Outcomes of upper extremity versus lower extremity placed peripherally inserted central catheters in a medical-surgical neonatal intensive care unit 1

Abstract

OBJECTIVE:

To compare outcomes of peripherally inserted central catheters (PICCs) placed in the upper extremity (UE) versus the lower extremity (LE) in a quaternary medical-surgical neonatal intensive care unit (NICU).

RESULTS:

We analyzed a total of 365 PICCs of which 250 (68%) were removed for end of therapy and 115 (32%) were removed due to complications. Patients who had UE insertions compared to LE insertions were of lower gestational age (median (IQR)), 30 (26, 35) vs. 32 (27, 37) weeks respectively (p = 0.014). UE PICCs were more likely to be removed due to complications compared to LE PICCs (39.9% vs. 26.4%, RR 1.51, 95% CI 1.12 –2.03, p = 0.007). UE PICCs were more likely than LE PICCs to be removed for the complications of malposition, dislodgement, and pleural or pericardial effusions; while LE PICCs were more likely to be removed for phlebitis. There were no differences in the rates of sepsis at 13.0% vs. 12.8% for UE vs. LE respectively, or causal organisms for sepsis. Survival analysis demonstrated that LE PICCs had a longer time to removal for a complication (p = 0.031).

CONCLUSIONS:

LE compared with UE PICCs were not associated with worse outcomes in a medical-surgical neonatal population that included a significant proportion of full-term neonates, and provide a valuable alternate site for central venous access. Increased awareness of the types of complications for UE compared with LE PICCs may help focus preventive and surveillance efforts based on PICC location, to improve safety and minimize the complications of NICU PICCs.

Keywords

Peripherally inserted central catheter upper extremity lower extremity complications neonate

1 Introduction

Peripherally inserted central catheters (PICCs) are increasingly used in neonatal intensive care units (NICUs) to provide secure access and to infuse hypertonic solutions [1]. Complications of PICCs compared with peripheral intravenous catheters inserted in premature infants are similar [2] and PICC insertions are recommended if the need for intravenous access is anticipated to be greater than six days [3].

Historically, the upper extremity (UE) has been recommended as the first choice for the site of PICC insertion [1]. However, limited UE sites or lack of success in the insertion of UE PICCs can frequently occur in the NICU, and alternate sites may be required to secure access. Experience with lower extremity (LE) PICCs in neonates is limited and reports on the comparison of outcomes of UE and LE PICCs has included mostly very premature infants [4, 5] or a relatively small sample size for LE PICCs [5]. The purpose of our study was to compare the outcomes and complications of PICCs placed in the UE compared with LE in a medical-surgical population of neonates.

2 Methods

We conducted a retrospective review of neonates who had PICC lines placed in the NICU at the Hospital for Sick Children (HSC) during January 2005 to August 2010 (67 months). The NICU at HSC is a quaternary children’s hospital in southern Ontario with approximately 800 admissions annually. There are no deliveries on-site, and 30% of admissions have a primary diagnosis of a surgical condition.

The majority of PICCs were inserted by specialized PICC nurses. The catheter choice and insertion site were guided by operator preference and vein availability. UE insertions were in the basilic, cephalic, median cubital or axillary veins; and LE insertions were in the greater saphenous vein, lesser saphenous vein, dorsal venous arch or the popliteal vein. Catheters used (in order of frequency) included Clinicath (2.0F, polyurethane, Smiths Medical), V-cath (1.9F and 2.0F, silicone, Neomedical), L-cath (1.9F and 3F, polyurethane, Argon Medical), or Argyle (1.9F, polyurethane, Covidien). The procedure was performed at the bedside and ultrasound guidance was not used. Post insertion X-rays were taken with the shoulder abducted at 30 degrees for UE PICCs and the hips in ‘frog’ position for LE PICCs. The target tip position was in the lower portion of the superior vena cava for UE PICCs and at the inferior vena cava-right atrial junction for LE PICCs. A repeat X-ray to confirm final tip position was done if the catheter was pulled by more than 1 cm. There was low tolerance for non-central tip locations, and the routine unit practice was to remove non-central PICCs within 24 hours of insertion. Blood products were not infused, and blood sampling was not done through PICCs with lumen size <3.0F. Heparin 0.5–1.0 unit per ml was routinely infused as per standard unit policy. PICC maintenance was provided by a dedicated team of nurses.

