Contaminated lithium heparin bottles as a source of pseudobacteraemia due to Pseudomonas fluorescens

~ournol

of Hospital

Infection

( 1999)

M

4 I : 23-28

Contaminated lithium heparin a source of pseudobacteraemia Pseudomonas fluorescens

bottles as due to

Simon Namnyak *, Sunny Hussaint, Jane Davalle*, Keith Raker* and Martin Strickland* *Department of Medical Microbiology., Havering Hospitals and j-Department of Paediatrics, Hovering Hospitals N-IS

NHS Trust Trust

Summary:

Pseudobacteraemia might be responsible for up to 50% of all positive blood cultures and its early recognition is important in order to avoid unnecessary treatment with antibiotics and delay in the search for the true cause of the fever. We describe pseudobacteraemia outbreak of Pseudomonas~fluorescens related to contaminated lithium heparin bottles in a paediatric ward. Twelve patients were involved in this outbreak from December 1996-January 1997. All patients had no clinical evidence of sepsis, nevertheless most children were treated with antibiotics. Blood collection bottles were suspected as source of pseudobacteraemia and only lithium heparin bottles were found to be contaminated with P. jZuorescencs indistinguishable from the blood isolates taken from these children. Withdrawal of these bottles led to the termination of the pseudobacteraemia. Following discussion with the manufacturer, the contaminated batch of lithium heparin bottles was sent back for testing, and replaced with bottles containing dried lithium heparin. A hazard report was sent to the Medical Devices Agency (MDA). In order to minimize the possibility of this problem occurring again, the manufacturer has informed MDA that all lithium heparin solution is to be filtered to 0.2 microns prior to issue, in order to minimize bacterial contamination. Continued monitoring after the pseudobacteraemia showed no isolates of P. fluorescens from the blood of paediatric patients. Keywords

Pseudobacteraemia;

Pseudomonas

jluorescens;

Introduction Pseudobacteraemia, defined, as presence of bacteria in a blood culture in the absence of clinical evidence of bloodstream infection,‘z2 might be Received 28 April 199X; revised manuscript September 10%. correspondence to: Dr Simon =2ddress Department of Medical Microbiology, Havering NHS Trust, Gubbins Lane, Harold Wood,

Essex RM3 ODE.

0195-6701/99/010023+06

$12.00/O

accepted Namnyak, Hospitals Romford,

23

Lithium

heparin

responsible for up to 50% of all positive blood cultures.” Recognition of the pseudobacteraemia is important because treatment with unnecessary and often toxic antibiotics may be instituted, and there may be delay in the search for the true cause of the fever. Investigations of outbreaks of pseudobacteraemia have identified contamination of blood cultures which may have occurred during venipuncture, in the preparation of culture media or during laboratory processing of the cultures.4 However, only a few

0

1999 The

Hospital

Infection

Society

24

S. Namnyak

of these investigations either included casecontrol study designs to identify the causes, for instance, of gram-negative pseudobacteraemia, an occurrence with low incidence,4,’ or utilized discriminatory molecular typing techniques for strain or clone identification and delineation to conclude for certain that cross-contamination occurred.6 While pseudobacteraemia due to the pseudomonas group of gram-negative bacteria is frequently reported,3 that due to P. jluorescens is rarely reported.7v8 In December 1996, a pseudobacteraemia outbreak occurred in our paediatric ward where P. j?uorescens was isolated from blood cultures of twelve patients and from several lithium heparin bottles. Withdrawal of these lithium heparin bottles interrupted the pseudobacteraemia. Subsequent typing by the Laboratory of Hospital Infection showed that all of the isolates produced the same banding pattern by pulsed-field gel electrophoresis of XbaI chromosomal digests, showing that these represent a single strain. We report the epidemiological and microbiological investigation of this outbreak, suggesting that faulty venepuncture technique was probably responsible for the blood culture contamination, although the case-control study did not identify a significant correlation between the recovery of P. Jluorescens in blood culture and the concomitant use of heparin lithium bottles.

Materials Description

and

Methods

of the outbreak

From 19 December 1996 to 27 December 1996, P. jhwescens was isolated from eight blood cultures that had been collected from 11 children hospitalized in the paediatric ward. The patients showed no evidence of septicaemia; nevertheless eight of the children were treated with penicillins, cephalosporins or gentamicin, singly or combined. Contamination of blood cultures during venepuncture involving commercial products and faulty technique, was a likely explanation. Several random samples of the current batch of lithium heparin bottles (batch number 4073) yielded P. jluorescens of the same

et a/.

susceptibility pattern as the patients’ isolates. Withdrawal of these bottles interrupted the pseudobacteraemia. When the same batch of lithium heparin bottles were again used in error, P. &orescens similar to the previous isolates, was isolated from 4 blood cultures that had been collected from 26 children between 1 January 1997 and 12 January 1997. Similarly, none of these children had septicaemia. Again when this batch of bottles was withdrawn from clinical use, no further cases of pseudobacteraemia occurred (Fig. 1).

