity for durationof catheters in predicting cholestasis with duration of total parenteral nutrition beingmore specific.

CONCLUSION. Lower extremity percutaneously inserted central venous catheters hadlower rates of catheter-related bloodstream infection, longer time to first complica-tion, and lower cholestasis despite longer duration of total parenteral nutrition.When possible, lower extremity inserted catheters should be used for the administration of total parenteral nutrition.

SINCE THE INTRODUCTION of total parenteral nutrition (TPN) in the 1960s1 and use in ill neonates in the 1970s,2

central venous access has become important to effectuate fluid and nutritional requirements while gastrointes-tinal tracts are temporarily inadequate. Percutaneously inserted central catheters (PICCs) offer a discernible route forcentral venous access via cannulation of a peripheral vein in either the upper or the lower extremity. PICCs havebecome a requisite instrument to deliver hyperosmolar solutions, medications, and TPN over a prolonged dwell time.

Currently, the strategy of where PICCs are placed within the upper or lower extremity is based on finding asuitable peripheral vein for cannulation rather than which extremity offers less complication. Reported complicationsof PICCs including thrombosis, infection, catheter occlusion, phlebitis, renal insufficiency, and pulmonary compli-

www.pediatrics.org/cgi/doi/10.1542/peds.2007-1962

doi:10.1542/peds.2007-1962

This work was presented at the annualmeeting of the American Academy ofPediatrics; October 26–30, 2007; SanFrancisco, CA.

KeyWordsPICC, CRBSI, septicemia, TPN, cholestasis

AbbreviationsTPN—total parenteral nutritionPICC—percutaneously inserted centralcatheterCRBSI—catheter-related bloodstreaminfection

Accepted for publication Sep 27, 2007

Address correspondence to Houchang D.Modanlou, MD, Chief, Division of Neonatology;Director, Neonatal-Perinatal MedicineFellowship Training Program. University ofCalifornia, Irvine, 101 City Dr South, UCIMC,Building 56, Suite 600, Orange, CA 92868.E-mail: modanlou@uci.edu

PEDIATRICS (ISSN Numbers: Print, 0031-4005;Online, 1098-4275). Copyright © 2008 by theAmerican Academy of Pediatrics

e1152 HOANG et al by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

cations are scarce in regard to characterizing the proba-ble differences in complications in upper extremity ver-sus lower extremity PICCs. Moreover, the role of centralPICCs compared with noncentral PICCs in the efficacyand safety profile has been questioned.3 Nevertheless,the consensus is that PICCs offer better risk–benefit ra-tio, and as a result, we compare only central lines ineither extremity.

The primary outcome of the study set forth was tolook at the complication rates of PICCs in the upperversus lower extremity in terms of infection, catheterocclusion, phlebitis, liver dysfunction, renal insuffi-ciency, and pulmonary complications. We hypothesizedthat PICCs placed either in the upper extremity or lowerextremity have no differences in complication rates.

METHODS

Study PopulationThis study was conducted at a 30-bed, tertiary-levelNICU of a university teaching hospital. The study wasapproved by our institutional review board, and in-formed consent was waived. Data were aggregated ret-rospectively from a neonatal database over a 48-monthinterval from June 2002 to June 2006. A total of 495PICC were placed during this study period. Neonateswith liver dysfunction and inborn errors of metabolismwere excluded. Liver dysfunction was defined as directhyperbilirubinemia (serum direct bilirubin of �2.0 mg/dL) and high alanine aminotransferase and alanine ami-notransferase levels. Of the 495 PICCs placed, 18 (4.3%)neonates were excluded, leaving 477 PICCs in 396 ne-onates. Sixteen were excluded secondary to liver dys-function before PICC placement and 2 for inborn errorsof metabolism. Total of 370 PICCs were placed in theupper and 107 in the lower extremity.

PICC PlacementAt our institution, PICCs are placed by specialized nurs-ing teams supervised by the neonatologists. Site selec-tion for insertion of PICCs includes the antecubital fossaand greater and lesser saphenous veins.4 Because theantecubital fossa is considered to be less colonized, oily,and moist, upper extremity PICCs are used before lowerextremity veins.5,6 Lower extremity PICCs were insertedbecause of failure to insert PICCs in the upper extremity,or it was the primary selection site. No patient had 2PICCs at the same time. Only 2 French Silastic catheters(L-CATH Peel Away System, Becton Dickinson InfusionTherapy Systems, Inc, Sandy, UT) are used at our insti-tution. Indications for a PICC are determined by theattending neonatologists and include the need for par-enteral nutrition with a dextrose concentration of�12.5%, continuous infusion of vesicant medications,therapies with variations in osmolarity or pH, and pro-longed antibiotic therapy. No blood products wereinfused via PICCs, and heparin was routinely added.Before the insertion of a PICC, informed consent wasobtained. Dwell time and removal of the PICC dependedon factors related to the patient’s clinical condition andavailable vascular access related to ongoing patient care.