The primary outcome was elective vs. non-elective removals for UE vs. LE PICCs. Elective removal was defined as removal for the end of therapy, and non-elective removal was defined as removal due to a complication related to the line.

Only PICCs inserted within the NICU were included, and those inserted by interventional radiology were excluded. Patients that were transferred out of the NICU with a PICC in situ, or died with a line in situ, were excluded from analysis since the primary outcome of the reason for removal could not be ascertained. PICCs that were malpositioned on the insertion X-ray that could not be used for infusion and removed immediately post X-ray, or PICCs removed within 24 hours of insertion for malposition were excluded from analysis. Data were collected on patient baseline demographics and line outcomes including the reason for removal. PICCs were removed for malposition if it migrated to a non-central location and a central location was required to infuse hyperosmolar solutions. Sepsis was defined when one or more blood cultures were positive with signs of clinical deterioration. Sepsis with the line in situ was defined when the blood culture taken a minimum 24 hours after catheter insertion and a maximum of 48 hours after catheter removal was positive.

The study was approved by our institutional research ethics board.

3 Statistical methods

Data were analyzed using SPSS Version 22.0. (Armonk, NY: IBM Corp.). Baseline characteristics and outcomes were compared between the groups using chi-square or Mann-Whitney U tests as appropriate. The chi-square test of independence was used to compare the complications between UE vs. LE PICCs. An adjusted residual value greater than ± 1.96 was considered significant for a p-value of <0.05. Duration of catheter and time to complication was compared between UE and LE PICCs using Kaplan-Meier survival analysis and the Cox proportional hazard model. A two-sided p-value of <0.05 was considered statistically significant.

4 Results

We analyzed a total of 365 PICCs of which 250 (68%) were removed for the end of therapy and 115 (32%) were removed due to complications. Patients who had UE compared to LE insertions were of lower GA with median (IQR) of 30 (26, 35) weeks for UE vs. 32 (27, 37) weeks for LE (p = 0.014). Otherwise, the groups were not significantly different in sex, age of admission, and age of line insertion (see Table 1). Types of surgical conditions included bowel obstruction and atresias, necrotising enterocolitis, bowel perforation, abdominal wall defects, tracheoesophageal fistula, congenital diaphragmatic hernia, genitourinary anomalies and choanal atresia and were not statistically different between UE and LE groups.

Table 1
Demographics of patients with upper extremity compared to lower extremity PICCs

Demographics Upper extremity N = 138 Lower extremity N = 227 p-value

GA in weeks; median (IQR) 30 (26, 35) 32 (27, 37) 0.014

GA > = 37 weeks; n (%) 27 (19.6%) 64 (28.2%) 0.080

BW in grams; median (IQR) 1246 (948, 2090) 1460 (960, 2720) 0.056

BW > = 1500 grams; n (%) 56 (40.6%) 112 (49.3%) 0.106

Male; n (%) 83 (60.1%) 130 (57.3%) 0.589

Primary diagnosis surgical; n (%) 60 (44.5%) 76 (33.5%) 0.06

Surgical groups; n (% of surgical) 0.660

Necrotising enterocolitis or bowel perforation 30 (50.0%) 39 (51.3%)

Bowel obstruction or atresia 14 (23.3%) 18 (23.7%)

Abdominal wall defect 7 (11.7%) 7 (9.2%)

Tracheoesophageal fistula 3 (5.0%) 8 (10.5%)

