Infants with extended stays in the newborn intensive care units (NICU) often receive all or a large portion of their nutritional needs through intravenous (IV) fluids. Many of these infants require life-sustaining IV medications administered by either bolus or continuous drip method. Due to the fragility of peripheral veins, central lines are often placed as a semipermanent method of providing nutrition or medications. Central lines (CLs) are inserted using sterile technique by skilled professionals in the NICU, interventional radiology, or in surgery. The length of the line, externally and internally, is carefully measured to ensure the terminal end of the line enters a central vein. The insertion site is covered with a dressing to protect the line and the site from possible bacterial contamination. Changing the dressing presents an increased risk for CL dislodgement. Dislodgement occurs when the CL is inadvertently pulled from its central location into a peripheral vein.
At a large tertiary children's hospital, CL dislodgement has been identified as a frequent occurrence and an evaluation of central line maintenance bundles was requested to minimize adverse events. The types of CL used within the NICU can be grouped into 2 different categories.1 The first group is tunneled central lines (tCLs). These are lines that are inserted through the skin either at an extremity, femoral vein, or chest/neck area and are tunneled through the skin entering a peripheral vessel and advanced until the terminal end reaches a central vessel. The second group is nontunneled (ntCL) where the line enters the vessel close to the insertion site and the line is threaded through the peripheral vessel until the terminal end is in the central vessel.2,3
In the prior 5 months, there were 19 incidences of CL dislodgements and 5 CL infections. Fourteen dislodgements and 4 infections occurred with nontunneled CL (ntCL). Currently, the policy is to change the dressing using a sterile, 2-person technique with a chlorhexidine-impregnated patch to ensure security of the line; however, even with implementation of this policy the number of dislodgements and infections continues to rise.
There is some debate as to the terminology used to describe these lines. Healthcare professionals may use the terms peripherally inserted central catheter (PICC) lines and percutaneous central venous catheter (PCVC) lines interchangeably, but for some, it depends on the manufacturer and the catheter lumen size. For this tertiary center, a PCVC is only an ntCL and has a smaller lumen size that delivers fluids and medications but is not used for drawing blood, whereas a PICC, either tCL or ntCL, has a larger lumen size and blood can be drawn if necessary for assessing laboratory values. Regardless of the type of catheter used, there is no difference in length of time a CL can be in place if the line is maintained properly and functioning. Currently, the CL maintenance bundle is a “one size fits all” approach across all patients within our hospital. According to our unit-based policy, the dressing is changed 24 hours after the initial insertion of the line and a chlorhexidine impregnated disc (BioPatch) is placed on the insertion site. The entire area is then covered with a bordered, transparent dressing (SorbaView). Furthermore, according to the policy the dressing is to be changed every 7 days and the BioPatch replaced using a 2-person sterile technique.4
PubMed, CINAHL, and Cochrane databases were systematically searched using the key words “central line,” “neonatal,” “dressing change,” “PICC line,” and “care bundle.” Articles included in the review were those published in English, in the last 10 years and specific to the NICU population. Literature and recommendations from the National Association of Neonatal Nurses, the Centers for Disease Control and Prevention (CDC), and the American Academy of Pediatrics were also included.
Although the CDC has placed reducing hospital-acquired infections as a top priority, and there have been countless studies regarding central line maintenance, there is little literature related to the specific NICU population found within a tertiary care center.3 These infants are often of greater gestational age as well as postmenstrual age and larger weight but still very fragile with complex needs. These infants often lack IV access sites, making CL management and maintenance even more critical. Although recent studies have shown that the BioPatch may help in reducing infections, in the NICU population, no statistically significant differences have been found.4,5 Use of the BioPatch requires frequent dressing changes every 7 days. However, the rates/risks of dislodgements or infection from dressing compromise are greater than the potential benefit.4,5 Furthermore, due to the fragility of the infant's skin, the risk of skin breakdown and/or skin irritation, it is now no longer recommended to use a BioPatch and regularly change the dressing but only use a transparent, semipermeable dressing over the site and change when necessary. Waiting to change the dressing when it is still intact has been found to not increase infection rates.1,6
Currently, the CDC only recommends using the BioPatch for patients 18 years and older.7 Previous recommendations for the NICU population included use of a transparent, semipermeable dressing to cover the insertion site.8–10 For the larger infant who is more active yet still has high probability for CL failure, a stronger bordered dressing (such as SorbaView) may be more beneficial.2,11 Evidence also shows there was no statistically significant increase in infection rates for ntCL when the line was in place for longer durations of time; however, maintenance practices were found to contribute to increases in infection and dislodgement rates.1,6,12 Evidence supports that the use of a standardized bundle that is specific for neonates while accounting for the unique needs of the larger, more complex infant has the potential to significantly reduce infection and dislodgment rates.2,13
Table 1 shows the new recommended CL maintenance bundle supported by current evidence. Implementation now includes no longer using a BioPatch for ntCL and after insertion, placing a transparent, semipermeable dressing over the site and only remove and change when the dressing is compromised.
