1.gif (1892 bytes)

Brief Reports

Indian Pediatrics 2002; 39:752-760

Central Venous Catheter Colonization and Related Bacteremia in Pediatric Intensive Care Unit

Anil Sachdev
Dhiren Gupta
Arun Soni
K. Chugh

From the Department of Pediatrics, Sir Ganga Ram Hospital, Rajinder Nagar, New Delhi 110 060, India.

Correspondence to: Dr. Anil Sachdev, 63/12 Old Rajinder Nagar, New Delhi 110 060, India.

E-mail: [email protected]

Manuscript received: March 15, 2001;

Initial review completed: April 24, 2001;

Revision accepted: February 4, 2002.

Central venous catheters (CVC) are employed to provide reliable and readily available venous access for administration of intravenous fluids, drugs, parenteral nutrition, blood products and for monitoring of central venous pressure (CVP)(1). Besides technical problems associated with insertion of CVC in pediatric patients, catheter related bacteremia (CRB) is a major and common clinical problem particularly in critically ill children requiring monitoring or prolonged nutritional support(2). Catheter related bacteremia rates of up to 43% are reported(3).

Confirmation of diagnosis of CRB involves removal of CVC(4) or catheter exchange over a guidewire(5). In pediatric patients however, these are not always practical(2) and cost effective. Moreover insertion of pediatric central venous catheters is technically more difficult than in adults. So there is a need to recognize certain clinical risk factors which help the clinician to predict CRB without removal of central line.

This retrospective study was conducted to determine (a) the incidence of CVC colonization and related bacteremia, (b) the microorganisms responsible for such infection, and (c) the clinical risk factors associated with positive catheter cultures and related bacteremia in a multidisciplinary pediatric intensive care unit (PICU).

Subject and Methods

Medical case records of seventy five consecutive children admitted in seven bedded multidisciplinary pediatric intensive care unit from July 97 to Dec. 99 and who had central venous cannulation were analyzed.

Patient data: Patients’ clinical and laboratory data including total and differential leukocyte counts, coagulation profile, arterial blood gas analysis and blood culture (blood culture A) at the time of admission were recorded. The median age of patients was 1.5 years (range 1 day to 14 years) with male predominance (M: F: 62:13). Forty-nine patients were febrile (>38º C) at the time of admission while clinical shock and respiratory distress were present in 34 (45.3%) and 43 (57.3%) patients respectively. Acidemia (PH <7.30), disseminated intravascular coagulation, abnormal total leukocyte count and neutrophils were recorded in 23 (30.6%), 22 (29.3%), 48 (64%) and 43 (57.3%) patients respectively. The primary clinical diagnosis in the study group were meningoencephalitis (n = 18), pneumonia or bronchopneumonia (n = 7), septicemia (n=5) and meconium aspiration syndrome (n=5). Infective polyneuritis, hypoxic-ischemic encehalo-pathy, subdural empyema, liver abscess, congenital gastrointestinal defects, aspiration pneumonia, acute renal failure and malignant disorders were present in three patients each. Degenerative brain disease, empyema thoracis, duodenal perforation and skin burns were present in two patients each. There was one patient each of the following – abdominal tuberculosis, acute gastroenteritis, dengue fever, cystic fibrosis, congenital lobar emphysema, severe laryngomalacia, acute asthma and drowning.

Reports of blood culture samples (blood culture B) sent during PICU stay from different site and not through the central line were also recorded. Positive culture of other specimens like urine, cerebrospinal fluid, tracheal aspirates, bone marrow and peritoneal fluid if done were also recorded to find out other sources of infection.

Patient’s clinical data including fever > 38ºC, total and differential leukocyte count done at the time of removal of catheter or within the last 24 hours were recorded. Blood sample drawn from a distant peripheral site was sent for culture (blood culture C) in all patients at the time of removal of central catheter. In case of death of a patient with central line in situ peripheral blood culture and leukocyte count done in the last 24 hours were recorded for analysis. Reference tables(6) were used to define normal or abnormal hematological parameters.