Care of PICCs was performed by our nursing staff withoutcome monitoring documented on a procedural log-book kept in the NICU. Upper extremity PICCs wereconsidered as central when the tips resided in the supe-rior vena cava before the right atrium. For low extremityPICCs, the tips were in high inferior vena cava at orabove T10. They were considered noncentral whenPICCs tips were located elsewhere.7–9 Removal of PICCswas accomplished because of complication, prolongedadministration of TPN was not required, or feedingswere advanced enough that supplemental fluids couldbe given by peripheral veins.

DefinitionsConfirmation of catheter colonization and catheter-re-lated bloodstream infection (CRBSI) was accomplishedfollowing Centers for Disease Control and Preventionguidelines.10 CRBSI is defined as a positive culture of anintravascular catheter with the same species as from �1peripheral blood cultures. For culture, �1.0 mL of bloodwas procured from both a peripheral site and the centrallines.11,12 A specialized team consisting of a neonatolo-gist, a nurse epidemiologist, and a clinical nurse special-ist provided surveillance for the occurrences of infectionin our NICU, including those related to PICCs, becausethis strategy can reduce the rate of nosocomial infec-tions.13,14 Phlebitis was defined as a physicochemical ormechanical complication not related to a proven infec-tion. Cholestasis and renal insufficiency were defined byelevated direct bilirubin �2 mg/dL and maximum serumcreatinine level of �1.6 mg/dL, respectively. Catheterocclusion was defined as pump occlusion or inability toflush and/or withdraw from the PICC and the cause to berelated to thrombotic event. Leakage was construed asfluid extravasation and/or pleural or pericardial effusion.Mechanical complications were determined whenever dis-lodgement of a PICC occurred. Time to complication wasdefined as the number of days until 1 of the followingcomplications occurred: phlebitis, occlusion, leakage, me-chanical problem, or septicemia.

Statistical MethodsData analysis was performed by using SAS 9.1 and SASJMP 6 (SAS Institute Inc, Cary, NC). Nonparametricanalysis, �2 test, and Fisher’s exact test were used forcategorical data, and Wilcoxon/Kruskal-Wallis test wasused for continuous variable. P � .05 was consideredsignificant. Multiple logistic regression analysis was usedto adjust for the patient’s gestational age, gender, andcatheter duration to compare the risk for complications.

RESULTS

Clinical CharacteristicsThe median gestational age and birth weight were 28weeks (25.5–30.0) and 937 g (760.0–1359.5) and 28weeks (25–31) and 946 g (740–1427) for the upper andlower extremities, respectively (Table 1). There was aslight predominance of male gender in both the upperand lower extremity groups. PICCs were placed on av-erage day of life 6 (3–12) for the upper and day of life 8

PEDIATRICS Volume 121, Number 5, May 2008 e1153 by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

(3–20) for the lower extremity. When the duration ofPICC placement was analyzed, the lower extremityPICCs were indwelling significantly longer (P � .004) ascompared with the upper extremity group. The medianduration of TPN was 27 days (16–48) for the upper and33 days (19.25–43.5) for the lower extremity, but themedians were not statistically significant. Both upperand lower extremity groups had a median (interquartilerange) of 1 (1–2) PICC placed. Lower extremity PICCswere indwelling for longer periods of time before time tofirst complication, at a median of 15 days for the lowerversus 9 days for the upper extremity. This difference didnot reach statistical significance. Using logistic analysis,as the numbers of PICCs were used, it did not contributeto the increasing risk for complications (data notshown).

