- Email: [email protected]
Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors
Clinical Nutrition xxx (2013) 1e5
Contents lists available at SciVerse ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors Judith Pichler a, b, *, James Soothill c, Susan Hill a a
Department of Paediatric Gastroenterology, Great Ormond Street Hospital NHS Trust, London WC1N 3JH, UK Department of Paediatric and Adolescent Medicine, Medical University of Vienna, Austria c Department of Microbiology, Great Ormond Street Hospital NHS Trust, London WC1N 3JH, UK b
a r t i c l e i n f o
s u m m a r y
Article history: Received 6 November 2012 Accepted 17 March 2013
Background & aims: Catheter-related-blood-stream-infection (CRBSI) might be prevented by optimal catheter connector antisepsis in children with intestinal failure on parenteral nutrition (PN). We changed the disinfectant used from isopropanol 70% to chlorhexidine 2% in 70% isopropanol, which leaves a residue of chlorhexidine on the connector. Methods: We conducted this before/after study in children treated with PN for >28 days. Episodes of CRBSI were recorded for all 42 children treated for >28 days during MayeNovember 2006 before introducing chlorhexidine and for all 50 children treated in MayeNovember 2007 after chlorhexidine was introduced. The number of hospital-acquired CRBSI and number of PN days was counted for each period. The rate of CRBSI/1000 catheter days and the proportion of patients that experienced at least one CRBSI during the two periods were compared. Results: There were 3.1 CRBSI/1000 catheter days prior to using chlorhexidine and 0.4 CRBSI/1000 catheter days after it was introduced, p ¼ 0.03. Prior to chlorhexidine 10/42 (24%) patients experienced at least one episode of CRBSI, compared to 3/50, (6%) after introducing it (p ¼ 0.02). The survival rate in both periods was similar, but after chlorhexidine significantly more children made a full recovery and a lower proportion of children had irreversible intestinal failure (p ¼ 0.01). Conclusions: Our results support the use of 2% chlorhexidine not only to reduce risk of sepsis for central venous catheter connector antisepsis in catheters used for intravenous nutrition, but also to improve the patients’ chances of recovering intestinal function. Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Catheter-related bloodstream infections Chlorhexidine 2% Parenteral nutrition Intestinal failure Children
1. Introduction Parenteral nutrition (PN) is a good medium for microbial and fungal growth due to constituents such as glucose, lipids and protein1 and this may increase the risk of infection in such patients. Central venous catheters (CVC) used for PN might be more prone to infection and less responsive to an improvement in catheter disinfection than catheters used for clear fluids and medication. We are not aware of any previous studies reviewing sepsis rates solely in patients with catheters used for PN.
Abbreviations: IF, intestinal failure; PN, parenteral nutrition; CRBSI, catheter related bloodstream infection; CVC, central venous catheter. * Corresponding author. Department of Paediatric Gastroenterology, Great Ormond Street Hospital for Children NHS Trust, WC1N 3JH London, United Kingdom. Tel.: þ44 2078138384; fax: þ44 2078138181. E-mail address: [email protected] (J. Pichler).
Children with severe intestinal failure (IF) require treatment with PN. PN should be infused through a central venous catheter in order safely to provide adequate amounts of calories and protein to meet the child’s daily needs.2 Catheter related blood stream infections (CRBSI) are a common cause of illness in such patients and can lead to death. Along with intestinal failure associated liver disease, sepsis is the most serious and life threatening complication.3e5 In addition, the underlying gastrointestinal disease is frequently exacerbated by each septic episode. In order to avoid life-threatening problems the catheter may need to be removed when sepsis occurs. Repeated replacement of catheters can damage the major blood vessels and lead to loss of venous access. This is a serious problem in a child with chronic severe intestinal failure who may need access for many months or even years. Unlike adults, young children commonly contaminate CVC with faeces6 and attempt to touch and pick at their catheter and exit site predisposing them to sepsis.6,7
0261-5614/$ e see front matter Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. http://dx.doi.org/10.1016/j.clnu.2013.03.021
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
2
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
Since the catheter hub8 and needle-less connectors (to attach the PN/other fluid infusions) are recognised sources of infection, disinfection of connectors has been recommended9,10 for many years. Chlorhexidine in isopropanol has a broad-spectrum microbicidal activity that is thought to be mediated by the disruption of microbial cell membranes.11 It was found to be more effective than isopropanol alone for the antisepsis of needle-less connectors in clinical use12 and a change from isopropanol to 2% chlorhexidine in isopropanol was followed by a substantial fall in CRBSI.13e15 Previous studies have looked at CVC in general; the sub-group of patients using the CVC for PN have not been investigated separately. The aim of this study was to investigate the efficacy of chlorhexidine in reducing infections solely in the high-risk group of patients receiving parenteral nutrition infusions. Secondary aims were to review length of hospitalisation, patient survival rate, proportion of children weaned from PN and the number who failed to wean and were discharged home on PN. 2. Methods This observational single centre study was conducted retrospectively during a period of routine surveillance of catheterrelated sepsis in a tertiary children’s hospital. We investigated children receiving PN for more than 28 days. Taurolodine, prophylactic antibiotics or other indwelling ‘line locks’ other than heparin, were not used. Children treated with PN for over 27 days have been defined to have ‘long-term’ IF.16 Positive blood cultures were documented during the study period, i.e. pre and post adding 2% chlorhexidine to 70% isopropanol for cleansing the catheter at the site of connections and disconnections. From FebruaryeApril 2007 the disinfectant used to clean the hubs of needle-less connectors and catheter hubs was changed from swabs containing 70% isopropanol (Iso-sachets from Griffiths & Nielsen Ltd., Billinghurst, UK) to 2% Chlorhexidine in 70% isopropanol (predominately ClinellÒ wipes from Gama Healthcare, London, UK). The recommended method of disinfection was to swab the catheter connector for 30 s, then to leave the connector port to dry completely before use. This method remained unchanged throughout the study. From hospital pharmacy records we obtained the identities of patients with severe IF who received PN for a period of more than 27 days16 as inpatients from MayeNovember 2006 and Maye November 2007 and the dates when they received it. This time interval was chosen not to interfere with an audit of hand hygiene within the institution that could have additionally contributed to a reduction in infection rate. PN constituents including fluid volume, calories, nitrogen and lipid prescribed/kg body weight were recorded. Energy and protein provision was concordant with the ESPGHAN/ESPEN guidelines of 2005.2 Demographic characteristics, gender, age, duration of hospitalisation and age of starting PN were obtained from the patients’ medical records. In addition patient outcome was recorded with regard to survival, weaning off PN and regaining intestinal function or persistent intestinal failure with discharge home on PN treatment.