The Iowa Model of Evidenced-Based Practice was used to guide this project.15 This model can be applied to an entire healthcare system. Prior to implementing this new CL bundle, meetings with the director of infection prevention, vascular access team, interventional radiology, and all medical and nursing staff in the NICU were convened to involve all the stakeholders in the process.15 Nursing education to support the implementation of the care bundle included development of a poster in the unit, which described the new dressing change process along with the training of 7 “super users” that were available for additional education and assistance with implementation. The 2-person sterile dressing technique was already in practice, so the new education was related to the changing the dressings as needed rather than at a regularly scheduled interval and no longer using the BioPatch for ntCL.
All NICU patients with a CL were included for data collection, but only ntCL included in implementation of the CL maintenance bundle. The type of CL included tCL (Broviac, internal jugular, PICC) or ntCL (PICC, PCVC). To provide baseline data, retrospective data for a 5-month period were collected on all CL patients. There were 36 patients for a total of 59 CLs. After the implementation of the new CL maintenance bundle, another 5 months of data were collected on 39 patients for a total of 51 CLs. Two patients, each with a CL, were excluded from data collection because part of their CL days were preimplementation and part were postimplementation.
Table 2 describes the type of data that were collected so that all factors could be evaluated for maintenance bundle effectiveness.
Of the total CL, 6.4% (n = 7) were Broviac, 0.9% (n = 1) were internal jugular, 34.5% (n = 38) were tunneled PICCs, 36.4% (n = 40) were nontunneled PICCs, and 21.8% (n = 24) were PCVCs. The location of CL placement was 5.5% (n = 6) central chest/neck, 26.4% (n = 29) right upper extremity, 13.6% (n = 15) left upper extremity, 35.5% (n = 39) right lower extremity, 18.2% (n = 20) left lower extremity, and 0.9% (n = 1) scalp. Most were double lumen lines 86.4% (n = 95) and only 13.6% (n = 15) were single lumen lines. For CL placement, 7.3% (n = 8) were placed by a surgeon, 32.7% (n = 36) by interventional radiology, 41.8% (n = 46) by the hospital vascular access team, 1.8% (n = 2) by trained NICU staff, and 16.4% (n = 18) by an outside facility. Central line placement was required for several reasons and included: 6.4% (n = 7) placed for prolonged nothing by mouth (NPO) status, 5.5% (n = 6) for delivery of medications, 17.3% (n = 19) for delivery of high osmolarity fluids, 59.1% (n = 65) were cardiac patients, 6.4% (n = 7) for difficult/failed IV access, and 5.5% (n = 6) for frequent assessment of laboratory values (see Table 3).
During this study, the intervention group was for all ntCLs. Meaning, the new CL maintenance bundle was only for the ntCL and all tCLs were maintained using previous methods (using the BioPatch and scheduled dressing changes). Upon examination of data from all CLs, tunneled and nontunneled, there was no statistical significance in the total number of nonscheduled dressing changes between the pre- and postimplementation of the CL bundle, with the preintervention mean 2.29 (standard deviation = 2.32) and postintervention mean 1.96 (standard deviation = 2.34).