Clinical OutcomesThe occurrence of adverse events was generally higherfor the upper extremity placed PICCs (Table 2). CRBSIwas the most frequently occurring life-threatening com-plication. Although the overall rate of CRBSI was notdifferent between the 2 groups, the incidence of themost common nosocomial infection, coagulase-negativeStaphylococcus, was significantly higher (P � .05) withupper extremity PICCs. Our rates of complication fellbetween the known reported rates of CRBSI.14–34 Simi-larly, the rate of diagnosis for cholestasis was signifi-cantly higher (P � .05) for the upper compared with thelower extremity PICCs. Occlusion and phlebitis were themost common complications noted. Similar to a previ-ous report,35 phlebitis was not associated with septicemiain the lower or the upper extremity groups (P � .45 andP � .49, respectively). There were 3 pleural effusionsfrom the upper extremity group and 1 from the lowerextremity group. The pleural effusions were consideredpotentially life-threatening because they required thora-centesis. There was no incidence of pericardial effusionnoted as a complication. The overall rates of leakagewere similar in both groups. There were also no differ-ences in maximum serum creatinine between thegroups. Last, the median length of hospitalization andsurvival were not statistically different between the up-per and lower extremity PICC groups (Table 1).

Risk Factors Leading to CRBSIAs noted in Table 2, CRBSI was the most commoncomplication noted in the study. There were a total of8235 catheter days in both the upper and lower extrem-ities: 6045 from the upper and 2190 from the lowerextremity. The rate of CRBSI in the study was 11.6%(7.1 infections per 1000 catheter days) in the upperextremity and 9.3% (4.8 infections per 1000 catheterdays) in the lower extremity (not statistically signifi-cant). Neonates with lower gestational age and birthweights had significantly higher rates of CRBSI in theupper extremity (P � .005 and P � .0001, respectively;Fig 1A and D). In contrast, lower extremity PICCsshowed no statistical difference in gestational age andbirth weight in neonates with and without CRBSI (Fig1B and D). Neonates with CRBSI had an increased num-ber of PICCs placed (P � .0001; Fig 1C). Time to CRBSIwas longer in the lower extremity (14.5 days) as com-

TABLE 1 Demographic of Patients With PICCs From the Upper and Lower Extremities

Parameter Upper Extremity (n � 370 PICCs) Lower Extremity (n � 107 PICCs) P

Birth weight, median (IQR), g 937.0 (760.0–1359.5) 946.0 (740.0–1427.0) NSGestational age, median (IQR), wk 28.0 (25.5–30.0) 28.0 (25.0–31.0) NSMale gender, % 54.6 55.7 NSDay of life PICC placed, median (IQR) 6 (3–12) 8 (3–20) NSDuration of PICC, median (IQR), d 13.0 (8.0–22.0) 16.0 (11.0–26.8) �.004Duration of TPN, median (IQR), d 27.00 (16.00–48.00) 33.00 (19.25–43.50) NSNo. of PICCs, median (IQR) 1.0 (1.0–2.0) 1.0 (1.0–2.0) NSTime to complication, median (IQR), d 9.0 (4.0–18.0) 15.0 (9.5–22.0) .050Length of hospitalization, median (IQR), d 76.0 (42.0–101.0) 73.0 (46.0–99.0) NSSurvival, % 95.4 92.3 NSInborn patients, % 73.7 79.4 NS

Therewere 294 infants in theupper and102 infants in the lower extremity. IQR indicates interquartile range; PICC, percutaneously inserted centralcatheter; TPN, total parenteral nutrition.

TABLE 2 Types and Rates of Complications With Upper and LowerExtremity PICCs

Parameter Upper Extremity(n � 370 PICCs)

Lower Extremity(n � 107 PICCs)

P

Septicemia (CRBSI), n (%) 43 (11.6) 10 (9.3) NSOrganisms, n (%)Coagulase-negativeStaphylococcus

37 (86.0) 5 (50.0) �.05

Staphylococcus aureus 2 (4.7) 2 (20.0) NSEnterobacter 0 (0.0) 1 (10.0) NSPseudomonas 1 (2.3) 0 (0.0) NSKlebsiella 0 (0.0) 1 (10.0) NSSerratia marcescens 0 (0.0) 1 (10.0) NSCandida 2 (4.6) 0 (0.0) NS

NEC, n (%) 92 (24.1) 20 (18.7) NSIVH grades 3 and 4, n (%) 54 (18.1) 17 (17.0) NSOcclusion, n (%) 25 (6.7) 8 (7.5) NSLeakage, n (%) 25 (6.7) 3 (2.8) NSPhlebitis, n (%) 21 (5.7) 6 (5.6) NSCholestasis 112 (30.0) 25 (21.5) �.05Maximum serum

creatinine, median(IQR), mg/dL

1.100 (0.975–1.400) 1.200 (1.000–1.500) NS

NEC indicates necrotizing enterocolitis; IVH, intraventricular hemorrhage; PICC, peripherallyinserted central catheter; CRBSI, catheter-related bloodstream infection.

e1154 HOANG et al by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

pared with 11 days in the upper; however, it did notreach statistical significance (P � .05; Fig 1A and B).Last, we examined the role of TPN with lipid emulsionand CRBSI. In the lower extremity, the duration of TPNadministration was 34 and 32 days, respectively, withand without CRBSI (Fig 1A and B); however, in theupper extremity, the median duration of TPN use was 46days as compared with 25 days without CRBSI (P �.0001; Fig 1A).