2. The patient was not known to have a clinical condition (other than the presence of the CVC) likely to have caused the bacteraemia. 3. The patient had developed at least one of the following signs within 24 h of the culture being taken: fever > 38 C, hypothermia < 36 C, chills, rigors, systolic blood pressure 20 mmHg below that expected, apnoea or bradycardia. 4. The bacterial strain grown was not similar to one grown from a blood culture taken before admission to the tertiary paediatric hospital or within 48 h of admission. 5. A second blood or CVC tip culture had been positive for a similar strain to that cultured from the blood or there was a resolution of signs of infection after appropriate antibiotic treatment or removal of the CVC. 6. In patients with a CRBSI who had a further positive blood culture, this was only counted as a new episode if the above criteria were met and a new bacterial strain was isolated from a new blood culture. 7. Hospital-acquired CRBSI were included in the analyses if they arose during the period of PN treatment. New episodes of hospital-acquired CVC related bacteraemia were classed as polymicrobial bloodstream infection if two or more strains were isolated. Specific microorganisms causing infections were recorded. 2.2. Central venous catheters Most of the catheters were Hickman lines that were inserted under radiological control and tunnelled under the skin into the jugular or another major vein. Peripherally inserted long lines were used in neonates. These were inserted on the ward through a peripheral vein and advanced so that the tip was centrally placed. In all cases the tip of the central catheter was placed in the superior vena cava or high in the right atrium. The nutrition was infused through the catheter on a daily basis for up to 24 h. Many patients also received other intravenous fluids, antibiotics and other medications as needed through the catheter. The nurses caring for PN patients had all been specifically trained in handling CVCs. 2.3. Ethical approval An outline of the study was sent to the Chair of the Local Research ethics committee who approved the retrospective study since he considered further formal ethical consideration was unnecessary. 2.4. Statistics Statistical analyses were performed with SPSS Software (version 18.0 Mac; SPSS, Chicago, IL). The Fisher’s exact test (2 tailed) was used for comparisons of proportions of infections between the two periods. The total number of days patients were treated with PN was also recorded. Hospital-acquired CRBSI/1000 CVC days for 2006 (before chlorhexidine) were compared with those for 2007 (after chlorhexidine) using the ManneWhitney-U test (2 tailed).
2.1. Definition of catheter related bloodstream infection 3. Results Patients with a positive blood culture were considered to have a new episode of hospital-acquired CRBSI if the following were the case: 1. The blood culture was taken from an in-patient with a CVC in place, more than 48 h after admission.
3.1. Patients demographic characteristics Altogether, in the two periods MayeNovember 2006 and 2007, 92 children fulfilled the inclusion criteria and received PN for more than 27 days. Demographics of patients during the study periods
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5 Table 1 Demographics and diagnoses of patients during the study periods MayeDecember of 2006 and 2007. 2006
2007
p
Number of patients 42 50 Number of male patientsa 21 (50%) 27 (54%) Weightb 13.6 10.6 11 11.3 Prematurea 6 (14%) 8 (16%) Gestational age of preterm infantsc 26 (24e26) 26 (25e29) Transplantation (stem cells or solid 8 (19%) 19 (38%) a organ) c 2.1 (birth-12.7) 1.1 (birth-16.4) Age at start of study period, years Duration of hospitalization, days 87 (40e386) 109 (30e353) Diagnosis Primary intestinal failure Gastroenterologya 9 (21%) 6 (12%) Secondary intestinal failure for GI conditions Surgical 11 (26%) 14 (28%) Secondary intestinal failure from other illnesses or its treatment Oncology 12 (29%) 12 (24%) Immunology 7 (17%) 16 (21%) Cardiology 2 (5%) 2 (4%) Others 1 (2%) 5 (10%) Outcome Enteral feeds 22 (53%) 37 (74%) Home parenteral nutrition 11 (26%) 1 (2%) Death 9 (21%) 12 (24%) a b c
3
chlorhexidine period, p ¼ 0.006. This was associated with a reduction in the infection rate from 3.1/1000 CVC days in the 2006 period to 0.4/1000 CVC days in 2007, p ¼ 0.03. The mean number of episodes of CRBSI per patient also fell with a mean of 0.92 in the pre-chlorhexidine period (range 0e7), and a mean of 0.15 (range 0e6 episodes) per patient in the 2007 period after chlorhexidine was introduced. Ten of 42 or 24% of children developed a CRBSI in Maye November 2006 whereas just three of 50 or 6% did so in Maye November 2007, p ¼ 0.02 (Table 3). In addition, the proportion of children with a CRBSI who had at least one polymicrobial blood stream infection was statistically different in the two periods. Six of 42 were affected in 2006 and 1/ 50 in 2007, p ¼ 0.04. 3.4. Microorganisms cultured
0.04 0.01
Results are given as number (%). Results are given as mean SD. Results are given as median (range).
are given in Table 1. There were 48 males (52%). The median age at start of PN was 1.4 (range 0.01e16.4) years and the median duration of PN was 52 (range 28e192) days. PN was administered for a total of 5735 patient days. Demographic data, including primary diagnosis that had led to intestinal failure, was similar between the two periods apart from a wider age range during the second period than the first. In addition there was a trend for more patients in the second group to have had either a stem cell or solid organ transplant either before or during the period of PN, p ¼ 0.06.
The microorganisms cultured from blood are described in Table 4. The most frequent isolate was coagulase-negative staphylococcus. There was only one Gram-negative infection in the 2007 period. 3.5. Hospitalisation and outcome The reduction of CRBSI was not associated with a fall in duration of hospital stay. In 2006, the median duration was 87 (range: 40e 386) days whereas in 2007 the median duration was 109 (range: 30e353) days. The overall survival rate was 77% with 21 deaths and 71 children surviving. It was similar in both 2006 and 2007 with 79% surviving prior to the use of chlorhexidine in 2006 and 76% in 2007. Twenty-two or 52% of the surviving children gained enteral autonomy in 2006 in that they were successfully weaned off PN and tolerated full enteral feeds compared to 37 (74%) children in 2007, p ¼ 0.04. Eleven children or 26% needed to be discharged home on long-term PN in 2006 whereas only one child or 2% needed PN at home in 2007, p ¼ 0.01. 4. Discussion
3.2. PN characteristics The PN regimen provided a mean volume of 89 ml/kg/day, 67 kcal/kg calories, 1.9 g/kg/day lipids and 0.32 g/kg/day of nitrogen for the whole cohort. The duration, daily volume of PN, calorific value and nitrogen content were very similar in the two periods (Table 2). The median daily lipid administration was higher in 2007 than in 2006 with 2.1 g per kg body mass infused each day rather than 1.6 g in 2006, p ¼ 0.04. 3.3. Hospital-acquired CRBSI There was a significant reduction in the total number of CRBSI from 35 in the pre-chlorhexidine period to 8 in the post-
This study demonstrated a significant reduction in CRBSI after introducing chlorhexidine as a catheter cleaning solution. There was not only a reduction in the proportion of patients infected, but also in the infection rate per 1000 CVC days and the mean number of infections per patient. The initial infection rate at 3.1/1000 CVC days was similar to the rate described in previous paediatric studies in which incidence has varied from 1.6 to 13.6 infections per 1000 catheter days.9,17e21 The infection rate of 0.4/1000 CVC days with use of chlorhexidine was (to our knowledge), lower than previous published infection rates in children. Table 3 Positive blood cultures, catheter related bloodstream infections (CRBSI) and infection rates shown as CRBSI/1000 CVC days of parenteral nutrition (CRBSI/1000 CVC days) in the two study periods.
Table 2 Characteristics of parenteral nutrition treatments in the two study periods. PN characteristics per infusion
2006
2007
Age of start PN in years Duration of PN in days Volume ml/kg Calories cal/kg Nitrogen g/kg Lipid g/kg
2.1 (birth-12.7) 52 (34e126) 75 (25e170) 61.2 (13e102) 0.3 (0.1e0.5) 2 (0e3)
1.1 (birth-16.4) 51 (28e192) 94 (25e180) 78 (16e110) 0.3 (0.1e0.5) 3 (0e3.5)
All values are given as median (range). No significant differences were seen between the groups for all given variables.
