Journal of Parenteral and Enteral Nutrition http://pen.sagepub.com/
Microbial Contamination of Enteral Feeding Tubes Occurring During Nutritional Treatment Véronique Bussy, François Marechal and Salvador Nasca JPEN J Parenter Enteral Nutr 1992 16: 552 DOI: 10.1177/0148607192016006552 The online version of this article can be found at: http://pen.sagepub.com/content/16/6/552
Published by: http://www.sagepublications.com
On behalf of:
The American Society for Parenteral & Enteral Nutrition
Additional services and information for Journal of Parenteral and Enteral Nutrition can be found at: Email Alerts: http://pen.sagepub.com/cgi/alerts Subscriptions: http://pen.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://pen.sagepub.com/content/16/6/552.refs.html
>> Version of Record - Nov 1, 1992 What is This?
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
Microbial Contamination of Enteral Feeding Tubes Occurring During Nutritional Treatment VÉRONIQUE BUSSY, PHD*; FRANÇOIS MARECHAL, MD†;
AND
SALVADOR NASCA, MD‡
From *Institut Jean-Godinot, Laboratoire de Microbiologie Médicale, Reims Cedex; †Hôpital Général, d’Oncologie Médicale, Macon Cedex; and ‡Institut Jean-Godinot, Département d’Oncologie
ABSTRACT. Enteral nutrition is an effective treatment for catabolic patients with normal intestinal absorption. However, tube-fed patients are at risk from superinfection. Our study is the first to evaluate in vivo the microbial contamination of solutions staying in the nasogastric tube of cancer patients receiving nutritional preparations. After daily feeding, the tube was rinsed with nonsterile tap water. Tap water staying overnight in the tube was considered as tube-rinsing solution. Microbial burden of nutritional preparations was determined on the fifth day of enteral nutrition, from opening the first container and throughout feeding. The next day, a sample of the tube-rinsing solution was cultured. All bacterial species were identified and antibiotic susceptibility pattern was assessed. Thirty-one cancer patients were included, 12 on the hospital’s preparations and 19 on commercial feeding. Seven of
Malnutrition occurs frequently in cancer patients and be found in all stages of the diseased The disease itself, or aggressive treatment, often leads to nutritional imbalance. Enteral nutrition is an effective treatment for catabolic patients with normal intestinal absorpcan
tion.2-5
The contamination of nutritional preparations and feeding systems has been associated with colonization or infections complicating enteral feeding.6-9 Colonization of enteric nutritional preparations or contamination with bacterial enterotoxins constitutes a risk for the weak, immunodeficient, and intensive care patient.10 Although nutritional preparations used comply to high microbiologic standards of quality, 11-1 infections due to contaminated preparations are still often reported.14-16 Numerous works have addressed the bacterial contamination of enteral feeding preparations,17-20 but no study has yet evaluated in vivo the microbial colonization of the nasogastric tube. In a prospective study, we determined the extent of the contamination of the tuberinsing solution occurring during nutrition treatment. The antibiotic susceptibility pattern of the cultured strains was also studied.
nowadays
MATERIALS AND METHODS
Patients
They were
patients (29 men, 2 women), (18 head and neck, 6 esophagus, 2
31 consecutive
suffering from
cancer
Reprint requests: Véronique Bussy, Laboratoire de Microbiologie Médicale, Institut Jean-Godinot, B.P. 171, 51056 Reims, France.
the
hospital—and none
Département de Médecine Interne, Unité Médicale, Reims Cedex, France
of the
commercial—preparations were
contaminated. Among the 48 samples collected during feeding, 16 were contaminated, including 10 hospital and 6 commercial preparations. All the 31 tube-rinsing solutions were contaminated and 102 strains cultured. Their median concentration was 10 6 colony-forming units/mL (range 10-10 10 ). The strains were 48 Enterobacteriaceae, 20 group D streptococci, 9 Candida albicans, 9 Pseudomonas aeruginosa, and 16 others. Multiple antibiotic resistance was present in 12 of the 102 strains. Lower resistance was present in 33. The predominant microorganism of the tube-rinsing solution caused a bacterial colonization for three febrile patients. In conclusion, the feeding tube is an important reservoir for multiple antibiotic-resistant bacteria. (Journal of Parenteral and Enteral Nutrition 16: 552-557, 1992)
1 stomach, 4 multiple) who had been hospitalized for at least 48 hours before the nasogastric feeding tube was placed. Median age was 56 (range 33-69). Important dysphagia and/or malnutrition prompted the institution of enteral nutrition by nasogastric tube. When included, they did not receive any antibiotic during the 6 days of the study. Six of them were treated by inhibitors of gastric acid secretion (ranitidine, Glaxo, Paris, France or cimetidine, Smith Kline & French and Beecham, Paris,
colon,
France). Enteral
Feeding
For the first 12 patients, the nutritional preparations were prepared by the hospital dietary service by mixing sterile commercial liquids (Roussel Uclaf-Sopharga, Puteaux, France) and different sterile ingredients such as glucides, oil, vitamins, or inorganic products with a blender. The blender was sterilized daily. For the 19 following patients, clinicians and dietitians decided to use sterile commercially complete preparations (Nutricia, Rueil Malmaison, France). For each patient, three to five containers (500 mL) were infused daily for 3 to 4 hours each, between 8 AM and 10 PM. The nutrition was given via a silicone nasogastric tube (Peters, Bobigny, France) 1250 mm long, with a lumen diameter of 4.2 mm. The container was hung up and connected to the nasogastric tube through a sterile tubule. Infusion flow was regulated with a manual wheel. Three to more than five disconnections occurred daily, whenever required, to chance a container or for taking care of the Datient. After
552 Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
553 each disconnection, or at the end of the day, the tube was rinsed with 20 mL of nonsterile tap water. It was then plugged while filled with tap water. Tap water staying overnight in the tube was considered as tuberinsing solution. Tap water was cultured weekly using a filter method and found free of pathogenic microorganisms throughout the study.
Bacteriological Samples Nutritional preparations. For each patient, a bacterial viable count of corresponding nutritional preparations was determined on the fifth day. Samples of 3 to 5 mL were aseptically taken from the first daily container immediately before beginning the infusion. Two hours after the beginning of the first and the third daily containers, the infusion was stopped for a few minutes. The tube was aseptically disconnected from the container and 3 to 5 mL of the nutritional preparations was collected for bacterial viable counting. All the samples were sent immediately to the laboratory. Rinsing solutions staying overnight in the tube. In a preliminary study with 15 patients, we showed that the same flora contaminated the rinsing solution staying overnight in the tube and the nutritional preparation infusing through the tube. The microbial concentrations obtained were lower in the latter case. They were between 10 and 105 colony-forming units (cfu)/mL (median 10~). In the interest of simplicity, we studied the rinsing solution staying overnight in the tube, before beginning the daily infusion. For each patient, a sample of nasogastric tube-rinsing solution was taken 6 days after insertion of the tube, before the daily feeding. A sterile catheter introduced into the tube allowed us to collect the sample. The end of the catheter was plugged using sterile agar to prevent contamination during mobilization to the esophageal portion of the tube. A needle was inserted into the exposed end of the catheter, and a syringe (20 mL) was adapted to the needle. A positive pressure allowed the expulsion of the agar plug and a 200- to 400-~L sample of liquid was then obtained. The sample was then injected into a sterile flask and sent immediately to the
laboratory.
