Strategies to Reduce Catheter-Related Bloodstream Infections in Pediatric Patients Receiving Home Parenteral Nutrition:

Transcription

1 Original Communication Strategies to Reduce Catheter-Related Bloodstream Infections in Pediatric Patients Receiving Home Parenteral Nutrition: The Efficacy of Taurolidine-Citrate Prophylactic-Locking Journal of Parenteral and Enteral Nutrition Volume 00 Number 0 xxx C 2018 American Society for Parenteral and Enteral Nutrition DOI: /jpen.1043 wileyonlinelibrary.com Cecile Lambe, MD 1 ; Catherine Poisson 1 ; Cecile Talbotec, MD 1 ; and Olivier Goulet, MD, PhD 1,2 Abstract Background: Catheter-related bloodstream infections (CRBSIs) remain a major issue in patients who are receiving home parenteral nutrition (HPN). The aim of this interventional study was to assess the impact of a new strategy using taurolidine-citrate (T-C) prophylactic locks on the CRBSI rate in children with intestinal failure who are receiving HPN. Methods: TherateofCRBSIswas monitored every calendar year in a prospective cohort of 195 children with intestinal failure. T-C locks were initiated from October 2011 in children with recurring CRBSIs ( 2 episodes per year). Results: In the whole cohort, the median annual CRBSI rate per 1000 catheter days decreased significantly from 2.07 in 2008 to 2010 to 1.23 in 2012 to 2014 (P <.05).T-Clockswereusedin40 patients. No adverse events were reported. In taurolidine-treated patients, the CRBSI rate per 1000 catheter days decreased from 4.16 to 0.25 (P <.0001). The cumulative percentage of patients free of CRBSI at 18 months was 92% (95% confidence interval [CI]: 71 98) on T-C lock vs 61% (95% CI: 49 72) in controls (P =.01). In multivariate analysis, factors associated with CRBSI were immune deficiency (adjusted hazard ratio 3.49; 95% CI: ) and the young age of the parents (adjusted hazard ratio 4.79, 95% CI: ), whereas T-C locks were protective (adjusted hazard ratio 0.22, 95% CI: ). Conclusion: This study confirms the efficacy of T-C catheter locks in decreasing the incidence of CRBSIs in children with intestinal failure who are receiving HPN. (JPEN J Parenter Enteral Nutr. 2018;00:1 9) Keywords taurolidine-citrate locks; intestinal failure; home parenteral nutrition; children; central venous catheter-related bloodstream infections Clinical Relevancy Statement Prevention of central venous catheter infection is critical in patients with intestinal failure. Only recently has prophylactic-locking of catheters been introduced in home parenteral nutrition patients with ethanol (in North America) and taurolidine (in Europe) with a subsequent reduction of catheter infections. Available studies in children are limited by sample size and by heterogenous patient selection. In this study, we report the experience of our center with a prospective follow-up of 195 children with intestinal failure who received home parenteral nutrition over 7 years and the effect of the introduction of taurolidine locks on reducing the catheter infection rate. In addition, we identify risk factors for repeated catheter infections in this specific population. Introduction Home parenteral nutrition (HPN) is the mainstay of nutrition therapy in children with intestinal failure. Although HPN management has improved over the past decades, the occurrence of catheter-related bloodstream infections (CRBSIs) still remains a major issue. The mechanisms and the risk factors of CRBSIs in children receiving HPN are not yet completely elucidated. Some children will never have From the 1 Department of Pediatric Gastroenterology, Hepatology and Nutrition, Rehabilitation Center for Intestinal Failure and Home Parenteral Nutrition, Necker-Enfants Malades Hospital, Paris, France; and 2 University Paris Descartes, Paris, France. Financial disclosure: None. Conflicts of interest: None. Received for publication September 1, 2017; accepted for publication October 25, This article originally appeared online on xxxx 0, Corresponding Author: Cecile Lambe, MD, Service de Gastro-entérologie, Hépatologie et Nutrition pédiatrique, Hôpital Necker-Enfants Malades, 149 Rue de Sèvres, Paris, France. cecile.lambe@aphp.fr