Number of positive blood culture Number of patients with CRBSI Number of patients with polymicrobial CRBSI Total number of CRBSI Infections per patientsa Range CRBSI/1000 CVC feeding daysa
2006
2007
p
64 10 (24%) 6
12 3 (6%) 1
0.02 0.04
35 0.92 1.7 0e7 3.1 8
8 0.15 0.8 0e6 0.4 2
0.01 0.006 0.03
CVC central venous catheter. a Results are given as mean SD.
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
4
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
Table 4 Microorganisms cultured from blood cultures in the study periods MayeDecember of 2006 and 2007. Micro- organism
2006
Total infections Gram- positive infections 44 Gram- negative Infections 13 Mixed (gram positive and negative) 8 Fungal infections 1 Gram- positive Cocci associated with CRBSI (n) Staphylococcus spp. 33 Coag. Neg. Staphylococcus 33 Enterococcus spp. 24 Streptococcus spp. 1 Moraxella spp. 1 Diphertoides 1 Bacteroides spp. e Gram- negative Bacilli associated with CRBSI (n) Klebsiella spp. 6 Enterobacter spp. 7 Escherichia spp. 2 Citrobacter spp. 2 Pseudomonas spp. e Fungal infections associated with CRBSI (n) Candida spp. 1
2007 10 e 1 1
0.5 0.1 1 0.3
10 10 e e e e 1
0.2
e e e e 1 1
As far as we are aware this is the first report of catheter sepsis rates solely in catheters used for parenteral nutrition as opposed to other fluids. We have described a significant reduction in sepsis when the catheter cleaning-method was changed to include chlorhexidine. Chlorhexidine in isopropanol was recommended for catheter connector antisepsis in the 2007 UK guidelines.22 All children selected were on PN treatment for more than 27 days for severe intestinal failure. Many of the children were immuno-suppressed due to the severity of their illness and/or the treatment. For example 8 patients, (or 19%) of the 2006 group and 19 (or 38%) of the 2007 group were post-bone marrow transplant and as a consequence at high risk of sepsis.23,24 An additional reason for a higher infection rate with PN is the high content of nutrients that might act as a culture medium for bacteria, hence promote bacterial growth and lead to bacteraemia.7 Children given PN via a central venous catheter would appear to be at high risk of developing CRBSI. Those at greatest risk are not only neonates with their immature immune systems, but also immunosuppressed children undergoing bone marrow transplant and surgical infants.13e15 Prevention of bloodstream infections in children with IF on intravenous nutrition therapy is essential for several reasons. These include 1. the symptoms and complications of the sepsis itself, 2. problems of replacing CVC when it is not possible to treat the sepsis with antibiotics and 3. association of liver disease with sepsis.2e4 In addition, this study demonstrated an improvement in development of intestinal autonomy in association with reduction in CRBSI. Children are frequently extremely unwell with sepsis and may rapidly develop life-threatening symptoms that need intensive care if the infection is not rapidly treated. At the same time in our experience the underlying intestinal failure is exacerbated and may take several days or even weeks to recover to the level at which it was prior to the infection. When the sepsis does not respond to appropriate antibiotic treatment, the central venous catheter may have to be removed and a new one inserted. If repeated CVCs are inserted the blood vessels used may become damaged and fibrotic and eventually block. One of the most serious complications of sepsis in children with IF is liver disease.2e4 Loss of venous access, recurrent episodes of sepsis and worsening liver disease are major causes of mortality,25 hence indications for evaluation for intestinal transplant.25
There were three differences between the two groups that might have predisposed to a higher infection rate in 2007 after introduction of chlorhexidine. Despite these factors the infection rate fell. Firstly, there was a higher use of lipid. PN formulas may predispose to infection by suppressing the immune response. Excess lipid administration can result in hypertriglyceridemia and impairment of leukocyte function.26,27 Lipid metabolites may accumulate in hepatic sinusoidal macrophages over time, thereby interfering with clearance and enhancing toxicity of bacteria and bacterial products in the bloodstream.28 Secondly, there were a higher proportion of immune-compromised patients after stem cell or solid organ transplantation and thirdly children studied were younger and would have had less mature immune systems. Both younger children and immune-compromised patients would be more vulnerable to infection. Younger children are also more likely to contaminate the line by pulling at it and playing with it, and by faecal material leaking from nappies, (particularly in children with diarrhoea), and have less well developed immunity than older children.29 Especially neonates are at high risk for developing bacterial infections due to quantitative and qualitative insufficiencies of innate immunity, particularly granulocyte lineage development and response to infection.29 Despite a greater potential for sepsis in the second group of patients the CRBSI rate fell in 2007. All these factors in 2007 make it more likely that the reduction in infection was attributable to the use of chlorhexidine. There was a similar finding in a prospective study of CVC hub care in neonates when using chlorhexidine along with additional education of medical staff with audio-visual presentation.15 In that study a significant decrease of CRBI rate was demonstrated with all catheter types and in all birth weight categories including extremely low birth weight infants that made up 40% of the cohort. In children with intestinal failure bacteraemia has been attributed either to the catheter when the hub is exposed for connections and disconnections or to translocation of bacteria from the gut.3e5 The profound fall in positive blood cultures following the change to chlorhexidine suggests that most of our infections originated from catheter manipulation and were not due to intestinal translocation. This supports the importance of adequate care of the CVC. It could be expected that the decrease in infection rate might have been associated with a reduction in length of hospitalisation in 2007 since less complications might be expected. Nevertheless patients in 2007 spent longer in hospital. Sicker infants appeared to be surviving in 2007 since the cohort of children were significantly younger and more had undergone transplantation; these two factors might have led to the longer length of in-patient stay after the introduction of chlorhexidine. More children gained enteral autonomy in 2007 than in 2006, in that they were weaned off PN onto full enteral feeding. The lower proportion of children with a primary intestinal diagnosis in 2007 could partly explain this. Small volumes of enteral feed, specific nutrients, growth factors or hormones may be useful tools for modulation of gut immunity.3e5,30 A non-used gut is prone to bacterial overgrowth that causes mucosal injury, malabsorption and bacterial translocation.3e5 A limitation of this study is that it was done in retrospect so other confounding factors might have been responsible, at least in part for the reduction in CRBSI. Decisions as to whether each positive blood culture signified a new episode of infection were made non-blindly and therefore open to observer bias. However, observer bias was kept to a minimum since a strict protocol was used as described in the methods section. In addition, initial blood culture readings were by machine so were unbiased and these fell five fold. It is possible that the introduction of chlorhexidine would have been associated with increased attention to hygiene policies, and
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
this could have contributed. However on other wards when major initiatives were made to increase attention to hygiene policies; the initiatives were not followed by falls in catheter related infections, so any such effect is very unlikely to account for the dramatic fivefold fall. In conclusion a change from the use of isopropanol alone to the use of 2% chlorhexidine in 70% isopropanol for the disinfection of CVC hubs was followed by a major fall in catheter related sepsis in infants and children with severe intestinal failure receiving PN via the catheter. The results support the UK guidelines on catheter handling which advocate the use of 2% chlorhexidine in isopropanol for catheter connector disinfection. In addition there was a significant improvement in outcome with more patients gaining enteral autonomy after the infection rate was reduced. Statement of authorship The paper was initiated by Susan Hill and James Soothill. The data collection and analyses were done by Judith Pichler and James Soothill. The manuscript was written by Judith Pichler with the provision of significant advice and consultation by Susan Hill and James Soothill. All authors critically appraised the manuscript. Funding sources There were funding sources. Conflict of interest statement There were no financial or personal relationships with any company or organization sponsoring the research at the time the research was done.