Bacteriological Procedures Microbial numerations. The samples were diluted by sequential 10-fold steps until a 10-6 dilution in sterile 0.9% sodium chloride was reached; 100 ~L of each dilu-
5% horse blood agar and on were incubated at 37°C for 16 hours. A colony count was then carried out. In the absence of published standards for bacteriological safety of enteral feeding formulas, the standards applicable to milk published by the French legislature were used.21 These regulations place an upper limit of 3.10’ cfu/mL. Identification of microbial species. Bacteria were identified using API 20E for Enterobacteriaceae, API Strep for streptococci, and API 20NE for Pseudomonas and
tion
was
Drigalski
inoculated
agar.
Acinetobacter
on
Agar plates
organisms (Laboratoires Biom6rieux, Lyon, France). We used Chapman agar and the deoxy-
ribonuclease production as
Staphylococcus
categorize staphylococci staphylococcus coagulase
test to
aureus
or
negative. Antimicrobial susceptibility. Antimicrobial susceptibility of the strains cultured from the tube-rinsing solution samples was performed according to a standardized disk method22.’3 (Institut Pasteur Production, Paris, France). Inhibition diameters were measured. A disk susceptibility test was done for each bacterial strain presented in Table I, and for other species, according to previously
described
procedures. 21
to methicillin was detected by the disk diffusion method with a l-llg oxacillin disk. Susceptibility testing was performed at 37°C in Miiller-Hinton agar (Institut Pasteur Production, Paris, France) and at 30°C in Miiller-Hinton agar containing 5% NaCl. For Haemophilus organisms and neisseriae, the search of ¡3-lactamase was done by the chromogenic cephalosporin test (c6finase, Laboratoires Biom6rieux,
Staphylococci susceptibility
Lyon, France). The resistant phenotypes to #-lactams and aminoglycosides were determined for Enterobacteriaceae and Pseudomonas aeruginosa according to previously described methods.25-30 Comparison of strains of the same species. Strains of the same species were compared by serotyping, when they were cultured from both a nutritional preparation and a corresponding tube-rinsing solution sample (service des ent6robact6ries, Institut Pasteur, Paris, France).
Relationship Between Infections
and
Tube-Rinsing
Solution Flora
The occurrence of infections was studied over a period of 10 days after insertion of the tube; criteria were fever above 38°C for at least 1 day and all clinical signs of infection. Cultures were made to search for the origin of an infection (blood, throat, feces, and swabs or samples from other infected sites). We attempted to identify a relationship between bacteria causing infections and corresponding tube-rinsing solution flora. RESULTS
Microorganisms Cultured From Nutritional Preparations Before beginning the daily infusion. Of the 31 samples, corresponding to the 31 patients, 24 were sterile and 7 were contaminated. All seven contaminated samples corresponded to hospital dietary service preparations. Ten different species were cultured as shown in Table II. The concentrations obtained were less than 103 cfu/mL, except in two cases, with 104 cfu/mL Acinetobacter calcoaceticus anitratus for one and l0a cfu/mL Moraxella sp for the other.
Throughout infusion. For 17 patients, 2 samples were on the fifth day. For the 14 others, only 1 sample was obtained from either the first or the third daily container. In total, among 48 samples cultured, 32
obtained
sterile and 16 grew bacteria. Ten of them were associated with hospital preparations. They concerned eight patients, of whom seven had received a preparation
were
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
554 TABLE I
Antibiotics tested for
*
species
most
frequently cultured from tube-rinsing solution samples
Amoxicillin + clavulanic acid. TABLE II Strains cultured from nutritional feeds
TABLE III
Family,
genus,
or
species cultured from the 31 tube-rinsing solutions and concentrations obtained
* Fourteen E
coli,
3 P
mirabilis,
15
Klebsiella,
15
Enterobacter,
1
Citrobacter freundii. t Fourteen S faecalis, 5 S durans, 1 S faecium. $ The other species were categorized in 2 Micrococcus, 2 staphylococcus
already contaminated before infusion. Six positive samples were associated with commercial preparations corresponding to three patients (three samples from the first container and three from the third). Thirty-one strains were cultured from the 16 contaminated samples (Table II). The median concentration was 102 cfu/mL (range: 10-10’). For 12, the microbial concentrations were less than 104 cfu/mL; they were between 105 and 10’ cfu/mL for the 4 others. As shown in Table II, two strains belonged to fecal flora (Klebsiella pneumoniae and Enterobacter cloacae) and one was skin and mucous membrane saprophyte (Candida albicans). These three strains were each from one different sample. Microorganisms Cultured From the Tube-Rinsing Solutions
coagulase negative, 1 bacillus, 1 Corynebacterium, 1 Haemophilus, neisseriae, 1 Saccharomyces cereuisiae, 1 Streptococcus pneumoniae. § Median 106 cfu/mL. =
Antimicrobial Strains
cocci.
Susceptibility of Tube-Rinsing Solution
Enterobacteriaceae, pseudomonads, and Acinetobacter organisms. The susceptibility pattern to {3-lactams is shown in Table IV. All but three Enterobacteriaceae were susceptible to all aminoglycosides; three Escherichia coli were resistant to kanamycin and neomycin. One pseudomonad was resistant to all aminoglycosides, six were resistant to both kanamycin and gentamicin, and one was resistant to kanamycin. One of the three Acinetobacter organisms was highly resistant to all antibiotics. The other two were
All 31 rinsing solutions studied were found to be contaminated with a total of 102 different strains (Table III). The median number of strains by tube was three (range 1-7). Strains belonging to fecal flora were the most frequently cultured. They composed 67%, categorized in 48 Enterobacteriaceae and 20 group D strepto-
1
susceptible.