Congenital diaphragmatic hernia 2 (3.3%) 2 (2.6%)

Genitourinary anomaly 2 (3.3%) 2 (2.6%)

Head and neck surgery (e.g. choanal atresia) 2 (3.3%) 0

GA > = 37 weeks; n (%) 14 (23.7%) 22 (28.9%) 0.559

BW > = 1500 grams; n (%) 30 (50.8%) 39 (51.3%) 1.000

Age admission in days; median (IQR) 1 (0, 3.25) 1 (0, 3) 0.636

Age line insertion in days; median (IQR) 7 (4, 13) 8 (3, 14) 0.633

Demographics	Upper extremity N = 138	Lower extremity N = 227	p-value
GA in weeks; median (IQR)	30 (26, 35)	32 (27, 37)	0.014
GA > = 37 weeks; n (%)	27 (19.6%)	64 (28.2%)	0.080
BW in grams; median (IQR)	1246 (948, 2090)	1460 (960, 2720)	0.056
BW > = 1500 grams; n (%)	56 (40.6%)	112 (49.3%)	0.106
Male; n (%)	83 (60.1%)	130 (57.3%)	0.589
Primary diagnosis surgical; n (%)	60 (44.5%)	76 (33.5%)	0.06
Surgical groups; n (% of surgical)			0.660
Necrotising enterocolitis or bowel perforation	30 (50.0%)	39 (51.3%)
Bowel obstruction or atresia	14 (23.3%)	18 (23.7%)
Abdominal wall defect	7 (11.7%)	7 (9.2%)
Tracheoesophageal fistula	3 (5.0%)	8 (10.5%)
Congenital diaphragmatic hernia	2 (3.3%)	2 (2.6%)
Genitourinary anomaly	2 (3.3%)	2 (2.6%)
Head and neck surgery (e.g. choanal atresia)	2 (3.3%)	0
GA > = 37 weeks; n (%)	14 (23.7%)	22 (28.9%)	0.559
BW > = 1500 grams; n (%)	30 (50.8%)	39 (51.3%)	1.000
Age admission in days; median (IQR)	1 (0, 3.25)	1 (0, 3)	0.636
Age line insertion in days; median (IQR)	7 (4, 13)	8 (3, 14)	0.633

GA = gestational age. BW = birth weight. IQR = interquartile range.

Table 2

PICC outcomes: Reason for removal

	Upper extremity N = 138	Lower extremity N = 227		p-value
Removal for complications; n (%)	55 (39.9%)	60 (26.4%)	RR 1.51 95% CI: 1.12 –2.03	0.007
Reason for removal; n (%)				<0.001
			Adjusted residual*
End of therapy	83 (60.1%)	167 (73.6%)	–2.7^†
Malposition	17 (12.3%)	4 (1.8%)	4.2^†
Dislodgement	8 (5.8%)	4 (1.8%)	2.1^†
Mechanical: occlusion or leaking	14 (10.1%)	28 (12.3%)	–0.6
Interstitial	3 (2.2%)	3 (1.3%)	0.6
Pleural or pericardial effusion	3 (2.2%)	0	2.2^†
Phlebitis	1 (0.7%)	10 (4.4%)	–2.0^†
Sepsis	9 (6.5%)	10 (4.4%)	0.9
Thrombus	0	1 (0.4%)	–0.8
Line duration for all removals in days; median (IQR)	17 (8, 27)	16 (9, 30)		0.978
Line duration for elective removals in days; median (IQR)	20 (8, 32)	16.5 (10, 29)		0.778
Line duration for non-elective removals in days; median (IQR)	16 (8, 25)	12.5 (5.25, 34.75)		0.989

*Adjusted residual for chi-square. ^†Adjusted residual values > ± 1.96 are statistically significant at p < 0.05.

