How to cite this article: García HJ, Torres-Yáñez HL. Survival and complication rate of central venous catheters in newborns. Rev Med Inst Mex Seguro Soc. 2015;53 Suppl 3:S300-9.
Received: August 14th 2014
Accepted: September 1st 2015
Heladia Josefa García,a Héctor Leonardo Torres-Yáñeza
aUnidad de Cuidados Intensivos Neonatales, Hospital de Pediatría, Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Distrito Federal, México
Communication with: Heladia Josefa García
Telephone: (55) 5627 6932
Background: In the current medical practice, central venous catheters (CVC) are very useful; however, their use involves certain risks, which increase morbidity and mortality, especially in newborns. The aim of this study was to describe both the frequency of complications and survival of CVC placed in newborns hospitalized in a third level neonatal intensive care unit (NICU).
Methods: A descriptive, observational and prospective study was carried-out in the NICU from the Hospital de Pediatría, Centro Médico Nacional Siglo XXI. Demographic, perinatal and CVC variables were recorded.
Results: We included 152 CVCs, which were inserted in 123 newborns. For the CVC insertion, the puncture technique [percutaneous and subclavian] was used in 56.6 % (n = 86). There was at least one complication in 48.7 %.(n = 74). The most frequent complications were colonization 32.4 % (n = 24) and CVC-related bacteremia in 27 % (n = 20). Survival probability for CVC was 93.4 % at 10 days and 91.4 % at 17 days. Kaplan–Meier survival analysis demonstrated significantly lower survival probability for non-central catheters.
Conclusions: Most of CVC complications occurred within the first two weeks after these CVC were installed. Infectious complications were the most frequent.
Keywords: Central venous catheters, Bloodletting, Bacteremia, Catheter-related infection, Newborn infant.
Venous catheters are plastic devices that allow central access to the intravascular compartment. They vary in design and structure depending on temporary (days) or permanent use (weeks, months), in the material they are made of, the number of lumens, and insertion technique (venesection or percutaneous). The use of these devices has been very useful in clinical practice because they allow quick and safe access to the bloodstream, and can be used to administer intravenous fluids, medications, blood products, total parenteral nutrition (TPN), monitoring of hemodynamic status, and procedures such as blood transfusion or hemodialysis. The use of venous access is a fundamental question of modern medical treatment. Different types of catheters have been installed in newborns from the 60’s to the present.1-3
Various methods have been employed to achieve central venous access, venesection having been the most widely used in pediatric patients. Another way to approach to CVC placement is by percutaneous puncture. In newborns the vessels used for this process are from both upper and lower extremities.4
The use of central vascular accesses has increased significantly in current medical treatment, especially in intensive care units, and although their use confers many benefits, they are not without risks. An unlimited number of complications have been reported, so the morbidity and mortality risk is increased, especially in newborns.5-7
Mechanical and infectious complications have been described that can occur in both percutaneous and venesection approaches, which include CVC malposition or aberrant localization, puncture of the artery or the vein itself resulting in hemothorax, hydrothorax or extravasation of fluid into peripheral tissues, pneumothorax, cardiac tamponade, thoracic duct injury (resulting in chylothorax), hematoma or local infection, phlebitis, embolisms, damage to adjacent structures (such as nerve or artery), perforation of the trachea, stroke due to carotid injury, obstruction, accidental mobilization, leakage, rupture, sepsis, venous thrombosis, superior vena cava syndrome, thrombosis of the catheter tip, and diaphragmatic hernia injury to the phrenic nerve, among others.7-17
CVC-related infection is one of the major complications of its use and is the primary cause of nosocomial bacteremia. The incidence of bacteremia attributable to CVC use is variable between different hospitals.5,18-24 The most commonly isolated organisms are Gram positive such as coagulase-negative Staphylococcus, Staphylococcus aureus, Enterococcus faecium, and faecalis, followed by Gram negative as Klebsiella pneumoniae, E. coli, Enterobacter cloacae and Pseudomonas aeruginosa.3,13,19,25-27
The duration of catheters is variable, some authors report an average of 13.5 days, and up to 98 days in percutaneous CVC and 57 days for those inserted by venesection.11,28-30 The most frequent reasons for CVC removal are: end of treatment, accidental exit, entry site infection, catheter-related sepsis, suspected infection, and death.24,31,32
The objectives of this study were to record the frequency and type of complications and the time free from complications, for CVC inserted by venesection or venipuncture in newborns hospitalized in the Unidad de Cuidados Intensivos Neonatales (Neonatal Intensive Care Unit, NICU) of the Hospital de Pediatría of the Centro Médico Nacional Siglo XXI (CMN SXXI), Instituto Mexicano del Seguro Social (IMSS).