Central Venous Catheter (CVC) Insertion, Maintenance and Removal

Central venous catheter (20-22 G diameter and 8-12 cm length) made of polyurethane material manufactured by Arrow (Arrow International Inc, Reading, Philadelphia), Viggo (Ohmed, Swindon, UK) and Vygon (BP7,95400, Ecouen, France) were used. Femoral and subclavian veins were cannulated percutaneously using standard Seldinger’s technique. For peripherally inserted central catheter (PICC), Nutricath (Vygon) made of silicon was used. Length of this catheter varied from 30 to 35 cm and diameter from 18 to 20 G. The cubital vein was cannulated with a short cannula-over-needle provided along with the silicon catheter. This long catheter was introduced through the short cannula after removing the needle. No CVC was inserted by the cutdown method. Pediatric intensivist at the bedside in the PICU using standard aseptic precautions (gown, face mask and gloves) performed all catheterizations. The selected site was cleaned with povidone iodine and sterile drapes were used before infiltrating 2% xylocaine at the entry point. All catheters were secured with two sutures. After successful insertion, the entry site was cleaned with hydrogen peroxide. Povidone iodine ointment was applied at the entry point and dry gauze and nontransparent elastic adhesive bandage was used. Femoral catheters were covered with a polythene sheet after dressing to prevent soiling from urine and stool. Extension tubing of 10 cm length with an attached three way connector was connected (Mediflexo, Eastern Medikit Ltd, Gurgaon, Haryana) to the central line hub to increase the number of working ports. CVC were used for multiple purposes including IV therapy, inotropes, parenteral nutrition, blood or blood product transfusion, CVP monitoring, drug and antibiotic therapy depending on the patient’s requirement. IV administration sets were changed every 48 hours. All central lines were examined for signs of local inflammation and were dressed using antiseptic ointment on alternate day by a senior resident or consultant intensivist. No catheter was exchanged over a guidewire and tunneling procedure was not used.

Catheter data including indication of insertion and removal, day of insertion after admission, duration of catheter in situ, site of insertion and type of catheter were recorded. Paucity of peripheral veins was the main indication for insertion of CVC (n = 41; 54.7%) followed by the need for multiple infusion sites (n = 24; 32%) and CVP monitoring (n = 10; 13.3%). Mechanical block of central catheter and local entry site problems (nonselective removal) were the indications for removal in 37(49.3%) and 7(9.3%) patients respectively. CVC was selectively removed in 31 patients because of death, patient leaving against medical advise or discharge on request or when the CVC was no longer needed.

Catheter colonization and Catheter related bactermia were defined as described by Brun -Buisson et al.(7).

Catheter Colonization: Significant catheter colonization was taken as a quantitative culture of catheter tip showing > 103 organisms /mL.

Catheter Related bacteremia (CRB): Defined as simultaneous catheter tip colonization showing 103 organisms/mL and peripheral blood culture positive for same organism, in the absence of secondary source of sepsis.

Statistical analysis. The rates of catheter colonization and CRB were calculated as proportions and compared using Chi-square or Fisher’s exact tests wherever appropriate. Means were compared using student’s t test. Association of risk factors individually and in combination with CVC colonization and CRB was assessed by univariate analysis and odd ratios (confidence intervals) were estimated(8). Kaplan-Meier survival analysis was performed to study the median life span of catheter placement. P value < 0.05 was considered statistically significant.

Results

Clinical and Central venous catheter characteristics

Each patient had only one CVC during his stay. Forty-eight (64%) patients had CVC inserted in the first 3 days and only 5 (6.6%) had the insertion done after 2 weeks of admission. Sixty (80%) cases had a femoral cannulation. Upper extremity central venous sites were used in rest of the patients. Subclavian vein was cannulated in 7 cases while PICC was inserted in the cubital vein in eight patients. Single lumen catheter was used in 62 (82.6%) patients while 4 had a double lumen and only one case (of acute leukemia that had bone marrow transplantation) received a triple lumen catheter in his subclavian vein. Parenteral nutrition was given to 17 children during their stay in the PICU but that was not the primary indication for insertion of CVC. No major complications were encountered at the time of insertion of the central catheter except in one case of disseminated intravascular coagulation who had more than usual bleeding.

