Epidemiology of Airborne Staphylococcal Infection · cially true of the forms of staphylococcal...

BACTERIOLOGICAL REVIEWS, Sept., 1966Copyright © 1966 American Society for Microbiology

Vol. 30, No. 3Printed in U.S.A.

Epidemiology of Airborne Staphylococcal InfectionR. E. 0. WILLIAMS

Wright-Fleming Institute of Microbiology, St. Mary's Hospital Medical School, London, England

INTRODUCTION ................................................................Material Reviewed..........................................................

DISPERSAL OF STAPHYLOCOCCI INTO THE AIR .......................................Nose and Skin Carriers......................................................Mechanism ofDispersal....................................................Frequency and Magnitude ofDispersal........................................Factors Influencing Dispersal................................................Infected Lesions............................................................Air Contamination Resultingfrom Dispersal ......................................

TRANSFER THROUGH THE AIR..................................................Viability in Air.............................................................

ACQUISITION OF AIRBORNE STAPHYLOCOCCI.....................................

Operating Room Infection...................................................Airborne transfer from without ..............................................Locally generated aerial contamination ....................................Air counts and infectionrates.......................................

Infection in Wards..........................................................Minimal infective dose.....................................................

Relevance ofNasal Acquisition ................................................CONCLUSION..................................................................LITERATURE CITED ..........................................................

660660661661661661663663663664665665665665666666667668669670670

INTRODUCTION

It is a characteristic of the airborne route ofinfection, in contrast to transfer by food or water,that whenever there is the possibility of aerialtransfer there is almost always also the possibilityof transfer by other routes. This is perhaps espe-cially true of the forms of staphylococcal infec-tion that have been most extensively studied,namely, those occurring in hospitals. But, duringthe last few years, there has been a great volumeof work based on the assumption that airbornespread is an important route in the spread of thehuman staphylococcal disease, and there is there-fore a considerable body of information for re-view.

It is logical and convenient to discuss first thestudies on dispersal of staphylococci into the airand, second, the survival of the cocci in, and theircarriage by, the air. These aspects can be pre-sented in some precise and quantitative detail.When we come to consider acquisition, we enteran area in which extrapolation and analogy loomlarge, but sufficient quantitative data have nowbeen accumulated to give some factual founda-tion to the discussion. Nevertheless, the finalsumming up as to the probable importance ofairborne transfer in relation to other modes ofspread is of necessity a product of judgmentrather than arithmetic.

Material ReviewedFor the most part this paper is based on a

selective review of published reports, with specialreference to those on a series of investigations(44, 45, 58, 59) carried out with R.A. Shooter atSt. Bartholomew's Hospital, London, England(referred to as S.B.H.).Use has also been made of a recent study of

my own at St. Mary's Hospital, London, England(S. A.H.), a report of which is in preparation. Inthis study, we sampled the air in two surgicalwards, one having a total of 14 or 15 patients infour rooms and the other having 22 beds in anopen ward. Petri dishes [diameter, 6 inches (15cm)] of serum-agar containing phenol-phthaleinphosphate (2) were exposed in the rooms for12 hr (or in part of the experiment for 24 hr) oneach of 5 days each week. Nasal cultures wereexamined from each patient weekly.The total number of staphylococcus-carrying

particles on the air plates was recognized by thephosphatase reaction and either all or, when thenumbers of colonies on the plate were larger thanthree to five, a portion were subcultured andtested for coagulase and for phage type. It wasthus possible to make some estimate of the num-ber of coagulase-positive staphylococci settlingfrom the air, and of the proportion with variousphage patterns, for attempted correlation with

660

on January 14, 2021 by guesthttp://m

mbr.asm

.org/D

ownloaded from

on January 14, 2021 by guest

http://mm

br.asm.org/

Dow

nloaded from


mbr.asm

.org/D

ownloaded from

http://mmbr.asm.org/



AIRBORNE STAPHYLOCOCCAL INFECTION

the strains isolated from the patients' nasal cul-tures.

For airborne particles with a diameter equiva-lent to those found to carry Staphylococcus aureus,that is about 14 IA (38), the settling rate in coloniesper square foot per minute is numerically approxi-mately equal to the volume count expressed ascolonies per cubic foot. A 6-inch petri dish hasan area of approximately 0.2 ft2; in round figures,therefore, the count on such a plate exposed for24 hr is about 60% of the number of particlesinhaled by an adult person in the same time, sincethe volume inhaled is ordinarily about 0.3 ft3 permin.