Acknowledgements We would like to thank for the assistance of Liz Wright for communicating her impression that chlorhexidine disinfected lines had less infections, Paul Lock and Bob Paget and for help with computer work, nursing staff for data entry and Microbiology staff for processing blood cultures.
References 1. Opilla M, Grove E, Richmond V. Epidemiology of bloodstream infection associated with parenteral nutrition. AJJIC 2008;36(10):173e6. 2. Koletzko B, Goulet O, Hunt J, Krohn K, Shamir R. Parenteral Nutrition Guidelines Working Group, European Society for Clinical Nutrition and Metabolism, European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN), European Society of Paediatric Research (ESPR). Guidelines on paediatric parenteral nutrition of the European Society of Paediatric Gastroenterology, Hepatology and nutrition (ESPGHAN) and the European Society for clinical nutrition and Metabolism (ESPEN), Supported by the European Society of paediatric research (ESPR). J Pediatr Gastroenterol Nutr 2005;41(Suppl. 2): S1e87. 3. Hermans D, Talbotec C, Lacaille F, Goulet O, Ricour C, Colomb V. Early central catheter infections may contribute to hepatic fibrosis in children receiving long- term parenteral nutrition. J Pediatr Gastroenterol Nutr 2007;44:459e63. 4. Kelly DA. Intestinal failure-associated liver disease: what do we know today? Gastroenterology 2006;130(2 Suppl. 1):S70e7.
5
5. Goulet O, Ruemmele F, Lacaille F, Colomb V. Irreversible intestinal failure. J Pediatr Gastroenterol Nutr 2004;38(3):250e69. 6. Johnson T, Sexton E. Managing children and adolescents on parenteral nutrition: challenges for the nutritional support team. Proc Nutr Soc 2006;65(3): 217e21. 7. Sizer T. Standards and guidelines for nutritional support of patients in hospital. Maidenhaed, Berks: BAPEN; 1996. 8. Sitges-Serra A, Puig P, Linares J, Perez JL, Farrero N, Jaurrieta E, et al. Hub colonization as the initial step in an outbreak of catheter-related sepsis due to coagulase negative staphylococci during parenteral nutrition. JPEN J Parenter Enteral Nutr 1984;8(6):668e72. 9. Salzman MB, Isenberg HD, Rubin LG. Use of disinfectants to reduce microbial contamination of hubs of vascular catheters. J Clin Microbiol 1993;31(3):475e9. 10. O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al., Healthcare Infection Control Practices Advisory Committee. Summary of recommendations: guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis 2011;52(9):1087e99. 11. Kuyyakanond T, Quesnel LB. The mechanism of action of chlorhexidine. FEMS Microbiol Lett 1992;79:211e5. 12. Casey AL, Worthington T, Lambert PA, Quinn D, Faroqui MH, Elliot TS. A randomized, prospective clinical trial to assess the potential infection risk associated with the PosiFlow needleless connector. J Hosp Infect 2003;54(4): 288e93. 13. Soothill JS, Bravery K, Ho A, Macqueen S, Collins J, Lock P. A fall in bloodstream infections followed a change to 2% chlorhexidine in 70% isopropanol for catheter connection antisepsis: a pediatric single center before/after study on a hemopoietic stem cell transplant ward. Am J Infect Control 2009;37(8):626e30. 14. Bishay M, Retrosi G, Horn V, Cloutman-Green E, Harris K, De Coppi P, et al. Chlorhexidine antisepsis significantly reduces the incidence of sepsis and septicemia during parenteral nutrition in surgical infants. J Pediatr Surg 2011;46(6):1064e9. 15. Sannoh S, Clones B, Munoz J, Montecalvo M, Parvez B. A multimodal approach to central venous catheter hub care can decrease catheter-related bloodstream infection. Am J Infect Control 2010;38(6):424e9. 16. Gowen H, Lloyd C. British Intestinal Failure Survey (BIFS): a referral registry to record and determine the outcome of childhood intestinal failure. Proc Nutr Soc 2009;68(OCE1):E15. 17. Onder AM, Chandar J, Billings A, Diaz R, Francoeur D, Abitol C, et al. Chlorhexidine- based antiseptic solutions effectively reduce catheter- related bacteremia. Pediatr Nephrol 2009;10:1154e5. 18. Onder AM, Chandar J, Chaokley S, Francouer D, Abitol C, Zilleruelo G. Controlling exit site infections: does it decrease the incidence of catheter-related bacteremia in children on chronic hemodialysis? Hemodial Int 2009;13(1): 11e8. 19. Newman C. Catheter- related bloodstream infections in paediatric intensive care unit. Semin Pediatr Infect Dis 2006;11:20e4. 20. Smith MJ. Catheter-related bloodstream infections in children. Am J Infect Control 2008;36(10):e1e3. S173. 21. Sarvikivi E, Lyytikainen O, Vaara M, Saxen H. Nosocomial bloodstream infections in children: an 8-year experience at a tertiary-care hospital in Finland. Clin Microbiol Infect 2008;14(11):1072e5. 22. Pratt RJ, Pellowe CM, Wilson JA, Loveday HP, Harper PJ, Jones SR, et al. Epic2: national evidence-based guidelines for preventing healthcare-associated infections in NHS hospitals in England. J Hosp Infect 2007;65(Suppl. 1):S1e64. 23. Bateman SL, Seed PC. Procession to pediatric bacteremia and sepsis: covert operations and failures in diplomacy. Pediatrics 2010;126(1):137e50. 24. Je˛ drzejczak-Czechowicz M, Kowalski ML. Effects of parenteral lipid emulsion on immune system response. Anestezjol Intens Ter. 2011;43(4):248e55. 25. Drews BB, Sanghavi R, Siegl JD, Metcalf P, Mittal NK. Characteristics of catheter-related bloodstream infections in children with intestinal failure: implications for clinical management. Gastroenterol Nurs 2009;32(6):385e90. quiz 391e382. 26. Caresta E, Pierro A, Chowdhury M, Peters MJ, Piastra M, Eaton S. Oxidation of intravenous lipid in infants and children with systemic inflammatory response syndrome and sepsis. Pediatr Res 2007;61(2):228e32. 27. Monson JR, Sedman PC, Ramsden CW, Brennan TG, Guillou PJ. Total parenteral nutrition adversely influences tumour-directed cellular cytotoxic responses in patients with gastrointestinal cancer. Eur J Surg Oncol 1988;14(5):935e43. 28. Cukier C, Waitzberg DL, Logullo AF, Bacchi CE, Travassos VH, Torrinhas RS. Lipid and lipid-free total parenteral nutrition: differential effects on macrophage phagocytosis in rats. Nutrition 1999;15(11e12):885e9. 29. Melvan JN, Bagby GJ, Welsh DA, Nelson S, Zhang P. Neonatal sepsis and neutrophil insufficiencies. Int Rev Immunol 2010;29(3):315e48. 30. Dhaliwal R, Heyland DK. Nutrition and infection in the intensive care unit: what does the evidence show? Curr Opin Crit Care 2005;11(5):461e7.
