Peripherally Inserted Central Catheters-associated blood stream infections—occurrence, risk factors, and pathogens, a single center study

Abstract

Background

Peripherally inserted central catheters (PICCs) are central venous catheters inserted peripherally but terminate in great vessels. PICCs are widely used for patients requiring long-term intravenous therapy in both in-patient and out-patient settings.

Aim

This study was carried out to understand PICC-related complications, specifically infections and causal pathogens, in a tertiary care hospital in Kerala, South India.

Methods

A retrospective analysis of PICC insertions and follow-up during a 9 years period to look at patient demographics and infections related to PICC was carried out.

Results

The overall PICC-related complication rate is 28.1% (4.98 per 1000 PICC days). Commonest complication was thrombosis followed by infection, either PICC-associated bloodstream infection (PABSI) or local infection (LI). PABSI noted in this study was 1.34 per 1000 catheter days. The majority (85%) of PABSI were due to Gram-negative rods. The average duration of PICC days for occurrence of PABSI was 14 days and the majority occurred in in-patients.

Conclusion

Thrombosis and infection were the commonest PICC-related complications. PABSI rate was comparable to that of previous studies.

Background

Peripherally inserted central catheters (PICCs) are unique in that they have the ability to provide venous access for many months in hospital and home, both in-patient and out-patient settings. PICCs are considered to be associated with lesser risk of blood stream infection than non-tunneled central venous catheter (CVC) (Maki et al., 2006). In recent years, PICC is being increasingly used. The growing use of PICC has led to recognition of the risk of PICC-associated bloodstream infection (PABSI), but not much data are available from Asian countries, notably from India. We sought to identify the incidence rates, pathogens, and risk factors associated with PABSI as well as other complications associated with PICC.

Methods

A single center retrospective chart review of electronic medical record of adult and pediatric patients who underwent PICC insertion between January 2013 and September 2021 was performed. The study was conducted in a 500-bedded tertiary care hospital in Kerala, South India, after institutional review board approval. PICC insertion was carried out either by a trained nurse or an anesthetist in PICC team, following sterile and ultrasound-guided technique or by interventional radiologist under fluoroscopic guidance. PICCs used were radiopaque polyurethane single lumen catheters by either Arrow PICC set (Teleflex, Mexico) or Groshong NXT ClearVue Catheter (Bard Access Systems, Inc. Utah.) PICCs were placed in the middle third of the upper arm, above the antecubital fossa, in either the cephalic or basilic vein, and the distal tip of the catheter was placed in the lower third of the superior vena cava. Nurses were trained to follow aseptic precautions and follow central line-associated blood stream infection (CLABSI) prevention bundle recommended by the Centre for Disease Control and Prevention (CDC), United States of America. Patients and caregivers were educated about PICC line care at home. The patients were followed up until the removal of PICC or death.

The variables collected included patient demographics, indication for PICC insertion, comorbidities, duration of catheterization (PICC days), reasons for catheter removal, complications, PABSI rate, pathogens, and outcomes. Complications were classified as either local infection (LI) or PABSI, thrombosis, breakage, pull-out, block, and limb edema. PABSI was defined according to Central Line-Associated Blood Stream Infection (CLABSI) definition by the National Healthcare Safety Network (NHSN). Patients who had other types of CVC in addition to a PICC were excluded from analyses as it would be difficult to identify the catheter-causing infection.

The variables were compared between patients with and without PABSI. Categorical variables were compared using χ2 and medians using Mann–Whitney U test. p-Value less than 0.05 was taken as statistically significant. Variables that were found to be significant risk factors for PABSI on univariate analysis were subjected to multivariate analysis to assess the independent risk factors for PABSI. All statistical analyses were performed using IBM SPSS version 23.

Results

During the 9 years period, 266 patients had undergone PICC line insertion. Six were lost to follow-up, and seven had line in situ at the time of analysis. These 13 patients were excluded from the final analysis. Of the 253 patients who were studied, 11 had revision of PICC twice and one patient had revision thrice. As the re-insertions were not carried out on the same calendar day as of the removal of previous catheter, these were counted as new insertions. Thus, 266 PICC insertions were analyzed. PICC insertion by interventional radiologist under fluoroscopic guidance was introduced in our institution in 2018. A majority of PICC insertions 207 (77.8%) were performed between 2018 and 2021. Interventional radiologist performed the maximum insertions (n = 167, 62.5%) followed by trained nurse (n = 73, 27.6%) and anesthetist (n = 26, 9.9%). Most of the insertions (n = 194, 73%) were carried out under fluoroscopic guidance. Total PICC line days was 14,853, and the median duration was 30 days (IQR: 15–55), the longest duration being 444 days. All PICCs were single lumen. Five belonged to the pediatric age group, rest were adults. The median age was 56.5 (IQR: 43–50), and 58.3% were males.

