Abstract
Introduction
Central line associated bloodstream infection (CLABSI) is a common complication of venous access. Rates of CLABSI are reportedly higher in middle-income countries compared with high-income countries. There is a paucity of data concerning CLABSI in South Africa. In this study we aimed to assess the prevalence of CLABSI and its associated factors at Dora Nginza Hospital.
Methods
We conducted a retrospective review of medical and laboratory records of all paediatric intensive care unit (PICU) patients who had central lines placed between January 2019 to December 2019. The CLABSI was defined according to Centers for Disease Control and Prevention guidelines. Eligible patients had to have had a central line for two days prior to CLABSI. Information concerning the type of the central line, associated factors, septic markers, and the organisms that were cultured was obtained from the records. Ethics approval was granted by the Institutional Review Board of Walter Sisulu University.
Results
In all, 289 patients were admitted in PICU during the period of the study, and 131 had central lines. The files of 107 children were used for the study with a total of 598 catheter days. Among 107 participants 19 CLABSI were identified yielding a CLABSI rate of 31.7 per 1000 catheter days. Patients with CLABSI had more central line days than those who did not develop CLABSI (p<0.01). Blood transfusion was also a risk factor for CLABSI in this study. Other known risk factors like antibiotics, catheter site, total parenteral nutrition, malnutrition were not found to be significantly associated in this study. Causes of CLABSI were Gram-negative (n=10, 52%) being predominant, followed by fungal at (26%, n=5) and Gram-positive (n=3, 15.8%).
Conclusion
The prevalence of CLABSI at Dora Nginza Hospital is high compared to rates reported in middle-income countries. Strict adherence to prevention bundles is required reduce CLABSI prevalence.
Introduction
Central line catheters are essential vascular access devices that are commonly used in critically ill patients. The use of central lines in intensive care is often necessary for patients requiring multiple infusions such as intravenous fluids, blood products, antibiotics, and total parenteral nutrition (TPN).1, 2 However, the use of central lines is not without risks, as they can lead to bloodstream infections.1
World-wide, central line associated bloodstream infections were the commonest cause of hospital acquired infections in 2001, with a reported mortality rate of 12-15%.3 Central line associated bloodstream infection (CLABSI) can lead to high mortality, morbidity, increased length of stay and increased hospital costs in intensive care unit.4
The Centers for Disease Control and Prevention (CDC) defines CLABSI as “a laboratory confirmed bloodstream infection that occurs when an eligible bloodstream infection organism is identified and an eligible central line is present when the bloodstream is laboratory confirmed”.5 An eligible central line is defined as a central line that has been in place for more than two consecutive calendar days. An eligible pathogen is defined as any pathogen that can be used to meet laboratory confirmed bloodstream infection criteria. The infection cannot be related to any other infection the patient might have and must not have been present or incubating when the patient was admitted to the facility.6
With good aseptic techniques, most cases of central line-associated bloodstream infections can be minimised.7 Introduction of CLABSI prevention guidelines reduced the incidence of CLABSI between 2001 and 2009 by 58%.1 Care bundles are a group of interventions used for patients with intravascular central catheters, that when implemented together, result in better outcomes than when implemented individually. Central line bundle elements include hand hygiene, maximal barrier protection, chlorhexidine skin antisepsis, optimal catheter site selection, and daily review of need for the central line.8
Data on CLABSI from low- and middle-income countries are limited,2 but it is reported that in low-income countries CLABSI rates are 3-5 times higher than in high-income settings.9 In South Africa (an upper middle-income country), the CLABSI rate in a tertiary neonatal intensive care unit was 5.9 per 1000-line days,2 and this was similar to that noted in some low- and middle-income countries, but higher than CLABSI rates in high-income settings. The rate of CLABSI noted in Iran (upper middle-income) was 10 per 1000 catheter days.10 In the United States of America, CLABSI rate in intensive care units is estimated to be 0.8 per 1000 central line days.11
The risk factors of CLABSI may be classified as host-related and catheter-related. Host related risk factors include malnutrition,12 leukopenia,13 administration of blood products,14 and TPN,15 haematological malignancies,12 prolonged use of broad-spectrum antibiotics,16 stem-cell transplantation,17 and male gender.18
Catheter related risk factors encompass types of catheters, including non-tunnelled, and multi-lumen catheters, as these carry a greater risk of CLABSI than tunnelled ones.19 Femoral catheterisation is associated with a higher risk of infection than internal jugular and subclavian access.20 The number of days since central line insertion is also a predictor for CLABSI.21
There is a paucity of data on CLABSI in critically ill children from low- and middle-income countries, particularly from Africa.2 Knowledge of the CLABSI rate in poorly resourced settings and of common organisms causing CLABSI and their sensitivity patterns will help improve management of central lines and assist in choosing the appropriate antibiotics to treat CLABSI. This will likely assist in reducing morbidity and mortality associated with CLABSI in these settings. This study aimed to determine the prevalence, aetiology and outcomes of CLABSI in children admitted to a paediatric intensive care unit (PICU) in the Eastern Cape Province of the Republic of South Africa.
