Archives of Disease in Childhood 1996; 74: F99-F104 F99 Reservoirs of coagulase negative staphylococci in preterm infants K Eastick, J P Leeming, D Bennett, M R Millar Bristol Public Health Laboratory, Bristol Royal Infirmary, Marlborough Street, Bristol BS2 8HW K Eastick J P Leeming University Department of Pathology and Microbiology D Bennett M R Millar Correspondence to: Dr J P Leeming. Accepted 22 November 1995 Abstract This investigation was undertaken to determine the magnitude of, and inter- relations between, reservoirs of coagulase negative staphylococci on infants' skin at various sites (including sites used for insertion of intravascular catheters) and in faeces during the first six months of life. Sites with large numbers of coagulase negative staphylococci were identified by sampling 16 skin sites and stools from 20 preterm neonates at 8-30 days of life. A more detailed survey of numbers and types of coagulase negative staphylococci in stool and at six skin sites of 10 preterm infants was then performed over the first six months of life. Isolates of coagulase negative staphylococci were collected and characterised by speciation, antibiotic susceptibility profiling, and plasmid restriction fragment length polymor- phism analysis. Large, relatively stable reservoirs were identified in the faeces, around the ear, and in the axilla and nares. Skin on the forearm and leg, sites at which peripheral catheters are frequently sited, carried small unstable numbers of coagulase negative staphylococci, which were usually indistinguishable from coagulase negative staphylococci isolated from other body sites on the same baby. Contamination of catheter insertion sites with coagulase negative staphylococci from reservoir sites on the same baby could explain these observations. These data suggest that interventions reducing cross-contamination between sites on the same baby might be as important in preventing coagulase negative staphylococcal bacteraemias as measures taken to prevent cross infection between babies. Procedures which are likely to result in heavy coagulase negative staphy- lococcal contaminaton of the hands of healthcare staff, such as changing soiled nappies, should receive particular atten- tion. (Arch Dis Child 1996; 74: F99-F104) Keywords: coagulase negative staphylococci, pretenn infants, skin sites, cross infection. Coagulase negative staphylococci are a fre- quent cause of bacteraemia in infants in inten- sive care units. 1-6 Treatment may require the use of expensive and potentially toxic glycopeptide antibiotics, because isolates of coagulase negative staphylococci are often resistant to multiple antibiotics. Although rarely associated with death, coagulase negative staphylococcal bacteraemia may considerably increase the duration of hospital stay.7 Most systemic infections with coagulase negative staphylococci are intravascular catheter associated infections.8-'4 Infection may arise as a result of contamination of the outer surface of the catheter at the time of insertion, or by subsequent invasion of the catheter tunnel by coagulase negative staphylo- cocci from the skin. Alternatively, the lumen of the catheter may become colonised, usually via the catheter hub, resulting in infusion of bacteria directly into the bloodstream. Thin, immature skin provides a limited barrier to the ingress of bacteria around catheters, and immunological deficits, including poor opsonic activity,15 increase the probability that invading bacteria will survive. These factors predispose preterm infants to bacteraemia with coagulase negative staphylococci. Coagulase negative staphylococci can be isolated from the skin of most preterm infants on special care units. Most investigators have reported low numbers, however, on the skin of premature infants relative to those found on adult skin. Several authors have observed unpredictable day to day variation in the numbers and antibiotic resistance profiles of coagulase negative staphylococci isolated from neonatal skin.1618 These data suggest that these staphylococci at the sites studied, includ- ing sites used for intravascular catheter inser- tion, constitute transient rather than stable resident populations. Although coagulase negative staphylococci on adult skin may be stable over many years, the transient residence of microflora is well established for many other bacteria. 