The resistance to sulfonamides and to tetracyclines was 25% for each. All the strains were susceptible to other antibiotics. Group D streptococci. All but two strains were susceptible to amoxicillin. Resistance to penicillin was always present. High resistance to gentamicin was not present. The susceptibility to macrolides, lincosamides, and sy-
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
555 ,~-Lactams susceptibility of the 48 Enterobacteriaceae, the
TABLE IV 9 pseudomonads, and the 3 Acinetobacter
*
The penicillin resistance was higher for 4 E coli and 1 K pneumoniae. (In vitro, t The cephalosporin resistance was higher for 4 strains of Enterobacter. $ Ticarcillin resistant and (3-lactamase negative. § Ticarcillin resistant and /3-lactamase positive.
nergistins
was
in accordance with the literature.31 All
susceptible to vancomycin and teicoplanin. Staphylococcus aureus. One of the three strains of S aureus was resistant to meticillin, all aminoglycosides, and ciprofloxacin. It was resistant to lincosamides and synergistins with lincosamides and streptogramin A phenotype. The two other strains were susceptible to all strains
were
antibiotics tested. Other germs and C albicans. Other microorganisms were susceptible to all antibiotics tested. The susceptibility of C albicans to antifungal agents was not studied.
Comparison of Strains For two patients, we found a relationship between strains of nutritional sample taken throughout infusion and corresponding tube-rinsing solution. For one patient it was a C albicans, and for the other it was a K pneumoniae.
Relationship Between Infections and Tube-Rinsing Solution Flora
The
of infections could be studied for 20 11 others, clinical data could not be obtained or infected sites were not examined. Ten patients had a temperature above 38°C for at least 1 day. It was related to a systemic candidosis for one patient. The corresponding tube-rinsing solution was free of Candida organisms but presented a strain of E coli with a higher viable count (>3.104 cfu/mL). This strain was related to the bacterial colonization of throat and nose. For nine patients, fever was not related to bacteremia. For three of them, throat and feces were colonized by the bacteria found with a higher viable count in the corresponding tube-rinsing solution. These bacteria cultured with a higher viable count in feces. They were P aeruginosa for one patient and K pneumoniae for two. The strain of P aeruginosa was highly resistant to antibiotics. One of the two patients with K pneumoniae had diarrhea. Diarrhea resolved after removal of the tube. For one of these febrile but nonbacteriemic patients, feces, throat, and nose were colonized by E coli. This strain was found in the corresponding tube-rinsing solution, but without a higher viable count. For two, the throat alone was colonized with bacteria found in the corresponding tuberinsing solution. It was a strain of E coli for the first and occurrence
patients; for the
no
potentiating
organisms cultured from the rinsing solutions
effect of clavulanic
acid.)
strain of C albicans for the second. One other patient had diarrhea, but feces were free of pathogenic microorganisms. The throat and the tube-rinsing solution were both colonized with a Proteus mirabilis, but without higher viable counts. For the last two patients, no relationship was shown. Seven of these febrile patients had empiric parenteral antibiotic treatment after the sixth day, but defervescence was not observed. For the patient with E coli, the antibiotic treatment was modified. Amoxicillin/clavulanic acid was given through the tube and fever resolved. Among the 10 afebrile patients, 2 presented a bacterial colonization related to microorganisms found with a higher viable count in the tube-rinsing solution. There were E cloacae in feces and C albicans in throat for one and Klebsiella oxytoca and E cloacae, both in feces for the other. Five patients presented a bacterial colonization related to microorganisms found in the tube-rinsing solution but without a higher viable count. There were three P aeruginosa and two C albicans, each one of them in feces and/or throat. Two patients did not present any bacterial colonization, and one presented a fecal colonization with C albicans without any relationship with the tube-rinsing solution flora. a
DISCUSSION
When cultured before beginning the infusion, the 31 nutritional preparations studied were sterile in 24 cases and contaminated in 7. These seven had all been prepared by the hospital dietary service. These results confirmed the previously described superiority of commercial preparations, especially regarding microbial quality.~~ 12, 13, 3z The cultured germs were not found in samples from corresponding tube-rinsing solutions (Tables II and
III). The contamination of nutritional preparations during their administration was more frequent than from the beginning of the infusion and also affected commercial preparations. A third of the samples (16/48) showed contamination. In four cases, high viable counts were found (105_10~, cfu/mL), suggesting a defect in administration methods, either by the prolonged stay of the preparations at room temperature or by contamination at the time of setting up the feeding containers. For two of these four highly contaminated samples, the presence of microorganisms usuallv transmitted by hand,‘j3~ 3~ (es-
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
556
pecially E cloacae and K pneumoniae) indicated that the contamination probably occurred when the feeding containers were set up. Among these four highly contaminated samples, only two presented a relationship with the corresponding rinsing solution flora. It concerned a K pneumoniae for one and a C albicans for the other. The two others did not present any relationship with the corresponding tube-rinsing solution flora, despite the presence of
an E cloacae cultured from one. A total of 102 different strains had been cultured from the 31 tube-rinsing solution samples obtained with the 31 patients. The median concentration was 106 cfu/mL (range 10-101°). The contamination of the tube-rinsing solution was most often not related to the contamination of the nutritional preparations. Most strains cultured from the tube-rinsing solutions belonged to fecal flora (67% of the strains), whereas fecal strains had been exceptionally cultured from nutritional preparation samples taken during infusion. Two fecal strains were cultured from two different nutritional samples, but only one was found in the corresponding tube-rinsing solution. Pseudomonas aeruginosa was frequently cultured from the tube-rinsing solutions but never from nutritional preparations. Candida albicans, also frequently cultured from the tube-rinsing solution, was cultured from only one nutritional preparation sample taken during infusion (Tables II and III). We can suggest several reasons for the contamination of the tube-rinsing solutions in our usual practice. On one hand, the contamination could occur during insertion of the tube. It could be caused by the oropharyngeal flora, which is often abundant and rich in