Table 3

PICC outcomes: Sepsis, Mortality and LOS

	Upper extremity N = 138	Lower extremity N = 227	p-value
Septic during line; n (%)	18 (13.0%)	29 (12.8%)	0.941
Organism			0.969
Gram positive; n (%)	13 (9.4%)	20 (8.8%)
Coagulase negative staphylococcus	12 (8.7%)	17 (7.5%)
Staphylococcus aureus	1 (0.7%)	1 (0.4%)
Group B streptococcus	0	1 (0.4%)
Enterococcus	0	1 (0.4%)
Gram negative; n (%)	5 (3.6%)	9 (4.0%)
Klebsiella	1 (0.7%)	3 (1.3%)
Eschericia coli	2 (1.4%)	1 (0.4%)
Enterobacter	1 (0.7%)	2 (0.9%)
Serratia marsescens	0	2 (0.9%)
Proteus	1 (0.7%)	0
Gram negative bacteria not specified	0	1 (0.4%)
Age sepsis in days; median (IQR)	21 (17, 25.5)	21 (10.5, 39)	0.878
Duration line at sepsis in days; median (IQR)	14 (10.75, 21.75)	8 (4, 23)	0.196
Died; n (%)	7 (5.1%)	14 (6.2%)	0.818
LOS in days; median (IQR)	38 (20, 63.5)	37 (18, 63)	0.773

IQR = interquartile range. LOS = length of stay.

PICC outcomes, NICU length of stay and mortality are shown in Tables 2 and 3. UE PICCs were more likely to be removed due to complications compared to LE PICCs (39.9% vs. 26.4%, RR 1.51 (95% CI 1.12, 2.03), p = 0.007). UE PICCs were more likely to be removed for the complications of malposition, dislodgement, and pleural or pericardial effusions; while LE PICCs were more likely to be removed for phlebitis. The median (IQR) duration of the line was 17 (8, 27) days for UE vs. 16 (9, 30) days for LE (p = 0.978). The duration of the line for elective removals was longer compared with non-elective removals, and was not significantly different for UE vs. LE PICCs.

The overall rate of the line associated sepsis was similar at 13.0% and 12.8% for UE and LE, respectively (p = 0.941). Central line-associated blood stream infection rates for UE vs. LE were equal at 5.8 per 1000 line days, with a total of 47 infections in 8046 line days for UE and LE PICCs combined. The most common organism was Coagulase negative staphylococcus, and there were no differences in the proportion of gram positive vs. gram negative organisms for UE vs. LE PICCs. The median (IQR) duration of the line at the onset of sepsis was 14 (10.75, 21.75) days for UE vs. 8 (4, 23) days for LE (p = 0.196). NICU length of stay and mortality was similar for patients with UE vs. LE PICCs.

Survival analysis of the time to line removal for a complication showed that UE PICCs had a shorter time to line removal for complications with a median (IQR) of 30 (1.82–58.18) days for UE, and 56 (48.55–63.45) days for LE (p = 0.031) (see Fig. 1). The Cox proportional hazard model showed that significant factors associated with a shorter time to line removal for complications were UE site (exp (B) 1.70, 95% CI 1.17–2.49, p = 0.006) and lower GA (exp (B) 1.07, 95% CI 1.03–1.11, p = 0.001).

Fig.1

Kaplan Meier survival analysis demonstrating lower probability of keeping PICCs without complications for upper extremity vs lower extremity PICCs.

