Peripheral Intravenous Catheter Complications in Critically Ill Children: A Prospective Study Jeffery S. Garland, Peter Havens, W.

Michael Dunne, Jr, Mary Hintermeyer, Mary Anne Bozzette, Jeni Wincek, Tina Bromberger and Michelle Seavers Pediatrics 1992;89;1145

The online version of this article, along with updated information and services, is located on the World Wide Web at:
http://pediatrics.aappublications.org/content/89/6/1145

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007. Copyright 1992 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August 26, 2013

Peripheral

Intravenous

Catheter Complications A Prospective Study

S. Garland, Havens, RN; and

in Critically

Ill Children:

Jeffery Peter Mary Anne Bozzette,

MD*; W. Michael Dunne, Jr. PhD; MD, MS*; Mary Hintermeyer,RN; Jeni Wincek, RN, Seavers, MSN; RN Tina Bromberger, RN; Michelle

ABSTRACT. Six hundred fifty-four peripheral Teflon catheters in 303 pediatric intensive care unit patients were examined to determine complication rates and associated risk factors. Phlebitis, extravasation, and baderial colonization occurred at rates of 13%, 28%, and 11%, respectively. Logistic regression of factors that increased phlebitis risk revealed infusion of hyperalimentation (odds ratio 2.9) or lorazepam (odds ratio 2.2) and catheter location (odds ratio 2.9) as the most important determinants of phlebitis risk. Age (1 year, odds ratio 2.0), catheter time in situ (72 hours, odds ratio 2.1), and infusion of antiepileptics (odds ratio 2.1) were the most important determinants of extravasation. Catheters were colonized most frequently with coagulase-negative Staphylococcus (51/54). Sepsis attributable to catheter colonization occurred in 1 patient. Duration of catheter placement (144 hours, odds ratio 5.8) was an important determinant of catheter colonization. Colonization risk increased from 11% in catheters that were in situ for 48 to 144 hours to 34% for catheters that were in for longer than 144 hours. Infusion of diazepam (odds ratio 11.0) or lipid emulsions (odds ratio 2.5) and age (1 year, odds ratio 2.2) were also important determinants of colonization risk. Replacing catheters in critically ill children every 72 hours would not decrease phlebitis, bacterial colonization, or catheter-induced sepsis and could increase extravasation risk. Catheters can be safely maintamed with adequate monitoring for up to 144 hours in critically ill children. Pediatrics 1992;89:1145-1150; intravenous catheters, catheter complications, pediatric intensive care, phlebitis, catheter colonization, extravasation. Phlebitis, extravasation, and catheter colonization are complications that may occur during peripheral intravenous therapy with Teflon catheters.4 Although several recent studies have examined peripheral intravenous complications in general pediatric patients,25 only one study has examined complications in children admitted to a pediatric intensive care unit.6 In that study, Damen and Tweel6 evaluated children recovering from cardiac surgery whose catheters remained in situ a median time of 27 hours.

Complications other than bacterial colonization were not evaluated. Complication rates and associated risk factors occurring during intravenous therapy in children on a general pediatric ward may not apply to children in an intensive care unit who may have different diseases, receive different medications, and are at a greater risk for nosocomial infections.7 This study was designed to determine the incidence of catheter complications and associated risk factors during peripheral intravenous therapy in critically ill children.

METHODS
All children admitted to the Pediatric Intensive Care Unit at Childrens Hospital of Wisconsin between February and October of 1988 who required peripheral intravenous therapy with a Teflon catheter for greater than 12 hours were eligible for the study. A maximum of four intravenous catheters per patient were evaluated.

Intravenous

lines

were

placed

by resident

or attending

physicians

or nursing staff in the operating room, intensive care unit, emergency department, or general ward. Intravenous teams and local antibiotic ointments were not used. Intravenous catheter sites were inspected hourly by the nursing staff and daily by one of the study nurses for signs of complications. Previously described definitions for phlebitis, suppurative phlebitis,

and extravasation
following insertion

were

used.48
pain,

Phlebitis

was present

if three

of the

were present: site. Suppurative

erythema, swelling, phlebitis was present

or a cord at the if pus could be

expressed

from

the insertion
in a catheter surrounding line was no

site of a phlebitic

catheter.