The most common indication for PICC insertion was chemotherapy (n = 123, 46.1%), followed by prolonged antibiotic therapy (n = 110, 41.2%), while in 30 patients (11.2%) PICC was inserted in view of difficult IV access. Total parenteral nutrition (TPN) was the indication in the rest 3 patients (1.1%). Among 110 patients with indication for IV antibiotic therapy, musculoskeletal and joint infections accounted for the most (n = 42) followed by infective endocarditis (n = 26). Other sites of infection needing prolonged antibiotic therapy were complicated skin and soft tissue infection, pulmonary, brain, renal, prostate, intra-abdominal infection, and skull base osteomyelitis. Rhino-orbital/cerebral mucormycosis was the indication in 5 patients. Common comorbidities noted were diabetes (n = 164, 61.4%), hypertension (n = 123, 46.1%), and cancer (n = 115, 43.1%).

Table 1 depicts PICC-related complications. Complications occurred in 76 patients (28.5%, 4.98 per 1000 PICC days). Over the years, there was no steady increase or decrease in the overall PICC-related complications. In majority of patients (n = 192, 71.9%), PICC was removed either due to end of therapy or death of the patient. The commonest complication noted was thrombosis (n = 29, 10.9%), followed by infection either PABSI (n = 20, 7.5%) or LI (n = 2, 0.7%). Displacement of line occurred in five patients (1.9%), and accidental removal occurred in two (0.7%) patients. PABSI and LI rates were 1.34/1000 PICC days and 0.13/1000 PICC days, respectively. None of them had absolute neutrophil count (ANC) < 500 at the time of PABSI, thus ruling out mucosal barrier injury related to BSI. The average duration of PICC dwell days when PABSI occurred was 14 days. The majority (65%) of the infections occurred whilst the patient was still in the hospital.

Table 1.

PICC outcome and complications.

	PICC outcome and complications	Frequency	Percentage
1	End of therapy/patient expired (no complication)	192	71.9
2	Thrombus	29	10.9
3	Bacteremia	20	7.51
4	Exit site infection (local infection)	2	0.7
5	Fever (not proven as PABSI)	7	2.6
6	Limb edema	7	2.6
7	Line displaced	5	1.9
8	Pain at the exit site	2	0.7
9	Accidentally pulled out	2	0.7
10	Patient request for the removal	2	0.7

A majority of PABSI were due to Gram-negative rods (n = 17, 85%), commonest being Klebsiella pneumoniae (seven isolates). Other organisms isolated are shown in Table 2 and are as follows: Acinetobacter baumanni (three), Pseudomonas aeruginosa (two), and Burkholderia cepacia (two), and one isolate each of E. coli, Proteus mirabilis, and Morganella morganii were identified. Candida auris was isolated in one patient. Methicillin-resistant Staphylococcus aureus and Staphylococcus epidermidis (grown in 4 bottles) were the Gram-positive organisms (GPC) isolated. One patient with PABSI belonged to pediatric age group who had S. epidermidis in 4 blood culture bottles, rest all were adults.

Table 2.

Organisms isolated in PABSI.

Gram-negative organisms = 17
Klebsiella pneumoniae	7
Acinetobacter baumanni	3
Pseudomonas aeruginosa	2
Burkholderia cepacia	2
Morganella morganii	1
E coli	1
Proteus mirabilis	1

Factors contributing to PABSI were analyzed and depicted in Table 3. Age, gender, comorbidities, indication for PICC insertion, personnel who inserted, place of insertion, and whether performed under fluoroscopic guidance were analyzed for risk factors for PABSI. Though on univariate analysis, insertion under fluoroscopic guidance and insertion in cath lab were noted as risk factors for PABSI; on multivariate analysis none of these factors were independent risk factors for PABSI.

Table 3.

Analysis of risk factors for PABSI.

Risk factors	Patients without PABSI (n = 246) (%)	Patients with PABSI (n = 20) (%)	p-Value
Age (median, IQR)	57 (43–66)	52 (37.75–67.75)	0.60
Sex, male	140 (56.9)	15 (75.0)	0.11
Diabetes mellitus	152 (61.8)	12 (60.0)	0.87
Hypertension	115 (46.7)	8 (40.0)	0.56
Dyslipidemia	53 (21.5%)	4 (20.0)	0.87
COPD	3 (1.2)	0 (0)	0.62
Active malignancy	107 (43.5)	8 (40.0)	0.76
CCF	3 (1.2)	0 (0)	0.62
CLD	25 (10.2)	1 (5)	0.45
CKD	22 (9.2)	1 (4.5)	0.70
CAD	26 (10.6)	0 (0)	0.12
Indication for PICC line
Antibiotic administration	103 (41.9)	7 (35.0)	0.85
TPN	3 (1.2)	0 (0)
Chemotherapy	113 (45.9)	10 (50.0)
Difficult IV access	27 (11.0)	3 (15.0)
PICC line inserted by
Interventional radiologist	149 (60.6)	18 (90.0)	0.03
Trained nurse	70 (28.5)	1 (5.0)
Others	27 (11.0)	1 (5.0)
PICC line inserted at
Ward	70 (28.5)	1 (5.0)	0.01
Cath lab	176 (71.5)	19 (95.0)