19 Transient micro-organisms are thought to be contaminants of the skin which are rapidly displaced because they are unable to establish a stable ecological niche. The source of the tran- sient coagulase negative staphylococci could either be exogenous (derived from other infants or adults, probably from the hands of carers) or endogenous (spread from elsewhere on the same infant, again probably through handling by adults). The latter possibility is more consis- tent with the observation that in the first one to two weeks of life a trend of increasing coagulase negative staphylococcal counts is superimposed on the fluctuations in bacterial numbers.1618 20 If this is so, it should be possible to identify one or more reservoirs with high and relatively stable populations of coagulase negative staphylococci, and to show that the bacteria on other body sites constitute, to some extent, a subset of the organisms present at the reservoir sites, probably combined with some coagulase on July 29, 2020 by guest. Protected by copyright. http://fn.bmj.com/ Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.74.2.F99 on 1 March 1996. Downloaded from
Transcript
Archives of Disease in Childhood 1996; 74: F99-F104 F99
Reservoirs of coagulase negative staphylococci inpreterm infants
K Eastick, J P Leeming, D Bennett, M R Millar
Bristol Public HealthLaboratory, BristolRoyal Infirmary,Marlborough Street,Bristol BS2 8HWK EastickJ P Leeming
University DepartmentofPathology andMicrobiologyD BennettM R Millar
Correspondence to:Dr J P Leeming.Accepted 22 November 1995
AbstractThis investigation was undertaken todetermine the magnitude of, and inter-relations between, reservoirs of coagulasenegative staphylococci on infants' skin atvarious sites (including sites used forinsertion of intravascular catheters) andin faeces during the first six months oflife.Sites with large numbers of coagulasenegative staphylococci were identified bysampling 16 skin sites and stools from 20preterm neonates at 8-30 days of life. Amore detailed survey of numbers andtypes of coagulase negative staphylococciin stool and at six skin sites of 10 preterminfants was then performed over the firstsix months of life. Isolates of coagulasenegative staphylococci were collected andcharacterised by speciation, antibioticsusceptibility profiling, and plasmidrestriction fragment length polymor-phism analysis. Large, relatively stablereservoirs were identified in the faeces,around the ear, and in the axilla andnares. Skin on the forearm and leg, sites atwhich peripheral catheters are frequentlysited, carried small unstable numbers ofcoagulase negative staphylococci, whichwere usually indistinguishable fromcoagulase negative staphylococci isolatedfrom other body sites on the same baby.Contamination of catheter insertion siteswith coagulase negative staphylococcifrom reservoir sites on the same babycould explain these observations.These data suggest that interventions
reducing cross-contamination betweensites on the same baby might be asimportant in preventing coagulase negativestaphylococcal bacteraemias as measurestaken to prevent cross infection betweenbabies. Procedures which are likely toresult in heavy coagulase negative staphy-lococcal contaminaton of the hands ofhealthcare staff, such as changing soilednappies, should receive particular atten-tion.(Arch Dis Child 1996; 74: F99-F104)
Coagulase negative staphylococci are a fre-quent cause of bacteraemia in infants in inten-sive care units. 1-6 Treatment may requirethe use of expensive and potentially toxicglycopeptide antibiotics, because isolates ofcoagulase negative staphylococci are oftenresistant to multiple antibiotics. Although
rarely associated with death, coagulase negativestaphylococcal bacteraemia may considerablyincrease the duration of hospital stay.7Most systemic infections with coagulase
negative staphylococci are intravascularcatheter associated infections.8-'4 Infectionmay arise as a result of contamination of theouter surface of the catheter at the time ofinsertion, or by subsequent invasion of thecatheter tunnel by coagulase negative staphylo-cocci from the skin. Alternatively, the lumen ofthe catheter may become colonised, usually viathe catheter hub, resulting in infusion ofbacteria directly into the bloodstream. Thin,immature skin provides a limited barrier tothe ingress of bacteria around catheters, andimmunological deficits, including poor opsonicactivity,15 increase the probability that invadingbacteria will survive. These factors predisposepreterm infants to bacteraemia with coagulasenegative staphylococci.