Extrava-

sation was present flashback, and had or the intravenous

if it could not be flushed, had no edema. If a complication occurred longer required, the appearance of

the site was recorded and a culture of the skin was obtained using a saline-soaked sterile swab. The skin around the catheter was then cleansed with alcohol before removal. Catheters were removed and placed in sterile tubes for transport to the microbiology laboratory.
If a complication investigator, until the occurred the complication site was resolved. followed If no daily, by complications an

occurred removal

the site was of the catheter.

examined Catheters

for at least 72 hours placed after discharge

hours of catheter

following from the

removal.

intensive care All cultures

unit were not followed. were done within 12

Catheter
blade and

tips with

were

separated
into distal

from
and

the hubs
proximal

with
halves.

a sterile
Segments

scalpel
were

divided

flushed
placed collected.

1 .0 mL of sterile

saline

using

a tuberculin
the saline flush surface of the

syringe
had hubs then

and
been with sub-

in screw-capped tubes in which After disinfecting the exterior

From

the

Department Milwaukee.

of Pediatrics

and

Department of Nursing,

of Pathology, Childrens

Medical Hospital of

alcohol,
with

the interior
a sterile

aspects

of the hubs
swab. The bath

were

thoroughly
were model

scoured tube.
450,

College
Wisconsin,

of Wisconsin;

and

Department

saline-moistened

swabs (EMC

merged
accepted Nov 12, 1991. Interscience Conference
TX, October 1989.

into
were

1.0 mL of sterile
placed into a sonic

saline
water

in a screw-capped

All
RAI

Received Presented,
crobial

for publication Aug 22, 1991; in part, at the 29th Annual

Agents and Chemotherapy,

tubes on Antimi-

Houston,

Reprint requests to (J.S.G.) 33 Southbourne PEDIATRICS (ISSN 0031 4005). Copyright

emy of Pediatrics.

Rd, Jamaica Plain, MA 02130. 1992 by the American Acad-

Research Corp, Hauppage, NY) and oscillated for 1 minute to remove adherent organisms. One hundred microliters of each eluant was spread evenly over the surface of a blood agar plate and incubated at 35#{176}C for a minimum of 12 hours before examination.9 Cultures were negative if there was no visible growth after

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on No. August 26, 2013 PEDIATRICS Vol. 89 6 June 1992

1145

72 hours of incubation. Catheter the colony counts exceeded 15 cultures) using the semiquantitative

tips were considered colonized if (sum of distal and proximal tip roll culture technique#{176} or 15

TABLE 1.

Study

Group

Characteristics
No.

(%)

colony-forming units/mL following sonic oscillation.9 Microorganisms were identified by Gram stain morphology, standard biochemical reactions, and commercial identification systems. Staphylococcal isolates were speciated using STAPHYLOS-

Sex
Male Race White Black Other Site Hand 365 (56) 493 (76) 125 (19) 36 (5) or foot fossa 442 (67) 58 (9) 55 (8) 52 (8) 45 (7) 156 186 89 91 57 75
280

LIDE

(BBL

Microbiology
system

Systems,
(Analytab

Cockeysville,
Products, Inc,

MD)
Plainview,

and

the
NY).

STAPH-INDENT

Meritec (Merdian Diagnostics, Inc, Cincinnati, OH) was used to serogroup fl-hemolytic streptoccocal isolates. Susceptibility patterns
of bacterial isolates were determined System or standard disk diffusion coagulase-negative Staphylococcus ing to the method of Christensen was present if there was a clinical using the Vitek Automicrobic testing. Slime production by (CONS) was evaluated accordet al Catheter-induced sepsis picture compatible with sepsis

Head
Antecubital Wrist Arm or leg Age 0-1 mo 1-12 mo 12-24 mo 2-5 y 5-10 y >10 y

(fever,
obvious organism

leukocytosis,
infection, grown and from

a change

in mental

status)

without
with Peripheral that

a focus
the same blood

of

the catheter a peripheral with clinical

tip was colonized blood culture.

cultures
Episodes

were

done

at the discretion

of the intensive
signs of sepsis

care unit team.

resolved with

associated

catheter removal in patients with positive catheter tip cultures and negative blood cultures were considered suspected sepsis episodes. Blood cultures, chest radiographs, and body fluid cultures were
obtained when septicemia or other infections were suspected.