Epidemiological

investigation

All policies regarding venepuncture technique, inoculation of blood culture bottles at the bedside and laboratory procedures for the processing of blood cultures, were reviewed in detail in the meeting involving the paediatric staff and members of the Infection Control Team. Actual practice patterns in these areas were directly observed where possible. Protocol guidelines dictated that when blood is taken for culture, it should not be used for any other purpose such as urea and electrolyte studies before being inoculated into the blood culture bottles. Paediatric House Officers took all blood for culture and other tests and admitted that this protocol could have been violated at times. Microbiology records from 1 November 1996 to 31 March 1997 were reviewed in order to identify cases of patients from whom P. fluorescens was isolated in blood culture. A case is defined as ‘a paediatric patient admitted in C2 and C4 Wards whose blood cultures yielded P. jiuorescens between 23 December 1996 and 15 January 1997 but without clinical evidence of sepsis’. For each case, a time- and ward-matched control (P. &orescens blood culture negative) was assigned by identifying the next patient to have a blood culture drawn on the same ward. The hospital charts of each case and control were reviewed to determine if any significant differences existed in respect of age, sex, underlying illnesses, reasons for blood culture, other blood tests drawn at the same time, the staff who drew the blood, duration of hospitalization

Pseudobacteraemia

and lithium

heparin

bottles

25

prior to the positive culture, presence and duration of intravascular lines, transfusions, antibiotic therapy and clinical outcome.’

Microbiology

(two-tailed), analysis of variance (ANOVA) Kruskal-Wallis test as appropriate for tinuous variables.

and con-

survey

Results The outbreak investigation involved microbiological sampling of environmental sites, with a focus on the use of commercial bottles for the collection of blood for a variety of tests or the processing of blood cultures. Environmental sites were selected for sampling on the basis either of experience from previously described P. jkwescens outbreaks (blood culture bottles, blood collection tubes, EDTA bottles, lithium heparin bottles, titrated bottles, antiseptic solutions”7’8) or whether they accorded with recommended investigations of pseudobacteraemia.” Techniques for sampling, culturing, bacterial identification and susceptibility testing were in accordance with recommended methods.’ Rates of growth of P. jluorescens in solutions containing lithium heparin, 3.8% sodium citrate, phosphate buffered saline and nutrient broth were studied at 4°C and 25”C, using standard tube dilution and plating techniques.

Macrorestriction electrophoresis

analysis by pulsed-field

gel

(PFGE)

Isolates of P.fluorescens were compared by DNA macrorestriction analysis as previously described.” Isolates were assigned to PFGE clonal groups according to the criteria by which one to three different bands constituted clonally related strains and more than three different bands suggested a different clonal group.“’

Epidemiological

Twelve blood cultures from 12 patients yielded P. jIuorescens between 23 December 1996 and 15 January 1997. There were five males and seven females with a mean age of 5.1 years (range, 1 .l-11.9 years). In addition to P. jluoyescens, two patients had Stenotrophomonas maltophilia isolates each and one lithium heparin bottle had Comamonas acidouorans. There were 22 P. fluorescens isolates from all hospitals wards including the Accident and Emergency department, but ward-specific rates (number of P. Juorescens isolates per ward/total of blood cultures per ward/per two months (during December and January) were higher from the paediatric wards. A review of the hospital charts of these patients indicated that P. jluorescens was not clinically significant in any of the cases although most of them received antimicrobial treatment. The characteristics of the patient and control populations were similar in terms of age and sex distribution, as was the range of diagnoses and procedures performed on admission, and the length of stay. The isolation of P. fiuoyescens and the concomitant urea and electrolyte studies were similar in both groups, as was the number of venepunctures performed by each doctor in both groups. Microbiological

Statistical

analysis

Univariate analysis was done using the Epi Info 6.05 a (CDC, Atlanta, Georgia) software packsignificant. Difage,” with PI .05 considered ferences in proportions (maximum likelihood estimate of odds ratio and 95% confidence intervals) between dichotomous variables, were analysed by x2 statistics and Fisher’s exact test when appropriate, and used Student’s t-test

investigation

investigation

P. jkorescens was not isolated from venepuncture equipment, skin disinfectants, uninoculated culture media, the external surfaces of any blood culture bottles or vents, laboratory surfaces or working solutions, EDTA bottles, fluoride bottles and plain bottles. The organism was isolated from all seven samples of liquid lithium heparin bottles from the current batch number 4073 and not from seven bottles in

26

S. Namnyak

et al.