Coagulase negative staphylococci can beisolated from the skin of most preterm infantson special care units. Most investigators havereported low numbers, however, on the skin ofpremature infants relative to those found onadult skin. Several authors have observedunpredictable day to day variation in thenumbers and antibiotic resistance profiles ofcoagulase negative staphylococci isolated fromneonatal skin.1618 These data suggest thatthese staphylococci at the sites studied, includ-ing sites used for intravascular catheter inser-tion, constitute transient rather than stableresident populations.
Although coagulase negative staphylococcion adult skin may be stable over many years,the transient residence of microflora is wellestablished for many other bacteria. 19Transient micro-organisms are thought to becontaminants of the skin which are rapidlydisplaced because they are unable to establish astable ecological niche. The source of the tran-sient coagulase negative staphylococci couldeither be exogenous (derived from other infantsor adults, probably from the hands of carers) orendogenous (spread from elsewhere on thesame infant, again probably through handlingby adults). The latter possibility is more consis-tent with the observation that in the first one totwo weeks of life a trend of increasing coagulasenegative staphylococcal counts is superimposedon the fluctuations in bacterial numbers.1618 20If this is so, it should be possible to identify oneor more reservoirs with high and relativelystable populations of coagulase negativestaphylococci, and to show that the bacteria onother body sites constitute, to some extent, asubset of the organisms present at the reservoirsites, probably combined with some coagulase
on July 29, 2020 by guest. Protected by copyright.
negative staphylococci from exogenous con-tamination. In this study we attempted toidentify large reservoirs of coagulase negativestaphylococci in preterm infants, to investigatethe stability of these reservoirs over the first sixmonths of life, and to compare reservoir iso-lates with those at transient colonisation sitessuch as catheter insertion sites.
MethodsPRELIMINARY LOCATION OF MAJOR RESERVOIRSOF COAGULASE NEGATIVE STAPHYLOCOCCISites harbouring large reservoirs were identifiedin a pilot survey of 20 neonates (gestational agerange 28-33 weeks and age at sampling 8-31days). Each was sampled on one occasion atnine to 13 of 18 sites, including the ear(anterior and posterior pinna separately);anterior nares, nose, axilla, umbilicus/peri-umbilicus, neck, antecubital fossa, upper arm,forearm, chest, popliteal fossa, inguinal folds,outer thigh, inner thigh, lower leg, toe cleftsand faeces. Sampling and enumerationmethods were as described below.
LONGITUDINAL SURVEYTen infants admitted within 24 hours of birthto the neonatal intensive care unit at StMichael's Hospital, Bristol, were recruitedbetween September 1993 and March 1994.Criteria for inclusion in the study weregestational age 30 weeks or less, informedconsent of parents, and a sufficiently stableclinical condition to allow sampling at 4 daysof age.Each infant was sampled on days 4, 8 ,12,
16, 21, and 28 of age and monthly thereafteruntil discharged home, and then at one to twomonthly intervals until 6 months of age. Sitessampled (selected on the basis of the pilotstudy data) were the ear (anterior and posteriorpinna); anterior nares, and axilla. The forearmand lower leg were also sampled because theseare common sites for peripheral catheter inser-tion.Rayon swabs moistened in lml phosphate
buffered 0-1% Triton X100 solution wereused to sample 1-3 cm2 of the skin; a 3 cm2plastic template was used when sampling theforearm and lower leg. The swab tip wasimmediately broken off into the wash solutionand vortex mixed for 30 seconds.18 Viablecounts were made2l on milk agar (CM2 1;Oxoid, Basingstoke), malassezia agar22 andReinforced Clostridial Agar (CM151; Oxoid)with 6 mg/l furazolidone for the enumerationof Propionibactenium acnes.23
Stool samples were collected when available,dispersed in 9 parts glycerol broth (1% LabLemco powder (L29; Oxoid), 10% glycerol)and stored at -70°C. Coagulase negativestaphylococci in stool samples collected closestto the skin sampling day were counted onDiagnostic Sensitivity Test agar (CM26 1;Oxoid) supplemented with (l-l) glycine (0 5g), lithium chloride (2 g), potassium thio-cyanate (22-5 g), and aztreonam (15 mg,Bristol-Myers Squibb).