(24) (28) (14) (28) (9) (11)

(43)

Location of placement Operating room

Intensive Emergency Ward Other care unit department

DATA
Data collected on admission formation, primary diagnosis,
tomic location of catheter, and

ANALYSIS
included patient location of catheter
catheter size. Data

239 (36) 109 (17) 20 (3)

6 (1)

demographic placement,
collected

inanadaily

included
intravenous pressure.

the appearance
fluid Length

of the intravenous

site,

medications

and

given through of time the

the catheter, and catheter remained

pump occlusion in situ and the

appearance
recorded.

of the

site

at the

time

of catheter

removal

was

also

Categorical and continuous data were analyzed using x2 and Students t tests, respectively. Relative risks and confidence intervals were calculated for factors associated with each complication. Factors were considered significantly associated with each complication if the relative univariate analysis.
used icant
1.5

risk was 1.5 and the For each complication,

importance analysis.

P value was <.01 logistic regression

in the was

to evaluate the by the univariate

of the factors identified as signifOnly factors with a relative risk of

in the univariate analysis were included in the model. For each of the complications the best logistic regression model was defined as the one that maximized the model x2 (calculated using

2-times Analysis

log likelihood ratio). All data System version 6.03.

were

analyzed

using

Statistical

RESULTS We studied 654 catheters in 303 children. Study group characteristics are shown in Table 1 . The average length of time catheters remained in situ was 83 47 hours (SD) with a range of 12 to 274 hours. Complications developed in 50% (164/325) of catheters remaining in situ for greater than 72 hours. This rate did not change for 128 catheters that remained in situ for greater than 120 hours. Since complication rates were similar for the first and last catheters placed in each patient, all catheters were combined for analysis. Phlebitis Phlebitis occurred mean time of onset 49 hours (SD) with Resolution of phlebitis (SD) with a range of Results of univanate phlebitis are shown tation with continuous infusion of lorazepam in 83 (13%) of 654 catheters. The of a phlebitis episode was 86 a range of 12 to 286 hours. occurred after 56 29 hours 13 to 120 hours. analysis of risk factors for in Table 2. Use of hyperalimenintravenous lipid emulsions, or aminophylline, catheter lo-

cation (antecubital fossa, arm or leg), and the length of time spent in the intensive care unit before catheter insertion increased phlebitis risk. Catheters placed in the operating room, inserted in surgical patients, or through which blood products were transfused had lower phlebitis rates. Phlebitis was not associated with the length of time catheters were in situ. Results of the logistic regression analysis are also shown in Table 2. Infusion of hyperalimentation with continuous intravenous lipid emulsions (odds ratio 2.9, P .002) or lorazepam (odds ratio 2.2, P .001), catheter location (antecubital fossa, arm, wrist, or leg: other sites [odds ratio 2.9, P .0003]), and length of time patients were in the intensive care unit before catheter insertion (>48 hours: 48 hours, odds ratio 1 .8, P .04) increased phlebitis risk.
= = = =

Extravasation Extravasation occurred in 179 (28%) of 654 catheters. The mean time to an extravasation episode was 70 37 hours (SD) with a range 13 to 202 hours. Extravasation episodes decreased with time. If catheters were in situ for at least 72 hours, the incidence of extravasation decreased from 34% to 24% (P < .0001). Two extravasations resulted in skin sloughing. Results of univariate analysis of factors associated with extravasation are shown in Table 3. Catheter location (hand, foot, head, and leg), patient age (1 year), time in situ (72 hours), patient service (medical > surgical), and infusion of antiepileptics (brazepam, diazepam, phenytoin, phenobarbitab) or brazepam alone all increased the risk of extravasation. Logistic regression showed antiepileptics (odds ratio 2.1, P .0001), age (1 year: >1 year, odds ratio 2.0, P .0003), and time in situ (72 hours: >72 hours, odds ratio 2.1, P .0001) as the most important determinants of extravasation risk (Table 3).
= = =

1146

PERIPHERAL INTRAVENOUS CATHETER COMPLICATIONS Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August 26, 2013

TABLE

2.