December

I996

5

Date Figure

I

The

‘outbreak’

curve

of paediatric

of isolation; patients

arrows

epidemiology

Isolates of P. fluorescens produced identical macrorestriction profiles by PGFE of Xbal chromosomal DNA digests, showing that the isolates represented a single strain.‘O Infection

when

bottles

who developed C fluorescens and January 1997.

batch number 4087 cultured. The antimicrobial susceptibilities of P. fluorescens isolates from the lithium heparin bottles and the blood cultures were identical. All were susceptible to cefuroxime, cefotaxime, ceftazidime, ceftriaxone, piperacillin-tazobactam, meropenem, ciprofloxacin and gentamicin, but resistant to amoxycillin, co-amoxiclav, trimethoprim and chloramphenicol. The organisms grew best at 25°C in nutrient broth, sodium citrate, phosphate buffered saline and lithium heparin. Molecular

indicate

control

As a result of our investigations, all lithium heparin bottles of the contaminated batch were withdrawn from all clinical areas and replaced by bottles from the sterile batch of lithium heparin. A report explaining the problem was

were

removed

pseudobacteraemia

between

sent to senior medical, nursing and managerial staff and all house officers were informed by telephone and instructed to inoculate blood culture bottles first whenever blood was taken for this and other purposes. Following discussion with the manufacturer (LIP Equipment and Services Ltd; Dockfield Road, Shipley West Yorkshire, England), the contaminated batch of lithium heparin bottles was sent back for testing, and a hazard report sent to the Medical Devices Agency (MDA). After further positive P. fh~orescens blood cultures from patients in paediatric wards, it was noted that some lithium heparin bottles from the contaminated batch were still in use. These were finally withdrawn. Despite the manufacturer having been informed in writing of the contaminated batch number, they sent a fresh supply of the same contaminated batch number. It was only due to the vigilance of the Phlebotomy Service Manager that this was observed. The manufacturer was again informed by telephone and in writing and the bottles were returned and replaced by equivalent bottles containing dried lithium heparin, which were tested and found to be sterile. In order to

Pseudobacteraemia

and lithium

heparin

bottles

minimize the possibility of this problem occurring again the manufacturer has informed MDA that all lithium heparin solution is to be filtered to 0.2 microns prior to issue, in order to minimize bacterial contamination. The MDA issued a ‘One Liner’ in January 1998 and addressed the issue of pseudobacteraemia in a recent Device Bulletin which covered aspects of medical device and equipment management in relation to in-vitro medical devices (including ‘order of draw’ for blood cultures and additive containing blood collection tubes).

Discussion The majority of reports of pseudobacteraemia are of Gram-negative bacteria especially coliforms and pseudomonads.” Three previous report&” of P. jluorescens pseudobacteraemia attributed this to cross-contamination of blood cultures with contaminated sodium citrate tubes used in coagulation studies, with one study2 observing this association in a case-control study that most blood isolates of P. jhorescens came from patients who had had concomitant coagulation studies done. We were alerted to the possibility of cross-contamination of blood cultures when we observed an increased isolation rate of P.$uorescens in children who had no symptoms or signs consistent with pseudomonal septicaemia. Based on previously published reports, our initial investigations were focused on blood collection tubes, skin disinfectants and blood culture bottles specifically used in the paediatric wards. Within two days of starting the investigations, P. jIuoyes(.elzs was isolated in heavy amounts from all randomly collected and cultured lithium heparin bottles currently in use and not from any other tubes and solutions. Withdrawal of these bottles led to a rapid decline in the number of new cases of P. jluorescens pseudobacteraemia, only for new cases to reappear when the bottles from the contaminated batch were again used in error, supporting the observation that there is a significant association between the concurrent use of lithium heparin and blood culture bottles, and the development of pseudobacteraemia in children.

27

Because the Paediatric House Officers took all blood for culture and other tests and admitted that recommended protocol could have been violated at times, a case-control study was instituted to compare the exposures of cases and controls to a variety of risk factors including the doctor as the phlebotomist, and the use of lithium heparin bottles in urea and electrolyte studies. Cases and controls were equally exposed to the individual doctors and to the concurrent use of lithium heparin bottles, suggesting that cross-contamination of blood cultures following prior inoculation of contaminated lithium heparin was a purely random phenomenon and non biased towards any one doctor in terms of not following recommended procedures for taking blood culture. No other significant associations were found with the variables tested. The subsequent molecular analysis explicitl!, implicated a common source of contamination with a single strain of P. jhovescens. ,411 isolates were recovered from several bottles in one batch of lithium heparin. Although we were unable to demonstrate conclusively the precise mechanism of contamination after changing to a different batch of liquid lithium heparin bottles from the same manufacturer, we hare not recovered P. jhorescens from this new batch or any other blood collection bottles. P. jhwescens, like other pseudomonads, is widespread in nature and is found in lvater, moist soil and vegetation. Although it is rarely associated with human disease, most clinical isolates are from the respiratory and urinary tracts, often in mixed culture, and occasionally described as a cause of empyema and soft tissue infections.12 However, there have been three reports of P. fluorescens bacteraemia and sepsis related to contaminated refrigerated blood products.‘“-” The organism is psychrophilic, is able to grow at refrigerator temperatures (4°C) and is able to use citrate as a carbon source for growth.‘” The presence of P. jhorescens as a contaminant in blood sampling tubes containing lithium heparin has not been reported, though incriminated for other bacteria.‘” Our investigations demonstrated that P. jluorescens grows best at 25’C in nutrient broth, sodium citrate, phosphate buffered saline and