2 2 Journal of Parenteral and Enteral Nutrition 00(0) an infection, whereas others will have repeated CRBSIs, even though the HPN protocol and the standard of care are the same. The different strategies that have been developed over the years to reduce the occurrence of CRBSIs, notably regular replacement of the catheter, antibiotic locks, pulsatile flushing, evaluation of caregivers and retraining, change of antiseptics, and introduction of positive pressure valves, have failed. Indeed, the incidence of CRBSIs in our patients who are receiving long-term HPN has been steady, around 2 infections per 1000 catheter days for the past 20 years. 1 The use of prophylactic-locking of catheters with antimicrobial solutions was proposed in the late 1990s. Taurolidine and taurolidine-citrate (T-C) locks have been used to prevent CRBSIs in patients receiving HPN. 2 Taurolidine is an antimicrobial agent and prevents biofilm formation, whereas citrate prevents clot formation. 3 In adults receiving HPN, both randomized controlled trials 4 and observational studies 5,6 reported a significantly decreased incidence of CRBSIs with the use of prophylactic taurolidine and T-C locks. In children, efficacy of T-C locks was demonstrated in patients with cancer. 7 However, studies in patients with intestinal failure who are receiving HPN are few and are limited by small numbers of patients. 8,9 T-C locks were introduced in 2011 in our center in the setting of an observational study with a prospective follow-up. From 2011 to 2014, the use of T-C locks was restricted to children on HPN with repeated CRBSIs. The aim of the present study was to assess the rate of CRBSIs in a large pediatric HPN population before and after the implementation of T-C locks. Methods Patients The study was performed in Necker-Enfants Malades Hospital, Pediatric Rehabilitation Center for Intestinal Failure and Home Parenteral Nutrition in Paris. Our center takes care of about 50% of children with intestinal failure in France. Consecutive patients with intestinal failure requiring HPN in our center from January 1, 2008, to December 31, 2014, were included in the study. T-C locks were initiated from October 2011 on patients who had recurring catheter infections diagnosed using Infectious Diseases Society of America (IDSA) criteria. 10 All parents gave informed consent to the initiation of the lock. The study was approved by the local medical ethical committee, and informed consent was obtained from the parents before inclusion. During the whole length of the study none of our procedures (ie, educational programs, antiseptic, catheter cleaning, or flushing procedures) were changed. HPN Protocol In Hospital. Before being admitted into the HPN program, the parents and their social situation were evaluated during a 1-hour interview with the treating physician and the training nurse. Sometimes a home visit took place to check that the living conditions were compatible with HPN. Parents were trained in handling the catheter and the administration of parenteral nutrition (PN) bags during a 3- to 4-week program, which systematically took place in our rehabilitation center before returning home with their child receiving PN. The training was divided into 2 consecutive parts: the first, theoretical, teaching the risks of central catheters, of infection, and of parenteral infusion; and the second practical, parents learned hands-on to connect and disconnect the catheter, first on a training doll and then on the child. The objective was for parents to become expert at these complicated procedures and to be aware of the risks. It was only when the parents had validated all the procedures that the child was sent home to start HPN. A final meeting was held with the physician, the training nurse, and the parents before discharge to alert them again to the risks of catheter infection and the urgent necessity to come back to the emergency department in case of fever for a physical examination, blood check, catheter blood cultures and peripheral blood cultures. The vital risk for severe sepsis due to CRBSIs was strongly emphasized. Only single-lumen silicone tunneled central venous catheters (CVCs; Broviac) were used. At Home. Parenteral compounded bags were delivered weekly to the homes by refrigerated trucks and stored in a PN-dedicated fridge with temperature monitoring. Caregivers wore full-body sterile gowns, masks, and sterile gloves. The catheter hubs were cleaned and manipulated with sterile gauze impregnated with chlorhexidine-alcohol (chlorhexidine digluconate 0.5%, ethanol 96%; Gifrer, Decines-Charpieu, France). Positive pressure valves were systematically used (Autoflush; Vygon, Ecouen, France). Valves were changed once a week and were systematically disinfected with chlorhexidine-alcohol on sterile gauze before connection for a minimum of 60 seconds. At disconnection, the catheter was flushed with saline (NaCl 0.9%, prefilled syringe; BD Saline XS; BD Medical, Erembodegem, Belgium). The catheter hub was then protected by a dry sterile gauze and a self-adherent bandage (3M; Coheban, Cergy-Pontoise, France).The catheter subcutaneous entry point was dressed once to twice a week with chlorhexidine and protected with semipermeable dressings IV3000 (Smith & Nephew, Neuillysur-Seine, France) or Mepore (Mölnlycke Health Care, Wasquehal, France). Assistance of a specialist nurse at home was recommended but was not always possible or compulsory.