During the course of the study, we collecteda number of clinical isolates of coagulase nega-tive staphylococci from catheter tips and bloodcultures of infants included in the study.Catheter tips were processed using the Makiroll method.24 Isolates were regarded asclinically important if more than 15 colonieswere recovered.A representative of each coagulase negative
staphylococcal colony type isolated from eachsampling site was stored in glycerol-citratemedium (40% glycerol, 60% 50 g/l trisodiumcitrate) at -70°C for subsequent characterisa-tion.
CHARACTERISATION OF COAGULASE NEGATIVESTAPHYLOCOCCAL ISOLATESAll coagulase negative staphylococci isolatesfrom the faeces, blood cultures, and line tipsand from 16 day skin samples were charac-terised by a variety of methods.
Antibiogram determination Susceptibility topenicillin G, tetracycline, erythromycin, trime-thoprim, chloramphenicol, gentamicin, netil-micin and streptomycin was determined by themodified rotary plating Stokes method.25
Speciation Isolates were tested for acidproduction from mannitol, trehalose, andN-acetylglucosamine, for susceptibility todesferrioxamine (1000 ,ug disc)26 and novo-biocin (5 jig disc), and production of IP-glucosidase (fluorescence at 366 nm aftergrowth on P agar containing 150 mg/l 4-methyl-umbelliferyl-j-D-glucopyranoside) andurease.27 Any isolates not speciated by thesetests were inoculated into an API Staph strip(bioMerieux).
Plasmid RFLP determination Plasmids wereextracted and purified by a modification of theprotocol of Voskuil and Chambliss,28 with50 U/ml lysostaphin added to the lysis mix toimprove cell lysis. Purified plasmids werecut with Cla 1 restriction endonuclease. Theresultant fragments were separated byelectrophoresis in 0-8% agarose, stained withethidium bromide, and photographed underultraviolet illumination.
Table I Median coagulase negative staphylococcal countsobtainedfrom 20 infants ofgestational age of less than 35weeks and postnatal age range 8-30 days
Site Count (cfiu)Faeces 1-1X107 g-IEar (posterior pinna) 4-6X 106 ml 1
To facilitate quantitative assessment bacteria were washed fromswabs by vigorous agitation in 0-1%/ Triton X100 solution.cfu=colony forming units; g-'=per g faeces; ml'1=per ml washsolution, area of skin not defined; cm-2=per cm2 skin, area ofswabbed skin defined by a plastic template.
on July 29, 2020 by guest. Protected by copyright.
Reservoirs of coagulase negative staphylococci in preterm infants F101
Ear (ant) .... + ... F/arm* Ear (post) --- Leg
--0-- Axilla --C-- Faeces8 - Nares
6
E ,(a
0 >2O 4 r'~\/
-ArA
04 8 12 16 21 28 60 90 180
Age (days)Figure 1 Coagulase negative staphylococci at various sitesduring the first six months of life. Note that counts aredisplayed as mean counts per gram forfaeces but meancounts per swab at a1l other sites.
ResultsPRELIMINARY LOCATION OF MAJOR RESERVOIRSOF COAGULASE NEGATIVE STAPHYLOCOCCIThe median population sizes found at variousskin sites and in faeces of neonates are given intable 1. The largest counts were found at theear (anterior and posterior pinna) anteriornares, axilla, and in the faeces. These sites wereselected for further study as were the forearmand the lower leg, where peripheral cathetersare frequently sited.
