Decanting—a source of contamination of enteral feeds?

CLINICAL NUTRITION (1990) 9: 157-162 I Longman Group UK Ltd 1990

Decanting-a Feeds?

0261-5614r90,0009_015~~SlOM)

Source of Contamination of Enteral

A. Anderton* and K. E. Aidoot *Department of Bioscience and Biotechnology, University of Strathclyde, Royal College Building, 204 George Street, Glasgow Gl lXW, U.K., and *Department of Applied and Life Sciences, The Queen’s College, 1 Park Drive, Glasgow G3 6LP, UK ! Reprint requests to A.A.)

ABSTRACT The techniques of opening and decanting ready-to-use enteral feeds packaged in bottles (crown-cap and screw-cap), cans and tetrapaks were evaluated as potential routes for the contamination of these feeds. It was found that the outsides of the feed containers, bottle openers, scissors and the experimenters’ hands all acted as sources of contamination during the transfer of feeds to the nutrient container. The main source of contamination appeared to be the experimenters’ hands with counts up to lo2 cfu ml-’ being recorded for feeds that had been decanted from screw-cap bottles, cans and tetrapaks by experimenters with either unprotected bare hands or hands experimentally contaminated with K. aerogenes. Levels of contamination and the number of samples contaminated after opening and decanting were consistently higher for cans and tetrapaks than for crown-cap or screw-cap bottles. Disinfection of feed containers followed by the use of sterile gloves and/or disinfected openers yielded bacteria-free feed from all the types of feed container studied.

INTRODUCTION Many authors have reported that the administration of contaminated enteral feeds can lead to colonisation and subsequent serious infections in hospital patients [l-5]. Sources of contamination include the feed ingredients, inadequately cleaned equipment, personnel, and the patients themselves [3, 6-101. According to a survey of enteral feeding carried out in hospitals in England and Wales, the proportion of ready-to-use liquid feeds administered has increased from 47O, in 1980 to 96O,, in I986 [ 1 I]. However, although these ready-to-use liquid feeds are less prone to contamination, because no inhospital preparation is required [9, 121, the procedures involved in the opening and decanting of the feed from the original container may lead to the contamination of the feed before it reaches the sterile nutrient container [13-l 51. The role of faulty handling procedures during the assembly of administration sets was highlighted by Schreiner et al. [16] and Schroeder et al 1171 who reported that the range of microorganisms isolated from the hands of nurses matched those detected in the patients’ enteral feeds. More recently, Anderton and Aidoo [ 151 have demonstrated that sterile feeds may become contaminated with bacteria during the assembly and delivery of feeds from a representative selection of enteral feeding systems. However, one aspect of the handling procedure that they did not examine was the potential entry of microbial contaminants into the feed when it was decanted from the

original container into the nutrient container. The present study was therefore carried out to examine the procedures used in the opening and decanting of a range of different types of pre-packed liquid feeds and to determine the resultant levels of contamination.

MATERIALS

AND METHODS

Feed containers The ready-to-use sterile liquid feeds used in this study were packaged in the following types of container:

(1) bottles with crown-cap covers (Fortison Standard, Cow and Gate Ltd, UK), with screw-cap covers (Ensure Liquid Nutrition, Abbot Laboratories, UK), (3) ring-pull cans (Clinifeed ISO, Roussel Laboratories Ltd, UK), (4) tetrapak waxed paper containers (Wyeth Nutrition Standard, Wyeth Laboratories, UK).

(2) bottles

The feeds packaged in tetrapaks were included in the study because, although the commercially prepackaged products are not widely used for enteral feeds in the UK, wax cartons may be used for the storage of reconstituted dried feeds [14] and tetrapaks are also frequently used for sip feeds. Patients receiving the latter may be subject to similar sources and consequences of microbial contamination as those on enteral feeds [ 181. 157

1%

DECANTINGA

Microbiological

SOURCE OF CONTAMINATION

OF ENTERAL FEEDS?

methods of examination

Viable counts were made on serial ten-fold dilutions of each sample in 0.1 O0 peptone water; 0.1 ml of each dilution was spread over the surface of plate count agar (Oxoid) and the plates incubated aerobically at 37°C for 24 h. Counts were made in triplicate and the mean count expressed as cfu ml-‘. Isolates obtained from experimenters’ hands and decanted feeds were identified using standard techniques [ 19,201.