Enterobacteriaceae or C albicans, especially in hospitalized cancer patients. In six cases, the patients were treated with inhibitors of gastric acid secretion, which have been shown to allow overgrowth of bacteria in the stomach.35 In these cases, backward contamination of the tube might occur. However, this seems to be a minor way of contamination, because all tube-rinsing solutions were highly contaminated, even those from patients without these medications. On the other hand, the contamination could have been transmitted by hand, as indicated by the large amount of Klebsiella and Enterobacter organisms (Table IIl).33,34 The nurses or the patients themselves could be the main causes of this contamination. In our study, an important number of strains presented a high degree of antibiotic resistance. In particular, 15 Enterobacteriaceae and 2 P aeruginosa strains were highly resistant to ~3-lactams. Thirty-one Enterobacteriaceae and 2 P aeruginosa were resistant to penicillins. One Acinetobacter organism was resistant to all antibiotics tested, and one S aureus was resistant to methicillin. The sensitivity to aminoglycosides seemed to be less often affected, except with P aeruginosa. Nine episodes of fever without bacteremia and one without relationship to the tube-rinsing solution flora occurred during the study. A relationship between the patient’s colonization and tube-rinsing solution flora occurred frequently for both febrile and afebrile patients. For this reason, a higher bacterial viable count could not be associated with induction of an infection. Neverthe-
less, for three febrile patients, throat and feces were colonized with the same bacteria as the predominant bacteria cultured from the corresponding tube-rinsing solution. For one patient, the bacteria, P aeruginosa, was highly resistant to antibiotics. For another patient diarrhea resolved after removal of the tube and, for the third, fever resolved after antibiotic treatment given through the tube. These clinical results, as well as our bacteriological results, allow us to consider that the tube may act, during hospitalization, as an important reservoir of resistant
microorganisms. ACKNOWLEDGMENTS
This work
supported in part by Roussel-Uclafthank C. Mestrude and M. Barouh for technical assistance, Professor M. Scavizzi for scientific discussion, M. Charlot and M. Ducrot, dietitians of our hospital, P. Coninx, MD, for assistance in editing the manuscript, and D. Gosnet for typing the manuscript. was
Sopharga. We
REFERENCES 1. Kern KA, Norton JA: Cancer cachexia. JPEN 12:286-298, 1988 2. Bozzetti F: Effects of artificial nutrition on the nutritional status of cancer patients. JPEN 13:406-420, 1989 3. Copeland EM, Daly JM, Dudrick SJ: Nutrition as an adjunct to cancer treatment in the adult. Cancer Res 37:2451-2456, 1977 4. de Vries EGE, Mulder NH, Houwen B, et al: Enteral nutrition by nasogastric tube in adult patients treated with intensive chemotherapy for acute leukemia. Am J Clin Nutr 35:1490-1496, 1982 5. Reynolds HM, Daly JM, Rowlands BJ, et al: Effects of nutritional repletion on host and tumor response to chemotherapy. Cancer
45:3069-3074, 1980 6. Thurn
Gerdts A, et al: Enteral hyperalimentation of nosocomial infection. J Hosp Infect 15:203-217, 1990 7. Casewell MW: Nasogastric feeds as a source of Klebsiella infection for intensive care patients. Res Clin Forums 1:101-105, 1979 8. Casewell MW, Cooper JE, Webster M: Enteral feeds contaminated with Enterobacter cloacae as a cause of septicaemia. Br Med J
J, Crossley K,
as a source
282:973, 1981 9. Freedland CP, Roller RD, Wolfe BM, et al: Microbial contamination of continuous drip feedings. JPEN 13:18-22, 1989 10. Anderton A: The potential of Escherichia coli in enteral feeds to cause food poisoning: a study under simulated ward conditions. J Hosp Infect 5:155-163, 1984 11. Bengoa JM, Hyde AL, Ducel G, et al: Sûreté bactériologique de la nutrition entérale à débit continu. Schweiz Med Wochenschr
115:903-906, 1985 C, Lipman TO, Geraghty M,
12. Hostetler
et al: Bacterial safety of reconstituted continuous drip tube feeding. JPEN 6:232-235, 1982 13. Grunow JE, Christenson JC, Moutos D: Contamination of enteral nutrition systems during prolonged intermittent use. JPEN 13:23-
14.
25, 1989 Levy J:
Enteral nutrition: An increasingly recognized cause of nosocomial bloodstream infection. Infect Control Hosp Epidemiol
10:395-397, 1989 15. 16.
17.
18.
19.
Pingleton SK: Enteral nutrition as a risk factor for nosocomial pneumonia. Eur J Clin Microbiol Infect Dis 8:51-55, 1989 Jacobs S, Chang RWS, Lee B, et al: Continuous enteral feeding: A major cause of pneumonia among ventilated intensive care unit patients. JPEN 14:353-356, 1990 Anderson KR, Norris DJ, Godfrey LB, et al: Bacterial contamination of tube-feeding formulas. JPEN 8:673-678, 1984 Anderton A: Microbiological aspects of the preparation and administration of naso-gastric and naso-enteric tube feeds in hospitals. A review. Hum Nutr Appl Nutr 37A:426-440, 1983 Bastow MD, Greaves P, Allison SP: Microbial contamination of
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
557 enteral feeds. Hum Nutr Appl Nutr 36A:213-217, 1982 20. Schroeder P, Fisher D, Volz M, et al: Microbial contamination of enteral feeding solutions in a community hospital. JPEN 7:364-
28.
368, 1983 Anonymous: Laits pasteurisés conditionnés destinés à la consommation humaine, Décret No. 74-21, 4 Janvier 1974. Journal officiel de la République Française, Lois et Décrets, 13 Janvier 1974, p526 22. Bauer AW, Kirby WMM, Sherris JC, et al: Antibiotic susceptibility testing by a single standardized disk method. Am J Clin Pathol 45:493-496, 1966 23. Ericson HM, Sherris JC: Antibiotic sensitivity testing: Report of an international collaborative study. Acta Pathol Microbiol Scand [B] (Suppl 217):1-90, 1971 24. Courvalin P, Goldstein F, Philippon A, et al: Choix des antibio-
29.
21.
tiques à étudier
selon les groupes bactériens. IN L’antibiogramme, Courvalin P, Goldstein F, Philippon A, et al (eds). mpc-videom, Paris, 1985, pp 177-188 25. Bush K: Recent developments in beta-lactamase research and their implications for the future. Rev Infect Dis 10:681-689, 1988 26. Davis BD: Mechanism of bactericidal action of aminoglycosides. Microbiol Rev 51:341-350, 1987 27. Michel-Briand Y, Alkhalaf-Haddad B, Bonnin M: Pseudomonas aeruginosa et antibiotiques. Bull Inst Pasteur 87:125-170, 1989
30.
31. 32.
Sykes RB, Matthew M: The beta-lactamases
of gram-negative bacteria and their role in resistance to beta-lactam antibiotics. J Antimicrob Chemother 2:115-157, 1976 Thabaut A, Philippon A, Meyran M: Beta-lactamases of Pseudomonas aeruginosa and susceptibility against beta-lactam antibiotics. Chemioterapia 4:36-42, 1985 van de Klundert JAM, Vliegenthart JS, van Doorn E, et al: A simple method for the identification of aminoglycoside-modifying enzymes. J Antimicrob Chemother 14:339-348, 1984 Duval J: Evolution and epidemiology of MLS resistance. J Antimicrob Chemother 16 (Suppl A):137-149, 1985 Keighley MRB, Mogg B, Bentley S, et al: "Home brew" compared with commercial preparation for enteral feeding. Br Med J 284:163,
1982 33. Casewell
34.