Risk of Phlebitis Factor Univanate Analysis Relative

(95%

P Value
Location (antecubital fossa, arm, leg) .00001
.00001 .0001 .003 .005 .005 .009 .01 .01 Logistic Parameter Estimate Regression SE

Risk
CI)

2.0 (1.3,

3.0)

Hyperalimentation Medical patient Time in PICU (>48 h) Intralipids No blood products through Where placed (outside OR) Lorazepam Aminophylline drip Variable

3.0 (1.8, 4.8) 1.7 (1.1, 2.6) 1.9 (1.3, 2.3 (1.3, 1.9 (1.2, 2.0 (1.1, 2.2 (1.2, 2.0 (1.1, Analysis Odds (95% Ratio CI) 2.9) 4.1) 3.1) 3.2) 4.1) 3.6)

line

Location
leg)

of catheter

(antecubital

fossa,

arm,

1.03 1.07 0.779 0.597 -2.55

0.2850 0.3420 0.3101 0.2852

2.9 (1.60, 2.9 (1.49, 2.2 (1.19, 1.8 (1.04,

4.90) 5.70) 4.0) 3.17)

Hyperalimentation Lorazepam Time in ICU prior

(>48 h)

to catheter

placement

Intercept
S

Univariate care

analysis

of factors

that

increased room; ICU,

the

risk

of phlebitis

(top)

and

logistic

regression

analysis

of variables
intensive tModelx2=37.05,P=

that
unit;

increased
OR,

the

risk

of phlebitis

(bottom).
care

CI,
unit.

confidence

interval;

PICU,

pediatric

operating

intensive

.0001,df=4.

TABLE 3.

Risk of Extravasation
Factor

P Value
(72 h) .0001 .001 .002

Relative (95% 1.7 (1.3, 1.6 (1.3, 1.5(1.2,2.0) 1.5 (1.2,

Risk CI) 2.2) 2.1)

Length catheter in situ Antiepileptic drugs Age(1y)

Lorazepam
Medical patient

.008
.007

1.5 (1.1, 2.2)

1.9)

Location

(hand, Variable

foot,

head,

leg)

.01 Logistic
Parameter Estimate

2.2 (1 .5, 3.3) Regression

SE

Analysis
Odds (95% Ratio

CI) 2.98)
3.16) 2.79)

Time

catheter

in situ (72
drugs

h)

0.7323
0.7501 0.6732

0.1840
0.2019 0.1858

2.1 (1.45,
2.1 2.0 (1.42, (1.36,

Antiepileptic Age (1 y)

Intercept
Univariate analysis analysis of variables
5

-1.9414
of factors that affected that increased extravasation
=

the risk (bottom).

of

extravasation CI, confidence

(top) and interval.

logistic

regression

t Model
Colonization

x2

41.040,

.0001, df

3.

459 cultured catheter tips were colonized. No factor, including length of time a catheter was in situ, increased or decreased the likelihood of a catheters being cultured. Seventeen of the 54 tips were colonized with at least two organisms and 4 of 54 were colonized with at beast three organisms. Coagulase-negative Staphylococcus cobonization gram, occurred most frequently. Using biotype, antibio-

Fifty-four

(1 1 .8%)

of

comprising 58 (74%) of Thirty (52%) of these Other species of CONS catheter tips included Staphylococcus warnerii molyticus (4).
Both Enterococcus

the 78 total CONS isolates. 58 isolates were 5+ strains. recovered from colonized Staphylococcus hominis (8), (8), and Staphylococcus haeand group A fl-hemolytic

faecalis

and slime production to differentiate strains of CONS, we found that 51 of the 54 colonized catheters were colonized by one or more strains of CONS. On 20 of the 51 CONS-colonized catheters, two or more strains with different sensitivities were isolated. Thirty-four (44%) of the 51 catheters were colonized with a slime-positive (5+) phenotype. Staphylococcus epidermidis was the most common species of CONS,