Experimental contamination of hands Klebsiella aerogenes (NCTC No. 8172) was grown in Nutrient Broth No. 2 (Oxoid) for 18 h at 37” C. The bacterial suspensions were then diluted in O.lO, peptone water (Oxoid), painted on the thumb, fingers and palms of both hands and air-dried to give a concentration of c.103 cfu cm-z of skin [7]. The level of contamination on the skin was checked by carrying out preliminary experiments in which the hands were contaminated as described above following which swab samples were taken; the hands were also pressed onto the surface of plates of agar.

20 500ml samples of feed were prepared for each experiment. (2) The outside of each bottle, can or tetrapak was not disinfected. Containers were then opened and the feed decanted as described in (i) and (ii) above. 20 500ml samples of feed were prepared for each experiment. (3) The outsides of 20 crown-cap bottles and 20 tetrapaks were disinfected by swabbing with 70°, alcohol and/or flaming. An experimenter wearing sterile gloves then opened 10 crown-cap bottles and 10 tetrapaks using (i) a non-disinfected bottle opener or scissors and (ii) an experimentally contaminated bottle opener or scissors. The contents of each feed container were then decanted into a sterile Erlenmeyer flask. This procedure was repeated for a further 5 days, using three feed containers for each procedure each day, a total of 25 containers being opened by each method. The bottle openers and scissors were not cleaned or disinfected over the 6 day period of this study. All the flasks were incubated at 37” C for 24 h in order to determine whether low levels of contamination had been introduced during opening and decanting the feed. Samples were taken for viable counts at 0, 4 and 24h.

Experimental contamination of bottle opener and scissors The bottle opener and a pair of scissors were dipped for 30 seconds in a lOOO-fold dilution of an 18h broth culture of K. aerogenes and then air-dried yielding a concentration of lo*-lo3 cfu cm-2. Levels of contamination were determined in preliminary experiments by taking surface swabs using a 1 cm2 template.

Opening and decanting of feeds In all of the following experiments the cans, bottles and tetrapaks of feed were shaken vigorously prior to opening and decanting 500ml of each feed into a sterile Erlenmeyer flask. The following procedures were employed: (1) The outside of each bottle, can or tetrapak was disinfected by swabbing with 700/;, alcohol and/or flaming. Containers were then opened and the feed decanted by an experimenter: (i) wearing sterile gloves and using a disinfected bottle opener or scissors. (ii) with bare hands and using a disinfected bottle opener or scissors. (iii) with hands, bottle opener and scissors experimentally contaminated with Klebsiella aerogenes as described above.

RESULTS Effects of wearing sterile gloves and using a disinfected bottle opener and scissors to open and decant the feed from the feed containers No organisms were detected in any of the feed samples from disinfected containers opened by an experimenter wearing sterile gloves and using a disinfected bottle opener or scissors (Table 1). When the feed containers were not disinfected no organisms were detected in any of the samples taken at 0 h. However, over the period of the experiment, the contaminants introduced from the outside of the containers multiplied in the feed and the results presented for 24 h give the true indication of the number of feeds that had been contaminated at the outset of the experiment. Samples collected at 0 h and 4 h help to give an indication of the comparative level of contamination introduced during the procedure of opening and decanting the feeds. Thus it can be seen that in the samples decanted from non-disinfected cans and tetrapaks contamination was detected in 12 of the 24 h samples from cans and 6 of those from tetrapaks. Initial levels of contamination were low, no organisms were detected in any tetrapak samples at 0 h or 4 h and lo*-lo2 cfu mll’ were detected in 3 of the 4h samples

CLINICAL

Table 1 Number of feeds (out of 20) in which contaminants were detected after feed containers decanted wearing sterile gloves and using a disinfected bottle opener or scissors* Time feed incubated at 37’C (h)

Crown-cap bottle n = 20

Feed container Screw-cap bottle n = 20

disinfected Can n = 20

Tetrapak n = 20

Crown-cap bottle n = 20

were opened

Feed container Screw-cap bottle n = 20

0 4

NC NC

NC NC

NC NC

NC NC

NC NC

NC NC

24

NC

NC

NC

NC

NC

NC

and the feed

not disinfected Can n = 20

Tetrapak n = 20

NC 3 (IO’-10’)

NC NC 6

t lW-10’)

(lOE0.) NC = no organisms detected in feed. * = figures in parenthesis indicate the range of levels of contamination

(cfu ml I).