M, Phillips I: Hands as route of transmission for Klebsiella species. Br Med J 2:1315-1317, 1977 Salzman TC, Clark JJ, Klemm L: Hand contamination of person-
nel as a mechanism of cross-infection in nosocomial infections with antibiotic resistant Escherichia coli and Klebsiella-aerobacter. Antimicrob Agents Chemother 7:97-100, 1967 35. Du Moulin GC, Paterson DG, Hedley-White J, et al: Aspiration of gastric bacteria in antacid-treated patients: A frequent cause of postoperative colonisation of the airway. Lancet 1:242-245, 1982
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
Microbial Contamination of Enteral Feeding Tubes Occurring During Nutritional Treatment Véronique Bussy, François Marechal and Salvador Nasca JPEN J Parenter Enteral Nutr 1992 16: 552 DOI: 10.1177/0148607192016006552 The online version of this article can be found at: http://pen.sagepub.com/content/16/6/552
Published by: http://www.sagepublications.com
On behalf of:
The American Society for Parenteral & Enteral Nutrition
Additional services and information for Journal of Parenteral and Enteral Nutrition can be found at: Email Alerts: http://pen.sagepub.com/cgi/alerts Subscriptions: http://pen.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://pen.sagepub.com/content/16/6/552.refs.html
>> Version of Record - Nov 1, 1992 What is This?
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
Microbial Contamination of Enteral Feeding Tubes Occurring During Nutritional Treatment VÉRONIQUE BUSSY, PHD*; FRANÇOIS MARECHAL, MD†;
AND
SALVADOR NASCA, MD‡
From *Institut Jean-Godinot, Laboratoire de Microbiologie Médicale, Reims Cedex; †Hôpital Général, d’Oncologie Médicale, Macon Cedex; and ‡Institut Jean-Godinot, Département d’Oncologie
ABSTRACT. Enteral nutrition is an effective treatment for catabolic patients with normal intestinal absorption. However, tube-fed patients are at risk from superinfection. Our study is the first to evaluate in vivo the microbial contamination of solutions staying in the nasogastric tube of cancer patients receiving nutritional preparations. After daily feeding, the tube was rinsed with nonsterile tap water. Tap water staying overnight in the tube was considered as tube-rinsing solution. Microbial burden of nutritional preparations was determined on the fifth day of enteral nutrition, from opening the first container and throughout feeding. The next day, a sample of the tube-rinsing solution was cultured. All bacterial species were identified and antibiotic susceptibility pattern was assessed. Thirty-one cancer patients were included, 12 on the hospital’s preparations and 19 on commercial feeding. Seven of
Malnutrition occurs frequently in cancer patients and be found in all stages of the diseased The disease itself, or aggressive treatment, often leads to nutritional imbalance. Enteral nutrition is an effective treatment for catabolic patients with normal intestinal absorpcan
tion.2-5
The contamination of nutritional preparations and feeding systems has been associated with colonization or infections complicating enteral feeding.6-9 Colonization of enteric nutritional preparations or contamination with bacterial enterotoxins constitutes a risk for the weak, immunodeficient, and intensive care patient.10 Although nutritional preparations used comply to high microbiologic standards of quality, 11-1 infections due to contaminated preparations are still often reported.14-16 Numerous works have addressed the bacterial contamination of enteral feeding preparations,17-20 but no study has yet evaluated in vivo the microbial colonization of the nasogastric tube. In a prospective study, we determined the extent of the contamination of the tuberinsing solution occurring during nutrition treatment. The antibiotic susceptibility pattern of the cultured strains was also studied.
nowadays
MATERIALS AND METHODS
Patients
They were
patients (29 men, 2 women), (18 head and neck, 6 esophagus, 2
31 consecutive
suffering from
cancer
Reprint requests: Véronique Bussy, Laboratoire de Microbiologie Médicale, Institut Jean-Godinot, B.P. 171, 51056 Reims, France.
the
hospital—and none
Département de Médecine Interne, Unité Médicale, Reims Cedex, France
of the
commercial—preparations were
contaminated. Among the 48 samples collected during feeding, 16 were contaminated, including 10 hospital and 6 commercial preparations. All the 31 tube-rinsing solutions were contaminated and 102 strains cultured. Their median concentration was 10 6 colony-forming units/mL (range 10-10 10 ). The strains were 48 Enterobacteriaceae, 20 group D streptococci, 9 Candida albicans, 9 Pseudomonas aeruginosa, and 16 others. Multiple antibiotic resistance was present in 12 of the 102 strains. Lower resistance was present in 33. The predominant microorganism of the tube-rinsing solution caused a bacterial colonization for three febrile patients. In conclusion, the feeding tube is an important reservoir for multiple antibiotic-resistant bacteria. (Journal of Parenteral and Enteral Nutrition 16: 552-557, 1992)
1 stomach, 4 multiple) who had been hospitalized for at least 48 hours before the nasogastric feeding tube was placed. Median age was 56 (range 33-69). Important dysphagia and/or malnutrition prompted the institution of enteral nutrition by nasogastric tube. When included, they did not receive any antibiotic during the 6 days of the study. Six of them were treated by inhibitors of gastric acid secretion (ranitidine, Glaxo, Paris, France or cimetidine, Smith Kline & French and Beecham, Paris,
colon,
France). Enteral
Feeding
For the first 12 patients, the nutritional preparations were prepared by the hospital dietary service by mixing sterile commercial liquids (Roussel Uclaf-Sopharga, Puteaux, France) and different sterile ingredients such as glucides, oil, vitamins, or inorganic products with a blender. The blender was sterilized daily. For the 19 following patients, clinicians and dietitians decided to use sterile commercially complete preparations (Nutricia, Rueil Malmaison, France). For each patient, three to five containers (500 mL) were infused daily for 3 to 4 hours each, between 8 AM and 10 PM. The nutrition was given via a silicone nasogastric tube (Peters, Bobigny, France) 1250 mm long, with a lumen diameter of 4.2 mm. The container was hung up and connected to the nasogastric tube through a sterile tubule. Infusion flow was regulated with a manual wheel. Three to more than five disconnections occurred daily, whenever required, to chance a container or for taking care of the Datient. After
552 Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
553 each disconnection, or at the end of the day, the tube was rinsed with 20 mL of nonsterile tap water. It was then plugged while filled with tap water. Tap water staying overnight in the tube was considered as tuberinsing solution. Tap water was cultured weekly using a filter method and found free of pathogenic microorganisms throughout the study.
Bacteriological Samples Nutritional preparations. For each patient, a bacterial viable count of corresponding nutritional preparations was determined on the fifth day. Samples of 3 to 5 mL were aseptically taken from the first daily container immediately before beginning the infusion. Two hours after the beginning of the first and the third daily containers, the infusion was stopped for a few minutes. The tube was aseptically disconnected from the container and 3 to 5 mL of the nutritional preparations was collected for bacterial viable counting. All the samples were sent immediately to the laboratory. Rinsing solutions staying overnight in the tube. In a preliminary study with 15 patients, we showed that the same flora contaminated the rinsing solution staying overnight in the tube and the nutritional preparation infusing through the tube. The microbial concentrations obtained were lower in the latter case. They were between 10 and 105 colony-forming units (cfu)/mL (median 10~). In the interest of simplicity, we studied the rinsing solution staying overnight in the tube, before beginning the daily infusion. For each patient, a sample of nasogastric tube-rinsing solution was taken 6 days after insertion of the tube, before the daily feeding. A sterile catheter introduced into the tube allowed us to collect the sample. The end of the catheter was plugged using sterile agar to prevent contamination during mobilization to the esophageal portion of the tube. A needle was inserted into the exposed end of the catheter, and a syringe (20 mL) was adapted to the needle. A positive pressure allowed the expulsion of the agar plug and a 200- to 400-~L sample of liquid was then obtained. The sample was then injected into a sterile flask and sent immediately to the
laboratory.