3 Lambe et al 3 Patient Follow-Up. Patients were seen in our outpatient clinic every 2 4 months. Their medical history was recorded, and they underwent a physical examination and a blood check. In the meantime, parents and local treating physicians kept the nurses or the physicians informed through telephone or . Any new event or change of situation was reported in the medical records and in the database. T-C Lock Prophylaxis T-C locks (taurolidine 1.35%, citrate 4%; TauroLock catheter lock solution; TauroPharm, Waldbüttelbrunn, Germany) were initiated in the cohort after October 2011 in patients who had 2 CRBSIs in a 12-month period (diagnosed according to the IDSA criteria). 10,11 A new infection (to exclude a relapse or a treatment failure) was defined as a CVC infection occurring >60 days after the first negative blood culture of the latest infection. When a child was admitted for a CRBSI, we looked at the date of the previous CVC infection: if the latest infection was <12 months before and the first negative blood culture of the latest infection was >60 days, the child was started on T-C lock after the end of the CRBSI treatment, which required a minimum 2-week IV antibiotic course. Blood cultures had to be sterile for >10 days. In most cases, the catheter was not removed. A new device was implanted only in case of Staphylococcus aureus or Pseudomonas aeruginosa catheter infection, when another vascular access was still available. Once the infection was correctly treated, parents followed a specific 2-hour training session to use T-C lock with our training nurse according to the following procedure: After disconnecting each overnight PN infusion, the catheter was flushed with 10 ml saline followed by a very slow instillation of T-C lock solution in 5 10 seconds into the positive pressure valve to fill the catheter lumen (from the hub, including the valve, to the tip, ml according to the size of the catheter). The T-C lock solution was prepared by the parents from a 3-mL single-use glass vial with a 1- ml syringe. The indwelling time was variable but was always between 2 cycles of parenteral infusions varying from 8 to 60 hours (median 12 hours). Before each perfusion, the T-C lock solution was withdrawn from the catheter with a 1-mL syringe. The catheter was then flushed with 10 ml saline 0.9%, and PN infusion was connected. Once on T-C lock, any new event in the patient situation was monitored at follow-up consultations and at any time by the nurses and physicians through phone calls. Signs evocative of adverse events (pain, nausea and vomiting, abnormal taste sensation, and allergic reaction, while manipulating the CVC) were systematically asked for at each consultation. In children who were not treated by T-C lock the protocol remained the same as usual: After disconnection, the catheter was flushed with saline into the positive pressure valve. At connection, the catheter was flushed with saline and the PN bag connected. Heparin locks have not been used since 2000 in our center. Data Collection Clinical information was collected prospectively: age, sex, cause of intestinal failure, fever episodes, CRBSIs, type of microorganisms, catheter exit site subcutaneous infection, catheter occlusions, and catheter rupture or breakage needing repair. Data regarding demographics, parental situation, age of the parents, social issues, living conditions, and assistance of a nurse for the catheter care were collected at admission and during follow-up. All this information was recorded in a database that was actualized every calendar year with a CRBSI s yearly incidence rate per 1000 catheter days. Outcome The primary outcome was the incidence of CRBSI diagnosed according to the IDSA definition. 10,11 CRBSI was suspected whenever a child presented with fever or chills. The child was admitted to the hospital in emergency. Physical examination and blood tests (blood count, C-reactive protein, procalcitonin, catheter and peripheral blood cultures) were systematically performed. A chest x-ray and a urine specimen were performed to exclude another focus of infection than the central line. Blood cultures were repeated on the central line at least twice before the initiation of empiric antibiotics (continuous vancomycin administrated through the CVC combined with cefotaxime and gentamicin in case of severe sepsis) and then at least once a day. CRBSI was confirmed when 1 CVC blood cultures + 1 peripheral blood cultures were positive to the same pathogen or in case of 2 CVC positive blood cultures to the same pathogen without another body site documented infection. Statistical Analysis Continuous data are expressed as a median (interquartile range [IQR]), and differences between the groups were tested for significance by Mann-Whitney test. Discrete data are given as percentages, and comparisons were made with Pearson χ 2 test. Incidence rates of CRBSI were calculated per 1000 catheter days, pooling data from patients groups. Comparison of 2 therapeutic periods in the same individuals was made with the paired Wilcoxon rank-sum test. Kaplan-Meier survival analysis was used to estimate the cumulative probability of remaining free from CRBSI from the date of T-C lock initiation as starting point. Patients receiving HPN who did not receive T-C lock during the same period served as control patients. In patients in this group who were already receiving HPN, the median date of T-C