DISPERSAL OF STAPHYLOCOCCI INTO THE AIRThe frequency with which normal individuals

harbor S. aureus in the nose and on the skin isnow well known (56), and such normal carriersare one important source from which the cocciare dispersed. The other source from which staph-ylococci may be dispersed, especially in hospitals,comprises patients with infected lesions-of skin,wounds, respiratory tract, or gut.

Nose and Skin CarriersHare and his colleagues were among the first

to define the frequency with which nasal carriersof S. aureus liberate the cocci into the environ-ment; they counted the numbers shed into the airof a very small cubicle during exercise. Hare andRidley (21) found that all but 6 of 19 carriersliberated staphylococci, and 7 gave substantialnumbers; this and subsequent work (41) pointedto the special importance as dispersers of personswho harbor staphylococci on the perineum. Onthe other hand, White (54, 55) emphasized therelation between dispersal of staphylococci andthe total numbers present in the nose and on theskin.One feature that also emerged from these and

other studies was the wide individual variation inthe number of staphylococci shed into the air bycarriers. The individuals at the upper end of thedistribution seemed to differ sufficiently fromthose at the lower end to justify the use of theterm "heavy disperser" for them, and the sug-gestion that such heavy dispersers might be re-sponsible for epidemics of hospital infectionstimulated further study of the mechanism ofdispersal.

Mechanism of DispersalHare and his colleagues showed that very few

staphylococci are liberated into the air directlyfrom the nose of carriers during ordinary activity;

Hare (19) described the liberation by other routesas "outflow" and emphasized the importance offriction with the skin. White (54) had found thatthe extent to which patients contaminated theirbedding was correlated with the numbers ofstaphylococci found in their nasal cultures. Sub-sequently, Davies and Noble (14) demonstratedthat large numbers of skin fragments are dispersedinto the air during the activities known to liberatebacteria; they suggested that most of the staphy-lococci are carried on such fragments, and theywere able to cultivate S. aureus from epithelialsquames liberated by a known carrier (15).

It thus seemed that the differences among in-dividual carriers in the number of staphylococcithat they disperse might be related to (i) thenumber of cocci on the skin, (ii) the particulararea of skin colonized, or (iii) the rate of desqua-mation. By parallel sampling of air for skinsquames and staphylococci, Noble and Davies(37) showed that the last of these was not likelyto be the explanation; they thought that the ex-tent of skin carriage was probably the most im-portant determinant. Hare and Ridley (21) hadpreviously suggested that carriage on the skin ofthe perineum was particularly likely to lead todissemination, and Solberg (49) not only con-firmed this but also showed that, in the absence ofperineal carriage, there is a correlation of num-bers of staphylococci disseminated with thenumber found in the nose or on the skin (in hisexperiments, of the fingers and hand). The impor-tance of the perineal skin as a source for dispersalreceives indirect support from observation that,in operating-room clothes, the greatest libera-tion of skin bacteria seems to be from below thewaist and especially through open trouser ends(5, 8). It may be noted that most observers havemeasured air contamination while the subjectswere exercising in some form of cubicle and gen-erally making quite vigorous leg movements; thismay perhaps over-emphasize the contribution ofthe perineum to aerial dispersal.

Frequency and Magnitude of DispersalIt was clear from the early work of Hare and

Ridley (21) that many nasal carriers shed staphy-lococci into the air while exercising. This has beenamply confirmed. On the basis of experiments insmall cubicles, Bethune et al. (5) reported that14 of 38 nasal carriers (from a group of150 normal people) generated an air contamina-tion level of one S. aureus particle per ft3 or more,corresponding to a total liberation of about 100particles or more in 2 min of walking-on-the-spot.Noble and Davies (37) examined 127 persons, 54of whom were normal adults whereas the rest

661VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACrERIOL. REV.