Bacteriological Procedures Microbial numerations. The samples were diluted by sequential 10-fold steps until a 10-6 dilution in sterile 0.9% sodium chloride was reached; 100 ~L of each dilu-
5% horse blood agar and on were incubated at 37°C for 16 hours. A colony count was then carried out. In the absence of published standards for bacteriological safety of enteral feeding formulas, the standards applicable to milk published by the French legislature were used.21 These regulations place an upper limit of 3.10’ cfu/mL. Identification of microbial species. Bacteria were identified using API 20E for Enterobacteriaceae, API Strep for streptococci, and API 20NE for Pseudomonas and
tion
was
Drigalski
inoculated
agar.
Acinetobacter
on
Agar plates
organisms (Laboratoires Biom6rieux, Lyon, France). We used Chapman agar and the deoxy-
ribonuclease production as
Staphylococcus
categorize staphylococci staphylococcus coagulase
test to
aureus
or
negative. Antimicrobial susceptibility. Antimicrobial susceptibility of the strains cultured from the tube-rinsing solution samples was performed according to a standardized disk method22.’3 (Institut Pasteur Production, Paris, France). Inhibition diameters were measured. A disk susceptibility test was done for each bacterial strain presented in Table I, and for other species, according to previously
described
procedures. 21
to methicillin was detected by the disk diffusion method with a l-llg oxacillin disk. Susceptibility testing was performed at 37°C in Miiller-Hinton agar (Institut Pasteur Production, Paris, France) and at 30°C in Miiller-Hinton agar containing 5% NaCl. For Haemophilus organisms and neisseriae, the search of ¡3-lactamase was done by the chromogenic cephalosporin test (c6finase, Laboratoires Biom6rieux,
Staphylococci susceptibility
Lyon, France). The resistant phenotypes to #-lactams and aminoglycosides were determined for Enterobacteriaceae and Pseudomonas aeruginosa according to previously described methods.25-30 Comparison of strains of the same species. Strains of the same species were compared by serotyping, when they were cultured from both a nutritional preparation and a corresponding tube-rinsing solution sample (service des ent6robact6ries, Institut Pasteur, Paris, France).
Relationship Between Infections
and
Tube-Rinsing
Solution Flora
The occurrence of infections was studied over a period of 10 days after insertion of the tube; criteria were fever above 38°C for at least 1 day and all clinical signs of infection. Cultures were made to search for the origin of an infection (blood, throat, feces, and swabs or samples from other infected sites). We attempted to identify a relationship between bacteria causing infections and corresponding tube-rinsing solution flora. RESULTS
Microorganisms Cultured From Nutritional Preparations Before beginning the daily infusion. Of the 31 samples, corresponding to the 31 patients, 24 were sterile and 7 were contaminated. All seven contaminated samples corresponded to hospital dietary service preparations. Ten different species were cultured as shown in Table II. The concentrations obtained were less than 103 cfu/mL, except in two cases, with 104 cfu/mL Acinetobacter calcoaceticus anitratus for one and l0a cfu/mL Moraxella sp for the other.
Throughout infusion. For 17 patients, 2 samples were on the fifth day. For the 14 others, only 1 sample was obtained from either the first or the third daily container. In total, among 48 samples cultured, 32
obtained
sterile and 16 grew bacteria. Ten of them were associated with hospital preparations. They concerned eight patients, of whom seven had received a preparation
were
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
554 TABLE I
Antibiotics tested for
*
species
most
frequently cultured from tube-rinsing solution samples
Amoxicillin + clavulanic acid. TABLE II Strains cultured from nutritional feeds
TABLE III
Family,
genus,
or
species cultured from the 31 tube-rinsing solutions and concentrations obtained
* Fourteen E
coli,
3 P
mirabilis,
15
Klebsiella,
15
Enterobacter,
1
Citrobacter freundii. t Fourteen S faecalis, 5 S durans, 1 S faecium. $ The other species were categorized in 2 Micrococcus, 2 staphylococcus
already contaminated before infusion. Six positive samples were associated with commercial preparations corresponding to three patients (three samples from the first container and three from the third). Thirty-one strains were cultured from the 16 contaminated samples (Table II). The median concentration was 102 cfu/mL (range: 10-10’). For 12, the microbial concentrations were less than 104 cfu/mL; they were between 105 and 10’ cfu/mL for the 4 others. As shown in Table II, two strains belonged to fecal flora (Klebsiella pneumoniae and Enterobacter cloacae) and one was skin and mucous membrane saprophyte (Candida albicans). These three strains were each from one different sample. Microorganisms Cultured From the Tube-Rinsing Solutions
coagulase negative, 1 bacillus, 1 Corynebacterium, 1 Haemophilus, neisseriae, 1 Saccharomyces cereuisiae, 1 Streptococcus pneumoniae. § Median 106 cfu/mL. =
Antimicrobial Strains
cocci.
Susceptibility of Tube-Rinsing Solution
Enterobacteriaceae, pseudomonads, and Acinetobacter organisms. The susceptibility pattern to {3-lactams is shown in Table IV. All but three Enterobacteriaceae were susceptible to all aminoglycosides; three Escherichia coli were resistant to kanamycin and neomycin. One pseudomonad was resistant to all aminoglycosides, six were resistant to both kanamycin and gentamicin, and one was resistant to kanamycin. One of the three Acinetobacter organisms was highly resistant to all antibiotics. The other two were
All 31 rinsing solutions studied were found to be contaminated with a total of 102 different strains (Table III). The median number of strains by tube was three (range 1-7). Strains belonging to fecal flora were the most frequently cultured. They composed 67%, categorized in 48 Enterobacteriaceae and 20 group D strepto-
1
susceptible.
The resistance to sulfonamides and to tetracyclines was 25% for each. All the strains were susceptible to other antibiotics. Group D streptococci. All but two strains were susceptible to amoxicillin. Resistance to penicillin was always present. High resistance to gentamicin was not present. The susceptibility to macrolides, lincosamides, and sy-
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
555 ,~-Lactams susceptibility of the 48 Enterobacteriaceae, the
TABLE IV 9 pseudomonads, and the 3 Acinetobacter
*
The penicillin resistance was higher for 4 E coli and 1 K pneumoniae. (In vitro, t The cephalosporin resistance was higher for 4 strains of Enterobacter. $ Ticarcillin resistant and (3-lactamase negative. § Ticarcillin resistant and /3-lactamase positive.
nergistins
was
in accordance with the literature.31 All
susceptible to vancomycin and teicoplanin. Staphylococcus aureus. One of the three strains of S aureus was resistant to meticillin, all aminoglycosides, and ciprofloxacin. It was resistant to lincosamides and synergistins with lincosamides and streptogramin A phenotype. The two other strains were susceptible to all strains
were
antibiotics tested. Other germs and C albicans. Other microorganisms were susceptible to all antibiotics tested. The susceptibility of C albicans to antifungal agents was not studied.