5 Discussion

Our results add to the limited literature comparing outcomes of NICU PICCs based on the site of insertion. Similar to Hoang [5] and Wrightson [4], we found that PICCs inserted in the LE compared with the UE may have significant advantages. Hoang et al. studied a population of mostly preterm neonates with median BW 940 g and GA 28 weeks, and reported that LE PICCs had a lower catheter-related infection rate and a longer complication-free interval [5]. In their study, the majority (78%) of PICCs were UE, and they had fewer LE PICCs (n = 107 LE PICCs). Wrightson reported on 374 UE and 252 LE PICCs among neonates with median BW 982 and 1062 g and median GA 28 and 29 weeks respectively [4], and showed non-elective rates of removal of PICCs of 27% for UE and 21% for LE and no differences in overall complication rates. Compared to these studies, our study population includes a more mature cohort and a significant proportion of surgical patients. Our rates of non-elective removal are slightly higher at 40% for UE PICCs and 26% for LE PICCs. Njere et al., also reported on catheter outcomes in medical-surgical neonates similar to our population, and had higher rates of non-elective removal at 46%, but they did not specify the site of insertion [6]. Our study is unique in that the full spectrum of PICC outcomes is reported, and compared by the site of insertion, for a medical-surgical population.

PICC related complications can be broadly categorized as catheter-related infection; complications of line migration with the tip remaining within the lumen (dislodgement, malposition), or line migration with the tip in the extra-luminal space (interstitial, pleural or pericardial effusion); mechanical problems (leaking, breakage, blockage); and irritation of the vessel by the catheter causing phlebitis or thrombus formation. We found that UE PICCs were more likely to have complications related to line migration including malposition, dislodgement, and pleural or pericardial effusions. The high rate of line malposition as the reason for removal seen in our study is likely related to our unit practice of close surveillance for all PICCs, and low tolerance for leaving in non-central lines. The practice of routine surveillance of PICCs is supported by Srinivasan et al. who reported that after adjustment for arm position, 47% of UE PICCs migrated at 24 hours post-insertion; and they recommended a follow-up X-ray after 24 hours to determine migration patterns [7]. Jain et al. reported higher malposition rates for UE vs. LE PICCs at 48% vs. 15% respectively [8]. They showed that non-central PICCs had a high rate of complications and recommended non-central catheter removal within four days after insertion.

We found that thrombus as a reason for PICC removal was rare with no significant differences between UE and LE PICCs; however, thrombi were likely underreported as we did not perform routine surveillance ultrasounds. It is well recognized that the rates for thrombi related to central lines that are symptomatic, compared with subclinical thrombi detected with routine surveillance ultrasounds, are highly discrepant; with rates of 1% for symptomatic thrombi, compared with 44% for subclinical thrombi [9]. There have been concerning reports about increased risk of complications with LE PICCs among surgical patients. Kisa et al., studied PICCs inserted at a tertiary children’s hospital and found that major thrombotic complications were found only in surgical patients with abdominal pathology who had LE PICCs placed below L1 and were running parenteral nutrition [10]. Ma et al., reported an increased risk of infiltration and phlebitis with LE PICCs in neonates with gastroschisis, especially if inserted during silo-reduction and within 5 days after abdominal closure; thought to be related to catheter migration and/or the effects of increased intra-abdominal pressure [11]. However, they did not detect an increased risk of thrombotic complications with LE PICCs. Of note, our one case of thrombus was in a 38 wk infant with omphalocele who had a LE PICC, which was removed 12 days after insertion. We also found that phlebitis was more common with LE PICCs, but did not find an increased rate of interstitial PICCs with LE PICCs.

Our rates of catheter-related sepsis were similar to others and also showed that there were no differences in rates between UE and LE PICCs [4 , 8]. We found no differences in the prevalence of gram positive or gram negative organisms between UE and LE PICCs. Although not shown in their analysis, Hoang et al. did state in their discussion, that there was a ‘slightly higher prevalence of gram negative organisms causing septicemia’ with LE PICCs [5].

There were several limitations to our study. Our study population was in a quaternary children’s hospital and included patients with more mature GA, had a higher BW, and a higher proportion of surgical patients, compared to most perinatal centers; and thus may not be generalizable to centers with different demographic characteristics.