4 4 Journal of Parenteral and Enteral Nutrition 00(0) lock initiation in the index patients was chosen arbitrarily as the starting point. All variables collected at the starting point suspected to be possible predictors of CRBSI (sex, age, prior HPN duration, gastrostomy, intestinal stoma, cause of intestinal failure, immunodeficiency [congenital immune deficiency, immunosuppressive treatment], actual training of both parents, parents age, single-parent family, social disadvantage [alcoholic parent, deplorable living conditions, childcare services, absence of income apart from social welfare], and assistance of a nurse at home) were tested by log-rank. A Cox proportional hazards regression model with a backward variable elimination procedure was used to assess the strength of the associations while controlling for possible confounding variables. Variables significant with P <.20 entered into the model. Calculations were performed using GB-stat statistical software (Silver Spring, MD). Results From January 2008 to December 2014, 195 children went through our HPN program representing a total of 203,468 catheter days. Among them, 67 were receiving HPN on January 1, 2008, for a prior median duration of 2.8 years (IQR 1.1 8), and 128 started HPN between 2008 and The number of patients entering the HPN program each year varied from 14 to 26. Patient characteristics are described in Table 1. Incidence of CRBSIs Over Time in the Whole Cohort Figure 1 illustrates the association between the introduction of T-C locks and the decrease of CRBSI rate over the whole cohort. Table 2 gives the incidence of CRBSIs per calendar year throughout the study.there was a significant decrease in CRBSIs over time. Meanwhile, the number of patients treated with T-C lock increased from 0 in 2008 to 2010 to 4 in 2011, 15 in 2012, 25 in 2013, and 36 in The introduction of T-C locks was associated with a 40% decrease in the annual incidence of CRBSIs, from 2.07 per 1000 catheter days during the 2008 to 2010 period, to 1.23 per 1000 catheter days in the subsequent period (year 2011 excluded; P <.05). The incidence of catheter occlusions leading to catheter replacement remained unchanged, from 0.10 per 1000 catheter days in 2008 to 0.09 per 1000 catheter days in 2014 (P =.92). Likewise, the incidence of catheter exit site subcutaneous infections did not change significantly over time (from 0.39 in 2008 to 0.18 per 1000 catheter days in 2014; P =.19). Outcome of Patients Treated With T-C Locks From October 2011 to December 2014, T-C locks were initiated in 40 patients. Five of those patients (12.5%) Table 1. Patient Characteristics. Male sex, n (%) 109 (56) Age a (y), median (IQR) 1.7 ( ) Cause of intestinal failure, n (%) Short bowel syndrome 95 (49) Long-segment Hirschsprung s disease 22 (11) Chronic intestinal pseudoobstruction 27 (14) Congenital enteropathy 33 (17) Immune deficiency with chronic diarrhea 5 (3) Digestive graft vs host disease after bone 4(2) marrow transplantation Crohn s disease 2 (1) Autoinflammatory disease 3 (2) Miscellaneous 4 (2) HPN total duration b (y), median (IQR) 2.5 (1 6) Broviac catheter, n (%) 195 (100) Gastrostomy, n (%) 66 (34) Permanent intestinal stoma, n (%) 51 (26) Still receiving HPN on December 31, 2014, n 106 (54) (%) Reasons for leaving the HPN program, n (%) c Autonomy via the enteral route 55 (28) Intestinal transplantation 12 (6) Transfer to adult HPN center 6 (3) Transfer to another parenteral nutrition 3(2) center Death 13 (7) Nine patients (3 with Hirschsprung s disease, 1 with chronic intestinal pseudoobstruction, and 5 with congenital enteropathy) underwent intestinal transplantation and graft had to be removed. a When starting HPN or in January 2008 in patients receiving home parenteral nutrition (HPN) before b Including years before c Eighty-nine patients eventually left the HPN program. IQR, interquartile range. No. CRBSI per 1000 catheter days (bars) P<0,05 Figure 1. Incidence of catheter-related bloodstream infections (CRBSI) and proportion of patients receiving taurolidinecitrate (T-C) locks prophylaxis during the study period. experienced development of a CRBSI after T-C lock initiation during a median observation period of 18 months. Cultures were positive in 2 cases for Staphylococcus aureus meti-s and in other cases for Pseudomonas aeruginosa, Percentage of pa ents on T-C locks (dots)