PICC = peripherally inserted central catheter, PABSI = PICC-associated blood stream infection, COPD = chronic obstructive pulmonary disease, CCF = congestive cardiac failure, CLD = chronic liver diseases, CKD = chronic kidney disease, CAD = coronary artery disease, TPN = total parenteral nutrition, IV = intravenous.

Discussion

PICC has been increasingly used and has led to the recognition of PICC-related complications, more importantly infections. Better understanding of the modifiable risk factors for PABSI and implementation of corrective measures can help in the safe use of PICC for long-term venous access, both as in-patient and out-patients. In this study on PICC-related complications from India, we analyzed 266 PICC insertions and 14,853 PICC days and is one of the largest single center data on PICC-related complications. Like many other centers, major indication was for chemotherapy followed by long-term antibiotic therapy. Out-patient antibiotic therapy (OPAT) for infections is becoming an important indication for PICC insertion.

A majority of patients did not have any PICC-related complications. The overall complication rate in our study was 28.1% (4.98 per 1000 PICC days). The commonest complication noted was thrombosis (10.9%) followed by infection (8.2%), either bacteremia (PABSI = 7.5%) or local infection (LI = 0.7%). Previous reports of PICC-related complications ranged from 14.4% to as high as 38.14% (Mielke et al., 2020; Grau et al., 2017; Parás-Bravo et al., 2016). In the study by Grau et al., the commonest complications were occlusion and accidental withdrawal, 8.9% each and infections including bacteremia and local infections occurred in 6.3%. In the study conducted in an oncology unit from India, the overall complication rate was higher to the range of 37.13% (Sundriyal et al., 2014). Mechanical complications were the commonest (19.7%), and thrombosis was the commonest mechanical complication noted (13.45%). This was followed by infection which occurred in 12.5%. In another study performed in an oncology unit in India by Saptoka et al., a total of 62,440 catheter days were followed up (Sapkota et al., 2020). The overall complications noted were low, 10.2% at the rate of 0.82/1000 PICC days. Infective complications were noted in 3.4% including LI and PABSI, while thrombotic complications were noted in 3%. Patients with hematological malignancies had more complications than those with solid malignancies.

Twenty PABSIs and two exit site infections were identified, and PABSI and LI rates per 1000 days were 1.34/1000 PICC days and 0.13/1000 PICC days, respectively. These rates are comparable to previous reports. In a study from France carried out in 2010, PICC-related infection rates were 2.3/1000 PICC days (0.57 PABSI and 1.72 LI per 1000 PICC days) (Grau et al., 2017). Another study conducted in a French university hospital with 222 patients reported PABSI incidence of 0.86/1000 PICC days and LI rates of 0.64/1000 PICC days (Leroyer et al., 2013). Similar PICC-related infection rates were noted in other studies as well; PABSI was 1.14/1000 catheter days in the study by Park et al., and as low as 0.17 cases per 1000 days in Parás-Bravo et al. (Park et al., 2020; Parás-Bravo et al., 2016). The incidence of PICC-related infections was slightly higher in the study by Mollee et al. in immune compromised hosts (onco-hematological patients), and incidence of infection was 2.5/1000 PICC days (Mollee et al., 2011). PICC infection rate was 2.69/1000 days in the study by Baxi et al. In 609 patients, they found 57 cases of bacteremia and noted that immunosuppression, use of multiple lumens, and use of power PICC were associated with higher risk of PICC-related infections (Baxi et al., 2013). In our study, though the immuno-compromised onco-hematological patients contributed for many of the insertions, almost equal numbers were for long-term antibiotic therapy and difficult IV access without major immunosuppression. This, along with dedicated PICC line team for insertion and follow-up, could be the reason for less incidence of PABSI compared to the previous studies.