InfusatesMore than 96% of PICCs that were placed had TPN as aninfusate, including lipids (Table 3). We considered infu-sates with extremes of pH, �6 or �8, and when theywere statistically different between the upper and lowerextremity groups. Within this category, there are somethat we identified for study: vancomycin (pH: �4), phe-nobarbital (pH: 8.5–10.5), and doxapram (pH: 3.5–5.5).The administration of Vancomycin and phenobarbitalwas similar in both groups; however, doxapram wasused more often in the lower extremity (P � .001).Furthermore, we looked at the effect of vasopressors and

hydrocortisone use related to the rate of complications.Dopamine and dobutamine were used more often in thelower extremity (P � .005), whereas hydrocortisone usewas similar in both groups.

Risk Factors Leading to CholestasisNeonates with lower gestational age and birth weighthad statistically higher rates of cholestasis in the upperextremity (P � .005 and P � .0001, respectively; Fig 2A

0

1

2

Med

ian

no. o

f PIC

Cs

no yes no yes

Lower Extremity Upper Extremity

Septicemia

0

100

200

300

400

500

600

700

800

900

1000

Med

ian

Birt

hwei

ght,

g

no yes no yes

Lower Extremity Upper Extremity Septicemia

A

D

Gestational age, wk

Duration PICC, d

Duration TPN, d

0

5

10

15

20

25

30

35

40

Med

ian

for

each

gro

up

no yesSepticemia

0

10

20

30

40

50

Med

ian

for e

ach

grou

p

no yesSepticemia

Upper Extremity Lower Extremity

B

a

b

aa

C

FIGURE 1A and B, Relationship between gestational age (weeks), duration of PICC (days), duration of TPN (days), and the incidence of septicemia in the upper extremity (A) and the lowerextremity (B). C, Relationship between themedian numbers of PICCs and the incidence of septicemia in the upper and lower extremities. D, Relationship between birth weight and theincidence of septicemia in the upper and lower extremity. a P � .0001. b P � .005.

TABLE 3 Percentage of Patients and Type of InfusatesAdministered Through PICCs

Infusates, n (%) Upper Extremity(n � 370 PICCs), n (%)

Lower Extremity(n � 107 PICCs), n (%)

P

TPN 358 (96.7) 103 (96.3)Vancomycin 107 (28.9) 36 (33.6)Phenobarbital 29 (7.8) 9 (8.4)Doxapram 10 (2.7) 19 (17.8) �.001Dopamine/dobutamine 62 (16.7) 31 (28.9) �.005Hydrocortisone 58 (15.7) 17 (15.9)

PICC indicates percutaneously inserted central catheter; TPN, total parenteral nutrition.

PEDIATRICS Volume 121, Number 5, May 2008 e1155 by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

and D). There was no such association seen in the lowerextremity (Fig 2A and B). Duration of TPN was signifi-cantly related to the development of cholestasis in bothextremities (P � .001 for both; Fig 2A and B). Neonateswith cholestasis had an increased number of PICCsplaced (P � .0001; Fig 2C). The duration of PICCs inplace was significant for the occurrence of cholestasis;however, this was observed only in the upper extremity(P � .0001; Fig 2A). For additional delineation of theroles of duration of PICCs versus TPN, a receiver oper-ating characteristic curve was drawn (Fig 3). The dura-tion of PICC use can predict cholestasis earlier than theduration of TPN administration. Between 20 and 30days, the sensitivity (true-positive) of predicting cho-lestasis ranged from 0.5 to 0.9, respectively. Conversely,for the duration of TPN to reach the same sensitivity, ittook 60 to 100 days. In the lower extremity, the durationof PICC was able to predict cholestasis quicker than theduration of TPN (Fig 4). Within the lower extremity, therange of PICC days was 20 to 90 days, signifying thatthe PICC was able to stay longer before sensitivityreached 0.9. These data are consistent with Table 1, inwhich the duration of PICC in place in the lower ex-

tremity was longer but had lower rates of cholestasis(Table 2).