5 Lambe et al 5 Table 2. Incidence of Catheter-Related Bloodstream Infections per Calendar Year Throughout the Study. Year Number of patients Catheter days 25,527 28,171 29,593 27,538 29,115 30,650 33,407 Number of CRBSIs/year Coagulase-negative Staphylococcus (%) 42 (69) 50 (81) 30 (61) 54 (78) 37 (82) 29 (85) 24 (67) Staphylococcus aureus (%) 9 (15) 2 (3) 10 (20) 6 (11) 4 (9) 2 (6) 5 (14) Number of children without a CRBSI Number of children with T-C lock at the end of the calendar year Number of CRBSIs with T-C lock Number of exit site infections Yearly infection rate/1000 catheter days CRBSI, catheter-related bloodstream infection; T-C, taurolidine-citrate. Table 3. Comparison of Patients Receiving Taurolidine-Citrate Lock Prophylaxis and Control Patients at Inclusion. a. T-C Lock (n = 40) Controls (n = 86) P Male sex, n (%) 24 (60) 41 (48).27 Median age, y (IQR) 2.2 ( ) 5.7 ( ).34 Median of prior HPN duration, mo (IQR) 28 (14 70) 38 (11 90).70 Prior rate of CRBSIs per 1000 catheter days <.0001 Prior number of patients with a CRBSI in the previous 18 mo (%) b 40 (100) 22 (37) <.0001 Gastrostomy, n (%) 18 (45) 43 (50).74 Permanent intestinal stoma, n (%) 15 (37) 30 (35).92 Short bowel syndrome, n (%) 20 (50) 39 (45).76 Motility disorder, n (%) 12 (30) 27 (31) 1 Congenital enteropathy, n (%) 8 (20) 17 (20) 1 Immunodeficiency, n (%) 2 (5) 5 (6).82 Training of both parents, n (%) 12 (30) 47 (55).02 Both parents <25 y, n (%) 9 (22) 13 (15).45 Single-parent family, n (%) 6 (15) 17 (20).69 Social disadvantage, n (%) 9 (22) 6 (7).03 Assistance of a nurse at home, n (%) 25 (62) 42 (49).22 CRBSI, catheter-related bloodstream infection; IQR, interquartile range. a Inclusion: date of taurolidine-citrate (T-C) locks initiation; date of starting home parenteral nutrition (HPN) or April, 2013, in patients with no T-C locks. b Calculated in patients already included in the HPN program. Serratia marcescens, and Staphylococcus capitis (1 case each). In 2 cases, the T-C locks had been interrupted 48 hours before the CRBSI occurrence, in 1 case there had been a blatant mistake in the catheter handling, and in 2 cases there were no obvious explanations. In the 36 patients who had 2 consecutive available (>3 months duration) HPN periods before and while on T-C lock, the overall CRBSI rate decreased from 4.16 per 1000 catheter days without T-C lock to 0.25 per 1000 catheter days with T-C lock (P <.001). Comparative Occurrence of CRBSI With Controls Table 3 gives the comparison of T-C lock treated and untreated (control) patients. The overall rate of CRBSI in the 86 control patients was 0.89 per 1000 catheter days (vs 0.25 in T-C lock treated patients; P =.006). Survival without CRBSI is shown in Figure 2. The cumulative percentage of patients free of CRBSI at 18 months was 92% (95% CI: 71 98) on T-C lock vs 61% (95% CI: 49 72) in controls (P =.01). Tested covariates are given in Table 4. Both in univariate and multivariate analysis only immune deficiency and the young age of the parents were associated with an increased risk for development of a CRBSI. Adjusted hazard ratios (95% CI) were 3.49 ( ) for immune deficiency, 4.79 ( ) for young parents, and 0.22 ( ) for T-C lock prophylaxis. Adverse Event No potential side effects were reported. Children did not experience perioral dysesthesia, nausea, vomiting, abnormal taste sensations, or discomfort. No allergic reaction

6 6 Journal of Parenteral and Enteral Nutrition 00(0) % of patients out of infection P=0, months of home TPN No. at risk T-C lock Controls Figure 2. Incidence of catheter-related bloodstream infections (CRBSIs) per calendar year throughout the study. T-C, taurolidine-citrate; TPN, total parenteral nutrition. was observed. No patient or parent asked to interrupt the treatment. There was no increase of catheter breakage with the use of T-C locks: From 2011 to 2014, there were 2 catheter repairs for breakage in 20,403 catheter days in children on T-C locks, whereas there were 25 in 99,774 catheter days in children without T-C locks (P =.18); we excluded breakage caused by traumatic reasons (such as a child incidentally pulling on the catheter). There were no episodes of catheter obstruction in the children on T-C locks. The only issue that was encountered in 2 cases was the inability to withdraw the T-C lock from the catheter because of an absence of a blood reflux. However, infusion was possible. Discussion The present observational study of a large cohort of children with intestinal failure who were receiving HPN demonstrated a significant decrease of CRBSIs after the introduction of T-C locks prophylaxis. This decrease over time was not linked to an improvement in HPN care or a modified selection of patients. Indeed, the annual CRBSI incidence rate was stable during the years before introduction of T-C locks prophylaxis, and the number of children entering the HPN program yearly remained unchanged. Finally, multivariate analysis showed that T-C locks after adjustment for other confounders had a major impact, dividing about 5 times the cumulative risk for CRBSI. No adverse event related to T-C locking was