The average duration of PICC dwell days when PABSI occurred was 14 days, and the majority occurred in in-patients. Park et al., in his study on PICC infections, found that the average duration was 25 days for PABSI occurrence (Park et al., 2020). Chopra et al. in their large study noted that the median time for PICC infection was 10 days which is shorter (Chopra et al., 2014). In many other studies, it was noted that hospitalized patients were more at risk of CLABSI, likely due to frequent handling of the vascular device for medication administration. In a meta-analysis of 57,250 patients from 23 studies by Chopra et al., 20 studies reported CLABSI episodes in patients with PICCs and CVCs (Chopra et al., 2013). The study found that CLABSI is commoner than PABSI (5.8% versus 5.2%), more so in hospitalized, thus mirroring our findings. Among outpatients, the risk of PABSI was low and was 0.5% versus 2.1% CLABSI. Pooled meta-analysis of these studies revealed that PICCs were associated with a lower risk of CLABSI than for CVC (relative risk [RR], 0.62; 95% confidence interval [CI], 0.40–0.94), and subgroup analysis demonstrated that CLABSI reduction was greatest in outpatients (RR [95% CI], 0.22 [0.18–0.27]) compared with hospitalized patients who received PICC (RR [95% CI], 0.73 [0.54–0.98]) (Chopra et al., 2013). But only 13 of the included 23 studies reported CLABSI per catheter days. Among these studies, PABSI occurred as frequently as CLABSI (incidence rate ratio [95% CI], 0.91 [0.46–1.79]) (Chopra et al., 2013). Again, in the study by Grau et al., PICC complication rate was higher in the hospitalization setting (36.1%) than in the outpatient setting (19.4%) (Grau et al., 2017). This is also consistent with observations by Maki et al., who noted that bloodstream infections occurred less frequently (0.2 per 1000 catheter days) in the outpatient setting than in the hospitalized patients (1.2 per 1000 catheter days) (Maki et al., 2006). Thus, the overall risk of PABSI is less in the outpatient setting.

A majority of PABSIs were due to Gram-negative rods (85%) and depicted in Table 2. The predominance of GNB causing PABSI is in contrast to previous studies which noted that GPCs were common. In a large retrospective study by Chopra et al., 58 PABSIs were noted; GPC (65%) was the commonest, (CONS 40%, enterococcus 14%, and S. aureus 7%), and GNB was noted in 24% of infections, Pseudomonas spp being the commonest (Chopra et al., 2014). Candida spp. was the etiology in one-fifth, C. albicans and C. glabrata being the commonest organisms (Mollee et al., 2011). Park et al., in his study carried out in Korea, also found that the GPC accounted for most of the PABSIs (41.7%) followed by Candida (37%); non-albicans Candida were commoner (Park et al., 2020). Data on PABSI from India are lacking. Among earlier reports of CLABSI from India, a majority were CONS (Chopdekar et al., 2011). In the study by Sundriyal et al. carried out in medical oncology unit over 1 year period, the bacteremia rate was 12.5%. Commonly identified organisms were C albicans, K pneumoniae, and S aureus. GNB accounted for 46.4%, GPC accounted for 28.6%, and the rest were candida (Sundriyal et al., 2014). The high number of vascular infections due to GNB noted in our study is in tune with recent study from India reporting infections related to CVC were predominantly due to Gram-negative organisms and may reflect the increasing burden of Gram-negative infections in this part of the world (Singhal et al., 2019).

Chopra et al. in their study looked at risk factors for PABSI (Chopra et al., 2014); on bivariate analysis, predictors for PABSI noted were intensive care unit (ICU) status, mechanical ventilation, length of stay, use of power PICCs, number of catheter lumens (triple lumen carries higher risk of PABSI than double lumen), and devices placed by interventional radiology. On multivariable analysis, only hospital length of stay, ICU status, and number of PICC lumens remained significantly associated with PABSI. More importantly, earlier time to infection among patients with multi-lumen PICCs was noted (Chopra et al., 2014). In our study, predictors for PABSI were analyzed and depicted in Table 3. Though on univariate analysis insertion under fluoroscopic guidance and catheter insertion carried out in cath-lab were noted as risk factors for PABSI, on multivariate analysis none of these factors were independent risk factors for PABSI.

While our study provides valuable insights into PICC-related practices and complications in the developing world, it is important to acknowledge its limitations. One major limitation is its retrospective design, which may have introduced bias and limited our ability to draw causal conclusions. Additionally, the findings from a single center may not be generalizable to other settings, and caution should be exercised when extrapolating the results to other populations. To further advance our knowledge on this topic, it is essential to conduct multicentric prospective studies which would allow for a more comprehensive assessment of PICC-related complications and provide a more robust evidence base for clinical practice.

Conclusion

PABSI noted in this study was 1.34 per 1000 catheter days, which is comparable with other published studies. The average dwell time of PICC for occurrence of PABSI was 14 days. A majority of infections occurred in the hospitalized patients and were due to Gram-negative rods. PICC will remain a good choice for patients needing long-term venous access and can be used in ambulatory care with less risk of infections if supported by a trained PICC insertion and maintenance team.

ORCID iD

Vettakkara Kandy Muhammed Niyas https://orcid.org/0000-0002-7255-6257