Additional analysis of confounding variables leadingto cholestasis was analyzed. Urinary tract infections,Gram-negative septicemia, CRBSI, and Gram-negativetracheal aspirates were statistically significant for con-tributing to cholestasis; however, this significance wasnot seen when upper and lower extremity groupswere compared. Multivariate logistic analysis revealedthat Gram-negative septicemia along with identifica-tion of a Gram-negative organism in the tracheal as-pirate, duration of PICC in place, and duration of TPNhad the most significant roles in contributing to cho-lestasis. Subsequent univariate analysis revealed thatonly the duration of TPN use was a significant con-tributor of cholestasis.

DISCUSSIONThis observational report represents the largest retro-spective study of PICC-related rates of complicationswith 396 infants with a median gestational age of 28weeks and birth weight of 940 g. We analyzed the com-

Gestataional age, wk

Duration PICC, d

Duration TPN, d

0

100

200

300

400

500

600

700

800

900

1000

1100

Med

ian

Birt

hwei

ght,

g

no yes no yes

Lower Extremity Upper Extremity

Cholestasis

0

1

2

Med

ian

no. o

f PIC

Cs

no yes no yes

Lower Extremity Upper Extremity

Cholestasis

0

10

20

30

40

50

60

70

Med

ian

for e

ach

grou

p

no yes CholestasisLower Extremity

0

10

20

30

40

50

60

Med

ian

for e

ach

grou

p

no yes Cholestasis

Upper Extremity

A

C D

B

a

ab

a

a

a

FIGURE 2A and B, Relationship between gestational age (weeks), duration of PICC (days), duration of TPN (days), and the incidence of cholestasis in the upper extremity (A) and the lowerextremity (B). C, Relationship between themedian numbers of PICCs and the incidence of cholestasis in the upper and lower extremities. D, Relationship between birth weight and theincidence of cholestasis in the upper and lower extremity. P � .0001 and � P � .005.

e1156 HOANG et al by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

plications rate between the upper and lower extremityinserted PICCs. Similar to findings of Freeman et al,36

decreasing birth weight, gestational age, and longerlength of stay increased CRBSI. The occurrence of CRBSIand cholestasis was significantly higher with PICCs in-serted from the upper extremities; therefore, we do notaccept our hypothesis that placement of PICCs in eitherthe upper extremity or the lower extremity has no dif-ference in catheter-related complications.

CRBSI is the major cause of morbidity and mortalityin NICU patients. The most common source is the skinflora contaminating either the catheter hub or the cath-eter itself. Janes et al37 conducted a prospective study ofextremely low birth weight infants and reported com-plication rates of CRBSI at 28% incidence of firstinfection and 34% for first and subsequent infections;however, their definition of CRBSI was different fromours, leading to different rates of its occurrence. In ourstudy, CRBSI was defined as a positive blood cultureresult from a PICC and a peripheral vessel, whereas thestudy by Janes et al37 defined CRBSI as any positiveblood culture result without simultaneous peripheral

vessel and PICC blood cultures. In another study ofPICCs, Cairns et al32 also reported a higher rate of CRBSIof 31.1%. These significantly higher reported rates ofsepticemia may be in part attributable to a smaller sam-ple size of 32 and 61 PICCs, respectively, as comparedwith 477 PICCs in our study. On additional analysis, wealso found that although coagulase-negative Staphylococ-cus was the major organism in both the upper and lowerextremity PICCs, lower extremity PICCs had a slightlyhigher prevalence of Gram-negative organisms causingsepticemia.

The type of solutions being infused via PICCs may con-tribute to the complication rate. The pH and osmolarity ofthe infusate, possible contamination of infusate, and therate of infusion are significant factors for the developmentof complications. For example, vancomycin, a commonlyused antibiotic for staphylococcal infections in the NICU, ishyperosmolar with an extremely low pH (2.4).

The pathogenesis of cholestasis of the neonate is mul-tifactorial. Processes in the development of cholestasisinclude increased bilirubin load from hemolysis, hepa-

A

Sen

sitiv

ity

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

20 days

.00 .10 .20 .30 .40 .50 .60 .70 .80 .90 1.00

1-Specificity

30 days

B

Sen

sitiv

ity

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

60 days

.00 .10 .20 .30 .40 .50 .60 .70 .80 .90 1.001-Specificity

100 days

FIGURE 3Receiver operator characteristic curve for upper extremity PICCs. A, Duration of PICC(days) and the sensitivity versus false-positive in predicting cholestasis. B, Duration of TPN(days) and the sensitivity versus false-positive in predicting cholestasis. Area under thecurve � 0.85.