28

S. Namnyak

lithium heparin. These properties may explain the ability of P. fluorescens to survive and multiply in lithium heparin solutions during storage at room temperatures and to multiply and travel on syringe needles into blood culture bottles during faulty or less than meticulous venepuncture and culture techniques. To prevent further pseudobacteraemia, the manufacturer replaced the contaminated batch with dried lithium heparin bottles, whereas the Infection Control Team again encouraged doctors to inoculate blood culture bottles before filling any other collection tubes. Stringent quality control and provision of sterile blood collection bottles on the part of the manufacturer, and an ongoing education of doctors and phlebotomists, can prevent potential hazardous consequences of pseudobacteraemia. In conclusion, this experience emphasises the potential for recurrent pseudobacteraemia in a clinical setting and demonstrates the value of molecular typing as an adjunct to clinical epidemiological investigations. In particular, PFGE provided critical evidence that the set of P. fluorescens isolates recovered resulted from a common source of contamination by a single strain, possibly at the industrial level.

Acknowledgements The authors would like to thank ME Kaufmann of the Laboratory of Hospital Infection at the Central Public Health Laboratory, Colindale for molecular typing the isolates.

References

Intern

4. Kusek JW. Nosocomial pseudobacteraemia and pseudo infections. An increasing problem. Am J Infect Control 1981; 9: 20-25. 5. Little JR, Trovillion E, Fraser V. High frequency of pseudobacteraemia at a university hospital. Infect Control Hasp Epidemiol1997; 18:

200-204. 6. Astagneau P, Gottot S, Gobin Y, et al. Nosocomial outbreak of Enterobacter agglomerans pseudobacteraemia associated with non-sterile blood collection tubes. J Hosp Infect 1994; 27:

73-74. 7. Gottlieb fluorescens

Med

1980;

140: 26-28.

2. Simor AE, Ricci J, Lau A, Bannatyne RM, Ford- Jones L. Pseudobacteremia due to Pseudomonas jluorescens.

Pediatric

Infect

Dis

1985;

4:

508-512. 3. Jumaa PA, Chattopadhyay B. Pseudobacteraemia. J Hosp Infect 1994; 27: 167-l 77.

T, Funnel G, Gorbell I. Pseudomonas pseudobacteraemia. MedJ Aus 1991;

155: 854-855. 8. Anderson M, Davey R. Pseudobacteraemia Pseudomonas

fluorescens.

Med

J Aus

1994;

with

160:

233-234. 9. Kaufmann

ME, Pitt TL. Pulsed-field gel electrophoresis of bacterial DNA. In: H Hart (ed). Methods in practical laboratory bacteriology. Florida: CRC Press Inc; Boca Raton; 1994:

83-92. 10. Tenover FC, Arbeit RD, Georing RV, et al. Interpreting chromosal DNA restriction patterns produced by pulsed-field gel electrophoresis: Criteria for bacterial strain typing. J Clin

Microbial

1995;

33: 2233-2239.

11. Dean AG, Dean JA, Coulombier D, et al. Epi Info, Version 6.04a: a word processing, database, and statistics program for epidemiology on microcomputers. Centres for Disease Control and Prevention, Atlanta, Georgia, USA, 1994. 12. Gilardi GL. Infrequently encountered Pseudomonas species causing infection in humans. Ann Intern Med 1972; 77: 211-215. 13. Felaby M, Munk-Andersen G, Siboni K. Simultaneous contamination of transfusion blood with Enterobacter agglomerans and Pseudomonas fluorescens, supposedly from the pilot tubes. J Med

1. Maki DG. Through a glass darkly, nosocomial pseudoepidemics and pseudobacteremias. Arch

et al.

Microbial

1978;

6: 413-416.

14. Khabbaz RF, Arnow PM, Highsmith AK, et al. Pseudomonas fkorescens bacteremia from blood transfusion. Am J Med 1984; 76: 62-68. 15. Tabor E, Gerety RJ. Five cases of Pseudomonas sepsis transmitted by blood transfusions. Lancet 1984;

1: 1403.

16. Washington JA. The microbiology of evacuated blood collection tubes. Ann Intern Med 1977;

86: 186-188.