LONGITUDINAL SURVEYPopulation sizeThe mean coagulase negative staphylococcalcounts at seven sites on 10 babies from 4 daysto 6 months of age are shown in fig 1. Asexpected from the preliminary investigations,the ear, nose, and axilla consistently carriedconsiderably higher numbers than the forearmand lower leg, which harboured low numbersthroughout the study. Bowel numbers werelargest around 12 days but progressivelydecreased thereafter. Mean bacterial counts inthe ear and nose also increased substantiallybetween 4 and 12 days, but did not changeconsistently beyond this age.
CompositionMost skin and faecal coagulase negativestaphylococci isolates were S epidermidis (76%of 191 strains speciated), S haemolyticus, Swarneri, or S capitis (table 2). The proportionsensitive to the antibiotics tested were peni-cillin G 3d1%, gentamicin 129%, netilmicin23-2%, trimethoprim 38a7%, erythromycin62-4%/, chloramphenicol 90 7%, strepto-mycin 91 8% and tetracycline 96@9%.Plasmid RFLPs were useful in the subdivisionof these species and revealed that many babies
carried one or more strains at several sites butthat these strains were rarely shared by differ-ent babies. Staphylococcus epidermidis plasmidtype 11 (resistant to penicillin, methicillin,gentamicin, netilmicin, erythromycin andtrimethoprim) was a notable exception, beingisolated from seven of the 10 babies studied.Eight of the nine line tip and blood cultureisolates collected were indistinguishable fromisolates recovered from the skin of the samebaby; two were also recovered from faeces.
StabilityCoagulase negative staphylococcal counts atmost skin sites on each individual showed widefluctuations between sampling days. Countsaround the ear fluctuated considerably less, asdid faecal coagulase negative staphylococcalnumbers. Plasmid typing of coagulase negativestaphylococci isolated from faeces on differentdays showed that there was also some qualita-tive stability, but a succession of differentisolates was also evident. For example, case 3yielded predominantly S epidermnidis on days 5and 7,S warneri on days 13, 15, and 18, a mix-ture of both species on day 27, and exclusivelyS epidermidis at 2, 3, and 4 months.
Table 3 shows the correlations betweenstaphylococcal counts obtained at sites usedfor catheter insertion (forearm and leg) andsites with large staphylococcal populations.Counts at the axilla correlated with those onboth the leg and the forearm. There were lessobvious correlations between the forearm andthe ear and nares. However, counts from thefaeces did not correlate with counts on theforearm and were inversely related to countson the leg.
OTHER SKIN MICRO-ORGANISMSPropionibacterium spp and Malassezia furfurpopulation densities are given in figs 2 and 3.Mean counts at all skin sites were low in thefirst three to four weeks of life but progressivelyincreased thereafter at the axilla, ear, and to alesser extent the nares. Patterns of colonisationwere broadly similar at all sites for these twogroups of organisms, but were different fromcoagulase negative staphylococcal colonisationpatterns.
DiscussionThe characteristic skin microbiota of adultsdevelops at puberty and thereafter can remainstable over many years,29 but the skinmicroflora in prepubescent children, includingneonates, has not been well described.Although the numbers of micro-organisms onskin are generally low in early childhood, cer-tain skin sites on the heads of infants in theirfirst year of life have been observed to carrylarge numbers of staphylococci, propionibac-teria, and Malasseziafjuffur.30 31 This investiga-tion confirms that skin at certain sites (ear,nares, and axilla) can harbour all of thesemicro-organisms. Staphylococci predominatedand colonised earlier than propionibacteria
on July 29, 2020 by guest. Protected by copyright.