Materials and Methods
A descriptive study, cross-sectional and retrospective review of records of patients who had central lines in the PICU from January 2019 to December 2019 was conducted at Dora Nginza Hospital. The hospital is based in Gqeberha (formerly Port Elizabeth), in the Eastern Cape province (the poorest province in the Republic of South Africa). The hospital is the only public health facility offering regional and tertiary level paediatric services for children in the Western Region of the Eastern Cape, which has a population of approximately 2 million people. The PICU potentially has 16 beds, but only 6 beds are functional due to a nursing staff shortage. The nurse-to-patient ratio in the unit is typically 1:2. All children with medical or surgical conditions that require intensive care are admitted to the unit. The ages of patients eligible for admission into the PICU range from one week to 12 years. There is one pediatric intensivist who supervises the medical officers and pediatric residents rotating in the PICU. Central lines are usually inserted by the medical officer/resident rotating in the unit. The indications for central line insertion include venous access for blood transfusion, administration of TPN, and inotropes.
All children with an eligible central line (two consecutive days) with a recovery of an eligible pathogen from blood culture by the time of infection or within 24 hours after the removal of the central line were enrolled. Children who did not meet the criteria of CLABSI, and those whose blood cultures were assessed as contaminants were excluded. Children who died or were discharged within 48 hours of PICU admission were also excluded.
The study was time-bound, but a priori sample size calculations indicated that at least 40 participants were required.
Paediatric intensive care files from January 2019 to December 2019 that met the inclusion criteria were evaluated by the principal researcher. The procedure included scrutiny of the PICU register to note the number of admissions for each month, the number of admissions who had central lines, and how many of them developed infections . The files were then retrieved from records storage, and demographic data (age and gender) and information on the indications of the central line, type of the central line, associated factors, septic markers, and the organisms that were cultured were reviewed by the researcher.
Known risk factors of CLABSI, underlying diagnoses, complications, antibiotics used, the duration of antibiotics, and the clinical outcome of each patient were identified. A data collection sheet was used for each patient.
The chi-square test was used to test for the association between organisms and complications, and for assessing association between demographic data, complications, antibiotics administered, and clinical outcomes. P-value of less than 0.5 was statistically significant.
Ethical clearance was obtained from the Institutional Review Board of Walter Sisulu University (certificate number: 009/2022, date: 30.03.2022).
Statistical Analysis
Data were analysed using the Statistical Package for the Social Sciences version 29.0 software. Data were assessed for normality by visual examination of histograms, followed by normality tests. Continuous data (including age and duration of central line) were described using frequencies and means with standard deviations if normally distributed, median with interquartile range was used for data that are not normally distributed. Frequencies were used to describe categorical data. Prevalence and common pathogens were described using frequencies.
Results
During the 12-month study period, 289 children were admitted to the PICU. One hundred and thirty-one patients had central lines, but 24 of them were excluded. Therefore, 107 patient files were fully analysed for the study, and 19 of them had central line-associated bloodstream infection. The assessment for eligibility and enrolment procedures is summarised in Figure 1.
Demographics, Human Immunodeficiency Virus (HIV), and Nutritional Status of the Patients
The gender distribution of the 107 children was 63 (58.9%) males; and the median age was 5.3 months. In terms of age distribution, more than half (52.3%) of the children were 6 months of age or younger and 18.7% were older than 5 years of age (60 months). Forty-two percent of children were between 1-6 months, 18% were above 60 months, and the remaining 40% were between 6 and 60 months.
The HIV status of all the children was known, and 8 (7.5%) were HIV positive with the remaining 99 (92.5%) being HIV negative. Seventeen percent of the children were malnourished, 13.1% (n=14) were moderately malnourished, 4.7% (n=5) were severely malnourished, and 82% (n=88) were well nourished. The demographic data is summarised in Table 1.