observed. Several limitations of this study should be considered. First, the comparison of incidence of CRBSIs in patients on T-C locks vs those receiving conventional HPN management was not based on similar groups of patients. Actually, patients on T-C locks tended to be younger and socially disadvantaged, and were selected by their high incidence of CRBSIs before initiation of T-C locks. Thus, they were at higher risk for subsequent infections, whereas we did observe the opposite result. We cannot exclude, however, that parents became more careful with their catheter handling technique because their child had experienced prior infections or because they were using a new procedure. A second limitation was that the finding of infection may have been modified by the parents and physicians knowledge of the current therapy. This is very unlikely because the diagnostic criteria of CRBSI were strict and objective. Third, the proportion of HPN children on T- C locks increased markedly over recent years, and a calendar Table 4. Variables Tested as Predictors of Catheter-Related Bloodstream Infections and Results of Univariate and Multivariate Analyses. n(%) P (Log-Rank) P (Cox Model) Male sex 65 (52).45 Age <36 mo 52 (41).19 NS Prior home parenteral nutrition duration <18 mo 83 (66).16 NS Gastrostomy 61 (48).52 Permanent intestinal stoma 45 (36).44 Short bowel syndrome 59 (47).41 Motility disorder 39 (31).80 Congenital enteropathy 25 (20).28 Immune deficiency 7 (6) Training of both parents 59 (47).66 Both parents <25 y 22 (17) Single-parent family 23 (18).66 Social disadvantage 25 (20).24 Assistance of a nurse at home 67 (53).29 Taurolidine-citrate lock prophylaxis 40 (32) NS, not significant.

7 Lambe et al 7 bias is possible. However, no other procedural changes other than the catheter lock strategy were implemented during the observation period: There was no changing of antiseptics and no changes of valve type; similar rules of CRBSIs treatment and replacement of catheters were applied, and the education of caregivers and parents was the same, given by the same providers. Moreover, we restricted the comparison with control patients to the most recent period: the date of inclusion in the control group corresponding to the median calendar date of inclusion in the T-C lock group. Finally, if one regards the whole cohort of our HPN patients, the decrease of the CRBSI rate was of relatively limited magnitude, from 2.07 to 1.23 per 1000 catheter days. This may be partly explained by the fact that a low proportion of patients (about one-third, only those at high risk) were put on T-C locks. On the other hand, the benefit of T-C locks in patients who have a low infection rate has been questioned. 12 Our results confirm data reported in 2 retrospective series of children receiving HPN. Chu et al 8 reported that in 19 children with intestinal failure, the introduction of T- C locks was followed by a dramatic drop in CRBSI rate, from 8.6 to 1.1 per 1000 catheter days. Fourteen of the 19 patients (74%) had no infections for up to 33 months after switching from heparin to T-C locks. Another short series of 7 children from the Netherlands reported a drop of CRBSI from 12.7 to 4.3 per 1000 catheter days after taurolidine locks introduction. 9 In adults receiving HPN, the prophylactic effect of taurolidine locks compared with heparin was demonstrated by randomized studies. 4,5 It was shown to decrease CRBSI incidence over time from 2008 to 2016 in these high-risk patients. 5 However, in HPN patients who have a low infection rate, the additional clinical value of T-C locks is limited, and cost-effectiveness is low. 12 Guidelines recommend the use of prophylactic antimicrobial lock solution only in patients with long-term catheters who have a history of multiple CRBSIs despite optimal adherence to aseptic technique. 13 T-C locks are of particular interest because of their action on the biofilm, with the catheter hub s microbial contamination being the main catheter infection pathway in case of long-term tunneled catheter, the absence of microbial adaptation to taurolidine, 8,14 and their efficacy in immune-compromised pediatric patients. 15 T-C locks do not protect from improper catheter handling and catheter entry site subcutaneous infections. Side effects have been reported in 20% of pediatric patients with hematological malignancies, including discomfort in the chest and neck, perioral dysesthesia, nausea, vomiting, and abnormal taste sensations 15 considered to be caused by the flushing of citrate in the blood circulation. Allergic reactions have been reported. 5,13 They are unlikely to be due to taurolidine itself, which is metabolized into taurine and carbon dioxide, nor to citrate, but to other constituents such as polyvinyl pyrrolidine (PVP 17) used for stability reasons. An excess of thrombotic events has also been reported in 1 randomized study using interdialytic T-C locks in patients receiving hemodialysis. 