Table 2 Number of feeds (out of 20) in which contaminants were detected decanted with bare hands and using a disinfected bottle opener or scissors* Time feed incubated at 37°C (h)

Crown-cap bottle n = 20

0

NC

Feed container Screw-cap bottle n = 20 NC

after feed containers

disinfected Can n = 20

Tetrapak n = 20

NC

NC

159

NUTRITION

Crown-cap bottle n = 20 NC

were opened

Feed container not disinfected Screw-cap Can bottle n = 20 n = 20

NC

NC

NC

NC

24

NC

NC

NC

NC

Tetrapak n = 20

Cl&

4 (10’)

(IO&, 18 (105-10’y

(IO& 15 (101-109)

NC

4

(10&06)

and the feed

detected in feed. * = figures in parenthesis indicate the range of levels of contamination (cfu ml ‘)

NC = no organisms

cans. No contamination was detected in any of the feed samples from non-disinfected crown-cap or screwcap bottles.

from

Effect of using bare hands and a disinfected bottle opener or scissors to open and decant the feed from feed containers No organisms were detected in any of the feed samples from either the disinfected or non-disinfected crowncap or screw-cap bottles opened by an experimenter with bare hands and using a disinfected bottle opener (Table 2). Contamination was detected in 8 of the samples from disinfected cans and 4 of the samples from disinfected tetrapaks. However, once again the initial levels of contamination were low with no organisms being detected at Oh and lOi-1Or cfu ml-’ being detected in two samples from cans and tetrapaks after 4 h. In both this experiment and a preliminary experiment in which two experimenters with bare hands each opened 12 disinfected cans and tetrapaks, the bacterial isolates from the contaminated feeds matched those that were present on the hands of the experimenters. Coliforms were the predominant organisms isolated but staphylococci,

including some coagulase-positive strains, were also found. When non-disinfected cans and tetrapaks were opened with bare hands the level of contamination and the number of samples contaminated were higher than when sterile gloves were worn (Tables 1 and 2) or when the containers were disinfected (Table 2). Thus, in the samples decanted with bare hands from non-disinfected cans and tetrapaks, contamination was detected in 18 of the samples from cans and 15 of those from tetrapaks. Initial levels of contamination in cans and tetrapaks were similar, 10’ cfu ml-’ being detected in 6 samples from cans and 4 samples from tetrapaks at 0 h.

Effect of opening and decanting the feedfrom disinfected feed containers with hands, bottle opener and scissors experimentally contaminated with K. aerogenes When the feed containers were opened with hands, bottle opener and scissors experimentally contaminated with K. aerogenes contamination was detected in 5 samples from crown-cap bottles, 10 from screw-cap bottles, 18 from tetrapaks and 20 from cans (Table 3).

160

DECANTING-A

SOURCE OF CONTAMINATION

OF ENTERAL

FEEDS?

Number of feeds (out of 20) in which K. aerogenerwas detected after the feed containers were opened and the feed decanted with hands, bottle opener and scissors experimentally contaminated with K. aerogenes*+ Table 3

Time feed incubated at 37°C (h)

Crown-cap bottle n = 20

0

NC

4

NC

Screw-cap bottle n = 20

(160’)

24 (1505)

(10~~10~) 10 (109-10’0)

Can n = 20

(lo&, (:& (lO9YOl.)

Tetrapak n = 20

(1901) (:09) 18 (109-10’0)

NC = no organisms detected in feed. l = figures in parenthesis indicate the range of levels of contamination (cfu ml I). + = feed containers were disinfected prior to opening.

