a b s t r a c t

Background: In some patients securing the peripheral intravenous cannula (PIVC) with a standard adhesive dressing can be difficult because of sweat or other body fluids. The aim of our study was to evaluate the use of Tissue adhesives alone as a means to secure PIVCs inserted in the emergency department.

Methods: We performed a prospective interventional pilot study from November 2019 to May 2020 in a medical emergency department of an urban tertiary hospital. Patients were randomized to two groups: tissue adhesives (TA) or adhesive dressing (AD) group. After randomization we followed them until day 4.

Results: There were no significant differences between TA and AD groups in the rate of unplanned removal of PIVCs in the first 72 h (57.1% vs. 45.8%, p = 0.29), the rate of unplanned removal of PIVCs in the ED (0% vs. 2.1%, p = 1.00), the rate of unplanned removal of PIVC in the first 24 h (42.8% vs. 35.4%, p = 0.52), as well as in the rate of phlebitis (7.1% vs. 14.6%, p = 0.34) and the rate of any blood-stream infection (0% vs. 0%, p = 1.00). Conclusion: We did not observe any significant differences when PIVCs inserted in the emergency department were secured with tissue adhesives alone, compared to standard adhesive dressings. We observed a high rate of unplanned removal of PIVCs, necessitating further research to determine more reliable ways of securing PIVCs.

(C) 2021

1. Introduction

Insertion of a peripheral intravenous catheter (PIVC) is a basic med- ical procedure in the emergency department (ED). The standard way of securing a PIVC is with adhesive dressings (AD) [1]. Before applying the dressing the skin around the puncture site needs to be dry and clean [1]. In some patients this can be difficult because the skin can be clammy and securement of PIVC can be challenging [2], which can contribute to unplanned removal of PIVC (affecting up to around 70% of PIVCs), dis- location of a PIVC and extravascular infusion (around 3% each) and phlebitis and blood-stream infections (occuring in at least around 10% and 3% of PIVCs, respectively) [3,4].

Abbreviations: AD, adhesive dressing; ICU, intensive care unit; PIVC, peripheral intravenous cannula; TA, tissue adhesive.

* Corresponding author at: Medical Intensive Care Unit, Medical Emergency Department, University Medical Centre Maribor, Ljubljanska 5, Maribor 2000, Slovenia.

E-mail address: [email protected] (A. Markota).

Using tissue adhesives (TA) for sutureless securing of central venous catheters is associated with decreased incidence of catheter-related Bloodstream infections and with unchanged incidence of unplaned re- moval of central venous catheters, even in patients with delirium [5]. Similarly, use of TA for radial arterial catheter securement has also been associated with a decreased incidence of blood-stream infections and similar rates of unplanned removal [6].

The aim of our study was to evaluate the use of TA alone as means to secure PIVCs inserted in the ED.

1. Methods
   1. Study design and setting

We performed a prospective interventional pilot study from Novem- ber 2019 to May 2020 in a medical ED of an urban tertiary hospital. In- stitutional ethics committee approval was obtained (No. UKC-MB-KME- 76/19) and informed consent was required.

https://doi.org/10.1016/j.ajem.2021.01.088

0735-6757/(C) 2021

• 1. Outcomes

Primary endopoint was the rate of unplanned removal of PIVC in the first 4 days or until hospital discharge or death, and secondary end- points were rate of unplanned removal of PIVC in the ED, rate of un- planned removal of PIVC in the first 24 h, rate of extravascular infusion, PIVC dwell-in time, rate of phlebitis at ED PIVC insertion site in the first 7 days or until discharge or death, rate of any blood-stream infection diagnosed during index hospital stay and rate of any positive blood-cultures in patients who developed phlebitis during index hospi- tal stay. Unplanned removal included any reason for removal before day 4, when removal of a PIVC is required as per institutional guidelines. The definition included removal of the PIVC by the patient themselves (“the PIVC fell out”) or removal as indicated by the clinical team, such as re- moval because of blockage, signs of phlebitis, pain, leaking or other. We hypothesized that there would be no differences between TA and AD groups in primary or secondary endpoints.

• 1. Study population

Inclusion criteria to the study were age 18 years or older, no intrave- nous access prior to arrival to the ED, intravenous access required dur- ing the diagnostic or therapeutic workup in the ED and high probability for hospital admission on presentation to the ED. Exclusion criteria were immediate life-saving procedures required upon arrival to ED precluding inclusion to the study and 3 or more unsuccessful at- tempts at obtaining a peripheral intravenous access.

• 1. Study intervention

After informed consent was obtained the patient was randomized to intervention (TA) or control (AD) group. Randomization was performed according to sequentially numbered, opaque sealed envelope method based on a randomization list generated by Microsoft Excel, Microsoft, Redmond, Washington, USA. After randomization a standard 16 or 18 gauge PIVC was inserted (BD Venflon Pro, BD, Franklin Lakes, New Jer- sey, USA), preferentially to cubital area. PIVCs in TA group were secured with SecurePortIV cyanoacrilate adhesive, Adhezion Biomedical, Wyomissing, Pennsylvania, USA, and PIVCs in AD group were secured using Cosmopor I.V. adhesive dressing, Hartmann Group, Heidenheim, Germany. PIVCs in the TA group were secured without placing addi- tional dressing over the PIVC, by applying the TA below wings of PIVC and on and around the puncture site. Puncture site was completely sealed with the TA. All staff involved in patient care received training in the use of tissue adhesives and adhesive dressings used in our study prior to the study. Research personnel were not blinded to intervention group as the appearance of tissue adhesives and adhesive dressings is clearly different. As per institutional policy PIVCs can remain in-situ maximally until day 4, after which removal is necessary.

• 1. Measurements

PIVC position was checked (by visual inspection and reflux of blood) and noted in the ED at 30 and 90 min after insertion, and twice daily (between 7 and 8 AM, and 7 and 8 PM) on hospital days 1, 2 and 3 after admission. On visits on days 1, 2 and 3 signs of phlebitis were checked according to Jackson Visual Inspection Phlebitis Scale. We noted the day of PIVC removal, the reason for PIVC removal and the need for microbiological sampling as per the treating physician. Basic demographical data was collected, as well as data on comorbidities and microbiological testing data.

• 1. Data analysis

Statistical planning and evaluation were performed by the IBM SPSS Statistics version 23 (Armonk, New York, USA). Independent samples

t-test was used for analysis of continuous data over two groups after Kolmogorov-Smirnov test of normality and Fisher’s exact test was used to analyze nominal categorical data. A p value of <0.05 was consid- ered statistically significant. Power analysis was performed using GPower 3.1.9.2 assuming power 80%, alpha level of 0.05, allocation ratio of >0.80% and combined geometrical mean proportions of un- planned removal reported elsewhere [7,8].

1. Results
   1. Baseline characteristics

Baseline characteristics of the patients are presented in Table 1. In all, 843 patients were screened for inclusion, and 100 patients consented and were enrolled into the study. 30 patients were not enrolled because immediate lifesaving procedures were necessary, 383 patients were not enrolled because of a preexistant PIVC and 330 patients were not en- rolled because of higher likelihood of Discharge home on presentation. Ten patients were lost to follow-up because of admission to dislocated departments, leaving 90 patients for final analysis, 42 in the TA group and 48 in the AD group. There were no significant differences between TA and AD groups in age, gender, need for hospital admission, basic de- mographic characteristics and Manchester Triage System (MTS) group (Table 1). Power analysis using proportions of PIVC failure (pain, block- age, leaking, accidental removal, catheter related infection) reported by Marsh et al. [7] and using proportions of PIVC failure (composite of one or more of infection, phlebitis, occlusion, or dislodgement) reported by Bugden et al. [8] has shown that to achieve 80% of statistical power, 77 patients per group have to be enrolled in the study.

• 1. Main results

There were no significant differences between TA and AD groups in the rate of unplanned removal of PIVCs in the first 72 h (57.1% vs. 45.8%, p = 0.29). Also, we observed no significant differences between TA and AD groups in the rate of unplanned removal of PIVC in the ED (0% vs. 2.1%, p = 1.00), the rate of unplanned removal of PIVC in the first 24 h (42.8% vs. 35.4%, p = 0.52), the rate of extravascular infusion (7.1% vs. 4.2%, p = 0.66), PIVC dwell-in time (1.7 +- 0.7 days vs. 1.8 +- 0.8 days, p = 0.39), the rate of phlebitis at ED PIVC insertion site in the first 7 days (7.1% vs. 14.6%, p = 0.34), the rate of any blood-stream infection diagnosed during index hospital stay (0% vs. 0%, p = 1.00) and the rate of any positive blood-culture in patients who developed phlebitis dur- ing index hospital stay (2.4% vs. 0%, p = 0.47). Blood cultures were pos- itive in one patient in the TA group. In that patient one bottle (of four) tested positive for S. epidermidis, and no diagnosis of blood-stream in- fection was made.

Seven patients in the TA group and 7 patients in the AD group reached hospital day 4 with index PIVC intact and functioning (16.7% vs. 14.6%, p = 1.00). One patient in the TA group and two patients in the AD group died with an intact and functioning index PIVC before hos- pital day 4 (2.4% vs. 4.2%, p = 1.00). Also, one patient in the TA group and 4 patients in the AD group were discharged home before hospital day 4 with an intact and functioning PIVC (2.4% vs. 8.3%, p = 0.37). We observed no skin complications, such as burns of eczema. Analysis of inter-operator variability was not possible because there were too many operators.

1. Discussion

In our study we compared TA to standard AD as a means to secure PIVCs in an ED setting. We did not observe any significant differences between PIVCs secured by TA compared to standard AD. We observed a high rate of unplanned removal of PIVCs inserted in the ED in both in- terventional and controlled group of patients.

Table 1

Basic demographic characteristics. List of abbreviations: CCU = coronary care unit, ICU = intensive care unit, MTS = Manchester Triage System, SD = standard deviation.

To our knowledge, there are only two additional prospective ran- domized trials comparing TA to standard AD for securement of PIVCs [8,9]. In both trials the study intervention comprised using TA in addi- tion to AD, compared to AD alone. Bugden at al performed a prospective randomized study in 369 patients, comparing cyanoacrylate based tis- sue glue and standard AD with AD alone. They observed a 10% decrease in rate of PIVC failure in intervention arm (17% vs. 27%), however, Failure rates were calculated at 48 h after insertion, when 57% of patients were already discharged home, and no data is presented on patient mortality or admission to intensive care units. Exclusion of Agitated patients from study also contributed to lower PIVC failure rate [8]. Ozkula et al. simi- larly compared cyanoacrylate based tissue glue and transparent poly- urethane AD with transparent polyurethane AD alone in 115 patients, and reported a 10% PIVC failure rate in the intervention arm. Patients with high likelihood for PIVC failure (agitated patients, history of phle- bitis or venous thrombosis, BMI > 40) were excluded from study. Final study visit was performed at 24 h after insertion. They did not re- port on patient mortality or admission rate to intensive care unit [9].

We observed much higher rates of PIVC removal compared to studies above [8,9]. However, the rates we observed are roughly comparable to failureratesof around 30% in previouslypublished data [7,10], andsubse- quent informal follow-up in our hospital revealed similar rates. Higher rate of unplanned PIVC removal can be explained by our study popula- tion. We enrolled a group of patients with significant morbidity on pre- sentation (around 50% red/orange group in Manchester Triage System), need for admission to intensive care or coronary care unit (around 35%) and hospital mortality (around 25%). Around 40% of enrolled patients re- quired intravenous opiates, intravenous sedatives or antipsychotics via any route during the first 24 h after admission to control pain, dyspnea or agitation. We enrolled patients with body mass index > 40 kg/m², however, they represented only a minority of patients (one patient in TA group, and 3 patients in AD group). Also, we used a definition of un- planned removal that encompassed all reasons for removal of a PIVC be- fore day 4, when removal is mandated as per local policy, which could have led to a higher rate of PIVC removal compared to other studies. Ad- ditionally, final study visits were performed on day 4 in our study, com- pared to day 2 [8] and at 24 h [9] in the studies above, with longer study period possibly contributing to higher rate of unplanned PIVC removal.

A number of studies were performed aiming to reduce the rate of

phlebitis and blood-stream infections associated with PIVCs, however, there is little data on techniques that could be used to deal with other causes of PIVC failure [11,12]. Takahashi et al. reported 18% reduction in PIVC failure (29% vs. 11%) after implementation of a bundle focusing on preventing mechanical irritation caused by PIVC [12]. Focus on preventing complications related to infection can be explained by nu- merous other factors that can contribute to PIVC failure, but cannot eas- ily be modified by any therapeutic interventions, such as concomitant diseases, site selection, choice of cannula, catheterisation technique, se- curement and intravenous therapy.

In contrast with other studies we decided to use TA alone, without additional AD in the intervention arm. Care was taken on insertion to firmly attach the PIVC to the skin by applying TA under PIVC wings, and then to apply TA over the puncture site, forming a protective film preventing any direct contact between outside and puncture site. The decision to use TA alone was mainly taken because the use of AD in our group of patients (critically ill adults) was expected to be compli- cated by clammy skin and presence of blood or other excretions in the ED setting that make the dressing non-adhesive. Also, changes of AD once it is applied to PIVC are difficult because of possibility of uninten- tional removal of PIVC. Inspite of widespread use of ADs to secure PIVCs there is little data to show superiority of ADs over more tradi- tional sterile gauze and adhesive tape [1]. Additionally, there are some reports on local antimicrobial activity of polyphenols in study TA [13]. We observed similar rates of phlebitis between intervention and control groups, no cases of any blood-stream infections, and only one positive blood-culture in both groups combined (S. epidermidis in only one bottle of four, no diagnosis of blood-stream infection in that patient).

• 1. Limitations

Our study was small (N = 100) and not blinded because of obvious differences between TA and AD groups. We are fully aware that sample size presents the main limitation of our study as final cohort did not meet the criteria calculated in power analysis. However, sample size in our study is comparable to the study performed by Edwards et al. with average of 49.5 individuals per group [6]. Moreover, sample size in the present study is also 2-fold higher than in the study performed by Marsh et al., where sample size per group was on average 21 [7]. We performed a single-center data collection, and data analysis did not include patients who were transfered to dislocated departments. Incidentaly, we lost 10 patients (10%) to follow-up, which further com- promised study population. We did not collect data on concomittant medications (e.g. anticoagulant or antithrombotic therapy), or intrave- nous therapy that was used in the ED or after admission (e.g. amioda- rone, blood transfusions, potassium, etc). Study visits were performed by study personnel, however, study PIVCs were used for normal clinical work. Finally, the single-center design of the study and the high rate of PIVC removal, which may be center-specific, prevent generalizability of our results.

1. Conclusion

Our study confirmed high rate of unplanned removal of PIVCs, even in a setting of a prospective trial. We did not observe any significant dif- ferences when PIVCs in the ED are secured with TA alone, compared to standard ADs. We advise against securing PIVCs with TA alone, how- ever, further research is required to determine more reliable ways of se- curing PIVCs.

Declaration of Competing Interest

All authors declare that there are no conflict of interests relating to the study entitled “Tissue Adhesives for Peripheral Intravenous Catheter Securement: a Propective Randomized Controlled Pilot Trial.”

References

1. Marsh N, Webster J, Mihala G, Rickard CM. Devices and dressings to secure periph- eral venous catheters to prevent complications. Cochrane Database Syst Rev. 2015; 12(6):CD011070. https://doi.org/10.1002/14651858.CD011070.pub2.
2. Fakih MG, Jones K, Rey JE, Takla R, Szpunar S, Brown K, et al. Peripheral venous cath- eter care in the emergency department: education and feedback lead to marked im- provements. Am J Infect Control. 2013;41(6):531-6. https://doi.org/10.1016/j.ajic. 2012.07.010.
3. Helm RE, Klausner JD, Klemperer JD, Flint LM, Huang E. Accepted but unacceptable: peripheral IV catheter failure. J Infus Nurs. 2015;38:189-203. https://doi.org/10. 1097/NAN.0000000000000100.
4. Pujol M, Hornero A, Saballs M, Argerich MJ, Verdaguer R, Cisnal M, et al. Clinical ep- idemiology and outcomes of peripheral venous catheter-related bloodstream infec- tions at a university-affiliated hospital. J Hosp Infect. 2007;67:22-9.
5. Ullman AJ, Cooke ML, Mitchell M, Lin F, New K, Long DA, et al. Dressings and secure- ment devices for central venous catheters (CVC). Cochrane Database Syst Rev. 2015; 10(9):CD010367. https://doi.org/10.1002/14651858.CD010367.pub2.
6. Edwards M, Rickard CM, Rapchuk I, Corley A, Marsh N, Spooner AJ, et al. A pilot trial of bordered polyurethane dressings, tissue adhesive and sutureless devices

compared with standard polyurethane dressings for securing short-term arterial catheters. Crit Care Resusc. 2014;16(3):175-83.

1. Marsh N, Webster J, Flynn J, Mihala G, Hewer B, Fraser J, et al. Securement methods for peripheral venous catheters to prevent failure: a randomised controlled pilot trial. J Vasc Access. 2015;16(3):237-44. https://doi.org/10.5301/jva.5000348.
2. Bugden S, Shean K, Scott M, et al. Skin glue reduces the failure rate of emergency de- partment-inserted peripheral intravenous catheters: a randomized controlled trial. Ann Emerg Med. 2016;68(2):196-201.
3. Ozkula U, Ozhasenekler A, Kurtoglu Celik G, Tanriverdi F, Pamukcu Gunaydin G, Ergin M, et al. Tissue adhesives to secure peripheral intravenous catheters: A ran- domized controlled trial in patients over 65 years. Turk J Emerg Med. 2018;19(1): 12-5.
4. Blanco-Mavillard I, Rodriguez-Calero MA, de Pedro-Gomez J, Parra-Garcia G, Fernandez-Fernandez I, Castro-Sanchez E. incidence of PEripheral intravenous cath- eter failure among inpatients: variability between microbiological data and clinical signs and symptoms. Antimicrob Resist Infect Control. 2019;22(8):124. https://doi. org/10.1186/s13756-019-0581-8.
5. Ray-Barruel G, Xu H, Marsh N, Cooke M, Rickard CM. Effectiveness of insertion and maintenance bundles in preventing peripheral intravenous catheter-related compli- cations and bloodstream infection in hospital patients: a systematic review. Infect Dis Health. 2019;24(3):152-68. https://doi.org/10.1016/j.idh.2019.03.001.
6. Takahashi T, Murayama R, Abe-Doi M, Miyahara-Kaneko M, Kanno C, Nakamura M, et al. Preventing peripheral intravenous catheter failure by reducing mechanical ir- ritation. Sci Rep. 2020;10(1):1550. https://doi.org/10.1038/s41598-019-56873-2.
7. Bouarab-Chibane L, Forquet V, Lanteri P, Clement Y, Leonard-Akkari L, Oulahal N, et al. Antibacterial properties of polyphenols: characterization and QSAR (quantita- tive structure-activity relationship) models. Front Microbiol. 2019;18(10):829. https://doi.org/10.3389/fmicb.2019.00829.