Only 10 (13.3%) catheters were removed within 6 days of insertion while in 24 cases central lines remained in situ for more than 14 days. Mean duration of CVC in situ was 12.78 ± 6.52 days (mean ± SD) (range 1.5 to 30 days). Total catheter days were 958 days. Kaplan-Meier survival analysis revealed that overall median life span was 11.00 ± 1.00 days (SE) while in patients with selective removal it was 10.00 ± l.67 days and in nonselective removal group it was 11.00 ± 1.47 days. Fever (>38ºC) was recorded in 23 (30.6%) patients at the time of removal of central lines. Abnormal total leukocyte count and neutrophil count were present in 27 (36%) and 37 (49.3%) patients respectively. Only 5 patients had leucopenia with neutropenia and rest had leucocytosis with neutrophilia. Local signs of inflammation were noticed in only 7 patients (only erythema 4, induration and discharge 1 and erythema with induration 2).

Microbiological Results

The details of micro-organisms isolated from blood cultures at various time points, and CVC tips are summarized in Table 1.

Table I-Details of Microorganism Isolated from Blood and Catheter Tip Culture
	
 
Blood culture A
n = 75)*
Blood culture B
(n = 66)**
CVC tip
(n = 75)***
Blood culture C
(n = 75)***
Positivity
8 (10.6%)
15 (22.7%)
35 (46.7%)
15 (20.0%)
No of Isolates
9
22
41
23
Type of Isolates
S. aureus
6
8
6
4
E.coli
2
4
3
Coagulase - ve Staphylococci
1
4
Pseudomonas sps
2
1
2
2
Enterobacter
1
2
3
3
Klebseilla sps
1
2
Strep. faecalis
2
Candida sps
7
18
11

Blood culture A: at admission, Blood culture B: subsequent sample from distant site while central venous 
catheter in site, Blood culture C: sample taken from peripheral site at the time of removal of CVC 
(central venous catheter)
* Two isolates in one patient. ** Same organism in CVC, blood culture B and C in only 7 patients.
 *** Mixed organism in 7 CVC and 6 blood culture C.

 

Medical records of patients who had catheter related bacteremias were also scrutinized for other sources of infection. Only 4 patients had grown the same organism from different sources. One case of pyogenic meningitis had Candida growth from the bone marrow and purulent discharge from the CVC entry site. Other 2 patients had Candida sps and one patient had grown E.coli from urine. Thirty-seven (49.3%) patients were mechanically ventilated for varying period of time during their stay in PICU. Only one patient had grown the same organisms (E. coli and Candida sps) from the tracheal aspirate, CVC tip and peripheral blood culture.

Insertion Site and Duration of Catheter Use

Half of femoral catheter tips were colonized and 21.6% was the CRB rate. Compared with CVC inserted in the upper extremity (subclavian vein and cubital vein), difference of colonization and CRB rates were not statistically significant ( P = 0.2 and 0.47 respectively).

Duration of catheterization had significant influence on catheter colonization and related bacteremia (Table II). In this study the catheter tip was colonized earliest on day 5 and related bacteremia was detected on day 7 of insertion. Mean ± SD duration of sterile catheter was 11.5 ± 6.21 days (range 1.5-30 days), as compared to that of colonized catheter tips (13.9 ± 6.79 days; range 5-30 days) (P > 0.05) and of catheter related bacteremia (17.4 ± 6.26; range 7-30 days) (P< 0.001).