16 Other controlled studies did not report relevant side effects or increased occurrence of thrombosis. 17 Discomfort and allergic reaction were not observed in our patients, probably because the T-C lock solution was very slowly instilled into the catheter at PN disconnection and systematically withdrawn from the catheter before infusion. Also, there was no increased occurrence of thrombosis. Ethanol locks have been widely used in North America to prevent CRBSIs in patients who are receiving HPN. 18 Complications have been reported, however, especially catheter occlusions. 19,20 There are no studies comparing taurolidine with ethanol locks regarding efficacy in reducing CRBSI incidence, side effects, and costeffectiveness. The taurolidine catheter-locking procedure is not available in the United States and has not been reviewed by the U.S. Food and Drug Administration. In addition to the bactericidal efficacy of prophylactic T-C lock in HPN children, we found that the risk for CRBSI was increased in children with young parents and in case of immunodeficiency. In adult patients who are receiving HPN, the risk factors of infections are related to the patient, his or her education, the venous access device (subcutaneous infusion ports instead of tunneled catheters, multiple-lumen catheters), HPN therapy (frequency of lipid emulsion infusion), and using the catheter to remove blood or giving other medications via the catheter. 21,22 In the pediatric HPN population, the identification of risk factors of infection is more difficult because series are smaller, parents are key actors of the HPN program, and causative factors are more complex. Our group previously did not recognize significant risk factors. 23 Other groups reported that age <3 years, 24 parental disadvantage, 24 short bowel syndrome, 25 and increased PN frequency 22 were associated with an increased risk for infection. The finding that a young age of both parents was associated with an increased risk for CRBSI in this study suggests that the active cooperation of the whole family is crucial in the management of HPN. Particular attention should be paid to the training and the implication of young parents. In agreement with Zamvar et al, 24 we believe that measures to support improved compliance may improve outcomes. In addition, immune deficiency is another expected but unavoidable risk factor. Actually, immune-deficient children who are receiving HPN are good candidates for primary T-C lock prophylaxis because T-C locks are efficient in immune-compromised children with malignancy. 7,15 Children with definitive intestinal failure and risk factors for Nutritional failure, such as limited central vascular access due to multiple thrombosis, associated liver disease, or psychosocial issues, are also good candidates for primary prophylaxis with T- C locks because a single CRBSI could precipitate their outcome. 26

8 8 Journal of Parenteral and Enteral Nutrition 00(0) In conclusion, this study confirms the efficacy of T- C locks for reducing catheter infection rate in children with intestinal failure who are receiving HPN. There was no sign of any adverse event with T-C locks, but experience is of limited duration in this population of young children. Furthermore, locking complicates the PN procedure and its sustained efficacy in the long term remains to be established. 14 Although taurolidine locks should be currently an important part of the daily management of HPN, 27 they should not replace other strategies to reduce CRBSIs such as rigorous hand hygiene, a strict catheter aseptic handling technique, and the education of the caregivers. In the future, finding the most efficient taurolidine formulation, 28 development of alternative costeffective methods of bactericidal locking, and long-term observational studies of HPN patients on T-C locks are desirable. Acknowledgments The authors wish to thank Jacques Cosnes for his assistance in writing this article, Malcolm Lambe for reviewing the English, and Virginie Colomb and Odile Corriol for making possible the treatment of children with T-C locks in our center. Statement of Authorship C. Lambe and C. Poisson contributed to the conception and the design of the research. C. Lambe, C. Poisson, C. Talbotec and O. Goulet contributed to the acquisition of the data. C. Lambe and C. Poisson contributed to the analysis and the interpretation of the data. C. Lambe drafted the manuscript. All authors critically revised the manuscript, agree to be fully accountable for ensuring the integrity and accuracy of the work, and read and approved the final manuscript. References 1. Abi Nader E, Lambe C, Talbotec C, et al. Outcome of home parenteral nutrition in 251 children over a 14-y period: report of a single center. Am J Clin Nutr. 2016;103: Jurewitsch B, Lee T, Park J, Jeejeebhoy K. Taurolidine 2% as an antimicrobial lock solution for prevention of recurrent catheterrelated bloodstream infections. JPEN1043 J Parenter Enteral Nutr. 1998;22: Shah CB, Mittelman MW, Costerton JW, et al. Antimicrobial activity of a novel catheter lock solution. Antimicrob Agents Chemother. 