Prevalence of CLABSI
Out of the 107 children who had central lines, 19 (17.8%) developed central line associated bloodstream infections; the rest did not. There were 19 patients with CLABSI and 598 catheter days. The overall CLABSI rate was calculated to be 31.7 per 1000 catheter days.
Admission Diagnosis of Study Population and of Children with CLABSI
The commonest admission diagnoses among the 107 enrolled children were pneumonia (30 children), congenital heart diseases (13 children), acute gastroenteritis (12 children), sepsis (11 children) and surgical conditions (10 children). The remaining 31 children had other conditions. Among the 19 children who developed CLABSI, the most common diagnoses were pneumonia (6, 32% of children), congenital heart diseases (3, 16%) and surgical conditions (3, 16%).
Risk Factors for CLABSI
Virtually all enrolled children (106/107) had non-tunneled central line catheters. All children who developed CLABSI had tunnelled central line catheters inserted in the femoral vein. There were 86 (97%) children who had femoral catheters, and 19 of these 86 children developed CLABSI (22%). With regard to duration of catheter days, 73 (83%) of the 88 children who did not have CLABSI had their central lines in situ for less than 7 days, while 9 (47%) of those who developed CLABSI had central lines in situ for less than 7 days. Findings related to catheter line days and other risk factors for CLABSI are summarised in Figure 2 and Table 2, respectively.
There were significantly more children with CLABSI who had central lines for more than 7 days (52.6%) compared to those children who did not develop CLABSI (16%), p<0.01. Eleven out of 19 patients who had CLABSI received a blood transfusion (57%), p<0.01.
All 19 patients who had CLABSI received antibiotics before the diagnosis of CLABSI, as opposed to 68 (77%) of those patients who received antibiotics on admission but did not develop CLABSI; however, this was not significant. There was also no difference in the prevalence of malnutrition and administration of TPN, between those who developed CLABSI and those who did not.
Laboratory Investigations on Non-CLABSI vs. CLABSI Patients
Most children with CLABSI (94.7%) had markedly or moderately elevated C-reactive proteins (CRPs). Only 5.3% had normal CRPs at the time of CLABSI diagnosis, and this was significantly different from children who did not have CLABSI. Fifty-two percent of all children with CLABSI had normal white cell counts. Of these, only 26.3% had low platelet counts, while 73.7% had normal or high platelet counts. This is shown in Table 2.
Organisms Causing CLABSI and Their Sensitivity
Out of the 19 children who developed CLABSI, 10 (52.6%) had Gram-negative bacilli responsible for the CLABSI. The Gram-negative bacilli were noted as follows: Five (26.3% of the CLABSI population) children had multidrug-resistant Escherichia coli (E. coli), which was sensitive to ertapenem and gentamicin; another two (10.5%) had Klebsiella pneumoniae, which was sensitive to meropenem, ertapenem, and gentamicin; and one (5%) who had Chryseobacterium gleum that was sensitive to ciprofloxacin.
The second most common organisms causing CLABSI were fungal infections, with 5 (26.3%) children having the following fungal infections: Candida albicans (two children), which was sensitive to fluconazole and amphotericin B; Candida glabrata (one child), which was only sensitive to amphotericin B; one child had Candida parapsilosis, which was sensitive to amphotericin B; and one child had Candida tropicalis (5% of CLABSI), sensitive to both fluconazole and amphotericin B.
The least common cause of CLABSI was Gram-positive organisms; 3 children (15.8% of CLABSI), and these were two children with Staphylococcus epidermidis (10%) and one with Enterococcus faecium (5%), both sensitive only to vancomycin (see Table 3).
Complications of CLABSI According to the Cultured Organisms
Out of 19 children who had CLABSI, 11 (59.9%) had septic shock as a complication. There was no evidence of other CLABSI complications. Fifty percent (n=5) of all the Gram-negative organisms had septic shock and needed inotropes, and the cultured organism was Acinetobacter baumannii in 80% of these cases. Thirty-three percent of the children in whom a Gram-positive organism was cultured developed septic shock, and the organism was Staphylococcus epidermidis in all cases. With respect to fungal infections, 5 (83%), children developed septic shock. Cerebrospinal fluid, hearing, eye, and bone assessments were not looked into in this study.