All isolates were S epidermidis except c=S capitis, W=S warneri, h=S haemolyticus, X=S xylosus.NT=not typable (no plasmids seen).
and Malassezia furfur, which have slowergrowth rates and might require greater matu-ration of skin structures than coagulase nega-tive staphylococci. The difference betweenpatterns of skin colonisation by staphylococciand the other major skin residents is also notedin adults.32 The limbs carried relatively fewmicro-organisms, and numbers fluctuatedconsiderably on a daily basis, as noted atseveral sites in previous studies.1618 Regularcontamination by bacteria originating fromsites harbouring large reservoirs may explainthis instability. The observation that differentsites on the same baby are much more likely toshare the same types of coagulase negativestaphylococci than are skin sites of differentbabies is consistent with this hypothesis. Itwould also be expected that there would be acertain degree of correlation between countsfound at 'source' and 'destination' sites. Thiswas the case to a certain extent when compar-ing candidate skin reservoirs with the leg andforearm, the most notable correlations beingbetween the axilla and these sites (table 3).These data must be interpreted with cautionbecause positive correlations may well be theresult of phenomena other than contaminationfrom one site to another, such as maturation ofboth sites simultaneously or response to sys-temic antibiotics (although no relation wasnoted between microbial density on the skinand antibiotic treatment).
Table 3 Correlation between staphylococcal counts at sites usedfor catheter insertion (forearm and leg) and sites withlarge staphylococcal populations
Figure 2 Mean propionibacteria counts at various sitesduring the first six months of life.
The bowel may be a particularly importantreservoir of coagulase negative staphylococci.Faeces contained large numbers, particularlyduring the first month when babies are atgreatest risk of acquiring coagulase negativestaphylococcal infection. During routine careof babies, particularly the changing of soilednappies, the hands of parents and unit nursingstaff are likely to become faecally contami-nated. This contamination will be spread tomultiple skin sites on the baby. Staphylococcitransferred to the skin of the infant by thisprocess may persist for prolonged periodsbecause of the resistance of staphylococci todesiccation. Furthermore, the large numbersof staphylococci present in faeces, in high con-centrations of organic matter, make it veryunlikely that contaminated hands could beeffectively disinfected by hand washing. Thiswill result in a high rate of cross contaminationbetween babies, ensuring that antibiotic resis-tant strains of coagulase negative staphylococcistrains endemic in many special care babyunits can readily be transferred to newbornpatients in whom they are likely to establishnew foci of infection under the selective pres-sure of intensive antibiotic use. Selection ofresistant isolates is more likely to occur in thebowel, into which many antibiotics penetratewell via excretion in the bile, than on the skin,where only low concentrations of hydrophilicantibiotics such as aminoglycosides and P-lactams are found following systemic use.Although the only significant correlationbetween population densities in faeces and onthe limbs was negative (faeces vs leg), contam-ination of the skin by bowel contents may only
on July 29, 2020 by guest. Protected by copyright.
Reservoirs of coagulase negative staphylococci in preterm infants F103
3.0 -0 Ear (ant) -A- Nares_ *-Ear (post)* + F/arm
--0-- AxilIa -- Leg
2.5-
co2.0
1.5-C0 t
2°10t
0.5 -
04 8 12 16 21 28 60 90 180
Age (days)Figure 3 Mean Malassezia furfur counts at various sitesduring the first six months of life.
occur to an appreciable extent after defecation,which can be infrequent in this patient group.Resultant 'spikes' in numbers of coagulasenegative staphylococci at contaminated skinsites would be difficult to detect without serialsampling before and after changing soilednappies. At other times other coagulase nega-tive staphylococci reservoirs would probablycontribute more bacteria to sparsely populatedsites than would the bowel.
Understanding the spread of coagulasenegative staphylococci between babies andfrom site to site on the same baby may facilitatethe design of effective interventions for reduc-ing the incidence of systemic coagulase nega-tive staphylococcal infections in these patients.Emphasis is frequently placed on theimportance of preventing the spread of micro-organisms between patients. However, suchmeasures are clearly inadequate in preventingcolonisation of infants by antibiotic resistantcoagulase negative staphylococci endemic to aunit.'7 33 34 In this study plasmid analysis sug-gested that a coagulase negative staphylococcalclone was carried by seven of 10 babies, andthat most of the other isolates were multiplyantibiotic resistant, implying that they maywell also be endemic to the unit.
