There is insufficient evidence from randomised trials to either support or refute the routine use of preventive antibiotics in newborn babies with umbilical vein catheters.
Sick newborn babies occasionally require the insertion of an umbilical vein catheter (a special tube) that goes into the vein in the umbilicus (belly button). This allows fluid and medicines to be given. Some people believe that antibiotics should be given to all babies with umbilical vein catheters in order to reduce the chance of infection occurring. However, antibiotics can have unwanted effects. The reviewers found insufficient evidence to either support or refute the routine use of antibiotics for all babies with umbilical vein catheters.
Patients requiring umbilical venous catheters may, by virtue of their underlying illness, have impaired defence mechanisms - both local and systemic. Prematurity is recognised as a risk factor for late onset sepsis (Dear 1999). Preterm neonates are at high risk of infection because of impaired immunity and umbilical venous catheters may further increase this risk because they are foreign bodies.
It is common practice in neonatal units to start antibiotics in infants with respiratory distress and suspected infection, or in those delivered following pre-term labour. Many of these infants will have an umbilical venous catheter inserted. It is not clear whether antibiotics should be discontinued if no infection is proven. It has been common practice in some neonatal units that if the infant has an umbilical venous catheter then antibiotics be continued in order to reduce the rate of colonisation of the umbilicus and likewise reduce the risk of acquired infection. Prophylactic antibiotics may be effective in preventing catheter-related blood stream infection (CRBSI), but may have the undesirable effect of promoting the emergence of resistant strains of micro-organisms (Freij 1999). A policy of prophylactic antibiotic use should take into account this possibility, and has been used as a basis for arguing against its implementation (Isaacs 2000; Isaacs 2003). Promotion of the emergence of resistant strains of organisms may vary between different antibiotics.
A recent Cochrane systematic review on the use of prophylactic antibiotics for neonates with umbilical artery catheters showed that there is no evidence from randomised trials to support or refute the use of prophylactic antibiotics when using umbilical artery catheters in newborn infants (Inglis 2004).
In separate comparisons, we planned to review two different policies regarding the prophylactic use of antibiotics in neonates with umbilical venous catheters:
1) Among neonates with umbilical venous catheters, a policy of prophylactic antibiotics for the duration of catheterisation (or other fixed duration of antibiotic treatment) versus placebo or no treatment. This addresses the question of whether or not neonates with umbilical venous catheters, who do not have clinical or laboratory evidence of infection at that time, should be routinely started on antibiotics at the time of catheterisation.
2) Among neonates with umbilical venous catheters who had been started on antibiotics at the time of catheterisation, but whose initial cultures to rule out sepsis are negative, a policy of continuing versus discontinuing prophylactic antibiotics. This addresses the question of whether or not antibiotics should routinely be stopped at the time rule out sepsis cultures are reported as negative.
Data permitting, subgroup analyses were planned to determine whether results differ by:
gestational age (e.g. preterm versus term, < 28 weeks gestational age (GA) or not);
type of antibiotic (e.g. penicillins, macrolides, aminoglycosides, cephalosporins, or combinations).
Trials where the unit of allocation is the catheter (in which case different catheters within the same patient might be managed differently) were not included.
Trials where the cluster unit is time were not included (as this would not allow the assessment of antibiotic resistance).
Secondary:
The standard search strategy for the Cochrane Neonatal Review Group was used. We searched MEDLINE from 1966 to April 2005, CINAHL from 1982 to April 2005, and the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 1, 2005) using the following strategy:
MeSH search terms ("Umbilicus" AND "Catheterization") OR the textwords
("umb$" AND ("cathet$" OR "cannul$")) OR "UVC" OR "umbilical vein catheter"
OR "umbilical venous catheter"
AND
MeSH search term "Infant, newborn" OR the textwords "neonat$" OR "infant"'
AND
MeSH search term "Antibiotics" OR the textword "antibiotic"
AND
MeSH search terms "Chemoprevention" OR "Antibiotic Prophylaxis" OR the textword "prophyl$".
We also searched previous reviews (including cross references). Searches were not restricted to publications in the English language or published data.
The two authors worked independently to search for and assess trials for inclusion and methodological quality. Studies were assessed using the following key criteria: allocation concealment (blinding of randomisation), blinding of intervention, completeness of follow up and blinding of outcome measurement assigning a rating of 'Yes', 'No' or 'Can't tell' for each. The authors extracted data independently. Differences were resolved by discussion. We contacted the second author of the study by Bhatt et al (Bhatt 1970) for additional information or data.
For pooled results: for continuous variables, weighted mean differences (WMD) and 95% confidence intervals were to be reported. For categorical outcomes, the relative risks (RR) and 95% confidence intervals were to be reported. For significant findings, the risk difference (RD) and number needed to treat (NNT) were also to be reported. Each treatment effect was to be tested for heterogeneity to help determine suitability for pooling of results in a meta-analysis. The fixed effects model was to be used for meta-analysis. If there were sufficient included studies, heterogeneity was to be assessed using the I squared test. If statistical heterogeneity was found the authors planned to look for an explanation. If studies with heterogenous results were thought to be comparable, a random effects model was to be used to combine the data.