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
Contents lists available at SciVerse ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors Judith Pichler a, b, *, James Soothill c, Susan Hill a a
Department of Paediatric Gastroenterology, Great Ormond Street Hospital NHS Trust, London WC1N 3JH, UK Department of Paediatric and Adolescent Medicine, Medical University of Vienna, Austria c Department of Microbiology, Great Ormond Street Hospital NHS Trust, London WC1N 3JH, UK b
a r t i c l e i n f o
s u m m a r y
Article history: Received 6 November 2012 Accepted 17 March 2013
Background & aims: Catheter-related-blood-stream-infection (CRBSI) might be prevented by optimal catheter connector antisepsis in children with intestinal failure on parenteral nutrition (PN). We changed the disinfectant used from isopropanol 70% to chlorhexidine 2% in 70% isopropanol, which leaves a residue of chlorhexidine on the connector. Methods: We conducted this before/after study in children treated with PN for >28 days. Episodes of CRBSI were recorded for all 42 children treated for >28 days during MayeNovember 2006 before introducing chlorhexidine and for all 50 children treated in MayeNovember 2007 after chlorhexidine was introduced. The number of hospital-acquired CRBSI and number of PN days was counted for each period. The rate of CRBSI/1000 catheter days and the proportion of patients that experienced at least one CRBSI during the two periods were compared. Results: There were 3.1 CRBSI/1000 catheter days prior to using chlorhexidine and 0.4 CRBSI/1000 catheter days after it was introduced, p ¼ 0.03. Prior to chlorhexidine 10/42 (24%) patients experienced at least one episode of CRBSI, compared to 3/50, (6%) after introducing it (p ¼ 0.02). The survival rate in both periods was similar, but after chlorhexidine significantly more children made a full recovery and a lower proportion of children had irreversible intestinal failure (p ¼ 0.01). Conclusions: Our results support the use of 2% chlorhexidine not only to reduce risk of sepsis for central venous catheter connector antisepsis in catheters used for intravenous nutrition, but also to improve the patients’ chances of recovering intestinal function. Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Catheter-related bloodstream infections Chlorhexidine 2% Parenteral nutrition Intestinal failure Children
1. Introduction Parenteral nutrition (PN) is a good medium for microbial and fungal growth due to constituents such as glucose, lipids and protein1 and this may increase the risk of infection in such patients. Central venous catheters (CVC) used for PN might be more prone to infection and less responsive to an improvement in catheter disinfection than catheters used for clear fluids and medication. We are not aware of any previous studies reviewing sepsis rates solely in patients with catheters used for PN.
Abbreviations: IF, intestinal failure; PN, parenteral nutrition; CRBSI, catheter related bloodstream infection; CVC, central venous catheter. * Corresponding author. Department of Paediatric Gastroenterology, Great Ormond Street Hospital for Children NHS Trust, WC1N 3JH London, United Kingdom. Tel.: þ44 2078138384; fax: þ44 2078138181. E-mail address: [email protected] (J. Pichler).
Children with severe intestinal failure (IF) require treatment with PN. PN should be infused through a central venous catheter in order safely to provide adequate amounts of calories and protein to meet the child’s daily needs.2 Catheter related blood stream infections (CRBSI) are a common cause of illness in such patients and can lead to death. Along with intestinal failure associated liver disease, sepsis is the most serious and life threatening complication.3e5 In addition, the underlying gastrointestinal disease is frequently exacerbated by each septic episode. In order to avoid life-threatening problems the catheter may need to be removed when sepsis occurs. Repeated replacement of catheters can damage the major blood vessels and lead to loss of venous access. This is a serious problem in a child with chronic severe intestinal failure who may need access for many months or even years. Unlike adults, young children commonly contaminate CVC with faeces6 and attempt to touch and pick at their catheter and exit site predisposing them to sepsis.6,7
0261-5614/$ e see front matter Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. http://dx.doi.org/10.1016/j.clnu.2013.03.021
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
2
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
Since the catheter hub8 and needle-less connectors (to attach the PN/other fluid infusions) are recognised sources of infection, disinfection of connectors has been recommended9,10 for many years. Chlorhexidine in isopropanol has a broad-spectrum microbicidal activity that is thought to be mediated by the disruption of microbial cell membranes.11 It was found to be more effective than isopropanol alone for the antisepsis of needle-less connectors in clinical use12 and a change from isopropanol to 2% chlorhexidine in isopropanol was followed by a substantial fall in CRBSI.13e15 Previous studies have looked at CVC in general; the sub-group of patients using the CVC for PN have not been investigated separately. The aim of this study was to investigate the efficacy of chlorhexidine in reducing infections solely in the high-risk group of patients receiving parenteral nutrition infusions. Secondary aims were to review length of hospitalisation, patient survival rate, proportion of children weaned from PN and the number who failed to wean and were discharged home on PN. 2. Methods This observational single centre study was conducted retrospectively during a period of routine surveillance of catheterrelated sepsis in a tertiary children’s hospital. We investigated children receiving PN for more than 28 days. Taurolodine, prophylactic antibiotics or other indwelling ‘line locks’ other than heparin, were not used. Children treated with PN for over 27 days have been defined to have ‘long-term’ IF.16 Positive blood cultures were documented during the study period, i.e. pre and post adding 2% chlorhexidine to 70% isopropanol for cleansing the catheter at the site of connections and disconnections. From FebruaryeApril 2007 the disinfectant used to clean the hubs of needle-less connectors and catheter hubs was changed from swabs containing 70% isopropanol (Iso-sachets from Griffiths & Nielsen Ltd., Billinghurst, UK) to 2% Chlorhexidine in 70% isopropanol (predominately ClinellÒ wipes from Gama Healthcare, London, UK). The recommended method of disinfection was to swab the catheter connector for 30 s, then to leave the connector port to dry completely before use. This method remained unchanged throughout the study. From hospital pharmacy records we obtained the identities of patients with severe IF who received PN for a period of more than 27 days16 as inpatients from MayeNovember 2006 and Maye November 2007 and the dates when they received it. This time interval was chosen not to interfere with an audit of hand hygiene within the institution that could have additionally contributed to a reduction in infection rate. PN constituents including fluid volume, calories, nitrogen and lipid prescribed/kg body weight were recorded. Energy and protein provision was concordant with the ESPGHAN/ESPEN guidelines of 2005.2 Demographic characteristics, gender, age, duration of hospitalisation and age of starting PN were obtained from the patients’ medical records. In addition patient outcome was recorded with regard to survival, weaning off PN and regaining intestinal function or persistent intestinal failure with discharge home on PN treatment.