The study was conducted in the NICU of the aforementioned hospital, a tertiary care hospital that receives patients from general hospitals in the southern area of Mexico City and the states of Guerrero, Queretaro, Chiapas, and Morelos, as well as private hospitals. The design was descriptive, prospective, and observational. We included newborns (NB) admitted to the NICU who had venous catheters placed while at the service. We did not include infants whose CVC were not placed in the NICU of the hospital where the study was conducted. 123 NB were included who received CVC during hospitalization in the NICU in the period between April 2010 and January 2011.
The following variables were recorded: gestational age, birth weight, sex, age at CVC insertion, CVC material, number of lumens, insertion technique, vein used for insertion, insertion procedure time, use of CVC, location of CVC tip (central: mouth of superior vena cava or inferior vena cava or right atrium entrance; not central: outside these anatomical sites), category of the doctor who placed it, number of cures during stay, reason for removal, CVC-related complications (infectious, mechanical, and thrombotic), microorganisms isolated in blood or CVC tip, CVC time in place without complications (survival), and NB condition upon discharge from NICU (dead or alive).
During the study period, one of the researchers daily identified the hospitalized patients in NICU who had CVC placed. General and catheter-related data were recorded and gathered in a sheet designed specifically for the study. Follow-up was then conducted to monitor patient and CVC progress to detect complications. Follow-up ended with catheter removal, or the patient’s NICU discharge in cases where the CVC was not removed.
For data analysis, descriptive statistics were used to calculate frequencies and percentages for qualitative variables, and median and range were calculated as measures of central tendency and dispersion for quantitative variables because the population had no distribution close to normal. Mantel-Haenszel chi-squared or Fisher’s exact tests were used for comparison between groups, and Kaplan-Meier estimator was used for analysis of duration of catheters.
Regarding ethics, according to the rules established in Mexico’s Ley General de Salud en materia de investigación, this is an investigation without risk.33 The protocol was approved by the local committee on health research of the Hospital de Pediatría del CMN SXXI.
123 infants were included in the study who had a total of 152 venous catheters placed through venesection, peripheral percutaneous puncture, or subclavian puncture. The catheter material was polyurethane and silicone, with one or two lumens.
56.9% (n = 70) of patients were male, median gestational age was 37 weeks (range 25-42), birthweight 2400 g (range 675-4080), and age at CVC placement was 9 days (range 1-112, interquartile range 3-24.5).
Table I shows diagnoses of patients upon admission to the NICU.
|Table I Diagnoses of newborns upon admission into NICU (n = 123)|
|Congenital heart disease||40||32.5|
|Coarctation of aorta||5||4.1|
|Interruption of aortic arch||5||4.1|
|Ventricular septal defect||5||4.1|
|Complex heart disease||4||3.3|
|Transposition of great arteries||2||1.6|
|Anomalous pulmonary venous connection||2||1.6|
|Patent ductus arteriosus||12||9.7|
|Injury of airway||5||4.1|
|Congenital diaphragmatic hernia||4||3.3|
|Pulmonary arterial hypertension||3||2.4|
|NICU = neonatal intensive care unit|
As regards the technique of CVC placement, 56.6% (n = 86) were done by puncture (percutaneous or subclavian) and 43.4% (n = 66) by venesection. All NB whose CVC were placed by subclavian puncture and who were under mechanical ventilation, were given sedation with midazolam (100 mcg/kg, one dose) and analgesia with buprenorphine (1-2mcg/kg, one dose), and those with spontaneous breathing were given local anesthesia with 2% lidocaine only. Percutaneous CVC were inserted into a vein in an upper limb in 86.6% of cases. The vein most commonly used for insertion by venesection was the internal jugular in 19.7%. All catheters were installed by medical residents, 85.6% in Neonatal and 14.4% in Pediatric Surgery. 89.5% of catheters were double lumen. All double lumen catheters were 4Fr gauge and those with one lumen were 2Fr gauge. In 58.1% of CVC inserted by puncture, only one puncture was made; two punctures were made in 27.9%, three punctures in 10.3%, four in 2.3%, and five punctures in 1.2%. The median time that the procedure took was 20 minutes (range 5-90) and the number of cures 2 (range 0 -13). The CVC were used to administer parenteral nutrition, solutions, medicines, blood, and blood products. The drugs administered were amines, antibiotics, analgesics, and anti-inflammatories.
The reason for CVC removal was suspected infection in 25% (n = 38); end of function in 19.7% (n = 30); patient death in 16.4% (n = 25); CVC dysfunction in 9.9% (n = 15); malposition in 4.6% (n = 7); and no removal because NB was transferred to another hospital in 24.3% (n = 37).
Table II shows the frequency and type of CVC complications. It stands out that 48.7% of CVC had complications and infectious complications were the most common. The incidence density of bacteremia related to colonization of central venous catheter (BRCCVC) was 9.9/1000 catheter days and the rate of catheter use was 7.5/100 CVC days.
|Table II Complications of central venous catheters (n = 74)|
|Colonization of CVC||24||32.4|
|Bacteremia related to CVC||20||27|
|Infection at insertion site||2||2.7|
|CVC = central venous catheter|
The microorganisms most frequently isolated in blood were Klebsiella pneumoniae with 36.4% and coagulase-negative Staphylococcus (CNS) with 27.3%; in isolates from the CVC tip, SCN predominated with 28.6% (Table III).
|Table III Microorganisms isolated in cultures|
|Type of microorganism||Frequency||%|
Table IV shows the frequency of complications and some features related to the CVC and its installation.
|Table IV Frequency of complications according to characteristics of CVC and insertion|
|Complication||Yes (n = 74)||No (n = 78)||Total (n = 152)|
|Type of placement|
|Number of lumens||0.14|
|Doctor who installed||0.29|
|CVC = central venous catheter
* Mantel-Haenszel Chi-squared or Fisher's exact test
Table V shows the frequency and type of complications according to the installation technique, i.e., percutaneous, subclavian puncture and venesection; Table VI shows the frequency and type of complications according to the vein used.
|Table V Type of complications according to technique of placement (n = 74)|
|Infection at CVC site||2||2.7||-|
|BRCVC = bacteremia related to colonization of central venous catheter; CVC = central venous catheter|
|Table VI Complications by anatomical site of CVC insertion|
|Complication||Subclavian||External jugular||Internal jugular||Saphenous||UE||LE||Total|
|Infection at insertion site||-||-||-||2||2.7||-||-||2||2.7|
|UE = upper extremity; LE = lower extremity; BRCCVC = colonization of central venous catheter-related bloodstream infection|
The duration of the CVC in place was a median of 9 days (range 1-60). The probability of CVC duration free of complications was 0.93 after 10 days and 0.91 after 17 days (Figure 1).
Figure 1 Kaplan-Meier survival analysis, showing that the longer the duration of the catheter, the less likely it is to be free of complications
Figure 2 shows the probability of CVC survival according to insertion technique. The probability of duration of CVC inserted by venesection was 0.93 after 10 days and 0.89 at 21 days; when inserted through the subclavian puncture it was 0.88 at 8 days and 0.80 at 21 days; and percutaneous 0.78 at five days and 0.56 at 11 days.
Figure 2 Kaplan-Meier survival analysis, showing that percutaneous CVC had lesser probability of duration, and that those inserted by venesection lasted the most, although the difference was not statistically significant
The probability of duration according to location of the CVC tip, that is, whether central or not, is shown in Figure 3. The probability of five days duration with non-central CVC was 0.61 and 0.15 at 12 days; for those with central location, the probability of five days duration was 0.96 and 0.94 at 12 days, with a statistically significant difference (p = 0.0001).
Figure 3 Kaplan-Meier survival analysis, showing that CVC with the tip in a non-central location had a lower probability of duration in the short term, with a statistically significant difference
Preterm infants and critically ill newborns require prolonged vascular access, which is achieved in modern medical practice with a central venous catheter. The most important indications for CVC installation are administering parenteral nutrition and prolonged administration of intravenous medications. However, central vascular access in infants has been associated with increased risk of mechanical, infectious, and thrombotic complications.3
This study reported the results of 152 catheters in newborns hospitalized in a tertiary neonatal intensive care, whose main diagnoses were diseases of surgical resolution such as malformations of the digestive tract and congenital heart disease, which requires prolonged fasting and administration of parenteral nutrition also for a long time, and administering medications as amines, antibiotics, blood, and blood products. A high frequency of complications (48.7%) was found, similar to that reported by Colacchio et al.34 of 44% and 47% by Jain et al. in percutaneous catheters32 as well as in non-central percutaneous CVC. Other authors have reported lower figures in recent studies with mainly percutaneous CVC.4,9 In the nineties, Hruszkewycz35 reported a complication rate of 88% and Pandit36 57% in CVC installed by venesection.
Percutaneous channeling through a peripheral vein catheter that progresses to a central position has become increasingly important, sometimes displacing the surgical technique for installing CVC; today it may be the most commonly used form of central channeling in NICUs.30,32 In the present study percutaneous CVC, that is those inserted in the upper or lower extremities, were less frequent, only 9.8% of total CVC. This is because the facility is a concentration hospital where patients come from other hospitals, and most of them have multiple punctures, making central venous access through the periphery difficult. However, nearly half (46.7%) of CVC were inserted by infraclavicular subclavian puncture. This technique is not often used in newborns, or at least not considered first choice for installing a CVC, as this technique is reported to have serious complications. In 1979, Filston37 reported using this technique even in tiny infants with low frequency of serious complications. Moon et al.38 recently reported using the subclavian vein for CVC insertion at three different times in a premature NB of 1280 g, without complications. In this report, the frequency of complications of CVC by subclavian puncture was similar to that of CVC inserted by venesection (21 and 24.3%, subclavian and venesection, respectively). However, with insertion in the subclavian vein serious complications were presented; two patients had pneumothorax and hemothorax, which resolved properly with immediate treatment. Hemothorax occurred in preterm infants, weighing 1300 to 1650 g. The two with pneumothorax were larger than 2000 g but with strong mechanical ventilatory support. It is noteworthy that of the 71 CVC installed by subclavian puncture, only 13 (18.3%) were in infants less than 2000 g. Based on these serious, life-threatening complications, CVC is only installed in the subclavian vein in patients > 2 kg without mechanical ventilation or when this has minimal variables, especially peak inspiratory pressure to avoid lung overdistension, and the procedure should be performed by personnel appropriately trained in this technique.
Central catheters are a potential source of nosocomial infection. Catheter-related sepsis is the most common nosocomial infection in NICUs.39 CVC-associated infection rates have been reported ranging from 1.7 to 49/1000 catheter days.13,15,18-20,22,23,32,36,40-42 This was one of the most frequent complications found in this report, at 9.9/1000 catheter days, similar to that reported in the literature, and most frequently observed in CVC installed by venesection.
Unlike Tsai,15 who reported a significantly higher incidence of BRCCVC in CVC installed in the femoral vein (10.9 and 6.8 episodes per 1000 catheter days in femoral and non-femoral, respectively), the present study did not find that infectious complications were more frequent in CVC inserted in the femoral or saphenous veins.
Newborns are particularly susceptible to thromboembolism, due to factors such as the small size of their blood vessels, immaturity of the thrombolytic system, and hemostatic disorders caused by asphyxia or perinatal complications such as congenital heart disease. Furthermore, CVC use is the most common cause of thrombosis in neonates. Intravascular catheters can cause thromboembolism by damage to the endothelium, as a foreign body is introduced with thrombotic properties, and the administration of parenteral nutrition for a long time can aggravate the endothelial damage.8,10,43
A frequency of central line-related thrombosis in neonates has been reported with 1% when only symptomatic cases are included, about 44% when intentionally looking for thrombi, and 65% in autopsy studies.44 The frequency of thrombotic complications in this study was low, at 5.4% (similar to the 6.8% reported by Saleem16), and these were cases in which echocardiography was done to intentionally search for thrombi. There was no difference between the CVC inserted by venesection or subclavian puncture, and no cases of thrombosis occurred in percutaneous CVC.
Regarding CVC survival, defined as duration free of complications, it was found that the probability of general CVC duration at 10 days was 93% and for the third week, 90%; this probability decreased over time. The main complications in the first week were non-infectious, although five of the 20 BRCCVC occurred in the first week.
Survival was lower in CVC inserted percutaneously, mainly due to mechanical complications. There was no statistically significant difference between the duration of CVC by subclavian puncture and venesection.
With respect to the location of CVC tip, it was found that non-central location had a lower probability of duration in the short term; at five days it was only 61%, compared with central CVC whose probability of duration at that time was 96.4%. These findings are similar to those reported by Jain et al.,32 although they found that the probability of duration was similar in both groups in the first four days, after which time a marked difference between central and non-central CVC was observed. Other authors4,17,31,34 have reported that non-central CVC have a higher frequency of complications; we found that of the 13 CVC whose location was not central, 10 had complications.
The frequency of CVC complications in newborns in a third level NICU is high but similar to that reported in other studies, and the complications are most frequently of infectious type, such as BRCCV. Potentially lethal complications were presented in the CVC inserted by subclavian puncture, although at low frequency. The probability of CVC duration in the first week is high and decreases over time. The CVC whose location is not central have less likely duration in the short term, i.e., the probability that they will have complications in the short term is greater than those with central location, so it is recommended that when installing a CVC in a NB, one should do everything possible so that it is well located to prevent further morbidity to the patient.
Furthermore, based on experience with CVC inserted by subclavian puncture, we recommend evaluating the risks and benefits of using this venous access, which should be done by well-trained staff and preferably in children weighing > 2000 g without or with only minimal mechanical ventilatory support and with adequate sedation and analgesia.
Conflict of interest statement: The authors have completed and submitted the form translated into Spanish for the declaration of potential conflicts of interest of the International Committee of Medical Journal Editors, and none were reported in relation to this article.