Streptococcus were Yeasts were recovered Candida parapsilosis,

lated from a single the time During

recovered from two catheters. from four catheters; three grew and Malassezia furfur was isocatheter. the catheters in situ, 545 study popufrom only catheter had been in a 10-monthwere The catheter was

blood lation.
one of boy

cultures were Catheter-induced

the colonized

obtained from the sepsis occurred

lines. epilepticus. The

placed
old

in the
with

emergency
status

department

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August ARTICLES 26, 2013

1147

removed 51 hours later because of phlebitis. A blood culture obtained 8 hours after the catheter was removed grew group A f3-hemolytic Streptococcus. The catheter tip (219 colonies) and hub (20 colonies) grew the same organism. The infant was treated for 7 days with ampicillin and discharged without sequelae. Six patients had cultures that were positive for S. epidermidis while catheter cultures were negative. Five of the six were considered skin contaminants and the patients were never treated. The sixth patient was symptomatic (fever, lethargy, and leukocytosis) and was treated with antibiotics for 7 days. The catheter tip grew only 13 colonies of S. epidermidis. In 38 colonized catheters the distal tip was cobnized and the hub was sterile. Hub colonization without proximal tip colonization occurred in only 1 catheter. Univariate analysis revealed length of time in situ as the most significant factor associated with cobonization (Table 4). Colonized lines (53/54 included in analysis) remained in place longer than sterile catheters (1 13 hours 54 hours vs 79 hours 38 hours [SDJ, P = .00001). As seen in the Figure, the prevalence of colonization increased with time (x2 for trend, 27.3, P < .00001). Catheter colonization rate was stable (1 1 %, 32/291) between 48 and 144 hours but increased threefold (34%, 18/53) in catheters in situ >144 hours. Other factors that significantly increased cobonization risk were age (1 year), patient admitting diagnosis (status epibepticus or trauma), and infusion of diazepam. However, only 15 catheters with diazepam infusions were cultured, and 1 4 of these catheters were started in the emergency department in patients who were seizing. Intravenous lipid emulsion use (relative risk 2.3, 95% confidence interval 1.1, 4.7) and hospital location of catheter insertion (emergency department [relative risk 1.6, 95% confidence interval 0.96, 2.67]) were associated with colonization but did not reach the predetermined level of statistical significance (Table 4). Phlebitis, extravasation, length of time the patient was in the intensive care unit before catheter insertion, and circumstances (emergent vs
TABLE 4. Risk of Colonization Factor

elective) of the catheter insertion were not associated with colonization. Logistic regression analysis (Table 4) revealed length of time in situ (>144 hours: 144 hours, odds ratio 5.8, P = .0001), patient age (1 year: >1 year, odds ratio 2.2, P = .0237), and infusion of lipid emulsions (odds ratio 2.5, P = .04) or diazepam (odds ratio 1 1 .0, P .0001) as the most important determinants of catheter colonization risk. Adding the variables admitting diagnosis (status epibepticus or trauma vs other diagnoses) and hospital location of catheter insertion (emergency department vs other locations) did not improve the model, but reduced the odds ratio for colonization associated with diazepam from 1 1 .0 to 4.9. Odds ratios for the other model variables were unchanged. DISCUSSION In this study of catheter complications during peripheral intravenous therapy with Teflon catheters in children in an intensive care unit, we found rates similar to previously reported complication rats in general pediatric patients.252 The rate of phlebitis (13%) in critically ill children was similar to the rate of phlebitis (10%) in general pediatric patients.2 Time to onset and resolution of phlebitis symptoms were greater than previous reports.23 Differences may have been due to study design or length of time catheters were in situ. As in our previous study, no patients hospital course was prolonged by an episode of phlebitis, nor was the there any episode of suppurative phlebitis. As in previous studies24 infusions of hyperabimentation fluids with continuous intravenous lipid emubsions increased phlebitis risk. The high osmobarity of hyperalimefltatiofl4 solutions may explain the association with phbebitis. Lipid emubsions have been shown to prolong catheter life in patients receiving hyperalimentation.5 Therefore, the phlebitis risk due to hyperalimentation alone may be even higher. The rate of phlebitis in intensive care patients older than 10 years of age (21%) was similar to rates reported in adults483 and older children.2 Phlebitis rates for chil-