If coagulase negative staphylococci causinginfection frequently gain access at catheterinsertion sites, and there is no stable residentbacterial population on surrounding skin, itmay be possible to reduce infection rates byreducing spread of organisms to these sitesfrom numerically important reservoirs ofcoagulase negative staphylococci on the samebaby during routine handling. Interventionswhich might be considered include improvedaseptic handling protocols for neonates, relat-ing particularly to 'high risk' procedures suchas nappy changing, or by antisepsis directed atthe reservoir sites. Alternatively, as transientpopulations are superficially located, it mightbe possible to eradicate most bacteria at thesite of catheter implantation using antiseptics,6and prevent ingress of coagulase negative
staphylococci by a combination of occlusivedressings and careful handling of catheters andinsertion sites witli gloved hands.
1 Battisti 0, Mitchison R, Davis PA. Changing blood cultureisolates in a referral neonatal intensive care unit. Arch DisChild 1981; 56: 775-8.
2 Munson DP, Thompson TR, Johnson DE, Rhame FS,VanDrunen N, Ferrieri P. Coagulase-negative staphylo-coccal septicemia: experience in a newborn intensive careunit. _J Pediatr 1982; 101: 602-5.
3 Freeman J, Platt R, Sidebottom DG, LeclairJM, Epstein MF,Goldmann DA. Coagulase-negative staphylococcal bac-teremia in the changing neonatal intensive care unit popula-tion: is there an epidemic?JAM4 1987; 258: 2548-52.
4 Schmidt BK, Kirpalani HM, Corey M, Low DE, PhilipAGS, Ford-Jones EL. Coagulase-negative staphylococcias true pathogens in newborn infants: a cohort study.Pediatr Infect Dis J 1987; 6: 1026-31.
5 Hall SL. Coagulase-negative staphylococcal infections inneonates. Pediatr Infect Dis J 1991; 10: 57-67.
6 Malathi I., Millar MR., Leeming JP, Hedges A, Marlow N.Skin disinfection in preterm infants. Arch Dis Child 1993;69: F312-6.
7 Freeman J, Epstein MF, Smith NE, Platt R, SidebottomDG, Goldmann DA. Extra hospital stay and antibioticusage with nosocomial coagulase-negative staphylococcalbacteremia in two neonatal intensive care unit population.AmjDis Child 1990; 144: 324-9.
8 Noel GJ, Edelson PJ. Staphylococcus epidermidis bacter-aemia in neonates: further observations and occurrence offocal infections. Pediatrics 1984; 74: 832-7.
9 de Cicco M, Chiaradia V, Veronesi A, Santini G, PanarelloG, Fracasso A, et al. Source and route of microbial colon-isation of parenteral nutrition catheters. Lancet 1989; ii:1258-61.
10 Kite P, Langdale V, MacKay P, Millar MR, Todd N. Directisolation of staphylococci from neonatal blood. J HospInfect 1989; 14: 135-40.
11 Cronin WA, Germanson TP, Donowitz LG. Intravascularcatheter colonization and related bloodstream infection incritically ill neonates. Infect Control Hosp Epidemiol 1990;11: 301-8.
12 Gellert GA, Ewert DP, Bendana N, Smith E, Beck-SangueC, Chin A, et al. A cluster of coagulase-negative staphylo-coccal bacteremias associated with peripheral vascularcatheter colonization in a neonatal intensive care unit. AmJ Infect Control 1993; 21: 16-20.
13 Salzman MB, Isenberg HD, Shapiro JF, Lipsitz PJ, RubinLG. A prospective study of the catheter hub as the portalof entry for microorganisms causing catheter-relatedsepsis in neonates. _lInfect Dis 1993; 167: 487-90.