were hospital patients, many with skin disease.The subjects removed all their clothing and thendressed again in a 100 ft3 cubicle from which theair could be sampled. Of the whole group, 30,including only 2 of the 54 normal adults, liberatedS. aureus to 1% of the total flora in the cubicle;this corresponded to the liberation of about 25staphylococcus-carrying particles or more. Eightliberated more than 10,000 S. aureus particles.More precise estimates of the numbers of

staphylococci liberated by carriers have been pro-vided by Solberg (49; personal communication),who estimated the aerial contamination resultingfrom a standardized agitation of a group of per-sistent carriers' bedding in a special chamber.Solberg found the air counts of staphylococci dis-persed during the making of the beds of his car-riers to be distributed in a log-normal fashion,and at least 20% of the 126 carriers (drawn from2,014 patients surveyed) dispersed more than10,000 staphylococcus-carrying particles in thestandard test (Fig. 1).Our own studies of air counts in a hospital

ward offer another basis for estimating the fre-quency of dispersal. In Fig. 2 are plotted themean daily counts of S. aureus of the phage typecarried by each of the patients who were carrierson admission to the subdivided S.M.H. ward,

1000

Staph/

/A/

10> 1 .

1 S l1 20 0 95isCunSAWl % at cwnts

FIG. 1. Air counts of Staphylococcus aureus. (A)Generated by disturbance of bedding ofpersistent car-

riers [after Solberg (49), supplemented by a personalcommunication]; (B) during undressing and redressing[after Noble and Davies (37)]. Plotted on a probabilityscale, so that a straight line represents a normal dis-tribution.

Aircountcots/sq.ft./24hr./

0.5 5 02 oo 0 go o goA 1 5 9Cumulative % of patients

FIG. 2. Air counts (particles per square foot settlingin 24 hr) ofStaphylococcus aureus generated by patientsadmitted as carriers to a hospital ward (S. M. H.).

excluding those who were carrying strains of typesalready known to be present in the ward; someof the patients, in contrast to Solberg's subjects,were only transient carriers. The counts are clearlyalso distributed log-normally. About 50% of car-riers generated air counts below 5 colonies per ft2per 24 hr, but 10% generated counts that averagedmore than 50 colonies per ft2 per 24 hr duringtheir stay in the ward, at times when they werethe only carriers known to be present. Somerough estimates as to the ventilation rate of theward suggest that this implies the liberation of101 to 107 staphylococcus-carrying particles in24 hr.

It seems likely, therefore, that the heavy dis-persers of staphylococci represent the top end ofa continuous distribution. This is compatible withthe idea that the degree of dispersal dependslargely on the extent of skin contamination withstaphylococci, and that the shedding of the staph-ylococci into the air is due to the continuousdesquamation of skin fragments carrying cocci,which either may be transients recently depositedthere from the reservoir area in the anterior nares,or may be actually multiplying in or on the skin.The rate of desquamation is presumably relatedin part to friction of the skin and clothes or otherskin areas.

It is, perhaps, remarkable how many bacteriaare shed on exercising, even when the subjects

/

/

662 WILLIAMS


mbr.asm

.org/D

ownloaded from



are naked (50). Also, in a few unpublished ob-servations in coal mines, O.M. Lidwell and Ifound that nearly naked miners distributed skinbacteria into the air in quite large numbers. It mayalso seem surprising that carriers liberate as manystaphylococci as they do when it is consideredhow relatively scarce staphylococci appear to bewhen carriage is determined by swabbing; how-ever, the area of skin generally examined is verysmall, and most methods for the bacteriologicalexamination of skin are known to be very ineffi-cient (62).

Factors Influencing Dispersal

Blowers and McCluskey (8) commented thatthey have not yet encountered a heavy disperserof staphylococci among the normal women theyhave examined, whereas they found nearly 10%of men to be dispersers. None of the other studieshas discussed the influence of sex, but, of the 10heavy dispersers reported by Solberg, 3 werewomen, and in general his results do not show anysignificant differences between men and womencarriers in the numbers of staphylococci dispersed.In my ward studies, one of the five heaviest dis-persers was a woman.

It has been found that treatment of a nasalcarrier of tetracycline-resistant staphylococci withtetracycline led to an increase in the number ofstaphylococci dispersed into the air, presumablyas a result of increased nasal carriage (17), orpossibly as a result of increased skin carriage re-sulting from a reduction in the normal flora anda consequent reduction in the fatty acid contentof sebum (18). A similar phenomenon was ob-served in debilitated or dying patients by Solberg:an increase in the number of organisms in thenose and a corresponding increase in the numbershed. M.T. Parker (personal communication) hasmade a similar observation. One observation thata concomitant virus infection might increase dis-persal (16) does not seem to have been confirmed.A very substantial increase in the number of

staphylococci liberated has been found to followthe taking of a shower bath (4, 62). The increasemay be 10-fold or more, and the effect persistsfor at least 60 min. The mechanism of this increaseis not known, though it is presumed that thewashing in some way allows an increased loss ofthe superficial squames.