Univariate

Analysis Relative (95% 4.0 4.6 3.8 2.2 (2.4, (1.7, (1.6, (1.2, Risk CI) 6.7) 12.8) 8.8) 4.2)

P
Time catheter Admit diagnosis trauma) Diazepam Age (s1 y) in situ (>144 h) (status epilepticus

Value

or multiple

.00001 .003 .004 .01

Variable Parameter Estimate Valium Time catheter Intralipids 2.4013 1.766 0.9071

Logistic

Regression SE

Analysist Odds (95% 11.0 5.8 2.4 (3.48, (2.86, (1.03, Ratio CI) 34.99) 12.0) 5.90)

in situ

(>144

h)

0.5887 0.3648 0.4460

Age (1
Intercept Univariate variables
S

y)
of factors that colonization

0.7885
-3.0892 increased colonization risk risk (bottom). CI, confidence
4.

0.3486
(top) and interval. logistic

2.2 (1.11,
regression

4.36)
analysis of

that

analysis increased 42.39,

tModel

x2

.0001,

df=

1148

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August 26, 2013 PERIPHERAL INTRAVENOUS CATHETER COMPLICATIONS

S
C

0 N

49-96

97-144

TIME Figure. Catheter < .00001).

CATHETER

IN SITU in situ

(HOURS)

colonization

vs time

(x2

for

trend,

27.3;

dren younger than 10 years were approximately onehalf this rate (1 1 %) and similar to the rate of younger general pediatric patients.2 Since phlebitis is probably the result of mechanical and physicochemical interactions,14 it is not surprising that certain medications, catheter location, and the length of time in the pediatric intensive care unit prior to catheter placement were all risk factors for phlebitis in this and other studies.34 Establishing and securing peripheral intravenous access is often difficult in criticably ill pediatric patients. Such patients often have a paucity of suitable, stable sites for catheter placement, especially as length of hospital stay increases. Movement of the catheter tip may damage the vessel wall and increase phlebitis risk. Extravasation rate (28%) in this population of intensive care patients was higher than the 16% rate in general pediatric patients2 but similar to rates reported in adults.4 Catheters in less than 72 hours were responsible for much of the increased rate of extravasation. The higher rate of extravasation in criticably ill children as well as the increased extravasation rate in catheters in less than 72 hours may be due to heightened nursing vigilance and caution with potenflabby toxic intravenous medications that are required in critically ill children. Leaving Teflon catheters in situ past 72 hours did not increase the risk of extravasation in critically ill children (20% extravasation rate after 72 hours). As in previous studies,53 children younger than 1 year of age were more likely to suffer an extravasalion. Phelps and Helms5 examined risk factors affecting extravasations in 151 peripheral catheters (56 steel needles) placed in children younger than 1 year of age. The rate of extravasation was 58%, more than the 33% (1 12/342) we found in children younger than 1 year of age. The 33% extravasation rate we noted in children younger than 1 year old is similar to the 25% rate reported in infants on a general ward who were younger than 1 year of age.2 Since catheter location is strongly associated with extravasation, difficulty in placing and securing catheters in this age group may explain the increased extravasation risk. Antiepileptic medications, drugs that are usually given rapidly in an emergent setting, increased the risk of extravasation in our study. Phelps and Helms5 showed an increased risk of extravasation if any medications were infused through the catheter. They