2. The patient was not known to have a clinical condition (other than the presence of the CVC) likely to have caused the bacteraemia. 3. The patient had developed at least one of the following signs within 24 h of the culture being taken: fever > 38 C, hypothermia < 36 C, chills, rigors, systolic blood pressure 20 mmHg below that expected, apnoea or bradycardia. 4. The bacterial strain grown was not similar to one grown from a blood culture taken before admission to the tertiary paediatric hospital or within 48 h of admission. 5. A second blood or CVC tip culture had been positive for a similar strain to that cultured from the blood or there was a resolution of signs of infection after appropriate antibiotic treatment or removal of the CVC. 6. In patients with a CRBSI who had a further positive blood culture, this was only counted as a new episode if the above criteria were met and a new bacterial strain was isolated from a new blood culture. 7. Hospital-acquired CRBSI were included in the analyses if they arose during the period of PN treatment. New episodes of hospital-acquired CVC related bacteraemia were classed as polymicrobial bloodstream infection if two or more strains were isolated. Specific microorganisms causing infections were recorded. 2.2. Central venous catheters Most of the catheters were Hickman lines that were inserted under radiological control and tunnelled under the skin into the jugular or another major vein. Peripherally inserted long lines were used in neonates. These were inserted on the ward through a peripheral vein and advanced so that the tip was centrally placed. In all cases the tip of the central catheter was placed in the superior vena cava or high in the right atrium. The nutrition was infused through the catheter on a daily basis for up to 24 h. Many patients also received other intravenous fluids, antibiotics and other medications as needed through the catheter. The nurses caring for PN patients had all been specifically trained in handling CVCs. 2.3. Ethical approval An outline of the study was sent to the Chair of the Local Research ethics committee who approved the retrospective study since he considered further formal ethical consideration was unnecessary. 2.4. Statistics Statistical analyses were performed with SPSS Software (version 18.0 Mac; SPSS, Chicago, IL). The Fisher’s exact test (2 tailed) was used for comparisons of proportions of infections between the two periods. The total number of days patients were treated with PN was also recorded. Hospital-acquired CRBSI/1000 CVC days for 2006 (before chlorhexidine) were compared with those for 2007 (after chlorhexidine) using the ManneWhitney-U test (2 tailed).
2.1. Definition of catheter related bloodstream infection 3. Results Patients with a positive blood culture were considered to have a new episode of hospital-acquired CRBSI if the following were the case: 1. The blood culture was taken from an in-patient with a CVC in place, more than 48 h after admission.
3.1. Patients demographic characteristics Altogether, in the two periods MayeNovember 2006 and 2007, 92 children fulfilled the inclusion criteria and received PN for more than 27 days. Demographics of patients during the study periods
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5 Table 1 Demographics and diagnoses of patients during the study periods MayeDecember of 2006 and 2007. 2006
2007
p
Number of patients 42 50 Number of male patientsa 21 (50%) 27 (54%) Weightb 13.6 10.6 11 11.3 Prematurea 6 (14%) 8 (16%) Gestational age of preterm infantsc 26 (24e26) 26 (25e29) Transplantation (stem cells or solid 8 (19%) 19 (38%) a organ) c 2.1 (birth-12.7) 1.1 (birth-16.4) Age at start of study period, years Duration of hospitalization, days 87 (40e386) 109 (30e353) Diagnosis Primary intestinal failure Gastroenterologya 9 (21%) 6 (12%) Secondary intestinal failure for GI conditions Surgical 11 (26%) 14 (28%) Secondary intestinal failure from other illnesses or its treatment Oncology 12 (29%) 12 (24%) Immunology 7 (17%) 16 (21%) Cardiology 2 (5%) 2 (4%) Others 1 (2%) 5 (10%) Outcome Enteral feeds 22 (53%) 37 (74%) Home parenteral nutrition 11 (26%) 1 (2%) Death 9 (21%) 12 (24%) a b c
3
chlorhexidine period, p ¼ 0.006. This was associated with a reduction in the infection rate from 3.1/1000 CVC days in the 2006 period to 0.4/1000 CVC days in 2007, p ¼ 0.03. The mean number of episodes of CRBSI per patient also fell with a mean of 0.92 in the pre-chlorhexidine period (range 0e7), and a mean of 0.15 (range 0e6 episodes) per patient in the 2007 period after chlorhexidine was introduced. Ten of 42 or 24% of children developed a CRBSI in Maye November 2006 whereas just three of 50 or 6% did so in Maye November 2007, p ¼ 0.02 (Table 3). In addition, the proportion of children with a CRBSI who had at least one polymicrobial blood stream infection was statistically different in the two periods. Six of 42 were affected in 2006 and 1/ 50 in 2007, p ¼ 0.04. 3.4. Microorganisms cultured
0.04 0.01
Results are given as number (%). Results are given as mean SD. Results are given as median (range).
are given in Table 1. There were 48 males (52%). The median age at start of PN was 1.4 (range 0.01e16.4) years and the median duration of PN was 52 (range 28e192) days. PN was administered for a total of 5735 patient days. Demographic data, including primary diagnosis that had led to intestinal failure, was similar between the two periods apart from a wider age range during the second period than the first. In addition there was a trend for more patients in the second group to have had either a stem cell or solid organ transplant either before or during the period of PN, p ¼ 0.06.
The microorganisms cultured from blood are described in Table 4. The most frequent isolate was coagulase-negative staphylococcus. There was only one Gram-negative infection in the 2007 period. 3.5. Hospitalisation and outcome The reduction of CRBSI was not associated with a fall in duration of hospital stay. In 2006, the median duration was 87 (range: 40e 386) days whereas in 2007 the median duration was 109 (range: 30e353) days. The overall survival rate was 77% with 21 deaths and 71 children surviving. It was similar in both 2006 and 2007 with 79% surviving prior to the use of chlorhexidine in 2006 and 76% in 2007. Twenty-two or 52% of the surviving children gained enteral autonomy in 2006 in that they were successfully weaned off PN and tolerated full enteral feeds compared to 37 (74%) children in 2007, p ¼ 0.04. Eleven children or 26% needed to be discharged home on long-term PN in 2006 whereas only one child or 2% needed PN at home in 2007, p ¼ 0.01. 4. Discussion
3.2. PN characteristics The PN regimen provided a mean volume of 89 ml/kg/day, 67 kcal/kg calories, 1.9 g/kg/day lipids and 0.32 g/kg/day of nitrogen for the whole cohort. The duration, daily volume of PN, calorific value and nitrogen content were very similar in the two periods (Table 2). The median daily lipid administration was higher in 2007 than in 2006 with 2.1 g per kg body mass infused each day rather than 1.6 g in 2006, p ¼ 0.04. 3.3. Hospital-acquired CRBSI There was a significant reduction in the total number of CRBSI from 35 in the pre-chlorhexidine period to 8 in the post-
This study demonstrated a significant reduction in CRBSI after introducing chlorhexidine as a catheter cleaning solution. There was not only a reduction in the proportion of patients infected, but also in the infection rate per 1000 CVC days and the mean number of infections per patient. The initial infection rate at 3.1/1000 CVC days was similar to the rate described in previous paediatric studies in which incidence has varied from 1.6 to 13.6 infections per 1000 catheter days.9,17e21 The infection rate of 0.4/1000 CVC days with use of chlorhexidine was (to our knowledge), lower than previous published infection rates in children. Table 3 Positive blood cultures, catheter related bloodstream infections (CRBSI) and infection rates shown as CRBSI/1000 CVC days of parenteral nutrition (CRBSI/1000 CVC days) in the two study periods.
Table 2 Characteristics of parenteral nutrition treatments in the two study periods. PN characteristics per infusion
2006
2007
Age of start PN in years Duration of PN in days Volume ml/kg Calories cal/kg Nitrogen g/kg Lipid g/kg
2.1 (birth-12.7) 52 (34e126) 75 (25e170) 61.2 (13e102) 0.3 (0.1e0.5) 2 (0e3)
1.1 (birth-16.4) 51 (28e192) 94 (25e180) 78 (16e110) 0.3 (0.1e0.5) 3 (0e3.5)
All values are given as median (range). No significant differences were seen between the groups for all given variables.