CLABSI Outcomes
Out of 19 children with CLABSI, 4 (21.1%) died, and all of them had a fungal infection. The fungal organisms cultured were Candida albicans (two children), Candida glabrata (one child), and Candida tropicalis (one child). The catheters were removed from all patients suspected to have CLABSI before organism identification. There was no use of antibiotic lock therapy for any of the patients due to its unavailability in the institution.
All 10 patients who had Gram-negative organisms cultured survived. Additionally, 3 patients who had Gram-positive organisms survived and were discharged from the intensive care unit. However, only two of the six children (33%) who had fungal infections survived.
Discussion
We conducted a descriptive study to determine the prevalence (rate) of CLABSI, identify common pathogens and sensitivity patterns, identify complications of CLABSI, and evaluate factors associated with CLABSI.
Out of the 289 patients admitted in PICU over the duration of the current study, only 131 (45%) had central lines inserted. The rate of central line insertion is lower than that in other intensive care units. A study conducted in the United States of America (USA) revealed that 50-80% of critically ill patients require central lines at some point during their stay in the intensive care unit (ICU). The lower rate may be because there is no paediatric high care unit at Dora Nginza Hospital. Consequently, some high care level patients are admitted to the ICU and these patients only need non-invasive ventilation and may not require central line insertion.
The ages of patients admitted to PICU in this study ranged between 0.1 months and 5 years of age. Young age has been shown to be a risk for developing CLABSI.23 This was also observed in a Swiss study, where children between the ages of 2 months and 5 years had high incidence rates of CLABSI.24 In the current study, 84% of the patients who developed CLABSI were younger than 5 years of age; this was in keeping with the above studies.
In the current study, there were only eight patients who were HIV positive (7.5% of the study population), and this was unlike the findings noted in an American study where HIV positive patients had increased risk of developing CLABSI compared with HIV negative patients.25 Children with malnutrition are reported to have an increased risk of developing CLABSI,23 but this was not the case in the current study. Children who are severely malnourished are not eligible for admission into the Dora Nginza Hospital PICU, and this probably reduced the expected effect of malnutrition on the development of CLABSI.
The study found a CLABSI rate of 31.7 per 1000 catheter days in a PICU, where both medical and surgical patients are admitted. This rate is higher than most studies that were previously done in Mumbai, India. For example, the rate was 4.3/1000-line days.26 In a study conducted in Cape Town, South Africa, the rate was 5.9/1000-line days,2 and in one conducted in Bloemfontein, South Africa, the rate was 26.3/1000-line days.27 These two studies were conducted in South African provinces that are better resourced than the Eastern Cape province. The rate noted in the current study is higher than what is reported in low- and middle-income countries, as CLABSI rates are reported to be 3-4 times higher than those in high-income countries.28 In Egypt (a middle-income country), CLABSI rates are reported to range between 2.9 and 14.3 per 1000 catheter days, with an overall rate of 9 per 1000 catheter days,29 while in India (a middle-income country), the overall CLABSI rate was 17 per 1000 catheter days, which is 16-fold higher than that in the USA.17 The rate of CLABSI noted in another middle-income country Iran, was 10 per 1000 catheter days.10
The possible explanation for very high CLABSI rates in our institution might be attributed to poor adherence to CLABSI reduction care bundles. The non-adherence might be due to a shortage of staff in the PICU, but the reasons for non-adherence were not assessed in the current study.
Males had a higher percentage of CLABSI compared to females, and this was in keeping with other observed findings where there was male predominance in patients with CLABSI.30 Age distribution for children with CLABSI was not different from those who did not develop CLABSI in the current study, and this was similar to studies conducted in Türkiye and India.17, 31 where there was no difference in age and sex between children who had CLABSI and those who did not develop it.
The common PICU diagnoses were acute respiratory conditions, congenital heart disease, sepsis, surgical patients post-operatively, trauma, and others in that order. This was in keeping with a study done in Australia where most indications of admission were respiratory; cardiac; surgical; trauma; sepsis, etc., but not necessarily in the same order.32
With regard to risk factors, the known risk factors associated withfor CLABSI in the current study were blood transfusion and long duration of the central line. Blood transfusion is known to be a risk factor for development of CLABSI.14 TPN and malnutrition are known to be independent risk factors for development of CLABSI.12, 15 However, in the current study, they were not significantly associated with the development of CLABSI.
All children who developed CLABSI had received broad spectrum antibiotics prior to developing CLABSI. However, there was no significant association when compared to the children who had also received broad spectrum antibiotics but did not develop CLABSI. This was not in keeping with studies conducted in India,17 America,11 and Japan,33 which reported that the use of broad spectrum antibiotics is a risk for developing CLABSI.