Our practice of catheter removal for malposition due to non-central line tip position may not be consistent with practices in other sites; thus, our results may not be generalizable to these sites that have a higher tolerance for non-central catheters. However, the practice of leaving in non-central catheters for a significant duration is not supported by evidence, as it has been shown to lead to significant complications [8 , 12–14]. Additionally, it is possible that we may have decreased downstream morbidities associated with catheter migration such as dislodgement, interstitial catheters, or pleural or pericardial effusions through the early removal of malpositioned catheters.

Due to the rarity of some of the major complications such as thrombi and gram negative sepsis, our sample size was insufficient to detect differences between UE and LE PICCs. Ongoing surveillance of the complications of PICCs is essential, especially in centers where practice trends are to increase the number of LE PICCs insertions, as has been the case in our NICU; as further experience may demonstrate emerging differences in complications between UE and LE PICCs.

6 Conclusion

LE PICCs were not worse than UE PICCs in a medical-surgical neonatal population and provide a valuable alternate site for central venous access. Increased awareness of the types of complications for UE PICCs compared with LE PICCs may help focus preventive and surveillance efforts based on PICC location, to improve safety and minimize the complications of NICU PICCs.

Disclosures

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests

None declared.

Footnotes

Acknowledgments

None.

References

Paulson

, Miller

. Neonatal peripherally inserted central catheters: Recommendations for prevention of insertion and postinsertion complications. Neonatal Netw. 2008;27(4):245–57.

Janes

, Kalyn

, Pinelli

, Paes

. A randomized trial comparing peripherally inserted central venous catheters and peripheral intravenous catheters in infants with very low birth weight. J Pediatr Surg. 2000;35(7):1040–4.

O’Grady

, Alexander

, Burns

et al., Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control. 2011;39(4 Suppl 1):S1–34.

Wrightson

. Peripherally inserted central catheter complications in neonates with upper versus lower extremity insertion sites. Adv Neonat Care. 2013;13(3):198–204.

Hoang

, Sills

, Chandler

, Busalani

, Clifton-Koeppel

, Modanlou

. Percutaneously inserted central catheter for total parenteral nutrition in neonates: Complications rates related to upper versus lower extremity insertion. Pediatrics. 2008;121(5):e1152–e1159.

Njere

, Islam

, Parish

, Kuna

, Keshtgar

. Outcome of peripherally inserted central venous catheters in surgical and medical neonates. J Pediatr Surg. 2011;46(5):946–50.

Srinivasan

, Tjin

ATA

, Galang

, Hecht

, Srinivasan

. Migration patterns of peripherally inserted central venous catheters at 24 hours postinsertion in neonates. Am J Perinatol. 2013;30(10):871–4.

Jain

, Deshpande

, Shah

. Peripherally inserted central catheter tip position and risk of associated complications in neonates. J Perinatol. 2013;33(4):307–12.

Revel-Vilk

, Ergaz

. Diagnosis and management of central-line-associated thrombosis in newborns and infants. Semin Fetal Neonatal Med. 2011;16(6):340–4.

10.

Kisa

, Ting

, Callejas

, Osiovich

, Butterworth

. Major thrombotic complications with lower limb PICCs in surgical neonates. J Pediatr Surg. 2015;50(5):786–9.

11.

, Garingo

, Jensen

, Bliss

, Friedlich

. Complication risks associated with lower versus upper extremity peripherally inserted central venous catheters in neonates with gastroschisis. J Pediatr Surg. 2015;50(4):556–8.

12.

Racadio

, Doellman

, Johnson

, Bean

, Jacobs

. Pediatric peripherally inserted central catheters: Complication rates related to catheter tip location. Pediatrics. 2001;107(2):E28.

13.

Kearns

, Coleman

, Wehner

. Complications of long arm-catheters: A randomized trial of central vs peripheral tip location. JPEN J Parenter Enteral Nutr. 1996;20(1):20–4.

14.

Jumani

, Advani

, Reich

, Gosey

, Milstone

. Risk factors for peripherally inserted central venous catheter complications in children. JAMA Pediatrics. 2013;167(5):429–35.