Number of positive blood culture Number of patients with CRBSI Number of patients with polymicrobial CRBSI Total number of CRBSI Infections per patientsa Range CRBSI/1000 CVC feeding daysa
2006
2007
p
64 10 (24%) 6
12 3 (6%) 1
0.02 0.04
35 0.92 1.7 0e7 3.1 8
8 0.15 0.8 0e6 0.4 2
0.01 0.006 0.03
CVC central venous catheter. a Results are given as mean SD.
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
4
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
Table 4 Microorganisms cultured from blood cultures in the study periods MayeDecember of 2006 and 2007. Micro- organism
2006
Total infections Gram- positive infections 44 Gram- negative Infections 13 Mixed (gram positive and negative) 8 Fungal infections 1 Gram- positive Cocci associated with CRBSI (n) Staphylococcus spp. 33 Coag. Neg. Staphylococcus 33 Enterococcus spp. 24 Streptococcus spp. 1 Moraxella spp. 1 Diphertoides 1 Bacteroides spp. e Gram- negative Bacilli associated with CRBSI (n) Klebsiella spp. 6 Enterobacter spp. 7 Escherichia spp. 2 Citrobacter spp. 2 Pseudomonas spp. e Fungal infections associated with CRBSI (n) Candida spp. 1
2007 10 e 1 1
0.5 0.1 1 0.3
10 10 e e e e 1
0.2
e e e e 1 1
As far as we are aware this is the first report of catheter sepsis rates solely in catheters used for parenteral nutrition as opposed to other fluids. We have described a significant reduction in sepsis when the catheter cleaning-method was changed to include chlorhexidine. Chlorhexidine in isopropanol was recommended for catheter connector antisepsis in the 2007 UK guidelines.22 All children selected were on PN treatment for more than 27 days for severe intestinal failure. Many of the children were immuno-suppressed due to the severity of their illness and/or the treatment. For example 8 patients, (or 19%) of the 2006 group and 19 (or 38%) of the 2007 group were post-bone marrow transplant and as a consequence at high risk of sepsis.23,24 An additional reason for a higher infection rate with PN is the high content of nutrients that might act as a culture medium for bacteria, hence promote bacterial growth and lead to bacteraemia.7 Children given PN via a central venous catheter would appear to be at high risk of developing CRBSI. Those at greatest risk are not only neonates with their immature immune systems, but also immunosuppressed children undergoing bone marrow transplant and surgical infants.13e15 Prevention of bloodstream infections in children with IF on intravenous nutrition therapy is essential for several reasons. These include 1. the symptoms and complications of the sepsis itself, 2. problems of replacing CVC when it is not possible to treat the sepsis with antibiotics and 3. association of liver disease with sepsis.2e4 In addition, this study demonstrated an improvement in development of intestinal autonomy in association with reduction in CRBSI. Children are frequently extremely unwell with sepsis and may rapidly develop life-threatening symptoms that need intensive care if the infection is not rapidly treated. At the same time in our experience the underlying intestinal failure is exacerbated and may take several days or even weeks to recover to the level at which it was prior to the infection. When the sepsis does not respond to appropriate antibiotic treatment, the central venous catheter may have to be removed and a new one inserted. If repeated CVCs are inserted the blood vessels used may become damaged and fibrotic and eventually block. One of the most serious complications of sepsis in children with IF is liver disease.2e4 Loss of venous access, recurrent episodes of sepsis and worsening liver disease are major causes of mortality,25 hence indications for evaluation for intestinal transplant.25
There were three differences between the two groups that might have predisposed to a higher infection rate in 2007 after introduction of chlorhexidine. Despite these factors the infection rate fell. Firstly, there was a higher use of lipid. PN formulas may predispose to infection by suppressing the immune response. Excess lipid administration can result in hypertriglyceridemia and impairment of leukocyte function.26,27 Lipid metabolites may accumulate in hepatic sinusoidal macrophages over time, thereby interfering with clearance and enhancing toxicity of bacteria and bacterial products in the bloodstream.28 Secondly, there were a higher proportion of immune-compromised patients after stem cell or solid organ transplantation and thirdly children studied were younger and would have had less mature immune systems. Both younger children and immune-compromised patients would be more vulnerable to infection. Younger children are also more likely to contaminate the line by pulling at it and playing with it, and by faecal material leaking from nappies, (particularly in children with diarrhoea), and have less well developed immunity than older children.29 Especially neonates are at high risk for developing bacterial infections due to quantitative and qualitative insufficiencies of innate immunity, particularly granulocyte lineage development and response to infection.29 Despite a greater potential for sepsis in the second group of patients the CRBSI rate fell in 2007. All these factors in 2007 make it more likely that the reduction in infection was attributable to the use of chlorhexidine. There was a similar finding in a prospective study of CVC hub care in neonates when using chlorhexidine along with additional education of medical staff with audio-visual presentation.15 In that study a significant decrease of CRBI rate was demonstrated with all catheter types and in all birth weight categories including extremely low birth weight infants that made up 40% of the cohort. In children with intestinal failure bacteraemia has been attributed either to the catheter when the hub is exposed for connections and disconnections or to translocation of bacteria from the gut.3e5 The profound fall in positive blood cultures following the change to chlorhexidine suggests that most of our infections originated from catheter manipulation and were not due to intestinal translocation. This supports the importance of adequate care of the CVC. It could be expected that the decrease in infection rate might have been associated with a reduction in length of hospitalisation in 2007 since less complications might be expected. Nevertheless patients in 2007 spent longer in hospital. Sicker infants appeared to be surviving in 2007 since the cohort of children were significantly younger and more had undergone transplantation; these two factors might have led to the longer length of in-patient stay after the introduction of chlorhexidine. More children gained enteral autonomy in 2007 than in 2006, in that they were weaned off PN onto full enteral feeding. The lower proportion of children with a primary intestinal diagnosis in 2007 could partly explain this. Small volumes of enteral feed, specific nutrients, growth factors or hormones may be useful tools for modulation of gut immunity.3e5,30 A non-used gut is prone to bacterial overgrowth that causes mucosal injury, malabsorption and bacterial translocation.3e5 A limitation of this study is that it was done in retrospect so other confounding factors might have been responsible, at least in part for the reduction in CRBSI. Decisions as to whether each positive blood culture signified a new episode of infection were made non-blindly and therefore open to observer bias. However, observer bias was kept to a minimum since a strict protocol was used as described in the methods section. In addition, initial blood culture readings were by machine so were unbiased and these fell five fold. It is possible that the introduction of chlorhexidine would have been associated with increased attention to hygiene policies, and
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021
J. Pichler et al. / Clinical Nutrition xxx (2013) 1e5
this could have contributed. However on other wards when major initiatives were made to increase attention to hygiene policies; the initiatives were not followed by falls in catheter related infections, so any such effect is very unlikely to account for the dramatic fivefold fall. In conclusion a change from the use of isopropanol alone to the use of 2% chlorhexidine in 70% isopropanol for the disinfection of CVC hubs was followed by a major fall in catheter related sepsis in infants and children with severe intestinal failure receiving PN via the catheter. The results support the UK guidelines on catheter handling which advocate the use of 2% chlorhexidine in isopropanol for catheter connector disinfection. In addition there was a significant improvement in outcome with more patients gaining enteral autonomy after the infection rate was reduced. Statement of authorship The paper was initiated by Susan Hill and James Soothill. The data collection and analyses were done by Judith Pichler and James Soothill. The manuscript was written by Judith Pichler with the provision of significant advice and consultation by Susan Hill and James Soothill. All authors critically appraised the manuscript. Funding sources There were funding sources. Conflict of interest statement There were no financial or personal relationships with any company or organization sponsoring the research at the time the research was done.