2002;46: Bisseling TM, Willems MC, Versleijen MW, Hendriks JC, Vissers RK, Wanten GJ. Taurolidine lock is highly effective in preventing catheter-related bloodstream infections in patients on home parenteral nutrition: a heparin-controlled prospective trial. Clin Nutr. 2010;29: Olthof ED, Versleijen MW, Huisman-de Waal G, Feuth T, Kievit W, Wanten GJ. Taurolidine lock is superior to heparin lock in the prevention of catheter related bloodstream infections and occlusions. PLoS One. 2014;9:e Toure A, Lauverjat M, Peraldi C, et al. Taurolidine lock solution in the secondary prevention of central venous catheter-associated bloodstream infection in home parenteral nutrition patients. Clin Nutr. 2012;31: Handrup MM, Moller JK, Schroder H. Central venous catheters and catheter locks in children with cancer: a prospective randomized trial of taurolidine versus heparin. Pediatr Blood Cancer. 2013;60: Chu HP, Brind J, Tomar R, Hill S. Significant reduction in central venous catheter-related bloodstream infections in children on HPN after starting treatment with taurolidine line lock. J Pediatr Gastroenterol Nutr. 2012;55: Hulshof EC, Hanff LM, Olieman J, et al. Taurolidine in pediatric home parenteral nutrition patients. Pediatr Infect Dis J. 2017;36: Manian FA. IDSA guidelines for the diagnosis and management of intravascular catheter-related bloodstream infection. Clin Infect Dis. 2009;49: ; author reply Mermel LA, Farr BM, Sherertz RJ, et al. Guidelines for the management of intravascular catheter-related infections. J Intraven Nurs. 2001;24: Klek S, Szczepanek K, Hermanowicz A, Galas A. Taurolidine lock in home parenteral nutrition in adults: results from an open-label randomized controlled clinical trial. JPEN1043 J Parenter Enteral Nutr. 2015;39: O Grady NP, Alexander M, Burns LA, et al. Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2011;52:e162-e Olthof ED, Rentenaar RJ, Rijs AJ, Wanten GJ. Absence of microbial adaptation to taurolidine in patients on home parenteral nutrition who develop catheter related bloodstream infections and use taurolidine locks. Clin Nutr. 2013;32: Dumichen MJ, Seeger K, Lode HN, et al. Randomized controlled trial of taurolidine citrate versus heparin as catheter lock solution in paediatric patients with haematological malignancies. J Hosp Infect. 2012;80: Solomon LR, Cheesbrough JS, Ebah L, et al. A randomized doubleblind controlled trial of taurolidine-citrate catheter locks for the prevention of bacteremia in patients treated with hemodialysis. Am J Kidney Dis. 2010;55: LiuH,DengJ,ChenL,YuanL,WuY.Preventingcatheter-related bacteremia with taurolidine-citrate catheter locks: a systematic review and meta-analysis. Blood Purif. 2014;37: Rahhal R, Abu-El-Haija MA, Fei L, et al. Systematic review and metaanalysis of the utilization of ethanol locks in pediatric patients with intestinal failure. JPEN1043 J Parenter Enteral Nutr. Published online August 2, Mokha JS, Davidovics ZH, Samela K, Emerick K. Effects of ethanol lock therapy on central line infections and mechanical problems in children with intestinal failure. JPEN1043 J Parenter Enteral Nutr. 2017;41: Wong T, Clifford V, McCallum Z, et al. Central venous catheter thrombosis associated with 70% ethanol locks in pediatric intestinal failure patients on home parenteral nutrition: a case series. JPEN1043 J Parenter Enteral Nutr. 2012;36: Dreesen M, Foulon V, Spriet I, et al. Epidemiology of catheter-related infections in adult patients receiving home parenteral nutrition: a systematic review. Clin Nutr. 2013;32: Buchman AL, Opilla M, Kwasny M, Diamantidis TG, Okamoto R. Risk factors for the development of catheter-related bloodstream infections in patients receiving home parenteral nutrition. JPEN1043 J Parenter Enteral Nutr. 2014;38: Colomb V, Fabeiro M, Dabbas M, Goulet O, Merckx J, Ricour C. Central venous catheter-related infections in children on long-term home parenteral nutrition: incidence and risk factors. Clin Nutr. 2000;19:

9 Lambe et al Zamvar V, Puntis JW, Gupte G, et al. Social circumstances and medical complications in children with intestinal failure. Arch Dis Child. 2014;99: Diamanti A, Basso MS, Castro M, Calce A, Pietrobattista A, Gambarara M. Prevalence of life-threatening complications in pediatric patients affected by intestinal failure. Transplant Proc. 2007;39: D Antiga L, Goulet O. Intestinal failure in children: the European view. J Pediatr Gastroenterol Nutr. 2013;56: Pittiruti M, Bertoglio S, Scoppettuolo G, et al. Evidence-based criteria for the choice and the clinical use of the most appropriate lock solutions for central venous catheters (excluding dialysis catheters): a GAVeCeLT consensus. J Vasc Access. 2016;17: Olthof ED, Nijland R, Gulich AF, Wanten GJ. Microbiocidal effects of various taurolidine containing catheter lock solutions. Clin Nutr. 2015;34: