Advertisement

Factors affecting the patency and complications of peripheral intravenous catheters in newborns

Open AccessPublished:October 29, 2022DOI:https://doi.org/10.1016/j.pedneo.2022.07.011

      Background

      Peripheral intravenous catheters (PICs) are necessary for medication, nutrient, and fluid administration in pediatric patients. However, PICs are uneasy to access and maintain in young infants. This study identified risk factors affecting the complications and patency of PICs.

      Methods

      This retrospective cohort study included neonates and infants aged <4 months. All PICs inserted in the neonatal intensive care unit and intermediate care nursery were analyzed more than 5 months. The variables included gestational age, age and body weight at PIC insertion, insertion site, methods to maintain PIC patency (continuous intravenous drip [CIVD] versus intermittent flushing), fluid infusion rate and osmolarity, and ampicillin and cefotaxime concentrations. The effects of these variables on PIC complications and lifespan were assessed using binary logistic regression analysis and a general linear model, respectively.

      Results

      In total, 315 PICs were analyzed. The mean indwelling time was 33.8 ± 21.5 h and complication rate was 82.2%. The most frequent complications were infiltration (55.9%) and leakage (22.2%). The infusion rate and method to maintain PICs significantly impacted PIC patency. A negative correlation was noted between the infusion rate and PIC patency, with the patency decreasing by 0.9 h (p = 0.047) on increasing the infusion rate by 1 mL/h. Notably, compared with intermittent flushing, CIVD using a hypertonic solution significantly decreased PIC patency by 14 h (p = 0.006). As the patients’ age increased by a month, the complication risk decreased by 35% (p = 0.027). However, as the infusion rate increased by 1 mL/h, the complication risk increased by 17% (p = 0.018).

      Conclusions

      Intermittent flushing may be preferred over CIVD to preserve PIC patency. An increased infusion rate is correlated with decreased PIC patency and increased complications. For the peripheral administration of ampicillin, we recommended preparing final concentrations below 50 mg/dL to prevent PIC complications.

      Key Words

      1. Introduction

      In pediatric hospitalization, peripheral intravenous catheters (PICs) are crucial for life support in extremely low-birth-weight preterm infants through antibiotic, nutrient, and fluid administration. Maintaining PIC patency in pediatric patients is crucial because their insertion procedure is time-consuming and associated with patient discomfort.
      • Reigart J.R.
      • Chamberlain K.H.
      • Eldridge D.
      • O'Brien E.S.
      • Freeland K.D.
      • Larsen P.
      • et al.
      Peripheral intravenous access in pediatric inpatients.
      The pediatric population is more susceptible to complications than adults because of their fragile vasculature, including capillary weakness, venous network with narrow vessels, and high adiposity.
      • Bitencourt E.S.
      • Leal C.N.
      • Boostel R.
      • Mazza V.
      • Felix J.V.C.
      • Pedrolo E.
      Prevalence of phlebitis related to the use of peripheral intravenous device in children.
      ,
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      The younger the patient, the longer is the time and greater are the attempts required to successfully insert PICs.
      • Reigart J.R.
      • Chamberlain K.H.
      • Eldridge D.
      • O'Brien E.S.
      • Freeland K.D.
      • Larsen P.
      • et al.
      Peripheral intravenous access in pediatric inpatients.
      Furthermore, vascular access in neonates is challenging, particularly in very-low-birth-weight (VLBW) infants and infants requiring prolonged intravenous access.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      PIC complications include phlebitis, infiltration, extravasation, clotting, infection, and accidental removal. In the neonatal population, the complication rate of PICs ranges from 55.6% to 86.6%, with the mean PIC lifespan being 23–84 h.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      • Legemaat M.
      • Carr P.J.
      • van Rens R.M.
      • van Dijk M.
      • Poslawsky I.E.
      • van den Hoogen A.
      Peripheral intravenous cannulation: complication rates in the neonatal population: a multicenter observational study.
      • Danski M.T.
      • Mingorance P.
      • Johann D.A.
      • Vayego S.A.
      • Lind J.
      Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
      • Dalal S.S.
      • Chawla D.
      • Singh J.
      • Agarwal R.K.
      • Deorari A.K.
      • Paul V.K.
      Limb splinting for intravenous cannulae in neonates: a randomised controlled trial.
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      Various risk factors affect the complication rate of PICs. These include the patient characteristics (gestational age [GA], postnatal age, and body weight),
      • Danski M.T.
      • Mingorance P.
      • Johann D.A.
      • Vayego S.A.
      • Lind J.
      Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
      ,
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      type of medication or fluid used,
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      ,
      • Ben Abdelaziz R.
      • Hafsi H.
      • Hajji H.
      • Boudabous H.
      • Ben Chehida A.
      • Mrabet A.
      • et al.
      Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
      concentration of medication,
      • Ben Abdelaziz R.
      • Hafsi H.
      • Hajji H.
      • Boudabous H.
      • Ben Chehida A.
      • Mrabet A.
      • et al.
      Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
      ,
      • Kuensting L.L.
      Treatment of intravenous infiltration in a neonate.
      osmolarity of intravenous fluid, method to maintain PIC patency (continuous intravenous drip [CIVD] versus intermittent flushing),
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      ,
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      ,
      • Stok D.
      • Wieringa J.W.
      Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
      and location of PIC insertion (lower extremity).
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      ,
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      However, few studies have assessed the effect of the volume of fluid exposure and use of a hypertonic solution in CIVD on PIC patency and complications.
      As in many other countries, registered nurses are the main health-care providers in charge of inserting and maintaining PICs in Taiwan. The responsibilities of registered nurses include inserting PICs by using an aseptic technique, regularly monitoring intravenous therapy and medication administered through PICs, checking the date of PIC insertion, and replacing PICs.
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      In our hospital, the policy of PIC replacement requires changing PICs every 72 h unless there are complications or specific instructions by clinical doctors.
      This study aimed to assess the incidence of complications associated with PICs and monitor the PIC lifespan in the relatively fragile pediatric population aged <4 months. Moreover, risk factors leading to complications and reducing the patency of PICs were identified.

      2. Methods

      2.1 Study design and setting

      This retrospective cohort study included patients hospitalized in the neonatal intensive care unit (NICU) and intermediate care nursery in a regional hospital that provides neonatal intensive care in the southern Taipei area, Taiwan.
      Data collection was performed between November 1, 2020, and March 31, 2021. The study was approved by the Institutional Review Board of Taipei Medical University (IRB-TMU N202108029). All 0-day- to 4-month-old patients who needed peripheral venous catheterization were included in the study. All PICs inserted in the study units were recorded from the time of insertion to the time of removal. PICs that were inserted before admission to the study units and were removed outside the units were excluded. The sample size was calculated considering 15–20 individuals per predictor in regression analysis. Eight variables, including three continuous and five categorical variables, and a total of 12 dummy variables were set in the regression model. A sample size of at least 240 PICs was considered large enough to overcome potential bias.

      2.2 Outcome variables

      Patient-related variables included the GA, age and weight at the time of PIC insertion, gender, and mean number of PICs inserted per patient. Catheter-related variables included the duration of PICs (time between PIC insertion and removal), reasons for PIC removal (infiltration [i.e., swelling documented on nursing records], phlebitis, leakage, occlusion, dislodgement, and end of medical need), insertion site (upper and lower extremities), type of medication, volume of fluid exposure, and method to maintain PIC patency (CIVD or intermittent flushing). The osmolarity of intravenous medication and intravenous fluid is a potential risk factor affecting the lifespan and complications of PICs. A hypotonic solution infiltrating the venous endothelial cells and blood cells may result in hemolysis, whereas a hypertonic solution may result in cell shrinkage.
      • Stranz M.
      • Kastango E.S.
      A Review of pH and osmolarity.
      Therefore, the intravenous fluid used to maintain PICs was categorized into hypertonic (>308 mOsm/L), isotonic (275–308 mOsm/L), and hypotonic (<275 mOsm/L). The osmolarity of the intravenous fluid was based on Drug Instruction (Supplementary Table 1).
      • Moritz M.L.
      Why 0.9% saline is isotonic: understanding the aqueous phase of plasma and the difference between osmolarity and osmolality.

      Kliegman, R. Nelson textbook of pediatrics. 21th ed. Philadelphia, PA: Elsevier; 2020, e5-e14

      Eight variables influencing the PIC lifespan and occurrence of complications were assessed. A complication was defined as an event necessitating PIC removal for reasons other than the end of medical need. The study variables included the GA (≥37, 34–36, and <34 weeks), age and body weight at PIC insertion, site of PIC insertion (upper and lower extremities), method to maintain PIC patency (CIVD with a hypertonic solution, CIVD with an isotonic solution, CIVD with a hypotonic solution, and intermittent flushing), volume of fluid exposure (expressed as the mean infusion rate, mL/min), and ampicillin and cefotaxime concentrations (≤50 and >50 mg/mL). Ampicillin and cefotaxime are frequently prescribed for early- and late-onset neonatal sepsis in patients aged <4 months. For peripheral intravenous infusion in fluid-restricted patients, the maximum concentration of ampicillin is 112 mg/mL in sterile water and that of cefotaxime is 147 mg/mL in sterile water.
      However, concentrations exceeding 50 mg/mL of both medications will increase osmolarity beyond 500 mOsm/L, which is the higher limit for peripheral vein tolerance recommended by the Infusion Nursing Society.
      Intravenous Nurses Society
      Intravenous nursing standards of practice.
      Therefore, we set 50 mg/mL was set as the cutoff point for ampicillin and cefotaxime to examine whether intermittent exposure to high-osmolarity medications is a risk factor affecting the lifespan and complications of PICs.

      2.3 Data analysis

      Categorical variables are expressed as the absolute frequency and percentage. Continuous variables are expressed as the mean, standard deviation (SD), median, and interquartile range (IQR). Intergroup differences were analyzed using Student’s t test or nonparametric analysis depending on the normality of quantitative variables. A general linear model was used to identify variables affecting the PIC lifespan. Variables that met two criteria, namely p < 0.1 and no evidence of collinearity, were included in the general linear model. Variables associated with a complication were identified through binary logistic regression analysis using the forward likelihood ratio method (p < 0.1 was the cutoff point). All statistical analyses were performed using SPSS 20 (SPSS Inc., Chicago, IL, USA), and p < 0.05 was considered statistically significant.

      3. Results

      3.1 Patient demography

      In total, 103 patients with 315 PICs were included in this study. The patients’ mean age was 0.66 ± 0.94 months, and 73% were neonates. Moreover, 60% of the patients were male. The mean GA was 37 ± 3 weeks (range, 28–41 weeks), with the GA being ≥37 weeks in 76.7% of the patients. The patients’ mean weight was 3.36 ± 1.16 kg (range, 0.69–6.91 kg), and 75% of them weighed more than 2.5 kg. On average, three PICs were inserted into a hospitalized patient per course (SD, 1.88; range, 1–11).

      3.2 PIC characteristics

      In total, 315 PICs were included in this study (Table 1). The size of all PICs was 24 gauge (Introcan Safety® IV Catheter 24 Ga × 0.75 in., PUR, Straight). Their mean indwelling time was 33.8 ± 21.5 h (range, 1–149.7 h). Most PICs (259 PICs, 82.2%) were removed because of complications, including infiltration, leakage, occlusion, and dislodgement. The most common complication was infiltration (176 PICs, 55.9%), followed by leakage (70 PICs, 22.2%), occlusion (12 PICs, 3.8%), and dislodgement (1 PIC, 0.3%). Only 56 PICs (17.8%) were removed without any complication (i.e., the end of medical need). PICs without complications had a longer lifespan than those with complications (41.6 ± 28.5 h versus 32.1 ± 19.3 h). In total, 300 PICs (95%) were inserted for administering antibiotics and other medications. On average, 1.7 ± 0.7 medications (range, 0–4) were administered per PIC. The most frequently administered medication was ampicillin (76.8%), followed by gentamicin (56.2%) and cefotaxime (17.1%). Moreover, 82% of PICs were inserted in the upper extremity and 90% of PICs were maintained using CIVD. Most PICs (62.5%) were maintained using CIVD with a hypertonic solution, while no PIC was maintained using CIVD with a hypotonic solution. CIVD with a hypertonic solution (4.1 ± 2.7 mL/h; range, 2–18 mL/h) had a similar infusion rate to CIVD with an isotonic solution (4.8 ± 3.3 mL/h; range, 2–20 mL/h). Only 10% of PICs (31 PICs) were maintained using intermittent flushing. Intermittent flushing maintained a longer PIC lifespan than CIVD with hypertonic and isotonic solutions (intermittent flushing, 46.4 ± 30.5 h; CIVD with an isotonic solution, 38.2 ± 23.6 h; CIVD with a hypertonic solution, 29.9 ± 17.4 h).
      Table 1Catheter characteristics.
      ItemN%PIC duration (hr)P-value
      N = 315MedianIQR
      Ward of IC insertion0.379
       NICU8125.7%25.323.0
       Intermediate care nursery23474.3%29.925.5
      Sex0.561
       Female12238.7%28.029.1
       Male19361.3%29.624.2
      Age at catheter insertion0.896
       <30 days23975.9%29.424.3
       1–4 months7624.1%29.230.7
      Weight at catheter insertion0.424
       <1000 g51.6%4642.6
       1000–1499 g72.2%32.221.5
       1500–2499 g5818.4%32.228.6
       2500–3999 g17555.6%28.123.9
       >4 kg7022.2%29.526.6
      Methods to maintain catheter0.001
       Hypertonic CIVD19762.5%25.323.0
       Isotonic CIVD8727.6%34.230.8
       Intermittent flush319.8%38.431.4
      Insertion site0.179
       Dorsum of the hand17555.6%26.225.3
       Forearm8426.7%31.126.5
       Ankle4113.0%34.323.7
       Dorsum pedis123.8%26.418.9
       Calf31.0%16.9
      Reason for catheter removal0.019
       End of medical need5617.8%38.737.1
       Infiltration17655.9%25.822.2
       Leaking7022.2%31.724.0
       Occlusion123.8%41.049.3
       Dislodgement10.3%
      CIVD: continuous intravenous drip; IQR: interquartile range; NICU: neonate intensive care unit.

      3.3 Risks factors affecting PIC lifespan

      Univariate analysis revealed that the PIC lifespan was influenced by the patients’ age, infusion rate, method used to maintain PICs, and cefotaxime concentration. As the patients’ age increased by a month, the PIC lifespan increased by 2.5 h (p = 0.048). However, as the infusion rate increased by 1 mL/h, the PIC lifespan decreased by 0.9 h (p = 0.020). The mean PIC lifespan using both CIVD methods was 8 h lower than that using the intermittent flushing method (CIVD with a hypertonic solution, −8.31 h, p = 0.002; CIVD with an isotonic solution, −8.149 h, p = 0.062). The patients’ body weight, PIC insertion site, and ampicillin concentration did not affect the PIC lifespan (Table 2).
      Table 2Factors influencing lifespan of peripheral intravenous catheters.
      ItemN%IC patency (hr)Simple regressionMultiple regression
      MeanSDβ95%CIP valueβ95% CIP value
      Age at catheter insertion (months)315100%33.821.52.537(0.021, 5.053)0.0483.439(−0.557, 7.435)0.091
      Weight at catheter insertion (kg)315100%33.821.5−0.182(−2.249, 1.855)0.863
      Infusion rate (mL/hr)315100%33.821.5−0.910(−1.678, −0.143)0.020−0.900(−1.788, −0.013)0.047
      Gestational age0.0580.463
       GA <34 weeks3410.8%42.127.2
       GA 34–36 weeks268.3%33.125.7−9.042(−19.992, 1.907)0.105−1.523(−13.250, 10.205)0.799
       GA ≧37 weeks25581.0%32.820.0−9.309(−16.982, −1.636)0.018−4.738(−13.674, 4.198)0.298
      Insertion site0.538
       Upper extremity25982.2%33.522.2
       Lower extremity5617.8%35.417.71.955(−4.282, 8.191)
      Methods to maintain catheter0.0000.012
       Intermittent flush319.8%46.430.5
       CIVD isotonic fluid8727.6%38.223.6−8.149(−16.722, 0.425)0.062−7.030(−18.101, 4.042)0.212
       CIVD hypertonic fluid19762.5%29.917.4−8.310(−13.586, −3.033)0.002−14.302(−24.469, −4.134)0.006
      Ampicillin0.0610.001
       ≦50 mg/mL16050.8%33.320.9
       50.1–100 mg/mL8226.0%38.125.64.824(−0.884, 10.533)0.097−3.545(−9.779, 2.690)0.264
       No ampicillin7323.2%30.116.6−3.217(−9.154, 2.720)0.287−13.324(−20.234, −6.415)0.000
      Cefotaxime0.0220.086
       ≦50 mg/mL4012.7%34.619.0
       50.1–125 mg/mL144.4%49.137.414.501(1.491, 27.511)0.02913.613(0.277, 26.949)0.045
       No cefotaxime26182.9%32.920.5−1.695(−8.809, 5.419)0.6400.587(−6.527, 7.700)0.871
      CIVD: continuous intravenous drip; NICU: neonate intensive care unit; hr: hour.
      After taking all variables into account, the infusion rate and method to maintain PICs were found to affect the PIC lifespan. A negative correlation was noted between the infusion rate and PIC lifespan, with a 0.9-h decrease in PIC lifespan being observed on increasing the infusion rate by 1 mL/h (p = 0.047). Compared with intermittent flushing, CIVD with a hypertonic solution significantly decreased the PIC lifespan by 14 h (p = 0.006), while CIVD with an isotonic solution no significantly decreased the patency of PICs by 7 h (p = 0.212). Moreover, ampicillin concentrations >50 mg/mL no significantly reduced the patency of PICs by 3.5 h in comparison with concentrations ≤50 mg/dL (p = 0.264; Table 2).

      3.4 Risk factors affecting PIC complications

      Univariate analysis revealed that younger age, the use of the CIVD method to maintain PICs, and the use of ampicillin contributed to PIC complications. As the patients’ age increased by a month, the complication risk decreased by 28% (p = 0.019). CIVD with a hypertonic solution had the highest complication rate (86.3%), followed by CIVD with an isotonic solution (79.3%) and intermittent flushing (64.5%). Compared with intermittent flushing, CIVD with a hypertonic solution led to a 3.4-fold increase in the complication risk (p = 0.004). Moreover, compared with ampicillin concentrations ≤50 mg/mL, ampicillin concentrations >50 mg/dL tended to increase the complication risk (odds ratio [OR], 1.148; p = 0.736). However, compared with the absence of ampicillin, higher concentrations of ampicillin led to higher risks of complications (ampicillin concentrations of 50.1–100 mg/mL: OR, 3.527; p = 0.003; ampicillin concentrations ≤50 mg/mL: OR, 3.072; p = 0.001). The patients’ body weight, infusion rate, GA, insertion site, and cefotaxime concentration did not affect the incidence of PIC complications (Table 3).
      Table 3Factors influencing complication of peripheral intravenous catheters.
      ItemComplicationNo complicationUnivariate analysisMultivariate analysis
      N%N%OR95%CIP valueOR95%CIP value
      Age at catheter insertion (months)25982.2%5617.8%0.720(0.548, 0.946)0.0190.648(0.441, 0.953)0.027
      Weight at catheter insertion (kg)25982.2%5617.8%0.901(0.705, 1.152)0.405
      Infusion rate (mL/hr)25982.2%5617.8%1.112(0.991, 1.248)0.0701.169(1.027, 1.330)0.018
      Gestational age0.118
       GA <34 weeks2470.6%1029.4%
       GA 34–36 weeks2076.9%623.1%1.389(0.430, 4.490)0.583
       GA ≧37 weeks21584.3%4015.7%2.240(0.995, 5.041)0.051
      Insertion site0.432
       Upper extremity21583.0%4417.0%
       Lower extremity4478.6%1221.4%0.750(0.367, 1.535)
      Methods to maintain catheter0.012
       Intermittent flush2064.5%1135.5%
       CIVD Isotonic fluid6979.3%1820.7%2.108(0.857, 5.187)0.104
       CIVD hypertonic fluid17086.3%2713.7%3.463(1.494, 8.024)0.004
      Ampicillin0.0010.027
       ≦50 mg/mL21583.0%4417.0%
       50.1–100 mg/mL4478.6%1221.4%1.148(0.516, 2.554)0.7361.591(0.688, 3.899)0.265
       No ampicillin4967.1%2432.9%0.325(0.168, 0.632)0.0010.555(0.241, 1.164)0.114
      Cefotaxime0.624
       ≦50 mg/mL3587.5%512.5%
       50.1–125 mg/mL1178.6%321.4%0.524(0.108, 2.552)0.424
       No cefotaxime21381.6%4818.4%0.634(0.236, 1.703)0.366
      CIVD: continuous intravenous drip; GA: gestational age; kg: kilogram.
      After taking all variables into account, the patients’ age, infusion rate, and ampicillin concentration (50.1–100 mg/mL) were found to affect the complication risk. Age had a positive impact on the complication risk. As the patients’ age increased by a month, the complication risk decreased by 35% (p = 0.027). By contrast, the infusion rate had a negative impact on the complication risk. As the infusion rate increased by 1 mL/h, the complication risk increased by 17% (p = 0.018). High ampicillin concentrations (>50 mg/mL) non-significantly increased the complication risk (1.5-fold increase) in comparison with low ampicillin concentrations (≤50 mg/mL, p = 0.265). However, compared with the absence of ampicillin, high ampicillin concentrations significantly increased the complication risk (threefold increase, p = 0.008). Low ampicillin concentrations also led to an increased risk, but with a wide confidence interval (CI) (95% CI, 0.859–4.115).

      4. Discussion

      4.1 PIC lifespan and risks factors

      Our findings revealed that PICs without complications had a longer lifespan (41.6 ± 28.5 h) than PICs with complications (32.1 ± 19.3 h). Previous studies have also revealed longer patency in PICs without complications.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      ,
      • Monasor-Ortolá D.
      • Cortés-Castell E.
      • Martínez-Pascual C.
      • Esteve-Ríos A.
      • Rizo-Baeza M.M.
      Factors influencing the success of peripheral venous access in neonates.
      The mean PIC lifespan in our study was 33.8 ± 21.5 h. The PICs included in our study were predominantly inserted in newborns (73%). In studies focusing on neonates, the mean catheter lifespan ranged from 23 to 84 h; the PIC lifespan in our study was comparable to previous findings.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      ,
      • Danski M.T.
      • Mingorance P.
      • Johann D.A.
      • Vayego S.A.
      • Lind J.
      Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
      • Dalal S.S.
      • Chawla D.
      • Singh J.
      • Agarwal R.K.
      • Deorari A.K.
      • Paul V.K.
      Limb splinting for intravenous cannulae in neonates: a randomised controlled trial.
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      ,
      • Stok D.
      • Wieringa J.W.
      Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
      ,
      • Monasor-Ortolá D.
      • Cortés-Castell E.
      • Martínez-Pascual C.
      • Esteve-Ríos A.
      • Rizo-Baeza M.M.
      Factors influencing the success of peripheral venous access in neonates.
      The PIC lifespan was clearly affected by the method used to maintain PICs and the osmolarity of CIVD solution after adjustment for all the included variables. The intermittent flushing method maintained a longer PIC lifespan than the two CIVD methods (intermittent flushing, 46.4 ± 30.5 h; CIVD with an isotonic solution, 38.2 ± 23.6 h; CIVD with a hypertonic solution, 29.9 ± 17.4 h). Similarly, Perez et al. and Hoff et al. found that intermittent flushing maintained a longer PIC lifespan than CIVD (Perez et al.: 92.8 versus 62.1 h, p = 0.01; Hoff et al.: 47.1 versus 35.4 h, p = 0.041).
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      ,
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      However, Stock et al. found that the PIC lifespan was similar between intermittent flushing and CIVD methods (55.42 versus 57.48 h, p = 0.797).
      • Stok D.
      • Wieringa J.W.
      Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
      In their study, many catheters were removed at approximately 48 h because of antibiotic discontinuation. This short duration could have disguised the actual effects on the statistics of PIC patency.
      Assessment of the effect of osmolarity revealed that compared with intermittent flushing, CIVD with a hypertonic solution reduced the PIC lifespan by 14 h. Of the 197 PICs that used a hypertonic solution in our study, 194 PICs used a solution in which the osmolarity was over the 500 mOsm/L limit defined by the Infusion Nursing Society.
      Intravenous Nurses Society
      Intravenous nursing standards of practice.
      Abdelaziz et al. also found that a high-osmolarity solution could reduce catheter lifespan (10% dextrose solution, 556 mOsm/L: 49.95 h versus normal saline, 308 mOsm/L: 96.67 h; p = 0.003).
      • Ben Abdelaziz R.
      • Hafsi H.
      • Hajji H.
      • Boudabous H.
      • Ben Chehida A.
      • Mrabet A.
      • et al.
      Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
      We considered the effects of ampicillin and cefotaxime concentrations >50 mg/mL; however, osmolarity >500 mOsm/L may be risky. We could not prove this assumption in our multiple regression analysis. Hoff et al. revealed that antibiotic use did not affect catheter lifespan.
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      Gupta et al. found that cefotaxime reduced catheter lifespan, whereas ampicillin did not.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      Few studies have examined the effect of osmolarity or concentration of intravenous medication on catheter lifespan. Further research is warranted to answer this question.
      Our findings revealed that CIVD with a hypertonic solution clearly reduced the PIC lifespan. Moreover, the PIC lifespan was negatively correlated with the infusion rate. As the infusion rate increased by 1 mL/h, the PIC lifespan decreased by 0.9 h (p = 0.047). This could be attributed to the larger volume of fluid exposed to blood vessels, higher pressure, or irritation of the capillary vasculature.
      The body weight, gestational age (GA), and insertion site had no impact on the PIC lifespan in our study, similar to the findings of Gupta et al.
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      Hoff et al. demonstrated that the body weight and insertion site did not affect the lifespan of catheters. However, they found that extremely preterm neonates had a significantly longer catheter lifespan.
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      In our study, less than 3% of PICs were inserted in extremely preterm neonates; therefore, the data were insufficient to examine the effect on extremely preterm neonates.

      4.2 PIC complications and risk factors

      The complication rate of PICs was high (82.2%) in our study, and the most frequent complication was infiltration (55.9%). Several studies focusing on neonates have also reported a high complication rate of PICs (range, 55%–86%) and high susceptibility of this population to infiltration/extravasation (range, 37%–68%).
      • Gupta P.
      • Rai R.
      • Basu S.
      • Faridi M.M.
      Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
      ,
      • Danski M.T.
      • Mingorance P.
      • Johann D.A.
      • Vayego S.A.
      • Lind J.
      Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
      • Dalal S.S.
      • Chawla D.
      • Singh J.
      • Agarwal R.K.
      • Deorari A.K.
      • Paul V.K.
      Limb splinting for intravenous cannulae in neonates: a randomised controlled trial.
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      ,
      • Monasor-Ortolá D.
      • Cortés-Castell E.
      • Martínez-Pascual C.
      • Esteve-Ríos A.
      • Rizo-Baeza M.M.
      Factors influencing the success of peripheral venous access in neonates.
      ,
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      In our study, treatment completion was achieved to a greater extent using the intermittent flushing method (35.5%) and by not administrating ampicillin (32.9%). Among all the variables studied, CIVD with a hypertonic solution had the lowest treatment completion rate (17.3%). Univariate analysis revealed that CIVD with a hypertonic solution led to a 3.5-fold higher complication risk than intermittent flushing. CIVD with an isotonic solution also increased the complication risk, but with a wide CI. Compared with the CIVD method, Stock et al.
      • Stok D.
      • Wieringa J.W.
      Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
      reported a lower complication rate and Hoff et al.
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      reported a lower extravasation rate when the intermittent flushing method was used. Although Perez et al. reported no difference in complications between the two methods, they noted an increase in the PIC lifespan when the CIVD method was used in neonates.
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      Our findings were similar to these previous findings. In contrast to our findings, Jacinto et al. found that the intermittent flushing method led to a higher occurrence of phlebitis than the CIVD method in older children with a median age of 7–8 years.
      • Jacinto A.K.L.
      • Avelar A.F.M.
      • Wilson A.M.M.M.
      • Pedreira M.L.G.
      Phlebitis associated with peripheral intravenous catheters in children: study of predisposing factors.
      The divergent results between our study and that of Jacinto et al. may be attributed to population differences. Our study population consisted of patients aged <4 months. Most of them were neonates who were physically more fragile and were administered fluid more carefully than the older pediatric population in the study of Jacinto et al.
      After taking all variables into account, we found that the complications of PICs were affected by the patients’ age, infusion rate, and ampicillin concentration. Age was negatively correlated with the complications of PICs in our study. Foster et al. found that neonates were 5.5 times more likely to have some degree of phlebitis than populations older than neonates.
      • Foster L.
      • Wallis M.
      • Paterson B.
      • James H.
      A descriptive study of peripheral intravenous catheters in patients admitted to a pediatric unit in one Australian hospital.
      However, Abdelaziz et al. and Sulinan et al. found that age was not a predictor of complications.
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      ,
      • Ben Abdelaziz R.
      • Hafsi H.
      • Hajji H.
      • Boudabous H.
      • Ben Chehida A.
      • Mrabet A.
      • et al.
      Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
      Therefore, age is still an equivocal factor in PIC complications. We found that the infusion rate was positively correlated with PIC complications. Few studies have discussed the relationship between the infusion rate and catheter complications in the pediatric population. Hecker and Mentin et al. found that a higher infusion rate and an infusion volume larger than 2000 mL/day, respectively, increased the incidence of catheter complications in adult patients.
      • Hecker J.F.
      Failure of intravenous infusions from extravasation and phlebitis.
      ,
      • Gök-metin Z.
      • Özdemir L.
      Determining intravenous complications in patients with infusion pumps.
      In the case of ampicillin, compared with no ampicillin administration, the administration of ampicillin at concentrations >50 mg/mL with osmolarity >500 mOsm/L conferred a clear complication risk (OR, 3.093; p = 0.008). Similarly, Jacinto et al. found that risk solutions or medications had a sevenfold higher chance of leading to complications.
      • Jacinto A.K.L.
      • Avelar A.F.M.
      • Wilson A.M.M.M.
      • Pedreira M.L.G.
      Phlebitis associated with peripheral intravenous catheters in children: study of predisposing factors.
      They defined risk solutions and medications as those with pH less than 5 or greater than 9 or with osmolarity above 350 mOsm/L.
      • Jacinto A.K.L.
      • Avelar A.F.M.
      • Wilson A.M.M.M.
      • Pedreira M.L.G.
      Phlebitis associated with peripheral intravenous catheters in children: study of predisposing factors.
      In our study, compared with the absence of ampicillin, ampicillin concentrations ≤50 mg/mL led to a 1.9-fold increase in the complication risk (p = 0.114). Of the 215 PICs administering ampicillin at concentrations of ≤50 mg/mL in our study, 210 used ampicillin at a concentration of 50 mg/mL. Preparing ampicillin at a concentration of 50 mg/mL by using normal saline as a diluent provides an osmolarity of 493 mOsm/kg, which is close to the risk concentration mentioned above. This may explain why ampicillin concentrations ≤50 mg/mL led to a 1.9-fold increase in the complication risk in our study (p = 0.114). The sample size of the high cefotaxime concentration group (>50 mg/dL, n = 14) was not sufficient to prove the effect of high osmolarity (osmolarity >500 mOsm/L) on the complication risk. Further studies are required to examine the effect of cefotaxime concentration on the complication risk.
      The body weight, GA, and insertion site did not affect the complication risk in our study. Similarly, Yuningsih et al. found that GA did not affect the risk of phlebitis.
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      However, as per the findings of Danski et al., the lower the weight of the newborn on the puncture day, the greater was the complication risk (risk ratio of 1, 1.25, and 1.29 for body weight >2500 g, between 1500 and 2400 g, and <1500 g, respectively; p = 0.0093).
      • Danski M.T.
      • Mingorance P.
      • Johann D.A.
      • Vayego S.A.
      • Lind J.
      Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
      Moreover, Yuningsih et al. found that VLBW infants (1000–1499 g) had a fivefold greater risk of developing phlebitis than low-birth-weight (LBW) infants (1500–2499 g).
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      In contrast to our findings, Abdelaziz et al. found that the insertion site did not affect the complication risk.
      • Ben Abdelaziz R.
      • Hafsi H.
      • Hajji H.
      • Boudabous H.
      • Ben Chehida A.
      • Mrabet A.
      • et al.
      Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
      However, Yuningsih et al. and Suliman et al. found that the lower extremities were linked to an increased incidence of phlebitis.
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      ,
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      Further studies are required to assess the effects of these factors on PIC complications.

      4.3 Strength

      This is the first cohort study to explore factors leading to complications and reducing the PIC lifespan in the Asian pediatric population. Our results provide pilot data for improving the maintenance of PICs in patients aged <4 months in clinical practice. Although studies have assessed whether CIVD or intermittent flushing is a better method for preserving PICs, these studies have used a small infusion rate (range, 0.5–2 mL/h).
      • Perez A.
      • Feuz I.
      • Brotschi B.
      • Bernet V.
      Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
      ,
      • Stok D.
      • Wieringa J.W.
      Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
      ,
      • Hoff R.
      • Vervisch K.
      • De Coen K.
      • Smets K.
      Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
      However, our study assessed the effects of a wide range of continuous infusion rates (range, 2–20 mL/min) on the complications and patency of PICs in the pediatric population.

      4.4 Limitations

      Our study has some limitations. First, the major population of our study consisted of neonates with body weight >1500 g. Only 5% of the patients were VLBW and LBW infants; therefore, our results could not be extrapolated to very and extremely preterm babies with body weight <1500 g. Second, because of limited data accessibility, we could not include all variables that may influence the PIC lifespan and incidence of complications. These variables include the PIC fixation method, effect of other medication or blood transfusion, and procedure time of catheter insertion.
      • Suliman M.
      • Saleh W.
      • Al-Shiekh H.
      • Taan W.
      • AlBashtawy M.
      The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
      ,
      • Dalal S.S.
      • Chawla D.
      • Singh J.
      • Agarwal R.K.
      • Deorari A.K.
      • Paul V.K.
      Limb splinting for intravenous cannulae in neonates: a randomised controlled trial.
      ,
      • Yuningsih R.
      • Rustina Y.
      • Efendi D.
      The related factors of phlebitis among low birth weight infants in perinatology ward.
      ,
      • Tripathi S.
      • Kaushik V.
      • Singh V.
      Peripheral IVs: factors affecting complications and patency - a randomized controlled trial.
      Finally, the lowest continuous infusion rate used to maintain PICs was above 2 mL/h. We could not examine whether using a small continuous infusion rate (less than 2 mL/h) could improve the patency of PICs and reduce complications in comparison with that by using the intermittent flushing method.

      5. Conclusion

      Based on our findings, we recommend using the intermittent flushing method and intravenous medication with osmolarity <500 mOsm/L to maintain PICs. If continuous intravenous fluid is indicated, we recommend using an isotonic solution with the lowest infusion rate that can satisfy patients’ needs.

      Declaration of Competing Interest

      None.

      Acknowledgment

      We would like to give special thanks to Pei-Hsuan Yang for helping with data collection, Dr. Yen-Kuang Lin for consultation on analytical methods, and all the medical and nursing staff at the neonatal intensive care unit/newborn center of Taipei Medical University-Shuang Ho Hospital for their support throughout the study. This manuscript was edited by Wallace Academic Editing.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Reigart J.R.
        • Chamberlain K.H.
        • Eldridge D.
        • O'Brien E.S.
        • Freeland K.D.
        • Larsen P.
        • et al.
        Peripheral intravenous access in pediatric inpatients.
        Clin Pediatr (Phila). 2012; 51: 468-472
        • Bitencourt E.S.
        • Leal C.N.
        • Boostel R.
        • Mazza V.
        • Felix J.V.C.
        • Pedrolo E.
        Prevalence of phlebitis related to the use of peripheral intravenous device in children.
        Cogitare Enferm. 2018; 23e49361
        • Suliman M.
        • Saleh W.
        • Al-Shiekh H.
        • Taan W.
        • AlBashtawy M.
        The incidence of peripheral intravenous catheter phlebitis and risk factors among pediatric patients.
        J Pediatr Nurs. 2020; 50: 89-93
        • Gupta P.
        • Rai R.
        • Basu S.
        • Faridi M.M.
        Life span of peripheral intravenous cannula in a neonatal intensive care unit of a developing country.
        J Pediatr Nurs. 2003; 18: 287-292
        • Legemaat M.
        • Carr P.J.
        • van Rens R.M.
        • van Dijk M.
        • Poslawsky I.E.
        • van den Hoogen A.
        Peripheral intravenous cannulation: complication rates in the neonatal population: a multicenter observational study.
        J Vasc Access. 2016; 17: 360-365
        • Danski M.T.
        • Mingorance P.
        • Johann D.A.
        • Vayego S.A.
        • Lind J.
        Incidence of local complications and risk factors associated with peripheral intravenous catheter in neonates.
        Rev Esc Enferm USP. 2016; 50 ([Article in Portuguese]): 22-28
        • Dalal S.S.
        • Chawla D.
        • Singh J.
        • Agarwal R.K.
        • Deorari A.K.
        • Paul V.K.
        Limb splinting for intravenous cannulae in neonates: a randomised controlled trial.
        Arch Dis Child Fetal Neonatal Ed. 2009; 94: F394-F396
        • Perez A.
        • Feuz I.
        • Brotschi B.
        • Bernet V.
        Intermittent flushing improves cannula patency compared to continuous infusion for peripherally inserted venous catheters in newborns: results from a prospective observational study.
        J Perinat Med. 2012; 40: 311-314
        • Yuningsih R.
        • Rustina Y.
        • Efendi D.
        The related factors of phlebitis among low birth weight infants in perinatology ward.
        Pediatr Rep. 2020; 12: 8691
        • Ben Abdelaziz R.
        • Hafsi H.
        • Hajji H.
        • Boudabous H.
        • Ben Chehida A.
        • Mrabet A.
        • et al.
        Peripheral venous catheter complications in children: predisposing factors in a multicenter prospective cohort study.
        BMC Pediatr. 2017; 17: 208
        • Kuensting L.L.
        Treatment of intravenous infiltration in a neonate.
        J Pediatr Health Care. 2010; 24: 184-188
        • Stok D.
        • Wieringa J.W.
        Continuous infusion versus intermittent flushing: maintaining peripheral intravenous access in newborn infants.
        J Perinatol. 2016; 36: 870-873
        • Stranz M.
        • Kastango E.S.
        A Review of pH and osmolarity.
        Int J Pharm Compd. 2002; 6: 216-220
      1. McEvoy G.K. Snow E.K. Miller J. Handbook on injectable drugs. American Society of Health-System Pharmacists, Bethesda, MD2018
        • Moritz M.L.
        Why 0.9% saline is isotonic: understanding the aqueous phase of plasma and the difference between osmolarity and osmolality.
        Pediatr Nephrol. 2019; 34: 1299-1300
      2. Kliegman, R. Nelson textbook of pediatrics. 21th ed. Philadelphia, PA: Elsevier; 2020, e5-e14

        • Intravenous Nurses Society
        Intravenous nursing standards of practice.
        J Intraven Nurs. 2000; 23: S37-S39
        • Monasor-Ortolá D.
        • Cortés-Castell E.
        • Martínez-Pascual C.
        • Esteve-Ríos A.
        • Rizo-Baeza M.M.
        Factors influencing the success of peripheral venous access in neonates.
        J Pediatr Nurs. 2019; 47: e30-e35
        • Hoff R.
        • Vervisch K.
        • De Coen K.
        • Smets K.
        Continuous infusion vs. intermittent flushing of peripheral cannulas in neonates using a needleless connector: a prospective cohort study.
        J Perinat Med. 2019; 47: 464-469
        • Jacinto A.K.L.
        • Avelar A.F.M.
        • Wilson A.M.M.M.
        • Pedreira M.L.G.
        Phlebitis associated with peripheral intravenous catheters in children: study of predisposing factors.
        Esc Anna Nery. 2014; 18: 220-226
        • Foster L.
        • Wallis M.
        • Paterson B.
        • James H.
        A descriptive study of peripheral intravenous catheters in patients admitted to a pediatric unit in one Australian hospital.
        J Infus Nurs. 2002; 25: 159-167
        • Hecker J.F.
        Failure of intravenous infusions from extravasation and phlebitis.
        Anaesth Intensive Care. 1989; 17: 433-439
        • Gök-metin Z.
        • Özdemir L.
        Determining intravenous complications in patients with infusion pumps.
        Hemşirelikte Araştırma Geliştirme Dergisi. 2016; 18: 22-30
        • Tripathi S.
        • Kaushik V.
        • Singh V.
        Peripheral IVs: factors affecting complications and patency - a randomized controlled trial.
        J Infus Nurs. 2008; 31: 182-188