Initial levels of contamination at 0 h ranged from i 10” cfu ml-l for the crown-cap bottles to lOi-lo* cfu ml-’ for the samples from cans. When the hands, bottle openers and scissors were experimentally contaminated with K. aerogenes the level of contamination and the number of samples contaminated after opening and decanting were consistently higher than with bare hands. K. aerogenes was the predominant organism detected in the feed in all of these experiments and the feed showed a characteristic ropiness in contaminated screw-cap, can and tetrapak samples after 8 h and in contaminated crown-cap samples after 24 h.

Effect of using non-disinfected and experimentally contaminated openers and scissors to open the feed containers Experimental contamination of both scissors and bottle openers resulted in more feeds being contaminated than when non-disinfected scissors or openers were used. Also both the number of feeds that were contaminated and the level of contamination in feeds from tetrapaks were consistently higher than for samples from bottles. Thus, over the 6 days of the experiment, 13 samples from tetrapaks opened with contaminated scissors were found to be contaminated with lo’-lOi K. aerogenes ml-’ at 0 h whereas K. aerogenes was only found in six samples from bottles (contamination was not detected until the samples had been incubated for 4 h). Similarly, the number of contaminated samples from tetrapaks opened with non-disinfected scissors was six as compared with only one from crown-cap bottles opened with a non-disinfected bottle opener. Levels of contamination were low, the tetrapak samples had to be incubated for 4 h and the bottle sample for 24 h before contaminants were detected.

DISCUSSION Many reports have shown that the administration of contaminated enteral feeds can lead to colonisation and subsequent serious infections in patients [l-5, 121. In the present study the outsides of the feed containers, the bottle openers, the scissors and the experimenters’ hands all acted as sources of contamination during the transfer of feeds to the nutrient container. Thus, when non-disinfected feed containers were opened and the feed decanted wearing sterile gloves and using disinfected bottle openers or scissors, no contamination was detected in samples from crown-cap or screw-cap bottles, but the feed from the cans and the tetrapaks was contaminated by organisms from their surfaces. More samples from cans were contaminated than from tetrapaks. This may be explained by the fact that during decanting, feed from the cans always collected in the part of the groove around the top of the can next to the opening and would have been contaminated by any dust or microorganisms trapped there. The role of equipment in the contamination of enteral feeds was demonstrated by the experiments in which non-disinfected and experimentally contaminated bottle openers and scissors were used to open the feed containers. Samples of feed from both crown-cap bottles and tetrapaks were contaminated in this way. Initial levels of contamination in the feed from tetrapaks were as high as 10’ cfu ml-i. The main source of contamination, however, seemed to be the experimenters’ hands, and counts up to lo2 cfu ml-’ were recorded for feeds that had been decanted from screwcap bottles, tetrapaks and cans by experimenters with either unprotected bare hands or experimentally contaminated hands. Samples from cans were the most frequently and the most highly contaminated. This may be because it is necessary to put a thumb on the rim next to the opening to provide leverage to open the ring

