Needleless connector decontamination for prevention of central venous access device infection: A pilot randomized controlled trial

Abstract

•Central venous access devices need physical and chemical ‘scrub the hub’ before use.•70% isopropyl alcohol wipes had 2% Central Line Associated Bloodstream Infection.•70% isopropyl alcohol caps had 2% Central Line Associated Bloodstream Infection.•Combined 70% isopropyl alcohol and 2% chlorhexidine gluconate wipe had no infection.•Large randomised controlled trials of cleaning methods are feasible and needed. Pilot randomized controlled trial (180 patients) of needleless connector decontamination. Central line-associated bloodstream infection occurred in 2% (1/61) of 70% isopropyl alcohol (IPA) wipe, 2% (1/59) of 70% IPA cap, and zero (0/58) infections in 2% chlorhexidine gluconate in 70% IPA wipe patients. Larger definitive trials are feasible and needed. Pilot randomized controlled trial (180 patients) of needleless connector decontamination. Central line-associated bloodstream infection occurred in 2% (1/61) of 70% isopropyl alcohol (IPA) wipe, 2% (1/59) of 70% IPA cap, and zero (0/58) infections in 2% chlorhexidine gluconate in 70% IPA wipe patients. Larger definitive trials are feasible and needed. Central venous access devices (CVADs) risk central line-associated bloodstream infection (CLABSI) which increase costs, morbidity and mortality.1O'Grady NP Alexander M Burns LA et al.Guidelines for the prevention of intravascular catheter-related infections.Clin Infect Dis. 2011; 52: e162-e193Crossref PubMed Scopus (1246) Google Scholar The intraluminal infection source can be minimized by needleless connector (NC) decontamination prior to each use using chlorhexidine gluconate (CHG), povidone-iodine, or 70% isopropyl alcohol (IPA).1O'Grady NP Alexander M Burns LA et al.Guidelines for the prevention of intravascular catheter-related infections.Clin Infect Dis. 2011; 52: e162-e193Crossref PubMed Scopus (1246) Google Scholar The optimal antiseptic is unknown, although povidone-iodine's slow dry-time presents challenges in clinical practice.2Slater K Fullerton F Cooke M Snell S Rickard CM Needleless connector drying time-how long does it take?.Am J Infect Control. 2018; 46: 1080-1081Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar Combination CHG/IPA wipes,3Flynn JM Rickard CM Keogh S Zhang L Alcohol caps or alcohol swabs with and without chlorhexidine: An in vitro study of 648 episodes of intravenous device needleless connector decontamination.Infect Control Hosp Epidemiol. 2017; 38: 1-3Crossref Scopus (7) Google Scholar,4Slater K, Cooke M, Fullerton F, et al. Peripheral intravenous catheter needleless connector decontamination study-Randomized controlled trial. Am J Infect Control. 2020;48:1013-1018Google Scholar or IPA in a cap format5Wright M-O Tropp J Schora DM et al.Continuous passive disinfection of catheter hubs prevents contamination and bloodstream infection.Am J Infect Control. 2013; 41: 33-38Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar,6Casey AL Karpanen TJ Nightingale P Elliott TSJ An in vitro comparison of standard cleaning to a continuous passive disinfection cap for the decontamination of needle-free connectors.Antimicrob Resist Infect Control. 2018; 7: 50Crossref PubMed Scopus (7) Google Scholar may be superior to traditional intermittent 70% IPA wipes, but no randomized controlled trials (RCTs) have been completed. Our aim was to generate feasibility and pilot data comparing 70% IPA wipes, 2% CHG in 70% IPA wipes, and 70% IPA caps. Three-arm pilot RCT at the Royal Brisbane and Women's Hospital and Gold Coast University Hospital in Australia. We had University and Hospital Ethics Committees approval (2016/410; HREC/15/QRBW/553) and Australian New Zealand Clinical Trials Registry registration: 12615001120561. The 4-week intervention had follow-up until 48 hours post study completion, hospital discharge or device removal. We surveyed registered nurses (RNs) for protocol compliance and satisfaction. Eligibility criteria: ≥18 years of age; CVAD (peripherally inserted central catheter or tunneled, cuffed CVAD) inserted <24 hours; CVAD required for ≥7 days; and written consent. Exclusions: baseline bloodstream infection, non-English speaking without interpreter, or previous enrolment. Research nurses (ReNs) screened daily, gave trial information, and