14 Freeman J, Goldmann DA, Smith NE, Sidebottom DG,Epstein MF, Platt R. Association of intravenous lipidemulsion and coagulase-negative staphylococcal bac-teremia in neonatal intensive care units. N Engl J Med1990; 323: 301-6
15 Fleer A, Gerards U, Aerts P, Westerdaal NAC, SendersRC, van Dijk H, et al. Opsonic defence to Staphylococcusepidermidis in the premature neonate. J Infect Dis 1985;152: 930-7.
16 Simpson RA, Spencer AF, Speller DCE, Marples RR.Colonisation by gentamicin-resistant Staphylococcusepidermidis in a special care baby unit. Jf Hosp Infect 1986;7:108-20.
17 D'Angio CT, McGowan KI, Baumgart S, St.Geme J,Harris MC. Surface colonisation with coagulase-negativestaphylococci in premature neonates. Jf Pediatr 1989; 114:1029-34.
18 Keyworth N, Millar MR, Holland KT. Development ofcutaneous microflora in premature neonates. Arch DisChild 1992; 67: 797-801.
19 Price PB. The bacteriology of normal skin: a new quantita-tive test applied to a study of the bacterial flora and thedisinfectant action of mechanical cleansing. Jf Infect Dis;1938; 63: 301-18.
21 Miles AA, Misra SS. The estimation of the bactericidalpower of the blood. Jf Hygiene 1938; 38: 732-48.
22 Leeming JP, Notman FH. Improved methods for isolationand enumeration of Malassezia furfur from human skin.J7 ClinMicrobiol 1987; 25: 2017-19.
23 Cove JH, Eady EA. A note on a selective medium for theisolation of cutaneous propionibacteria. J Appl Bacteriol1982; 53: 289-92.
24 Maki DG, Weise CE, Sarafin HW. A semiquantitativeculture method for identifying intravenous-catheter-related infection. N Engl JfMed 1977; 296: 1305-9.
25 Holt A, Brown D. Antimicrobial susceptibility testing. In:Hawkey PM, Lewis DA, eds. Medical bacteriology, a prac-tical approach. Oxford: IRL Press, 1989: 167-94.
26 Lindsay JA, Riley TV. Susceptibility to desferrioxamine: anew test for the identification of Staphylococcus epider-midis. JMed Microbiol 1991; 35: 45-8.
27 Kloos WE. Lambe DW. Staphylococcus. In: Balows A,Hausler WJ, Herrmann KL, Isenberg HD, Shadomy HJ,eds. Manual of clinical microbiolog. 5th edn. Washington,DC: American Society for Microbiology, 1991: 222-37.
28 Voskuil MI, Chambliss GH. Rapid isolation and sequencingof purified plasmid DNA from Bacillus subtilis. ApplEnviron Microbiol 1993; 59: 1138-42.
on July 29, 2020 by guest. Protected by copyright.
29 Evans CA, Strom MS. Eight year persistence of individualdifferences in the bacterial flora of the forehead. J InvestDermatol 1982,79: 51-2.
30 Leyden JJ, McGinley KJ, Mills OH, Kligman AM. Age-related changes in the resident bacterial flora of thehuman face. IJnvest Dernatol 1975; 65: 379-81.
31 Leeming JP, Sutton TM, Fleming PJ. Neonatal skin as areservoir of Malassezia species. Pediatr Infect Dis J 1995;14: 719-21.
32 Leeming JP, Notman FH, Holland KT. The distribution
and ecology of Malassezia furfur and cutaneous bacteriaon human skin. JAppl Bacteriol 1989; 67: 47-52.
33 Bialkowska-Hobrzanska H, Jaskot D, Hammerberg 0.Molecular characterization of the coagulase-negativestaphylococcal surface flora of premature neonates. J GenMicrobiol 1993; 139: 2939-44.
34 Lyytikainen 0, Saxen H, Ryhanen R, Vaara M, Vuopio-Varkila J. Persistence of a multiresistant clone ofStaphylococcus epidermidis in a neonatal intensive-careunit for a four-year period. Clin Infect Dis 1995; 20: 24-9.
on July 29, 2020 by guest. Protected by copyright.