Acknowledgements We would like to thank for the assistance of Liz Wright for communicating her impression that chlorhexidine disinfected lines had less infections, Paul Lock and Bob Paget and for help with computer work, nursing staff for data entry and Microbiology staff for processing blood cultures.
References 1. Opilla M, Grove E, Richmond V. Epidemiology of bloodstream infection associated with parenteral nutrition. AJJIC 2008;36(10):173e6. 2. Koletzko B, Goulet O, Hunt J, Krohn K, Shamir R. Parenteral Nutrition Guidelines Working Group, European Society for Clinical Nutrition and Metabolism, European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN), European Society of Paediatric Research (ESPR). Guidelines on paediatric parenteral nutrition of the European Society of Paediatric Gastroenterology, Hepatology and nutrition (ESPGHAN) and the European Society for clinical nutrition and Metabolism (ESPEN), Supported by the European Society of paediatric research (ESPR). J Pediatr Gastroenterol Nutr 2005;41(Suppl. 2): S1e87. 3. Hermans D, Talbotec C, Lacaille F, Goulet O, Ricour C, Colomb V. Early central catheter infections may contribute to hepatic fibrosis in children receiving long- term parenteral nutrition. J Pediatr Gastroenterol Nutr 2007;44:459e63. 4. Kelly DA. Intestinal failure-associated liver disease: what do we know today? Gastroenterology 2006;130(2 Suppl. 1):S70e7.
5
5. Goulet O, Ruemmele F, Lacaille F, Colomb V. Irreversible intestinal failure. J Pediatr Gastroenterol Nutr 2004;38(3):250e69. 6. Johnson T, Sexton E. Managing children and adolescents on parenteral nutrition: challenges for the nutritional support team. Proc Nutr Soc 2006;65(3): 217e21. 7. Sizer T. Standards and guidelines for nutritional support of patients in hospital. Maidenhaed, Berks: BAPEN; 1996. 8. Sitges-Serra A, Puig P, Linares J, Perez JL, Farrero N, Jaurrieta E, et al. Hub colonization as the initial step in an outbreak of catheter-related sepsis due to coagulase negative staphylococci during parenteral nutrition. JPEN J Parenter Enteral Nutr 1984;8(6):668e72. 9. Salzman MB, Isenberg HD, Rubin LG. Use of disinfectants to reduce microbial contamination of hubs of vascular catheters. J Clin Microbiol 1993;31(3):475e9. 10. O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al., Healthcare Infection Control Practices Advisory Committee. Summary of recommendations: guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis 2011;52(9):1087e99. 11. Kuyyakanond T, Quesnel LB. The mechanism of action of chlorhexidine. FEMS Microbiol Lett 1992;79:211e5. 12. Casey AL, Worthington T, Lambert PA, Quinn D, Faroqui MH, Elliot TS. A randomized, prospective clinical trial to assess the potential infection risk associated with the PosiFlow needleless connector. J Hosp Infect 2003;54(4): 288e93. 13. Soothill JS, Bravery K, Ho A, Macqueen S, Collins J, Lock P. A fall in bloodstream infections followed a change to 2% chlorhexidine in 70% isopropanol for catheter connection antisepsis: a pediatric single center before/after study on a hemopoietic stem cell transplant ward. Am J Infect Control 2009;37(8):626e30. 14. Bishay M, Retrosi G, Horn V, Cloutman-Green E, Harris K, De Coppi P, et al. Chlorhexidine antisepsis significantly reduces the incidence of sepsis and septicemia during parenteral nutrition in surgical infants. J Pediatr Surg 2011;46(6):1064e9. 15. Sannoh S, Clones B, Munoz J, Montecalvo M, Parvez B. A multimodal approach to central venous catheter hub care can decrease catheter-related bloodstream infection. Am J Infect Control 2010;38(6):424e9. 16. Gowen H, Lloyd C. British Intestinal Failure Survey (BIFS): a referral registry to record and determine the outcome of childhood intestinal failure. Proc Nutr Soc 2009;68(OCE1):E15. 17. Onder AM, Chandar J, Billings A, Diaz R, Francoeur D, Abitol C, et al. Chlorhexidine- based antiseptic solutions effectively reduce catheter- related bacteremia. Pediatr Nephrol 2009;10:1154e5. 18. Onder AM, Chandar J, Chaokley S, Francouer D, Abitol C, Zilleruelo G. Controlling exit site infections: does it decrease the incidence of catheter-related bacteremia in children on chronic hemodialysis? Hemodial Int 2009;13(1): 11e8. 19. Newman C. Catheter- related bloodstream infections in paediatric intensive care unit. Semin Pediatr Infect Dis 2006;11:20e4. 20. Smith MJ. Catheter-related bloodstream infections in children. Am J Infect Control 2008;36(10):e1e3. S173. 21. Sarvikivi E, Lyytikainen O, Vaara M, Saxen H. Nosocomial bloodstream infections in children: an 8-year experience at a tertiary-care hospital in Finland. Clin Microbiol Infect 2008;14(11):1072e5. 22. Pratt RJ, Pellowe CM, Wilson JA, Loveday HP, Harper PJ, Jones SR, et al. Epic2: national evidence-based guidelines for preventing healthcare-associated infections in NHS hospitals in England. J Hosp Infect 2007;65(Suppl. 1):S1e64. 23. Bateman SL, Seed PC. Procession to pediatric bacteremia and sepsis: covert operations and failures in diplomacy. Pediatrics 2010;126(1):137e50. 24. Je˛ drzejczak-Czechowicz M, Kowalski ML. Effects of parenteral lipid emulsion on immune system response. Anestezjol Intens Ter. 2011;43(4):248e55. 25. Drews BB, Sanghavi R, Siegl JD, Metcalf P, Mittal NK. Characteristics of catheter-related bloodstream infections in children with intestinal failure: implications for clinical management. Gastroenterol Nurs 2009;32(6):385e90. quiz 391e382. 26. Caresta E, Pierro A, Chowdhury M, Peters MJ, Piastra M, Eaton S. Oxidation of intravenous lipid in infants and children with systemic inflammatory response syndrome and sepsis. Pediatr Res 2007;61(2):228e32. 27. Monson JR, Sedman PC, Ramsden CW, Brennan TG, Guillou PJ. Total parenteral nutrition adversely influences tumour-directed cellular cytotoxic responses in patients with gastrointestinal cancer. Eur J Surg Oncol 1988;14(5):935e43. 28. Cukier C, Waitzberg DL, Logullo AF, Bacchi CE, Travassos VH, Torrinhas RS. Lipid and lipid-free total parenteral nutrition: differential effects on macrophage phagocytosis in rats. Nutrition 1999;15(11e12):885e9. 29. Melvan JN, Bagby GJ, Welsh DA, Nelson S, Zhang P. Neonatal sepsis and neutrophil insufficiencies. Int Rev Immunol 2010;29(3):315e48. 30. Dhaliwal R, Heyland DK. Nutrition and infection in the intensive care unit: what does the evidence show? Curr Opin Crit Care 2005;11(5):461e7.
Please cite this article in press as: Pichler J, et al., Reduction of blood stream infections in children following a change to chlorhexidine disinfection of parenteral nutrition catheter connectors, Clinical Nutrition (2013), http://dx.doi.org/10.1016/j.clnu.2013.03.021