obtained consent. The target was 60 per group (1 CVAD per patient) with recruitment July 31, 2017 to April 5, 2019.7Whitehead A Julious S Cooper C Campbell M Estimating the sample size for a pilot randomised trial to minimise the overall trial sample size for the external pilot and main trial for a continuous outcome variable.Stat Methods Med Res. 2016; 25: 1057-1073Crossref PubMed Scopus (434) Google Scholar Centralized, computer-generated randomization (https://randomisation.griffith.edu.au) using randomly varying permuted blocks of 3 and 6 (1:1:1 ratio): (1) 70% IPA wipes, (2) 2% CHG in 70% IPA wipes, or (3) 70% IPA caps. Clinical outcome assessors and data analysts were masked. •Seventy percent IPA wipes: 0.6 mL Alcohol Prep Pads (Reynard, New Zealand) applied vigorously to NC for 5 seconds (manufacturer recommended and hospital policy), visibly dry prior to CVAD access;•Two percent CHG in 70% IPA wipes: 0.6 mL Alcohol and CHG Prep Pads (Reynard, New Zealand), applied vigorously to NC for 15 seconds (guideline recommendation8Loveday HP Wilson JA Pratt RJ et al.epic3:national evidence-based guidelines for preventing healthcare-associated infections.J Hosp Infect. 2014; 86: S1-70Abstract Full Text Full Text PDF PubMed Scopus (637) Google Scholar), visibly dry prior to CVAD access;•Seventy percent IPA cap: Luer access valve cap Swabcap (intensive care unit [ICU] Medical, San Clemente) screwed onto NCs for minimum 5 minutes (manufacturer-recommended) prior to each access (70% IPA wipes were also used), then replaced with a new cap. NCs were Smartsite Needle-Free Valve or Max Plus (both Carefusion/BD, San Diego), attached to the CVAD hubs and all entry points of infusion systems. ReNs provided education (clinical staff undertook the intervention) and visited twice weekly to collect data, supply products, and reinforce the protocol. Decisions to culture blood/CVAD tips, or remove CVADs were made by medical staff (not investigators). Protocol feasibility was assessed as: (1) eligibility, (2) retention and attrition, (3) protocol adherence, (4) missing data, and (5) RN satisfaction. (i)CLABSI9NHSN. National healthcare safety network (NHSN) patient safety component manual. CDC, Ed. Atlanta, GA; 2018:1–38Google Scholar (2018 National Health and Safety Network definition) assessed by masked infectious diseases specialist;(ii)Mortality (all-cause) during trial;(iii)Primary bloodstream infection (laboratory confirmed bloodstream infection)9NHSN. National healthcare safety network (NHSN) patient safety component manual. CDC, Ed. Atlanta, GA; 2018:1–38Google Scholar;(iv)CVAD (tip) colonization (≥15 colony-forming units, semi-quantitative culture).1O'Grady NP Alexander M Burns LA et al.Guidelines for the prevention of intravascular catheter-related infections.Clin Infect Dis. 2011; 52: e162-e193Crossref PubMed Scopus (1246) Google Scholar We captured all potentially intervention-related events, and all-cause ICU admission (serious adverse event). Research Electronic Data Capture (REDCap, Nashville, TN) and Stata 15 (College Station, TX) were used. Feasibility outcomes were analyzed against predetermined criteria (>80% of screened patients eligible and >80% eligible patients recruited; ≥95% retention and attrition (not withdrawn/lost to follow-up); >90% study visits with correct products in use, and self-reported RN adherence to application/dry times; 5% missing data (CLABSI endpoint); RN satisfaction on 1-10 numerical rating scale. Clinical outcomes were compared using Fisher's exact and log-rank tests, incidence rates and Kaplan-Meier survival estimates (P < .05 statistically significant; patients censored at discharge). A modified intention-to-treat analysis excluded only randomized patients who never received a CVAD. Patient/device characteristics are presented in Table 1 and Supplementary Table 1. Average CVAD dwell-times were 11.3, 9.3, and 7.4 days in the 70% IPA, 2% CHG in 70% IPA, and 70% IPA cap groups, respectively.Table 1Participant (N = 180) and device (N = 178) characteristics at baseline70% IPA2% CHG in 70% IPA70% IPA capTotalParticipants per study groups*Row percentage shown.61 (34)59 (33)60 (33)180 (100)Age (years)†Median and interquartile range (25th and 75th percentiles) shown.61 (50-67)60 (47-67)63 (50-72)61 (50-70)Sex: male31 (51)28 (47)37 (62)96 (53)Cancer treatment‡In previous 6 months.19 (31)18 (31)17 (28)54 (30)Admission type - surgical47 (77)46 (78)49 (82)142 (79) - haematology12 (20)10 (17)10 (17)32 (18) - medical1 (2)3 (5)1 (2)5 (3) - medical oncology1 (2)0 (0)0 (0)1 (1)Comorbidities - nil or one17 (28)17 (29)16 (27)50 (28) - two or three20 (33)16 (27)20 (33)56 (31) - four or more24 (39)26 (44)24 (40)74 (41)Leucocytes§Absolute, within 72 hours of trial entry. <500/µl (n=179)5 (8)5 (9)5 (8)15 (8)Pre-existing infection27 (44)32 (54)34 (57)93 (52)Devices by study groups*Row percentage shown.61 (34)58 (33)59 (33)178 (100)Device type - PICC57 (93)54 (93)56 (95)167 (94) - TC4 (7)4 (7)3 (5)11 (6)No. of lumens - one16 (26)21 (36)20 (34)57 (32) - two45 (74)37 (64)39 (66)121 (68)Location - upper arm57 (93)54 (93)56 (95)167 (94) - chest4 (7)4 (7)3 (5)11 (6)IV medications - antibiotics43 (70)39 (67)42 (71)124 (75) - fluids24 (39)25 (43)21 (36)70 (39) - blood product9 (15)13 (22)5 (8)27 (15) - antiemetic9 (15)7 (12)9 (15)25 (14) - parenteral nutrition12 (20)6 (10)6 (10)24 (13) - potassium chloride6 (10)6 (10)4 (7)16 (9) - chemotherapy4 (7)5 (9)5 (8)14 (8) - antifungal/antiviral4 (7)1 (2)2 (3)7 (4) - other medication29 (48)25 (43)17 (29)71 (40)No medications (fluids only)5 (8)6 (10)7 (12)18 (10)Frequencies and column percentages shown unless otherwise noted. Row percentage shown.† Median and interquartile range (25th and 75th percentiles) shown.‡ In previous 6 months.§ Absolute, within 72 hours of trial entry. Open table in a new tab Frequencies and column percentages shown unless otherwise noted. Seventy percent (211/303) of screened patients were eligible and 85% (180/211) were randomized (31 declined, missed, or had CHG allergy; Fig 1). Two patients were excluded postrandomization due to CVAD insertion failure. There was 100% retention, 0% attrition, and 0% missing CLABSI endpoints (Fig 1). Thus, 178 patients were analyzed. Observed protocol adherence was 98% (174/178); all but three 2% CHG in 70% IPA wipe and two 70% IPA cap patients commenced the correct intervention. 70% IPA wipe patients had no protocol deviations. At least one incorrect product use occurred in 5% (3/58) 2% CHG in 70% IPA, and 10% (6/59) 70% IPA cap patients. Of 35 RNs (40 surveyed, response rate 88%), protocol-adherent scrub times were reported by 31 (89%) for 70% IPA wipe, and 26 (74%) for 2% CHG in 70% IPA wipe. Median satisfaction was 9 (interquartile range: 2), 10 (2), and 9 (2) for 70% IPA wipes, 2% CHG in 70% IPA wipes, and 70% IPA caps, respectively (N = 22 for 70% IPA caps; not all RNs had used these). CLABSI occurred in 1/61 (2%) 70% IPA wipe, 0/58 (0%) 2% CHG in 70% IPA wipe, and 1/59 (2%) 70% IPA cap patients (P = 1.0, Fig 2). CLABSI incidence per 1,000 catheter-days was 1.38 (95% confidence interval [CI]: 0.19-9.81), nil (no outcomes), and 1.70 (95% CI: 0.24-12.1) for 70% IPA wipes, 2% CHG in 70% IPA wipes, and 70% IPA caps, respectively (P = .637). Primary bloodstream infections occurred in 2/61 (3%) 70% IPA wipe, 2/58 (3%) CHG in 70% IPA wipe (1 of these was a mucosal barrier infection), and 1/59 (2%) 70% IPA cap patients. There were no deaths and no positive catheter tips (N = 10 cultured). Two 70% IPA cap NCs became opaque (IPA appeared to seep between the rubber inner and outer plastic, denaturing the plastic but with no effect on patients). Four patients required transfer to ICU for unrelated reasons (n = 3, 70% IPA wipe; n = 1, 70% IPA cap). NC decontamination is a high-volume, high-value practice that urgently needs high-quality evidence to prevent CLABSI. This pilot RCT confirms the feasibility of large RCTs, with acceptable recruitment, protocol adherence, and RN satisfaction, as well as high retention, low attrition and no missing data. Eligibility at 70% could be improved with amplified research nurse availability at device insertion to promote recruitment. CLABSI incidence was low in both groups using 70% IPA, and 0 when this antiseptic was combined with CHG. These results are consistent with laboratory data,3Flynn JM Rickard CM Keogh S Zhang L Alcohol caps or alcohol swabs with and without chlorhexidine: An in vitro study of 648 episodes of intravenous device needleless connector decontamination.Infect Control Hosp Epidemiol. 