A

Sen

sitiv

ity

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

20 days

.00 .10 .20 .30 .40 .50 .60 .70 .80 .90 1.00

1-Specificity

90 days

40 days

B

Sen

sitiv

ity

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

60 days

.00 .10 .20 .30 .40 .50 .60 .70 .80 .90 1.001-Specificity

110 days

FIGURE 4Receiver operator characteristic curve for lower extremity PICCs. A, Duration of PICC(days) and the sensitivity versus false-positive in predicting cholestasis. B, Duration of TPN(days) and the sensitivity versus false-positive in predicting cholestasis. Area under thecurve � 0.83.

PEDIATRICS Volume 121, Number 5, May 2008 e1157 by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

tocellular injury, septicemia-induced cholestasis, drugssuch as penicillin/cephalosporins and sulfa-containingantibiotics, prolonged use of TPN, and decreased bileflow such as extrahepatic cholestasis from obstruction ofthe hepatic or common bile duct. The relationship be-tween septicemia and cholestasis first described as pneu-monia biliosa was reported as early as 1837.38 Septicemia-induced cholestasis, especially urinary tract infection andGram-negative septicemia, in neonates has the highestassociation, although meningitis, omphalitis, and Gram-positive bacteria have also been reported.39–42 Similar toour findings, neonates with urinary tract infections,Gram-negative septicemia, and colonization of trachealaspirate with Gram-negative bacteria statistically hadhigher rates of cholestasis, but when analyzed for thedifference in extremities, they did not contribute to thehigher rate of cholestasis seen with upper extremityPICCs; therefore, the cause of higher rates of cholestasisin the upper extremity must lie with another explana-tion.

In preterm infants between 32 and 36 weeks of ges-tation, a reduction in enterohepatic circulation of bilesalts can lead to cholestasis by intestinal stasis andbacterial overgrowth with subsequent increase in litho-cholic acid and impairment of bile flow.43 With Gram-negative septicemia, there is a decrease in bile secretionsecondary to endotoxin production,44 hemolysis of redblood cells,45 and hepatic dysfunction secondary to de-creased hepatic blood flow from hypotension or pro-longed hypoxia. As a result, hepatic blood flow alongwith intestinal stasis and bacterial translocation can in-crease the risk for cholestasis. In animal studies, TPN hasbeen known to impair intestinal mucosal immunity andacquiesce bacterial translocation, leading to an increasedrisk for cholestasis.46,47 There may be corroborating evi-dence in that although the duration of TPN correlateswith the rate for cholestasis, the duration of PICCs hadbetter sensitivity in predicting patients who were at riskfor cholestasis as seen by the receiver operating charac-teristic curve.

CONCLUSIONSUpper extremity and lower extremity PICCs have statis-tically significant different rates of complications for co-agulase-negative Staphylococcus septicemia and the rate ofcholestasis, both of which were higher with the upperextremity PICCs. There was a slightly higher prevalenceof Gram-negative CRBSI with lower extremity PICCs.PICCs inserted from the lower extremity remained func-tional for a longer period and had lower rate of overallcomplication compared with PICCs inserted from theupper extremity. When technically possible, use ofPICCs inserted from the lower extremity should be con-sidered for prolonged administration of TPN in ill neo-nates.

REFERENCES1. Dudrick SJ, Wilmore DW, Vars HM, Rhoads JE. Long-term

total parenteral nutrition with growth, development and pos-itive nitrogen balance. Surgery. 1968;64(1):134–142

2. Shaw JCL. Parenteral nutrition in the management of sick lowbirth weight infants. Pediatr Clin North Am. 1973;20(2):333–358

3. Thiagarajan RR, Bratton SL, Gettmann T, Ramamoorthy C.Efficacy of peripherally inserted central venous cathetersplaced in noncentral veins. Arch Pediatr Adolesc Med. 1998;152(5):436–439

4. Peripherally inserted central catheters. Intravenous Nurses So-ciety. J Intraven Nurs. 1997;20(4):172–174

5. Noble WC. Skin microbiology: coming of age. J Med Microbiol.1984;17(1):1–12

6. Roth RR, James WD. Microbiology ecology of the skin. AnnuRev Microbiol. 1988;42:441–446

7. Infusion nursing standards of practice. Intravenous Nurses So-ciety. J Infus Nurs. 2000;23(suppl 6):S1–S88

8. National Association of Vascular Access Networks. Tip locationof peripherally inserted central catheters. J Vasc Access Devices.1998;2:8–10