Table II- Effect of Duration of Catherization on Cathether Colonization and Related Bacteremia
Duration
(days)
CVC
colonization*
Blood culture
positive**
1–5 (n = 6)
1 (16.7%)
0
6–10 (n = 29)
11 (37.9%)
2 (6.9%)
11–15 (n = 19)
11 (57.9%)
4 (21%)
>15 (n = 21)
12 (57.1%)
9 (42.8%)
* P <0.05, ** P <0.001

 

Analysis of Risk Factors

Association of four risk factors namely fever (>38ºC), total leukocyte count, neutrophil count, and use of parenteral nutrition with catheter colonization and related bacteremia was studied at the time of CVC removal (Table III). Patients were categorized according to the number of risk factors present. Statistical analysis revealed that progressive increase in the number of risk factors was associated with increased risk of CVC colonization and related bacteremia. (Table IV).

Table III-Association of Risk Factors with Central Venous Catheter Colonization and 
Related Bacteremia
Risk factor
CVC
colonization
n (%)
Odds
ratio
(OR)
P
value
 
CR
bacteremia
n (%)
Odds
ratio
(OR)
P
value
Fever (n = 23)
18 (78.2)
7.4
0.001
10 (43.4)
7.2
< 0.001
No fever (n = 52)
17 (32.7)
 
 
5 (9.6)
Abnormal TLC (n = 27)
16 (59.2)
2.2
0.1
11 (40.7)
7.6
< 0.001
Normal TLC (n = 48)
19 (39.6)
 
 
4 (8.3)
 
Abnormal neutrophil count (n = 37)
23 (62.1)
3.6
< 0.01
13 (35.1)
9.7
< 0.001
Normal neutrophil Count (n = 38)
12 (31.6)
 
 
2 (5.3)
Parenteral nutrition yes (n = 17)
12 (70.6)
3.6
< 0.05
5 (29.4)
2.0
0.27
   no (n =58)
23 (39.4)
 
 
10 (17.2)

TLC: total leucocyte count. 

 

Eight deaths were recorded in this study. Half of the deceased had Candida sps isolated from catheter tips and 3 (37.5%) patients had catheter-related candidemia. All these patients were receiving antifungal therapy at the time of death. One patient of hypoxic ischemic encephalopathy with aspiration pneumonia had grown S.aureus in CVC tips and peripheral blood culture.

Table IV__ Effect of Combination of Risk Factors on Central Venous Catheter Colonization (CVC) and Catheter Related Bacteremia

 
CVC colonization n (%)
*Odds ratio
( 95% CI)
Catheter Related
bacteremia**
n (%)
Odds ratio
(95% CI)
No risk factor (n = 23)
6 (26.1)
1.0
1 (4.3)
1.0
Single risk factor (n = 20)
6 (30)
1.2 (0.26-5.6)
1 (5)
1.2 (0.00-46.2)
Two risk factors (n = 16)
10 (62.5)
4.7 (0.99 - 24.1)
4 (25)
7.3 (0.6 - 194)
Three risk factors (n = 13)
10 (76.9)
9.4 (1.56 - 65.8)
6 (46.1)
18.9 (17- 497)
Four risk factors (n = 3)
3 (100)
 
3 (100)

CI - Confidence Interval, * P < 0.001, ** P< 0.001

 

Discussion

The results of the present study have highlighted the utility of central venous catheters and related problems especially colonization and related bactermia in Indian scenario. When catheter infection is suspected, it is important to differentiate bacterial colonization of catheter from clinically significant infection. Maki et al.(9) described semiquantitative and CIeri et al.(10) used quantitative culture methods to identify significant catheter infection. The later method has some distinct advantages including the ability to detect organism within the lumen of the catheter and to evaluate relative number of organisms from different species from mixed infection (11). However Moyer et al.(12) compared the two methods and did not find significant difference in the results.

Our technique differs slightly from that described by Cleri et al.(10) and Brun Buissen et al.(7). No attempt was made to flush the lumen of the catheter. Trypticose-soy instead of saline was added to the container of catheter tip and thoroughly vortexed. It was done to allow the recovery of most of the bacteria attached to its external and luminal surfaces and the use of trpticose-soy and incubation for 6 hours prior to plating to improve the yield of isolates.