2017; 38: 1-3Crossref Scopus (7) Google Scholar and a large RCT on pre-CVAD insertion skin decontamination which both favored combination CHG and IPA10Mimoz O Lucet JC Kerforne T et al.Skin antisepsis with CHG-alcohol vs povidone iodine-alcohol, with and without skin scrubbing, for prevention of intravascular-catheter-related infection (CLEAN): an open-label, multicentre, two-by-two factorial RCT.Lancet. 2015; 386: 2069-2077Abstract Full Text Full Text PDF PubMed Scopus (172) Google Scholar; a larger RCT would be needed to substantiate these findings in NCs. Although scrub times differed (15 seconds for 2% CHG in 70% IPA wipe as per guidelines,8Loveday HP Wilson JA Pratt RJ et al.epic3:national evidence-based guidelines for preventing healthcare-associated infections.J Hosp Infect. 2014; 86: S1-70Abstract Full Text Full Text PDF PubMed Scopus (637) Google Scholar and 5 seconds for 70% IPA wipes as per manufacturers and hospital policy), recent data indicates no difference in effectiveness with 5, 10, or 15 second scrub times.4Slater K, Cooke M, Fullerton F, et al. Peripheral intravenous catheter needleless connector decontamination study-Randomized controlled trial. Am J Infect Control. 2020;48:1013-1018Google Scholar CLABSI was infrequent, however as >50% were patients were discharged during follow-up, future RCTs should study the entire CVAD dwell (including home care) to ensure adequate sample size to test hypotheses and generalizability. Nevertheless, our CLABSI of approximately 1 per 1,000 catheter-days, is similar to reported USA rates, but may not be generalizable where rates are higher.11Rosenthal VD Maki DG Mehta Y et al.International Nosocomial Infection Control Consortium (INICC) report, data summary of 43 countries for 2007–2012. Device-associated module.Am J Infect Control. 2014; 42: 942-956Abstract Full Text Full Text PDF PubMed Scopus (182) Google Scholar Despite low frequency, CLABSI remains the most appropriate outcome to assess NC disinfection efficacy. Other methods such as routine CVAD tip culture have poor positive predictive value.12Peterson LR. Smith BA. Nonutility of catheter tip cultures for the diagnosis of central line–associated bloodstream infection.Clin Infect Dis. 2015; 60: 492-493Crossref PubMed Scopus (8) Google Scholar Insertion bundles have reduced CLABSI, with focus now needed on techniques to prevent postinsertion, intraluminal bacterial entry. Currently, 70% IPA wipes are dominant due to low cost, availability and rapid drying2Slater K Fullerton F Cooke M Snell S Rickard CM Needleless connector drying time-how long does it take?.Am J Infect Control. 2018; 46: 1080-1081Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar however the addition of CHG likely increases efficacy,3Flynn JM Rickard CM Keogh S Zhang L Alcohol caps or alcohol swabs with and without chlorhexidine: An in vitro study of 648 episodes of intravenous device needleless connector decontamination.Infect Control Hosp Epidemiol. 2017; 38: 1-3Crossref Scopus (7) Google Scholar,4Slater K, Cooke M, Fullerton F, et al. Peripheral intravenous catheter needleless connector decontamination study-Randomized controlled trial. Am J Infect Control. 2020;48:1013-1018Google Scholar and nonrandomized studies support 70% IPA caps.5Wright M-O Tropp J Schora DM et al.Continuous passive disinfection of catheter hubs prevents contamination and bloodstream infection.Am J Infect Control. 2013; 41: 33-38Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar,6Casey AL Karpanen TJ Nightingale P Elliott TSJ An in vitro comparison of standard cleaning to a continuous passive disinfection cap for the decontamination of needle-free connectors.Antimicrob Resist Infect Control. 2018; 7: 50Crossref PubMed Scopus (7) Google Scholar Pilot RCTs are not designed to test statistical differences in outcomes or for the effect of potential confounders or covariates such as NC/device type or patient factors. Large RCTs are needed to examine various modes and strengths of antiseptics, NC materials/designs, and monitor possible new adverse events as solutions are exposed to NCs and potentially the bloodstream. We thank Marie Cooke, Peter Mollee, Paul Scuffham, and Joan Webster for assistance in obtaining funding and Aidan Menzies for formatting assistance. We thank Christine Woods, Alyson Eastgate, Elise Sturgeon and Melissa Williams for assistance with patient recruitment and data collection. We thank the patients, relatives and staff of the participating hospitals. Download .docx (.0 MB) Help with docx files

Topics & Concepts