CLINICAL

pull and the feed is then poured over this contaminated rim. Skin contaminants were also frequently isolated from samples from the tetrapaks which were of a design that required to be opened by lifting the side flap and flattening it with the fingertips before cutting with scissors. However, it should be noted that the design of some types of paper containers is such that there is a tendency to use the finger to pull out the spout in order to decant the feed, thus posing a greater risk of skin contaminants being transferred to the feed. Both Bastow et al. [9] and Anderton [21] demonstrated that if the initial level of contamination introduced into the feed was lo’-lo* cfu ml-’ the organisms would multiply to yield 105-lo6 cfu ml-’ after 8 h. Various estimates have been made of the number of bacteria that need to be ingested for bowel colonization to occur, but a number of authors have reported that in healthy volunteers the ingestion of 104-106 E. coli, Klebsiella spp. or Pseudomonas spp. organisms produced detectable faecal counts [22-241. Subsequently, Pottether et al. [2] suggested that lo4 organisms ml-’ of feed are sufficient to colonise the digestive tract of hospital patients. Thus, compromised patients who receive contaminated enteral feeds are at risk of gastrointestinal colonisation and even septicaemia from Gram negative bacilli introduced during the preparation and administration of these feeds [3, 5, 25-271. Also, Casewell [ 121 has suggested that it may not only be individual patients who suffer gastroenteritis, colonisation or infection, since it is also possible that contamination of feeds with Gram negative bacilli bearing plasmids that confer multiple antibiotic resistance may provide new faecal reservoirs for the dissemination of these organisms amongst groups of high-risk patients. The results of the present study showed that contamination of the feeds occurred even when the sterile feed was transferred directly from the can, tetrapak or bottle into a sterile flask. This is contrary to the findings of Bastow et al. [9] who reported that there was no contamination of feed during the procedure of filling nutrient containers. Possible reasons for these apparently contradictory results may include the fact that the nutrient containers filled by Bastow et al. [9] were part of a special study in which it seems likely that particular care would be taken with this procedure. However, in the present study some of the feed containers were deliberately not disinfected and the outsides were handled in a way that simulated the faulty handling procedure that might occur under ward conditions [28]. Thus, under carefully controlled experimental conditions it is possible to avoid contamination of feeds, whereas this is not so easy for the nursing staff on a busy ward. This view is confirmed by the findings of Schroeder et al. [ 171 who, in an evaluation of the procedures for care and preparation of feeding solutions, observed much

NUTRITION

161

higher contamination rates under conditions of actual use than in a careful simulation. Contamination of feeds appears to be cumulative and is related to the many manipulations of the feed and feeding systems between preparation of the feed and the end of administration. The role of faulty handling procedures during the assembly of administration sets has already been discussed by a number of workers [ 1% 171 and the introduction of microorganisms during system manipulation has been highlighted by Cracker et al. [29]. The present study demonstrates that the opening and decanting of feeds is another potential source of contamination. However, it is important to emphasize that in the present study disinfection of the feed containers followed by the use of sterile gloves and/or disinfected openers yielded bacteria-free feed from all the types of feed container examined. It can therefore be concluded that the disinfection of feed containers and openers and the use of gloves in all procedures involved in the opening and decanting of feeds will help to minimize the risk of contamination of the feeds. It should, however, be stressed that ideally the sterile feeds should be packaged in a way that requires minimal manipulation [18] and, as suggested by Casewell and Philpott-Howard [5], ‘there is a need for a commercially available bacteriologically clean feed that is presented in a closed system that includes the tubing that connects to the nasogastric tube’.

ACKNOWLEDGEMENT

This study was supported by a research grant from Cow and Gate Clinical Products Division/Nutrition.

REFERENCES

[II Casewell M W 1979 Nasogastric feeds as a source of

PI

[31

[41

PI

klebsiella infection for intensive care patients. Research Clinical Forums 1: 101-105 Pottecher B, Goetc M L, Jacquemaire M A, Reeb E, Lavillaureix J 1979 Enterocolites infectiuses chez des malades de reanimation alimentes par sonde nasogastrique. Annales de L’Anesthesiologie Francaise 20: 595-602. Casewell M W, Cooper J E, Webster M 1981 Enteral feeds contaminated with Enterobacter cloacae as a cause of septicaemia. British Medical Journal 282: 973 Baldwin B A, Zagoren A J, Rose N 1983 Bacterial contamination of continuous infused enteral alimentation with needle catheter jejunostomy. Journal of Parenteral and Enteral Nutrition 8: 30-33 Casewell M W, Philpott-Howard J 1983 Septicaemia from inadvertent intravenous administration of enteral feeds. Journal of Hospital Infection 4: 403-405

162

DECANTING-A

SOURCE OF CONTAMINATION

161Gutman

OF ENTERAL FEEDS?