9. Pettit J, Wyckoff MM. Peripherally Inserted Central Catheters:Guideline for Practice. Genview, IL: National Association of Neo-natal Nurses; 2001

10. Pearson ML. Guideline for prevention of intravascular device-related infections: part I—intravascular device-relatedinfections: an overview. The Hospital Infection Control Prac-tices Advisory Committee. Am J Infect Control. 1996;24(4):262–277

11. Schelonka RL, Chai MK, Yoder BA, Hensley D, Brockett RM,Ascher DP. Volume of blood required to detect common neo-natal pathogens. J Pediatr. 1996;129(2):275–278

12. Brown DR, Kutler D, Rai B, Chan T, Cohen M. Bacterialconcentration and blood volume required for a positive bloodculture. J Perinatol. 1995;15(2):157–159

13. Gibb AP, Hill B, Chorel B, Brant R. Reduction in blood culturecontamination rate by feedback to phlebotomists. Arch PatholLab Med. 1997;121(5):503–507

14. Golombek SG, Rohan AJ, Parvez B, Salice AL, LaGamma EF.“Proactive” management of percutaneously inserted centralcatheters results in decreased incidence of infection in theELBW population. J Perinatol. 2002;22(3):209–213

15. Klein JF, Shahrivar F. Use of percutaneous silastic centralvenous catheters in neonates and the management of infec-tious complications. Am J Perinatol. 1992;9(4):261–264

16. Evans M, Lentsch D. Percutaneously inserted polyurethanecentral catheters in the NICU: one unit’s experience. NeonatalNetw. 1999;18(6):37–46

17. Racadio JM, Johnson ND, Doelman DA. Peripherally insertedcentral venous catheters: success of scalp-vein access in infantsand newborns. Radiology. 1999;210(3):858–860

18. Kelly RE, Croitoru DP, Nuss D, Flemmer LS, Bass WT. Choos-ing venous access in the extremely low birth weight (ELBW)infant: percutaneous central venous lines and peripherally in-serted catheters. Neonatal Intensive Care. 1997;10(Sept/Oct):15–18

19. Oellrich RG, Murphy MR, Goldberg LA, Aggarwal R. The per-cutaneous central venous catheter for small of ill infants. MCNAm J Matern Child Nurs. 1991;16(2):92–96

20. Goutail-Flaud MF, Sfez M, Berg A, et al. Central venous cath-eter-related complications in newborns and infants: a 587-casesurvey. J Pediatr Surg. 1991;26(6):645–650

21. Chathas MK, Paton JB. Sepsis outcomes in infants and childrenwith central venous catheters: percutaneous versus surgicalinsertion. J Obstet Gynecol Neonatal Nurs. 1996;25(6):500–506

22. Durand M, Ramanathan R, Martinelli B, Tolentino M. Prospec-tive evaluation of percutaneous central venous silastic cathe-ters in newborn infants with birth weights of 510 to 3,920grams. Pediatrics. 1986;78(2):245–250

23. Puntis JWL. Percutaneous insertion of central venous feedingcatheters. Arch Dis Child. 1986;61(11):1138–1140

e1158 HOANG et al by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

24. Loeff DS, Matalak ME, Black RE, Overall JC, Dolcourt JL,Johnson DG. Insertion of a small central venous catheter inneonates and young infants. J Pediatr Surg. 1982;17(6):944–949

25. Harms K, Herting E, Kron M, Schiffmann H, Schulz-Ehlbeck H.Randomized, controlled trial of amoxicillin prophylaxis forprevention of catheter-related infections in newborn infantswith central venous silicone elastomer catheters. J Pediatr.1995;127(4):615–619

26. Klein MD, Rudd M. Successful central venous catheter place-ment from peripheral subcutaneous veins in children. Anesthe-siology. 1980;52(5):447–448

27. Leick-Rude MK. Use of percutaneous silastic intravascularcatheters in high-risk neonates. Neonatal Netw. 1990;9(1):17–25

28. Nakamura KT, Sato Y, Erenberg A. Evaluation of a percutane-ously placed 27-gauge central venous catheter in neonatesweighing less than 1200 grams. JPEN J Parenter Enteral Nutr.1990;14(3):295–299

29. Rudin C, Nars PW. A comparative study of two different per-cutaneous venous catheters in newborn infants. Eur J Pediatr.1990;150(2):119–124

30. Sherman MP, Vitale DE, McLaughlin GW, Goetzman BW.Percutaneous and surgical placement of fine silicone elastomercentral catheters in high-risk newborns. JPEN J Parenter EnteralNutr. 1983;7(1):75–78