Risk factors related to the catheter site and the type of catheter inserted could not be compared for risk assessment because almost all (99%) study children had non-tunnelled central line catheters, Furthermore, out of all the non-tunnelled catheters, only one child had a subclavian central line; the rest, had femoral catheters. It is known that non-tunnelled catheters carry a greater risk of CLABSI than tunnelled,19 Femoral catheterisation is associated with a higher risk of CLABSI than internal jugular and subclavian access.20
Central line days since insertion is also a predictor for CLABSI, and this was also true in the current study. There were significantly more patients who had the lines in situ for more than 7 days among children with CLABSI compared with those who did not develop CLABSI. This was in keeping with a systematic review and meta-analysis which showed that prolonged catheterisation is a major risk factor for developing CLABSI.34
Ninety-four percent of all the patients with CLABSI had markedly and moderately elevated CRP levels, and only 5% had normal CRP levels at the time of CLABSI diagnosis. Only 47% of the CLABSI had elevated white cell count, while 53% had normal white cell count. CRP is an inflammatory protein used in clinical practice to guide antibiotic therapy, and the discrepancy between its increase and normal WCC in the current study might be due to the prior use of antibiotics before the diagnosis of CLABSI. This was not in keeping with a study done in Japan where patients with CLABSI showed a low white cell count and CRP compared to bloodstream infections without central line catheter.35
The most common causes of CLABSI in the current study were Gram-negative organisms like multi-resistant Acinetobacter baumannii, E. coli and Klebsiella pneumoniae.. Fungal infections were the second most common cause, and these were Candida albicans, Candida glabrata, Candida parapsilosis, and Candida tropicalis. The least common cause was Gram-positive organisms such as Staphylococcus epidermidis and Enterococcus faecium.
These findings were not consistent with findings from the CDC, where Gram-positive organisms like Coagulase negative staphylococci, Enterococci, and Staphylococcus aureus were the common causes of CLABSI, followed by Gram-negative organisms like Klebsiella, Enterobacter, Pseudomonas, E. coli, and Acinetobacter, with Candida species being the least common. It was however, in keeping with South African studies where Gram-negative organisms were found to be a common cause of bloodstream infections and CLABSI rather than Gram-positive organisms.2, 27 Multi-resistant Acinetobacter baumannii, requires that there be an emphasis on infection prevention measures and on strengthening antibiotic stewardship.
Complications of CLABSI include port pocket infections, endovascular infections such as endocarditis and thrombophlebitis, and metastatic infections such as osteomyelitis and liver, spleen, and brain abscesses. The only complication of CLABSI recorded in the current study was septic shock, and there was no evidence of other complications during the clinical examination.
Catheter tip cultures were not done routinely in all the patients who had central lines; the positive tips did not help in the diagnosis of CLABSI. Although 63% of the patients had positive catheter tips, with negative blood cultures. Peterson and Smith36 did not find any clinical significance of central line catheter tip cultures in the diagnosis of CLABSI. Huang et al.37 noted that central line catheter cultures had no impact on antimicrobial therapy for CLABSI.
The CLABSI mortality rate was 21% in the current study. The mortality rate for CLABSI is estimated to be 20-25% globally by the CDC, and 12-15% in Southern California.38 The cause of death among all the CLABSI patients was fungal infection. Fifty percent of them were HIV positive and malnourished, adding more risk of mortality, and the other two had cardiac issues. The cause for fungal CLABSI mortality is due to late diagnosis as most patients who deteriorate in the ICU are empirically commenced on antibiotic treatment for Gram-negative and Gram-positive organisms.
Study Limitations
This was a retrospective record review study in a hospital with no electronic filing system for patients, and approximately one fifth of the hospital files were either missing or had insufficient information. The findings of the study may not be generalizable to hospitals in well-resourced settings and may be compared to similarly resourced settings. Dora Nginza PICU only admits children 12 years, but in the studies used for comparison, the ages of participants were up to 18 years, and this should be considered a limitation.
Conclusion
This study found a very high CLABSI incidence of 31/1000 line days at Dora Nginza PICU; blood transfusion and long catheter line days were significant risk factors. The multi-drug resistant Acinetobacter baumannii was the most common cause of CLABSI, and fungal infection was associated with high mortality. Further studies focusing on fungal causes of CLABSI are recommended.