Comparison of Strains For two patients, we found a relationship between strains of nutritional sample taken throughout infusion and corresponding tube-rinsing solution. For one patient it was a C albicans, and for the other it was a K pneumoniae.
Relationship Between Infections and Tube-Rinsing Solution Flora
The
of infections could be studied for 20 11 others, clinical data could not be obtained or infected sites were not examined. Ten patients had a temperature above 38°C for at least 1 day. It was related to a systemic candidosis for one patient. The corresponding tube-rinsing solution was free of Candida organisms but presented a strain of E coli with a higher viable count (>3.104 cfu/mL). This strain was related to the bacterial colonization of throat and nose. For nine patients, fever was not related to bacteremia. For three of them, throat and feces were colonized by the bacteria found with a higher viable count in the corresponding tube-rinsing solution. These bacteria cultured with a higher viable count in feces. They were P aeruginosa for one patient and K pneumoniae for two. The strain of P aeruginosa was highly resistant to antibiotics. One of the two patients with K pneumoniae had diarrhea. Diarrhea resolved after removal of the tube. For one of these febrile but nonbacteriemic patients, feces, throat, and nose were colonized by E coli. This strain was found in the corresponding tube-rinsing solution, but without a higher viable count. For two, the throat alone was colonized with bacteria found in the corresponding tuberinsing solution. It was a strain of E coli for the first and occurrence
patients; for the
no
potentiating
organisms cultured from the rinsing solutions
effect of clavulanic
acid.)
strain of C albicans for the second. One other patient had diarrhea, but feces were free of pathogenic microorganisms. The throat and the tube-rinsing solution were both colonized with a Proteus mirabilis, but without higher viable counts. For the last two patients, no relationship was shown. Seven of these febrile patients had empiric parenteral antibiotic treatment after the sixth day, but defervescence was not observed. For the patient with E coli, the antibiotic treatment was modified. Amoxicillin/clavulanic acid was given through the tube and fever resolved. Among the 10 afebrile patients, 2 presented a bacterial colonization related to microorganisms found with a higher viable count in the tube-rinsing solution. There were E cloacae in feces and C albicans in throat for one and Klebsiella oxytoca and E cloacae, both in feces for the other. Five patients presented a bacterial colonization related to microorganisms found in the tube-rinsing solution but without a higher viable count. There were three P aeruginosa and two C albicans, each one of them in feces and/or throat. Two patients did not present any bacterial colonization, and one presented a fecal colonization with C albicans without any relationship with the tube-rinsing solution flora. a
DISCUSSION
When cultured before beginning the infusion, the 31 nutritional preparations studied were sterile in 24 cases and contaminated in 7. These seven had all been prepared by the hospital dietary service. These results confirmed the previously described superiority of commercial preparations, especially regarding microbial quality.~~ 12, 13, 3z The cultured germs were not found in samples from corresponding tube-rinsing solutions (Tables II and
III). The contamination of nutritional preparations during their administration was more frequent than from the beginning of the infusion and also affected commercial preparations. A third of the samples (16/48) showed contamination. In four cases, high viable counts were found (105_10~, cfu/mL), suggesting a defect in administration methods, either by the prolonged stay of the preparations at room temperature or by contamination at the time of setting up the feeding containers. For two of these four highly contaminated samples, the presence of microorganisms usuallv transmitted by hand,‘j3~ 3~ (es-
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
556
pecially E cloacae and K pneumoniae) indicated that the contamination probably occurred when the feeding containers were set up. Among these four highly contaminated samples, only two presented a relationship with the corresponding rinsing solution flora. It concerned a K pneumoniae for one and a C albicans for the other. The two others did not present any relationship with the corresponding tube-rinsing solution flora, despite the presence of
an E cloacae cultured from one. A total of 102 different strains had been cultured from the 31 tube-rinsing solution samples obtained with the 31 patients. The median concentration was 106 cfu/mL (range 10-101°). The contamination of the tube-rinsing solution was most often not related to the contamination of the nutritional preparations. Most strains cultured from the tube-rinsing solutions belonged to fecal flora (67% of the strains), whereas fecal strains had been exceptionally cultured from nutritional preparation samples taken during infusion. Two fecal strains were cultured from two different nutritional samples, but only one was found in the corresponding tube-rinsing solution. Pseudomonas aeruginosa was frequently cultured from the tube-rinsing solutions but never from nutritional preparations. Candida albicans, also frequently cultured from the tube-rinsing solution, was cultured from only one nutritional preparation sample taken during infusion (Tables II and III). We can suggest several reasons for the contamination of the tube-rinsing solutions in our usual practice. On one hand, the contamination could occur during insertion of the tube. It could be caused by the oropharyngeal flora, which is often abundant and rich in Enterobacteriaceae or C albicans, especially in hospitalized cancer patients. In six cases, the patients were treated with inhibitors of gastric acid secretion, which have been shown to allow overgrowth of bacteria in the stomach.35 In these cases, backward contamination of the tube might occur. However, this seems to be a minor way of contamination, because all tube-rinsing solutions were highly contaminated, even those from patients without these medications. On the other hand, the contamination could have been transmitted by hand, as indicated by the large amount of Klebsiella and Enterobacter organisms (Table IIl).33,34 The nurses or the patients themselves could be the main causes of this contamination. In our study, an important number of strains presented a high degree of antibiotic resistance. In particular, 15 Enterobacteriaceae and 2 P aeruginosa strains were highly resistant to ~3-lactams. Thirty-one Enterobacteriaceae and 2 P aeruginosa were resistant to penicillins. One Acinetobacter organism was resistant to all antibiotics tested, and one S aureus was resistant to methicillin. The sensitivity to aminoglycosides seemed to be less often affected, except with P aeruginosa. Nine episodes of fever without bacteremia and one without relationship to the tube-rinsing solution flora occurred during the study. A relationship between the patient’s colonization and tube-rinsing solution flora occurred frequently for both febrile and afebrile patients. For this reason, a higher bacterial viable count could not be associated with induction of an infection. Neverthe-