L T, Idriss Z H, Gehlbach M D, Blackman MD 1976 Neonatal staphylococcal enterocolitis: association with indwelling feeding catheters and S. aureus colonization. Journal of Paediatrics 8: 836-839 [71 Casewell M W, Phillips I 1977 Hands as a route of transmission for Klebsiella spp. British Medical Journal 2: 1315-1317 enteral feed. British PI Gill K J. Gill P 1981 Contaminated Medical Journal 282: 1971 I91 Bastow MD, Greaves P, Allison S P 1982 Microbial contamination of naso-gastric feeds. Human Nutrition: Applied Nutrition 36A: 213-217 quality of products [lOI Anderton A 1986 Microbiological used in enteral feeds. Journal of Hospital Infection 7: 68-73 r111 Green C, Tredger J, Dickerson J W T 1987 Enteral feeding: A survey to investigate current practices and attitudes of dietitians. Human Nutrition: Applied Nutrition 41A: 360-363 hazards of [121 Casewell M W 1982 Bacteriological contaminated enteral feeds. Journal of Hospital Infection 3: 329-331 u31 White W T III, Acuff T E, Sykes T R, Dobbie R P 1979 Bacterial contamination of enteral nutrition solution: a preliminary report. Journal of Parenteral and Enteral Nutrition 3: 459-461 1141 Nugent M, Hansel1 D T, Gray G R 1987 Bacterial contamination of reconstituted and commercially prepared enteral feeds. Clinical Nutrition 6: 21-24 1151 Anderton A, Aidoo K E 1988 The effect of handling procedures on microbial contamination of enteral feeds. Journal of Hospital Infection 11: 364-372 I161 Schreiner R L, Eitzen H, Gfell MA, Kress S, Gresham E L, French M, Moye L 1979 Environmental contamination of continuous drip feedings. Paediatrics 63: 232-237 1171 Schroeder P, Fischer D, Volz M, Plaoucek J 1983 Microbial contamination of enteral feeding solutions in a community hospital. Journal of Parenteral and Enteral Nutrition 7: 364-368 WI Anderton A, Howard J P, Scott D S 1986 Microbiological Control in Enteral Feeding: A Guidance

Submission date: 25 August

Document; Parenteral and Enteral Nutrition Group of the British Dietetic Association, UK Commission on Microbiological t191 International Specifications for Food 1978 Microorganisms in Food. 1 Their significance and methods of enumeration. Toronto: University of Toronto Press WI Cowan S T 1984 Cowan and Steel’s Manual for the Identification of Medical Bacteria 2nd Ed. Cambridge, Cambridge University Press PI Anderton A 1984 The potential of Escherichia coli in enteral feeds to cause food poisoning: a study under simulated ward conditions. Journal of Hospital Infection 5: 155-163 WI Buck A C, Cooke E M 1969 The fate of ingested Pseudomonas aertcginosa in normal persons. Journal of Medical Microbiology 2: 521-525 J Z, Doak P B, Taylor D E M, North [231 Montgomerie J K D, Martin W J 1970 Klebsiella in faecal flora of renal transport patients. Lancet ii: 787-792 v41 Shooter R A, Faiers M C, Cooke E M, Breaden A L, O’Farrell S M 1971 Isolation of Escherichiu coli, Pseudomonas aeruginosa and Klebsiella from food in hospitals, canteens and schools. Lancet ii: 390-392 1251 Shooter R A, Cooke E M, Gaya H, Kumar P, Pate1 N, Parker M T, Thorn B T, France D R 1969 Food and medicaments as possible sources of hospital strains of Pseudomonas aeruginoso. Lancer i: 1227-1229 WI Cooke EM, Shooter R A, Kumor P J, Rousseau A S, Foulkes A C 1970 Hospital food as a possible source of Escherichia coli in patients. Lancet i: 436-437 i271 Casewell M W, Phillips I 1978 Food as a source of Klebsiella species for colonization and infection of intensive care patients. Journal of Clinical Pathology 31: 841-849 of intravenous P81 Walter C W 1978 Bacterial contamination infusions due to faulty technique. In Advances in Parenteral Nutrition, I D A Johnson, ed. Proc. Intemat. Symp. Bermuda 1977 MTP Press Ltd, UK 1291 Cracker K S, Krey S H, Markovic M, Steffee W P 1986 Microbial growth in clinically used enteral delivery systems. American Journal of Infection Control 14: 250256

1988. Accepted after revision: 17 April 1989