When evaluating the frequency of CL dislodgements, in the preintervention group there were a total of 19 occurrences (14 from nontunneled catheters) and in the postintervention group there was only 1 (from a nontunneled catheter). A Levene's Test for Equality of Variances showed a significance of .00 and a t test showed t =3.07 and α = .003, a 95% decrease in dislodgements.
Infants weighing between 3000 and 3999 g had the most dislodgements (α = .45). In the postbundle implementation group, a descriptive analysis was completed to see whether there was any difference in weight groups when the dressing was changed to an only as needed basis. Although this was not an initial question in the project, incidental findings showed that the majority of dressing changes (27%) occurred in infants who weighed between 2000 and 2999 g. An analysis of variance analysis did not show a statistical significance in weight groupings, with α = .05, the significance between groups was .06.
Infection occurrences in the prebundle implementation group were a total of 5 and in the postbundle implementation group were a total of 4; a decrease from 1.17 per 1000 CL days to 1.12 per 1000 CL days. In all 4 cases in the postbundle implantation group, infection occurred in those with tunneled catheters, α = .01. A 2-tailed t test showed that all infections were in catheters placed by the hospitalwide vascular access team, α = .38.
Minimal data exist to perform a true correlation with patient care days due to infants being removed from the study if they were transferred to another unit, for example a transfer to the cardiovascular intensive care unit.
For this evidence-based quality improvement project, we found no statistical difference in the incidence of infection or line dislodgement associated with location of insertion site or provider placing the CL. With implementation of the new CL, maintenance bundle dislodgement rates reduced by 95% and the only occurrence of infection was when the 2-person sterile technique was not followed. Prior to implementation, there was a concern that although the CL bundle was changed to only as needed dressing changes, the larger infants who are more active would still require frequent dressing changes. Although not statistically significant, most dressing changes, when the dressing was compromised, were during the care of smaller infants (<2 kg), not supporting the concern of providers.
Patients in the NICU often bring unique challenges for providers. Due to the constantly changing and developing body systems, fluctuating glomerular filtration rates, and delicate balance of oxygen delivery to critical organs, treatments can affect each infant differently.9,10,12 In a recent review involving 413 different studies, infants with increased incidences of central line-associated bloodstream infections (CLABSIs) had a 6 times greater likelihood of infection if there was an intra-abdominal pathology, or surgery, or had greater than 3 heel stick blood samples prior to the infection.16 A similar pattern was also seen in our patients. In evaluating the infants who did have CLABSIs both before and after the intervention, all had an abdominal pathology or cardiac diagnosis where oxygen delivery to the gut could have been compromised.
This project was completed over a 10-month period. Although the results were significant in the number of dislodgements and patient care days, a longer follow-up is needed to determine whether the results are sustainable. Although the evidence supports this CL maintenance bundle with both ntCL and tCL, this bundle was implemented with only nontunneled catheters to evaluate results from a control group (tCL) and the intervention group (ntCL). However, the CL bundle could to be transferrable to tCL and the results observed carefully for similar or different results.
Recommendations for Clinical Development
The NICU infant at a tertiary care center requires a unique approach to CL maintenance.17 The complex physiological nature of their condition often demands the need for a CL. Once these lines are placed, it has been found that it is best to leave the dressing intact to protect the line and only change the dressing if absolutely necessary. The risk for line compromise and infection far outweighs the benefit of frequent, scheduled dressing changes.
With recent discoveries comparing infections with infant pathologies, it is beneficial to not only consider CL daily practices and maintenance but also careful consideration in care management when there could be an abdominal pathology or oxygen compromise in addition to needing a CL.14
Many of the NICU infants with complex cardiac conditions require surgery and are then transferred to the cardiovascular intensive care unit (CVICU) for postoperative management. These infants still have the same unique needs, and whether or not this new CL bundle could be utilized in CVICU is yet to be determined.