Data permitting, a sensitivity analysis was planned to see if results differed by quality of included studies i.e. adequacy of randomisation - quasi randomised versus randomised.
The study by Pulido et al (Pulido 1985) was included. The study attempted to address the question of whether term infants with umbilical venous catheters had lower rates of infection when given antibiotic prophylaxis, compared with untreated controls. The study was non-randomised (quasi-randomised, using alternate group allocation), was small, of short duration, and involved infants with very specific indications for umbilical venous catheter insertion: i.e. catheters were inserted for transfusion procedures in infants with hyperbilirubinaemia or polycythaemia. There was a wide range of age at enrolment (1 - 10 days). Twenty-nine term infants with umbilical venous catheters were allocated non-randomly (quasi-randomised - alternate allocation) to treatment (n = 15) or control (n = 14) groups. Those in the treatment group received penicillin and gentamicin for three days, but no mention is made of duration of catheterisation. It is possible that duration of treatment exceeded duration of catheterisation or vice versa. Two peripheral blood cultures were drawn from all study infants three days after catheter insertion. Results were presented as number of positive blood cultures and number of true-positive blood cultures.
For primary outcomes:
A major limiting factor in trying to determine the place of prophylactic antibiotics in infants with umbilical venous catheters is that catheter placement is quite often undertaken, for ease of fluid and drug administration, in the context of clinical circumstances (e.g. respiratory distress, preterm delivery) which may reflect infection. Newborn infants in such circumstances are usually commenced on antibiotics because their clinical circumstances may indicate infection at the same time that they may lead to the decision to insert an umbilical venous catheter. Because the majority of newborns in whom umbilical venous catheters are placed would be treated in this way, the first scenario described above would be relevant to relatively few newborns. The second scenario described above would be the more common one encountered.
One non-randomised (quasi-randomised) trial was found for inclusion in this review. Pulido et al (Pulido 1985) performed a small study on the use of antibiotic prophylaxis in infants undergoing transfusion procedures for hyperbilirubinaemia or polycythaemia via an umbilical venous catheter. The authors conclude that no infant in the study developed septicaemia following the procedure, and that the use of antibiotic prophylaxis is not indicated. A study of this size would have been underpowered to detect anything other than a very large effect. The study covered a period of only two months. It has been noted previously that nosocomial infections can occur in clusters (Adams-Chapman 2002). If the study under consideration here coincided with a nadir in nosocomial infection, then the resultant underestimation of septicaemia rates in one or both arms of the study could have affected the conclusions. It is difficult to generalise the findings of this study for a number of reasons. Since its publication there have been significant changes in the practice of neonatal medicine, including use and maintenance of vascular access devices. The use of umbilical venous catheters in this study was for specific indications and the background risk of infection in the study subjects may have been low. The average age at catheter insertion in this study was probably significantly greater than would be seen in most units today. Given the poor methodological quality of the study, we cannot rely on the results provided with regard to effects on infection rates.
Quasi-randomised trials are inherently prone to bias, and their results should be interpreted with caution. The alternate group assignment makes the upcoming treatment group allocation predictable, and that is a problem in the case of every eligible infant. Also, if two equally eligible infants present at the same time with different risks for infection a clinician might (consciously or not) enter them into the study in the order that would allow the infant that they believed should receive antibiotics to get antibiotics. If a large number of infants were enrolled in this way, serious imbalance in the treatment groups with respect to factors affecting the outcome would result (Hennekens 1987).
In order to justify the use of prophylactic antibiotics (rather than treatment of infection as it arises) in infants with umbilical venous catheters, there should be evidence that the benefit outweighs the harm. This should include an adequate assessment not only of short term outcomes such as infection rate and duration of hospital admission, but also of long term outcomes such as mortality, long term respiratory morbidity and neurodevelopmental outcome.
Theoretical concerns about the potential harm of prophylactic antibiotic use include emergence of resistant strains of bacteria, superinfection and drug toxicity. Altered antibiotic resistance patterns may be of consequence not only to the individual in whom prophylactic antibiotics are used but also to other patients within the hospital setting and to the broader community.
Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
Pulido 1985 | This non-randomised (quasi-randomised - alternate allocation) study took place between July and August 1984 at a regional neonatal intensive care unit in Chile. Enrolled infants were allocated alternately into intervention and control groups. All subjects had 2 peripheral blood cultures and a full blood count (FBC) drawn three days following the procedure. At this stage antibiotics were discontinued if there was deemed to be no clinical or laboratory (i.e. FBC) evidence of infection. Blood cultures were read at 7 days. Septicaemia was defined as positive blood culture combined with clinical and laboratory evidence of infection. Intervention was probably not blinded. Completeness of follow up is not addressed. It is unclear whether outcome assessment was blinded. | Twenty-nine infants were studied. All were term. Twenty-three underwent exchange transfusion for hyperbilirubinaemia, and 6 underwent globuloforesis (partial exchange transfusion) for haematocrit greater than 0.70. All infants had their procedures performed via umbilical venous catheter. Infants requiring repeat procedures were excluded from the study (the number of such infants, if any, is not specified). There were 15 infants in the intervention group and 14 in the control group. | Infants in the intervention group (n = 15) received penicillin and gentamicin for 3 days following the procedure. Control infants (n = 14) received no antibiotics. No placebo was used. Other care was similar. | Septicaemia: based on positive blood culture (3 days after UVC insertion) in conjunction with clinical and haematological evidence of infection. | D |
UVC = umbilical venous catheter
Study | Reason for exclusion |
Bard 1973 | Study of umbilical artery catheters. |
Bhatt 1970 | Published in Abstract form only. The second author was contacted and could offer no further data, except that the study involved infants with arterial, rather than venous, catheters. |
Cowett 1977 | Study of umbilical artery catheters. |
Pulido N, Montesinos A, Arriaza M, Esparza P. Uso profilactico de antibioticos en cateterismo umbilical en recien nacidos [Prophylactic use of antibiotics in umbilical catheterization in newborn infants]. Revista Chilena de Pediatria 1985;56:247-9.
Bard H, Albert G, Teasdale F, Doray B, Martineau B. Prophylactic antibiotics in chronic umbilical artery catheterization in respiratory distress syndrome. Archives of Disease in Childhood 1973;48:630-5.
Bhatt 1970 {unpublished data sought but not used}
Bhatt DR, Hodgman JE, Tatter D. Evaluation of prophylactic antibiotics during umbilical catheterization in newborns. Clinical Research 1970;18:217.
Cowett 1977 {published data only}
Cowett RM, Peter G, Hakanson DO, Stern L, Oh W. Prophylactic antibiotics in neonates with umbilical artery catheter placement: a prospective study of 137 patients. The Yale Journal of Biology and Medicine 1977;50:457-63.
* indicates the primary reference for the study
Adams-Chapman I, Stoll BJ. Prevention of nosocomial infections in the neonatal intensive care unit. Current Opinion in Pediatrics 2002;14:157-64.
Chien L, MacNab Y, Aziz K, Andrews W, McMillan DD, Lee SK; Canadian Neonatal Network. Variations in central venous catheter-related infection risks among Canadian neonatal intensive care units. Pediatric Infectious Disease Journal 2002;21:505-11.
Dear P. Infection in the newborn. In: Rennie JM, Roberton NRC, editor(s). Textbook of Neonatology. 3rd edition. Edinburgh: Churchill Livingstone, 1999:1109-202.
Freij BJ, McCracken Jr GH. Acute infections. In: Avery GB, Fletcher MA, MacDonald MG, editor(s). Neonatology: pathophysiology and management of the newborn. 5th edition. Philadelphia: Lippincott Williams & Wilkins, 1999:1189-230.
Hennekens CH, Buring JE. Intervention studies. In: Mayrent SL, editor(s). Epidemiology in medicine. Boston: Little, Brown & Company, 1987:178-212.
Inglis GDT, Davies MW. Prophylactic antibiotics to reduce morbidity and mortality in neonates with umbilical artery catheters. In: The Cochrane Database of Systematic Reviews, Issue 3, 2004.
Isaacs D. Rationing antibiotic use in neonatal units. Archives of Disease in Childhood. Fetal and neonatal edition 2000;82:F1-2.
Isaacs D; Australasian Study Group for Neonatal Infections. A ten year, multicentre study of coagulase negative staphylococcal infections in Australasian neonatal units. Archives of Disease in Childhood. Fetal and neonatal edition 2003;88:F89-93.
Nagata E, Brito AS, Matsuo T. Nosocomial infections in a neonatal intensive care unit: incidence and risk factors. American Journal of Infection Control 2002;30:26-31.
Ogawa Y, Shimizu H, Takasaki J, Nakamura T. Strategy for the prevention and treatment of chronic lung disease of the premature infant. Pediatric Pulmonology 1999;Supp 18:212-5.
Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA, Ehrenkranz RA et al. Late-onset sepsis in very low birth weight neonates: The experience of the NICHD Neonatal Research Network. Pediatrics 2002;110:285-91.
Stoll BJ, Hansen NI, Adams-Chapman I, Fanaroff AA, Hintz SR, Vohr B, Higgins RD. Neurodevelopmental and growth impairment among extremely low-birth-weight infants with neonatal infection. JAMA 2004;292:2357-65.
Comparison or outcome | Studies | Participants | Statistical method | Effect size |
---|---|---|---|---|
01 Prophylactic antibiotics versus no antibiotics | ||||
01 Septicaemia | RR (fixed), 95% CI | No numeric data |
The review is published as a Cochrane review in The
Cochrane Library, Issue 4, 2005 (see http://www.thecochranelibrary.com for
information). Cochrane reviews are regularly updated as new evidence emerges
and in response to comments and criticisms, and The Cochrane Library should
be consulted for the most recent version of the Review. |