31. Shulman RJ, Pokorny WJ, Martin CG, Petitt R, Baldaia L,Roney D. Comparison of percutaneous and surgical placementof central venous catheters in neonates. J Pediatr Surg. 1986;21(4):348–350

32. Cairns PA, Wilson DC, McClure BG, Halliday HL, McReid M.Percutaneous central venous catheter use in very low birthweight neonate. Eur J Pediatr. 1995;154(2):145–147

33. Aggarwal R, Downe L. Use of percutaneous silastic centralvenous catheters in the management of newborn infants. In-dian Pediatr. 2001;38(8):889–892

34. Neubauer AP. Percutaneous central IV access in the neonate:experience with 535 silastic catheters. Acta Paediatr. 1995;84(7):756–760

35. Safdar N, Maki DG. Inflammation at the insertion site is not

predictive of catheter-related bloodstream infection with short-term, noncuffed central venous catheters. Crit Care Med. 2002;30(12):2632–2635

36. Freeman J, Platt R, Epstein MF, Smith NE, Sidebottom DG,Goldmann DA. Birth weight and length of stay as determinantsof nosocomial coagulase-negative staphylococcal bacteremia inneonatal intensive care unit populations: potential for con-founding. Am J Epidemiol. 1990;132(6):1130–1140

37. Janes M, Kalyn A, Pinelli J, Paes B. A randomized trial com-paring peripherally inserted central catheters and peripheralintravenous catheters in infants with very low birth weight.J Pediatr Surg. 2000;35(7):1040–1044

38. Garvin IP. Remarks on pneumonia biliosa. S Med Surg. 1837;1:536–544

39. Bernstein J, Brown AK. Sepsis and jaundice in early infancy.Pediatrics. 1962;29:873–882

40. Hamilton JR, Sass-Kortsak A. Jaundice associated with severebacterial infection in young infants. J Pediatr. 1963;63:121–132

41. Ng SH. Rawston JR. Urinary tract infections presenting withjaundice. Arch Dis Child. 1971;46(246):173–176

42. Rooney JC, Hill DJ, Danks DM. Jaundice associated with bac-terial infection in the newborn. Am J Dis Child. 1971;122(1):39–41

43. Watkins JB, Szczepanik P, Gould JB, Klein P, Lester R. Bile saltmetabolism in the human premature infant: preliminary ob-servations of pool size and synthesis rate following prenataladministration of dexamethasone and phenobarbital. Gastroen-terology. 1975;69(3):706–713

44. Utili R, Abernassa CO, Zimmerman HG. Endotoxin effects onthe liver. Life Sci. 1977;20(4):553–568

45. Shander A. Anemia in the critically ill. Crit Care Clin. 2004;20(2):159–178

46. Alverdy JC, Aoys F, Moss GS. Total parenteral nutrition pro-motes bacterial translocation from the gut. Surgery. 1988;104(2):185–190

47. Spaeth G, Specian RD, Berg RD, Dietch EA. Bulk preventsbacterial translocation induced by the oral administration oftotal parenteral nutrition solution. JPEN J Parenter Enteral Nutr.1990;14(5):442–447

PEDIATRICS Volume 121, Number 5, May 2008 e1159 by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from

DOI: 10.1542/peds.2007-1962; originally published online April 7, 2008; 2008;121;e1152Pediatrics

Houchang D. ModanlouViet Hoang, Jack Sills, Michelle Chandler, Erin Busalani, Robin Clifton-Koeppel and

InsertionNeonates: Complications Rates Related to Upper Versus Lower Extremity Percutaneously Inserted Central Catheter for Total Parenteral Nutrition in

  

ServicesUpdated Information &

htmlhttp://pediatrics.aappublications.org/content/121/5/e1152.full.including high resolution figures, can be found at:

References

html#ref-list-1http://pediatrics.aappublications.org/content/121/5/e1152.full.at:This article cites 46 articles, 10 of which can be accessed free

Citations

html#related-urlshttp://pediatrics.aappublications.org/content/121/5/e1152.full.This article has been cited by 4 HighWire-hosted articles:

Subspecialty Collections

nd_metabolismhttp://pediatrics.aappublications.org/cgi/collection/nutrition_aNutrition & Metabolismthe following collection(s):This article, along with others on similar topics, appears in

Permissions & Licensing

mlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhttables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on June 4, 2013pediatrics.aappublications.orgDownloaded from