Evidence shows that the use of a BioPatch can help in reducing infection; however, for the NICU infant there is a significant risk of CL compromise during the dressing changes and their immature skin predisposes them to higher incidences of skin breakdown from frequent removal of the dressing.1 According to the CDC and Healthcare Safety Network, current recommendations are to discontinue using the BioPatch and change CL dressings as needed.1,16 Units that have adopted this practice have seen fewer incidences of CLABSI and a decrease in dislodgement rates.1,18
1. Marschall J, Mermel LA, Fakih M, et al Strategies to prevent central line
–associated bloodstream infections in acute care hospitals: 2014 Update. Infect Control Hosp Epidemiol. 2014;35(7):753–771. doi:10.1086/676533.
2. Ullman AJ, Kleidon T, Gibson V, et al Central venous access device securement and dressing effectiveness (CASCADE) in paediatrics: protocol for pilot randomized controlled trials. BMJ Open. 2016;6(1):e011197. doi:10.1136/bmjopen-2016-011197.
3. Greenberg RG, Cochran MA, Smith PB, et al Effect of catheter dwell time on risk of central line-associated bloodstream infection
in infants. Pediatrics. 2015;136(6):1080–1086. doi:10.1542/peds.2015-0573.
4. Lai NM, Taylor JE, Tan K, et al Antimicrobial dressings
for the prevention of catheter-related infections in newborn infants with central venous catheters. Cochrane Database Syst Rev. 2016;3:DC011082. doi:10.1002/14651858.CD011082.pub2.
5. Centers for Disease Control and Prevention. Healthy People 2020. Rockville, MD: Office of Disease Prevention and Health Promotion; 2017.
6. Ullman AJ, Cooke ML, Mitchell M, et al Dressings
and securement devices for central venous catheters. Cochrane Database Syst Rev. 2015;9:CD010367. doi:10.1002/14651858.CD010367.pub2.
7. Centers for Disease Control and Prevention.Central line
-Associated Bloodstream Infections (CLABSI): Checklist for Prevention of Central Line
Associated Blood Stream Infections. Atlanta, GA: Centers for Disease Control and Prevention; 2011.
8. Gavin NC, Webster J, Chan RJ, Rickard CM. Frequency of dressing changes for central venous access devices on catheter-related infections. Cochrane Database Syst Rev. 2016;2:CD009213. doi:10.1002/14651858.CD009213.pub2.
9. Marsh N, Webster J, Mihala G, Rickard CM. Devices and dressings
to secure peripheral venous catheters to prevent complications. Cochrane Database Syst Rev. 2015;6:CD011070. doi:10.1002/14651858.CD011070.pub2.
10. Kime T, Mohsini K, Nwankwo MU, Turner B. Central line
“attention” is their best prevention. Adv Neonatal Care. 2011;11(4):242–248. doi:10.1097/ANC.0b013e3182256680.
11. Sharpe EL. Tiny patients, tiny dressings
: a guide to the neonatal PICC dressing change. Adv Neonatal Care. 2008;8(3):150–162.
12. Milstone AM, Reich NG, Advani S, et al Catheter dwell time and CLABSIs in neonates with PICCs: a multicenter cohort study. Pediatrics. 2013;132(6):e1609–e1615. doi:10/1542/peds.2013-1645.
13. McMullin R, Gordon A. Impact of a central line
infection prevention bundle in newborn infants. Infect Control Hosp Epidemiol. 2016;37(9):1029–1036. doi:10.1017/ice.2016.127.
14. The Joint Commission. Preventing Central Line
-Associated Bloodstream Infections: Useful Tools. Oakbrook Terrace, IL: The Joint Commission;2013.
15. Neill S, Haithcock S, Smith PB, et al Sustained reduction in bloodstream infections in infants at a large tertiary care neonatal intensive care unit
. Adv Neonatal Care. 2016;16(1):52–59. doi:10.1097/anc.0000000000000164.
16. Dahan M, O'Donnell S, Hebert J, et al CLABSI risk factors in the NICU: potential for prevention: a PICNIC study. Infect Control Hosp Epidemiol. 2016;37(12):1446–1452. doi:10.1017/ice.2016.203.
17. Takashima M, Ray-Barruel G, Ullman A, Keogh S, Rickard CM. Randomized controlled trials in central vascular access devices: a scoping review. PLoS One. 2017;12(3):e0174164.
18. Melnyk BM, Fineout-Overholt E. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2014.