Central venous catheter colonization and catheter related bacteremia rates vary widely. Catheter colonization rate varies from 5% to 61% (5,7,13) and catheter related bacteremia rate varies from 2% to 43% (3,5,14). Such wide variation in these rates is due to differences in the definitions of CVC colonization and catheter related infection used and patient population studied. In the present study, CVC colonization (46.7%) and CRB rates (20%) are rather high. The possible reasons are use of central lines in critically sick patients with or without sepsis, multipurpose use of single lumen catheter with a 3 way connection, post insertion care by attending pediatric intensivist and not by specialized IV care team and no scheduled protocol for central catheter replacement over guide wire or reinsertion at new central vein site. Similar reasons have been cited in literature(15). The use of local antimicrobial agents, type of dressing and frequency of catheter site care are crucial factors in preventing CRB since most of the infections develop at the skin site(16).

Role of routine blood culture to detect catheter colonization and related sepsis is questionable(11). In present study only 10.6% of blood cultures (B) samples grew same organism as from the catheter tip and peripheral blood culture sample (C). Bozzetti et al.(4) reported yield of same organism from the catheter tip and blood while CVC was in situ in 36% of patients and 13 blood culture samples were required to identify a single catheter sepsis.

Candida was the commonest organism isolated from catheter tip and blood cultures in the present study. Prager et al. (1) reported colonization due to Candida in 30% of CVC out of which 53.3% had candidemia in a study of 179 cathters. Other microbials responsible for catheter colonization and related sepsis include S. aureus, Coagulase negative staphylococci, Klebsiella, E. Coli and Pseudomonas(2,4). Multiple organism isolation up to 19% has been reported(14). In the present study 8% of CRB had more than one organism.

Though the number of femoral vein catheters was more than those inserted in the upper extremity, there was no significant difference in CVC colonization and CRB. Similar observation has been reported in literature (4,16).

Brun-Buisson et al.(7) reported increase in incidence of catheter related sepsis from 2.5% at 3 days to 12% at 6 days and 22% after seven or more days of use. In the present study, no CVC colonization was detected in the first four days while first CRB was detected after 6 days of insertion of central catheter. Increase in duration of CVC use revealed statistically significant increase in colonization and CRB rate. Our analysis is supported by previous studies done in adult patients which showed that the earliest occurrence of CRB was detected after the catheter had been in place for 4-5 days(10,17).

Central venous catheters are commonly used in acutely ill patients among whom the manifestations of catheter infection may be subtle or confusing and diagnosis needs high index of suspicion. Any fever, leucocytosis or deterioration in function should cause the clinician to consider the possibility of catheter infection(15). Fever as a reliable and consistent clinical symptom of catheter infection has been reported (1,18). In our analysis, 78.3% of patients with CVC colonization and 43.4% with CRB had fever. However Civetta et al.(17) removed 164 catheters for the presence of fever but only 18% had catheter related infection.

King et al.(18) followed the total and differential leukocyte count during catheter sepsis episode. The average WBC was 12000/Cu mm during the three day period prior to the onset of bacteremia and this increased to a mean of 13130/cu mm during the first three days of the septic course. The differential count changed significantly in the infected patients. The leftward shift (band forms) was observed 24 to 48 hours before the clinical recognition of sepsis. White cell count above 15,274 / cu mm was reported in 64% of patients with positive CVC cultures and in 8 out of 15 patients with demonstrable bacteremia by Prager et al.(1). In our study, 40.7% of patients with abnormal total leucocyte counts and 35.1% with abnormal neutrophil count had catheter related bacteremia.

Rose et al.(19) reported 55 cases of venous catheter associated candidemia with mortality rates of 36.3%. In the present study, the overall mortality was 10.7% but 37.5% among these had candidemia. We believe that candidemia was only a contributory factor to the serious underlying disease that the deceased were suffering from. This study reveals the utility of simple clinical and laboratory parameters to suspect CRB and alert the clinician to take necessary action. There is a need to perform prospective study in a larger group of patients with common underlying illness (e.g. oncology patients) or with common indication (e.g. parenteral nutrition) to validate the use of these risk factors.

Contributors: AS contributed to concept, design and analysis of data and is guarantor for the study. DG and ArunS contributed to data collection, literature search and drafted the article. KC did critical review of the manuscript.