Within broad limits, clothing makes remark-ably little difference to the liberation of skinbacteria, and indeed Speers et al. (50) found thatsome of their subjects liberated as many bacteriawhen exercising naked as they did when fullydressed, either in street clothes or in a sterile oper-ating room suit. The only practicable method so

far described for reducing the rate of liberationis the use of very closely woven clothing, with atrouser suit tightly closed at the ankles (4, 8).

Infected Lesions

The discussion to this point has been concernedwith healthy carriers of staphylococci, withoutany staphylococcus-infected lesions. As would beexpected, patients with staphylococcal infectionsof the skin tend to be especially heavy dispersers(1, 20, 37). Thom and White (52) found, however,that there was little dispersal from septic woundsduring the performance of wound dressing, and itseems likely that the effect of skin lesions is, partlyat least, to increase the load of staphylococci onthe skin. There may also be an increase in the rateof desquamation, for example, in some patientswith psoriasis, and one such has been implicatedas the source of an epidemic of surgical wound in-fection in an operating room (32a), thoughgenerally in such patients many of the skinparticles dispersed are too large to remain air-borne (37). Our own observations (59) indicatedthat carriers can be as important as sources ofcross infection as patients with septic lesions.Burke and Corrigan (13), on the other hand,found patients with septic lesions to dispersemore staphylococci than healthy carriers; butthey studied only 44 carriers. Patients with chestinfections have been thought from time to time tobe especially dangerous as dispersers (e.g., 44),but there is little direct evidence on this point.The possible effect of antibiotic treatment ondispersal needs to be considered when patientswith septic lesions are being compared withhealthy carriers.

Air Contamination Resulting from Dispersal

It can thus be concluded that most persons whocarry staphylococci in the nose, all of whom mustfrom time to time contaminate their skin, liberatetheir staphylococci into the air around them. Asmall proportion of the carriers are especiallyheavy dispersers and give rise to a high level ofaerial contamination. It is not surprising, there-fore, that there are considerable variations in thecounts of staphylococci in hospital ward air (Fig.3). The variations in the air counts are, of course,directly related to the presence or absence ofheavy dispersers in the ward (34). When the aircount in the ward was high, it was virtually al-ways found that the air staphylococci were al-most all of one phage type and usually attribut-able to spread from one person. In the large ward,there were occasions when two dispersers con-tributed significantly to the air count, but thoseoccasions were relatively uncommon (34).

663VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACTERIOL. REV.

2I2

X 10tC

"t

20M

100-

10

40

15 25 5 15 25 5FEB. MAR. 1965

is 25APR.

5MAY

FIG. 3. Air counts (particles per square foot per 24hr) of Staphylococcus aureus in three rooms of the di-vided S. M. H. ward.

But, as in other situations, the air counts inhospital wards have been found to conform to alog-normal distribution. In Fig. 4, the air countsfrom the divided ward are plotted as logarithmson a probability scale and are seen to fall closeto a straight line. For comparisons between wards,therefore, the median is the most appropriatestatistic. Lines depicting the distributions in fivedifferent wards are shown in Fig. 5, and the medi-ans from some of them are given in Table 3.

In the S.B.H. open ward, the median countof S. aureus was about 0.1 colonies per ft3. Thiscount was derived from two periods of 2-hr sampling each week, and it might be thought that thiscould be no more than generally indicative of thetotal daily exposure of the particles to airbornestaphylococci. However, a very similar mediarand distribution of staphylococcal counts wereobserved in the open ward at S.M.H., tested b)exposure of 12-hr sedimentation plates.

It is instructive to present the counts in termqof the numbers of staphylococcus-carrying particles that might be inhaled by ward patients in 2ahr. In the two open wards, the median numberthat would be inhaled per day were about 18 anc23 particles; the daily dose exceeded 100 particleon about 15 to 22% of days. In the divided ward

at S.M.H. and S.B.H., the median was about 4,and a dose of 100 was exceeded on only about3% of days.

It is interesting to note that in a small series oftests in a ward at the Queen Elizabeth II Hospital,Welwyn, which consists of four-bed bays openingoff a wide corridor, the air counts are intermediatebetween those of the open and the divided wards(data kindly supplied by R. W. Payne). The ex-planation of these differences clearly demandsfurther investigation, and it is of obvious rele-vance to the acquisition of nasal carriage of staph-ylococci, discussed below.