less, for three febrile patients, throat and feces were colonized with the same bacteria as the predominant bacteria cultured from the corresponding tube-rinsing solution. For one patient, the bacteria, P aeruginosa, was highly resistant to antibiotics. For another patient diarrhea resolved after removal of the tube and, for the third, fever resolved after antibiotic treatment given through the tube. These clinical results, as well as our bacteriological results, allow us to consider that the tube may act, during hospitalization, as an important reservoir of resistant
microorganisms. ACKNOWLEDGMENTS
This work
supported in part by Roussel-Uclafthank C. Mestrude and M. Barouh for technical assistance, Professor M. Scavizzi for scientific discussion, M. Charlot and M. Ducrot, dietitians of our hospital, P. Coninx, MD, for assistance in editing the manuscript, and D. Gosnet for typing the manuscript. was
Sopharga. We
REFERENCES 1. Kern KA, Norton JA: Cancer cachexia. JPEN 12:286-298, 1988 2. Bozzetti F: Effects of artificial nutrition on the nutritional status of cancer patients. JPEN 13:406-420, 1989 3. Copeland EM, Daly JM, Dudrick SJ: Nutrition as an adjunct to cancer treatment in the adult. Cancer Res 37:2451-2456, 1977 4. de Vries EGE, Mulder NH, Houwen B, et al: Enteral nutrition by nasogastric tube in adult patients treated with intensive chemotherapy for acute leukemia. Am J Clin Nutr 35:1490-1496, 1982 5. Reynolds HM, Daly JM, Rowlands BJ, et al: Effects of nutritional repletion on host and tumor response to chemotherapy. Cancer
45:3069-3074, 1980 6. Thurn
Gerdts A, et al: Enteral hyperalimentation of nosocomial infection. J Hosp Infect 15:203-217, 1990 7. Casewell MW: Nasogastric feeds as a source of Klebsiella infection for intensive care patients. Res Clin Forums 1:101-105, 1979 8. Casewell MW, Cooper JE, Webster M: Enteral feeds contaminated with Enterobacter cloacae as a cause of septicaemia. Br Med J
J, Crossley K,
as a source
282:973, 1981 9. Freedland CP, Roller RD, Wolfe BM, et al: Microbial contamination of continuous drip feedings. JPEN 13:18-22, 1989 10. Anderton A: The potential of Escherichia coli in enteral feeds to cause food poisoning: a study under simulated ward conditions. J Hosp Infect 5:155-163, 1984 11. Bengoa JM, Hyde AL, Ducel G, et al: Sûreté bactériologique de la nutrition entérale à débit continu. Schweiz Med Wochenschr
115:903-906, 1985 C, Lipman TO, Geraghty M,
12. Hostetler
et al: Bacterial safety of reconstituted continuous drip tube feeding. JPEN 6:232-235, 1982 13. Grunow JE, Christenson JC, Moutos D: Contamination of enteral nutrition systems during prolonged intermittent use. JPEN 13:23-
14.
25, 1989 Levy J:
Enteral nutrition: An increasingly recognized cause of nosocomial bloodstream infection. Infect Control Hosp Epidemiol
10:395-397, 1989 15. 16.
17.
18.
19.
Pingleton SK: Enteral nutrition as a risk factor for nosocomial pneumonia. Eur J Clin Microbiol Infect Dis 8:51-55, 1989 Jacobs S, Chang RWS, Lee B, et al: Continuous enteral feeding: A major cause of pneumonia among ventilated intensive care unit patients. JPEN 14:353-356, 1990 Anderson KR, Norris DJ, Godfrey LB, et al: Bacterial contamination of tube-feeding formulas. JPEN 8:673-678, 1984 Anderton A: Microbiological aspects of the preparation and administration of naso-gastric and naso-enteric tube feeds in hospitals. A review. Hum Nutr Appl Nutr 37A:426-440, 1983 Bastow MD, Greaves P, Allison SP: Microbial contamination of
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
557 enteral feeds. Hum Nutr Appl Nutr 36A:213-217, 1982 20. Schroeder P, Fisher D, Volz M, et al: Microbial contamination of enteral feeding solutions in a community hospital. JPEN 7:364-
28.
368, 1983 Anonymous: Laits pasteurisés conditionnés destinés à la consommation humaine, Décret No. 74-21, 4 Janvier 1974. Journal officiel de la République Française, Lois et Décrets, 13 Janvier 1974, p526 22. Bauer AW, Kirby WMM, Sherris JC, et al: Antibiotic susceptibility testing by a single standardized disk method. Am J Clin Pathol 45:493-496, 1966 23. Ericson HM, Sherris JC: Antibiotic sensitivity testing: Report of an international collaborative study. Acta Pathol Microbiol Scand [B] (Suppl 217):1-90, 1971 24. Courvalin P, Goldstein F, Philippon A, et al: Choix des antibio-
29.
21.
tiques à étudier
selon les groupes bactériens. IN L’antibiogramme, Courvalin P, Goldstein F, Philippon A, et al (eds). mpc-videom, Paris, 1985, pp 177-188 25. Bush K: Recent developments in beta-lactamase research and their implications for the future. Rev Infect Dis 10:681-689, 1988 26. Davis BD: Mechanism of bactericidal action of aminoglycosides. Microbiol Rev 51:341-350, 1987 27. Michel-Briand Y, Alkhalaf-Haddad B, Bonnin M: Pseudomonas aeruginosa et antibiotiques. Bull Inst Pasteur 87:125-170, 1989
30.
31. 32.
Sykes RB, Matthew M: The beta-lactamases
of gram-negative bacteria and their role in resistance to beta-lactam antibiotics. J Antimicrob Chemother 2:115-157, 1976 Thabaut A, Philippon A, Meyran M: Beta-lactamases of Pseudomonas aeruginosa and susceptibility against beta-lactam antibiotics. Chemioterapia 4:36-42, 1985 van de Klundert JAM, Vliegenthart JS, van Doorn E, et al: A simple method for the identification of aminoglycoside-modifying enzymes. J Antimicrob Chemother 14:339-348, 1984 Duval J: Evolution and epidemiology of MLS resistance. J Antimicrob Chemother 16 (Suppl A):137-149, 1985 Keighley MRB, Mogg B, Bentley S, et al: "Home brew" compared with commercial preparation for enteral feeding. Br Med J 284:163,
1982 33. Casewell
34.
M, Phillips I: Hands as route of transmission for Klebsiella species. Br Med J 2:1315-1317, 1977 Salzman TC, Clark JJ, Klemm L: Hand contamination of person-
nel as a mechanism of cross-infection in nosocomial infections with antibiotic resistant Escherichia coli and Klebsiella-aerobacter. Antimicrob Agents Chemother 7:97-100, 1967 35. Du Moulin GC, Paterson DG, Hedley-White J, et al: Aspiration of gastric bacteria in antacid-treated patients: A frequent cause of postoperative colonisation of the airway. Lancet 1:242-245, 1982
Downloaded from pen.sagepub.com at SHANTOU UNIV on May 13, 2014