speculated that administration technique and not the medication may be the most important determinant of extravasation. Pushing a medication rather than giving it slowly could result in more extravasations. The bacterial colonization rate of catheters we report (1 1 .8%) is similar to the rate (1 0.6%) we noted in general pediatric patients whose catheters were in situ a mean of 59 hours. Batton et a13 reported a 4.0% rate in 25 preterm infants. Cronin et al6 reported a 13% rate of catheter colonization in neonates in a special care nursery. In the only study done of peripheral catheter colonization in children admitted to a pediatric intensive care unit, Damen and Tweel6 reported a 0.9% colonization (4/424). Their study included only postoperative congenital heart patients and the median time catheters were in situ was 27 hours. In our study catheters in situ less than 48 hours had a low rate of colonization (3/1 15, 2.6%). In postoperative heart patients, colonization did not occur in catheters in situ less than 48 hours. Similar to the results of a previous study performed in our institution, CONS represented the most frequentby recovered bacteria from colonized catheters. Staphylococcus epidermidis was the most common spedes isolated. Multiple strains of CONS with distinct biotypes, susceptibility patterns, or slime phenotypes were recovered from 20 of the 5 1 catheters colonized with this organism. To our knowledge, colonization of catheters with multiple strains of CONS has not been previously reported. Slime-producing CONS were more common (34/78 isolates of CONS) than in our earlier study of children on general wards. Although none of the lines colonized by CONS resulted in catheter-induced sepsis, the increased rate of slime-producing strains is alarming since virulence has been directly correlated with slime production. Diazepam infusion and continuous intravenous lipid emulsions were both associated with an increased risk of colonization. Others have reported an increased risk of catheter colonization and bacterial sepsis in patients, especially ?reterm neonates, who are receiving lipid infusions. An increased risk of sepsis in critically ill children receiving lipid emulsions has not been reported and requires further study because of their association with colonization. Although there was a strong association between catheter colonization and diazepam infusion, the attributabbe risk for colonization is small since only 15 catheters with diazepam infusions were cultured. Controlling for admitting diagnosis and hospital location of catheter insertion reduced the colonization odds ratio associated with diazepam infusion, suggesting that much of the colonization risk attributed to diazepam could be explained by the fact that these catheters were placed in acutely ill children who required rapid intravenous access in an emergency setting. Sitges-Serra et a12#{176} suggested that hub contamination may be an important cause of the central venous catheter-related bacteremia . However, Cooper and Hopkins2 reported that most bacteremias are derived from infections along the percutaneous tract. The results of our study support Cooper and Hopkins findings. In 38 of 54 colonized catheters in our study,

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on AugustARTICLES 26, 2013

1149

the proximal and distal tips of the catheter were colonized with organisms cultured from the skin while the hub was sterile. None of the catheters had distal tip and hub colonization without proximal tip colonization. Catheter-induced sepsis, as reported earlier, was uncommon (1/459). The single episode occurred in a catheter that was in situ less than 72 hours. Although the rate of colonization increased threefold in catheters that were in situ more than 144 hours, episodes of catheter-induced sepsis did not occur. Although the Centers for Disease Control22 recommends catheter rotation in adults receiving peripheral intravenous therapy, our results suggest that in critically ill children catheter-induced sepsis could not be reduced by rotating catheters to new sites at 72 hours, and perhaps extravasation rates would increase. Since the phlebitis rate does not increase with time, properly monitored peripheral intravenous Tefbon catheters can remain in situ for up to 144 hours in critically ill children. REFERENCES

8.

Tomford and Med.

JW, peripheral DJ,

Hershey venous

CO,

McLaren catheter-associated

CE,

et al. Intravenous complications.

therapy Arch

team Intern

1984;144:1191-1194 corrado ML, Seligman SJ. Quantitative and other intravascular inserts.
Weise CE, Sarafin HW. A semiquantitative infection.

9.

Cleri

culture

of intravenous
Dis. 1980;141: method for 1977;

catheters 781-786
10. MaId DB,

Infect

culture N Engl

identifying

intravenous Parisi

catheter-related JT, Bisno AL, negative TM,

Med.

296:1305-1309 I 1 . Christensen GD,

significant 1983; 12. Garland strains JS, Nelson 18:258-269

et al Characterization staphylococci. et al. Complications

of clinically

of coagulase DB, Johnson

Clin

Microbiol. intraRes. of

during Pediatr history

venous
1986;265A. 13. Hershey

therapy
Abstract CO.

with
Tomford

Teflon
JW,

catheters
McLaren CE,

or steel

needles.
natural Med.

et al The Arch Intern Br

intravenous 1373-1375
14. 15. Lewis 220-223 Phelps SJ, GB,

catheter-associated
Hecker Cochran JF. Infusion EB. Effect

phlebitis.
thrombophlebitis. of the

1984;144: 1985;57: of fat

I Anaesth.

continuous

administration

emulsion
peripheral 1989;13:628-632 16. Cronin nization Control 1 7. McKee

on

the

infiltration
nutrition

of intravenous
solutions. JPEN

lines

in infants
Parenter

receiving
Nutr.

parenteral

Enteral

WA, and Hosp

Germanson related Epidemiol. MA,

TP, bloodstream 1990;1 Green

Donowitz infection 1:301-308 HL, Schaffner

LG.