TRANSFER THROUGH THE AIRFor a proper understanding of the mode of

spread of airborne staphylococcal infection, aknowledge of the size of the airborne particlesand of their load of staphylococci is needed.Studies with the size-grading sampler devised byLidwell (26) indicated that the mean "equivalentdiameter" of particles carrying S. aureus wasabout 14 ,i (the "equivalent diameter" is the diam-eter of a sphere of unit density settling in air atthe same rate as the particle in question); the in-terquartile range was about 8 to 20 A (38). Amuch smaller proportion of large particles wasobserved by Walter et al. (53), using the Andersensampler, but this is doubtless attributable to thecharacteristics of that instrument (29). Earlierwork by Lidwell and his colleagues (27) indicatedthat, on the average, airborne staphylococcus

sq.ft.per24

50O

20

10

5-

2-

1.S

I

S1S

0-1

cots.

-O5 sq-ft.peran.

2*b 5!b t to 9b SCurboative % of cuts

FIG. 4. Distribution of air counts of Staphylococcusaureus in the divided S. M. H. ward, based on a total of1,037 12-hr sedimentation plates.

4w

300

1

... II 1.11. .... II... 11 .-I 'III

664 WILLIAMS

--A -

Ilk I.--. ... III, 11


mbr.asm

.org/D

ownloaded from



CurU~ahV% Of cont

FIG. 5. Distributions of air counts ofStaphylococcusaureus in different wards. A = S.M.H., 22-bed openward; B = S. B. H., open wards; C = Queen ElizabethII Hospital, 4-bed open bays; D = S. M. H., dividedward; E = S. B. H., divided ward.

particles carried about 4 viable cocci, the rangebeing from 6 for the particles greater than 18 Mi

in diameter to about 1 for those less than 4 As.These sizes and numbers of bacteria are conso-nant with the idea that most airborne S. aureuscells are associated with desquamated fragmentsof skin (37).

In normally turbulent air and in a room 10 fthigh, particles 14 IA in diameter settle at a rateequivalent to about six air changes per hour, sothat 50% of the particles remain suspended for6 min and 20% for 15 min. Directional air cur-rents of 40 to 50 ft per min are not uncommonin occupied buildings, so that transfer of staphy-lococci for considerable distances is clearlypossible.

In a few studies, in an open surgical ward inwhich we have found large numbers of staphy-lococci being dispersed near one sampling point,the mean counts at sampling points about 20 andabout 70 ft distant were, respectively, 26 and 11%of the count at the point nearest to the disperser.When a heavy disperser was present in one of

the rooms in the four-room S.M.H. ward, thecount in the other rooms has been on averageabout 5% of that in the room with the disperser.Lidwell and his colleagues (27a) have studied award divided into nine rooms and found that thecount in rooms other than that containing asource of staphylococci is about 10% of that inthe source room.

In studies of staphylococcal infection in a surgi-cal operating room, Shooter et al. (46) demon-strated what appeared to be aerial transfer over

the distance of 90 ft that separated the wardsfrom the operating room.The actual extent to which staphylococci can

be conveyed within a ward or between roomsmust depend on the local circumstances of struc-ture, site, and ventilation, but enough has beensaid to show that aerial conveyance over consider-able distances is quite possible.The aerial route is not, of course, the only way

by which staphylococci can be conveyed from oneroom to another, and hospital routines commonlyprescribe quite elaborate rituals for dealing withpotentially infected dust on floors, shoes, trolley(cart) wheels, and the like, which is thought togenerate secondary airborne spread. But, thoughmany workers have estimated the bacterial con-tent of floors, few have made any useful studiesof actual transfer by this route (see 53).

Viability in Air

There is a considerable amount of laboratorywork to show that staphylococci commonly sur-vive in the dried state for periods measured indays or weeks. Whether there is any significantalteration in their infectivity on storage is notcertain. Indications of some loss of infectivitywere obtained by Hinton et al. (23) and by Tayloret al. (51); other workers have found no sucheffect (e.g., 28, 42). Noble's (35, 36) experimentalwork in animals has suggested that any effect oninfectivity from desiccation is limited to an ex-tension of the lag period and that, if staphylococciare protected from body defenses immediatelyafter introduction into the tissues, they are asvirulent as fresh organisms.