Funding: None.

Competing interests: None stated.

 

Key Messages

• Central venous catheters though very useful in the management of critically ill neonate and children carry a high risk of colonization and catheter related bacteremia.

• Prolonged duration of catheter use (more than 5 days) and presence of fever, abnormal total leukocyte and neutrophil counts and use of parenteral nutrition significantly increase the risk of catheter colonization and related bacteremia and candidemia.

• When more than one of the above listed risk factors are present, risk of catheter related bacteremia is high.

 

References


1. Prager RL, Silva J Jr. Colonization of central venous catheters. South Med J 1984; 77: 458-461.

2. Prince A, Heller B, Levy J, Heird WC. Management of fever in patients with central vein catheters. Pediatr Infect Dis J 1986; 5: 20-24.

3. Bentley DW, Lepper MH. Septicemia related to indwelling venous catheters. JAMA 1968; 206: 1749-1752.

4. Bozzetti F, Terno G, Camerinii E, Baticci F, Scarpa D, Pupa A. Pathogenesis and predictability of central venous catheter sepsis. Surgery 1982; 91: 383-389.

5. Cobb DK, High KP, Sawyer RG, Sable CA, Adams RB, Lindley DA et al. A controlled trial of scheduled replacement of central venous and pulmonary artery catheters. N Engl J Med 1992; 327: 1062-1068.

6. Lubin BH. Reference values in infancy and childhood. In: Hematology of infancy and childhood, vol 2, 4th edition. Eds. Nathan DG, Oski FA. Philadelphia, W B Saunders Co. 1993; Appendix XI.

7. Brun-Buisson C, Abrar F, Legrand P , Huet Y, Larabi S, Rapin M. Diagnosis of central venous catheter related sepsis. CriticaI level of quantitative tip cultures. Arch Intern Med 1987; 147: 873-877.

8. Indrayan A, Satyanarayana L. Essentials of biostatistics: 9. Statistical inference from qualitative data: Proportions, relative risks, and odd ratios. Indian Pediatr 2000; 37: 967-981.

9. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture method for identifying intravenous catheter related infection. N Engl J Med 1977; 296: 1305-1309.

10. Cleri DJ, Corrado ML, Seligmat J. Quantitative culture of intravenous catheters and other intravascular inserts. J Infect Dis 1980; 141: 781-786.

11. Bozzetti F. Central venous catheter sepsis. Surg Gynaecol Obstet 1985; 161: 293-301.

12. Moyer MA, Edward L D, Farley L. Comparative culture methods on 101 intravenous catheters~utine,semiquantitative and blood cultures. Arch Intern Med 1983; 143: 66-69.

13. Bernard RW, Stanl WM, Chase RM. Subclavian vein catheterization-A prospective study. II Infectious complications. Ann Surg 1971; 173: 191-200.

14. Yeung CY, Lee HC, Huang FY, Wang CS. Sepsis during total parenteral nutrition: exploration of risk factors and determination of the effectiveness of peripherally inserted central venous catheters. Pediatr Infect Dis J 1998; 17: 135-142.

15. Decker MD, Edward KM. Central venous catheter infections. Pediatr Clin N Amer 1988; 35: 579-612.

16. Norwood S, Ruby A, Civetta J, Cortes V. Catheter related infections and associated septicemia. Chest 1991; 99: 968-975.

17. Civetta JM, Civetta JH, Bail S. Decreasing catheter-related infections and hospital costs by contitus quality improvement. Crit Care Med 1996; 24: 1660- 1665.

18. King DR, Komer M, Hoffinan J, Ginne-Pease ME, Stanley ME, Powel D, et al. Broviac catheter sepsis: The natural history of an iatrogenic infection. J Pediatr Surg 1985; 20: 728-733.

19. Rose RD. Venous catheter associated candidemia. Am J Med Sci 1978; 275: 265-269.

Home

Past Issue

About IP

About IAP

Feedback

Links

 Author Info.

  Subscription