Intravascular in critically

catheter ill neonates.

cobInfect bacillary

KT, Melly

W. Gram-negative

1 . Garland

JS, Nelson DB, Cheah TE, et al. Infectious complications during peripheral intravenous therapy with Teflon catheters: a prospective study. Pediatr Infect Dis. 1987;6:918-921
DB, DG, Garland Maisels JS. The AJDC. M, natural history P. a controlled LM, Use of Teflon catheter associated phlebitis in children. in premature GH,

sepsis
AJDC. 18. JaMs Infect 19. Freeman lipid neonatal 20. Sitges-Serra step

associated
1979;133:649-650 WR, Dis. Highsmith

with

use
AK,

of lipid

emulsion
RW.

in parenterab
Polymicrobial

nutrition.
bacteremia

2. 3.

Nelson Batton

Allen

JR, Haley

1987;141:1090-1092 of peripheral
study. Pediatrics.

associated

with

lipid

emulsion DA, Smith

in a neonatal NE,

intensive

care unit.

Pediatr

Appelbaum infants: Baldini

intravenous
1982;70:

1983;2:203-208

cannulas 487-490
4. Tully

J, Goldman
emulsion intensive A, Puig during GL, Hopkins and care

et al. Association
staphylococcab

of intravenous
bacteremia as the Enteral initial Nutr. catheter N Engl in

coagulase-negative units. Perez N Engl JL,

JL, Friedland

et al. Complications

of intravenous Am

I Med.
Hub sepsis

1990;323:301-308 colonization due to coagulase

therapy
702-706 5. 6. Phelps venous Damen ated with

with
SJ, lines

steel

needles
RA. Risk

and
factors

Teflon

catheters.

Med. of

1981;70:
peripheral

P,

et al. nutrition.

in an outbreak

of catheter-related parenteral

negative

Helms

effecting

infiltration risk cardiac

staphylococci 1984;8:668-672 21. Cooper

JPEN of

Parenter intravascular

in infants.

J Pediatr.

J, Tweel IV. Positive

intravascular

1987;111:384-389 tip cultures and related

in pediatric

factors
patients.

associCrit

cc.

Rapid

diagnosis

catheterization

1988;16:221-228 7. Donowitz LG. High risk of nosocomial infection care patient. Crit Care Med. 1986;14:26-28
Care Med.

associated infection by direct JMed. 1985;312:1142-1147

22. Centers for Disease Control of intravenous therapy-rebated

Gram
Working

starting
Groups.

of catheter
Guidelines Infect Control.

segments.
for

in the pediatric

critical

prevention

infections.

1981;3:62-79

CONSENT

BASED

ON

TRUST

. .Granted informed on
.

the limits to medical matters, trust.

the possibility the real and

of non-medical effective basis

persons of consent

being is not

fully infor-

mation

but

Dunstan GR, Press; 1983.

Sellers

MJ. Consent

in Medicine:

Convergence

and Divergence.

Oxford:

Oxford

University

Submitted

by Student

1150

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August 26, 2013 PERIPHERAL INTRAVENOUS CATHETER COMPLICATIONS

Peripheral Intravenous Catheter Complications in Critically Ill Children: A Prospective Study Jeffery S. Garland, Peter Havens, W. Michael Dunne, Jr, Mary Hintermeyer, Mary Anne Bozzette, Jeni Wincek, Tina Bromberger and Michelle Seavers Pediatrics 1992;89;1145
Updated Information & Services Citations Permissions & Licensing including high resolution figures, can be found at: http://pediatrics.aappublications.org/content/89/6/1145 This article has been cited by 10 HighWire-hosted articles: http://pediatrics.aappublications.org/content/89/6/1145#related-urls Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://pediatrics.aappublications.org/site/misc/Permissions.xhtml Information about ordering reprints can be found online: http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Reprints

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007. Copyright 1992 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at Hospital Clinico Univ de Valladolid on August 26, 2013