ACQUISITION OF AIRBORNE STAPHYLOCOCCIThere are two important ways in which air-

borne staphylococci might infect patients in hos-pitals: by inhalation or by settling directly intosome susceptible area, such as a wound, or ontoinstruments or dressings that subsequently comeinto contact with the wound. Inhalation infectionmay occur anywhere and at any time; sedimenta-tion infection is of particular importance in opera-ting rooms and treatment rooms where surgicalwounds are exposed, often for long periods oftime. It will be convenient to deal with sedimenta-tion infection in operating rooms first.

Operating Room InfectionAirborne transfer from without. Recent studies

of air hygiene in surgical operating rooms datelargely from the work of Bourdillon and Cole-brook (10) in a Burns Unit treatment room, butit was the application of their work to the controlof a high incidence of postoperative staphylococ-

665VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACTERIOL. REV.

cal wound infection by Blowers et al. (6) and byShooter et al. (46) that brought the subject togeneral attention. Shooter et al. estimated that,in an 8-month period, the incidence of operating-room infections was 9% of 427 wounds; 0.07particles per ft3 containing S. aureus were foundin samples from the air during operations. Asimple alteration of the ventilation so as to gen-erate a positive pressure in the operating roomand exclude staphylococcus-contaminated airfrom the wards was followed by a substantial re-duction in the general air bacterial count (thenumber of S. aureus was not reported), and by areduction to less than 1% in the incidence ofsepsis of presumed operating theater origin in532 wounds. It is reasonable to assume that thisreduction in sepsis was attributable to a reductionin the number of staphylococci settling from theair into the wounds and onto the sterile instru-ments and equipment. No other investigation hasbeen reported in which alteration of the ventila-tion was the only change made, but the publishedreport of Blowers et al. (6) and subsequent un-published experience (Blowers, personal commun-ication) supported the general idea that theprevention of contamination of operating roomair with bacteria from other parts of the hospitalby the introduction of positive-pressure ventila-tion has often been associated with a reductionin the incidence of postoperative sepsis. Blowersand Crew (7) recorded a mean S. aureus count of0.6 colonies per ft3 in an exhaust-ventilated opera-ting room, compared with 0.03 colonies per ft3in a plenum-ventilated operating room.

Locally generated aerial contamination. Thework just cited concerned contamination ofoperating room air with staphylococci from otherparts of the hospital, drawn into the operatingroom by air currents. It is this form of transferthat is controllable by positive-pressure ventila-tion. But aerial contamination can also begenerated within the operating room, either bydisturbance of the patients' bedclothes and drapesor from the skin of the operating room personnel,as discussed already.

Air counts and infection rates. The bacterialcount observed in the air of an operating roomis clearly the sum of that produced by infiltrationof contaminated air from without and that gen-erated locally. It would be of great value for themonitoring of operating room hygiene if it werepossible to relate the staphylococcal (or even thetotal bacterial) count in the air to the risk of post-operative sepsis. The difficulties of deriving sucha relationship are, however, very great. The inci-dence of wound sepsis is in any case generallyvery low-perhaps between 1 and 5%. Only a

part of the septic cases are infected during opera-tion, and this portion is difficult to estimate, andin any case is not all attributable to sedimentationof airborne staphylococci. In addition, the num-bers of staphylococci actually settling ontosusceptible areas are so small as to be difficult tomeasure.

Burke's (11) study is in many ways the mostdetailed available. By using a very sensitive tech-nique, he was able to recover S. aureus from 46of 50 wounds examined at the end of operation;most wounds yielded two or more differentstrains, and the mean number of viable units ofstaphylococci was 14 per wound. Potential sourcesfor the staphylococci found in the wounds were:air, 68%; carrier site on patient, 50%; hands ornasopharynx of the surgical team, 20%. (In somecases, there were two or more potential sources.)Only 2 of the 50 wounds developed any clinicalsign of postoperative infection; the rate forwounds that had not been carefully washed outfor bacteriological examination was not pre-sented.

In a comparable study of the sources of infec-tion in 35 patients who developed wound sepsisapparently resulting from operating room infec-tion, Bassett et al. (3) thought that a member ofthe surgical team was concerned in 31% and thepatient himself in 17%, the source of the remain-der being untraced.There are several other published studies in

which an attempt was made to relate postopera-tive infection to airborne staphylococci found inthe operating room (e.g.,24,53,60), but they donot allow easy summary. The general impressionis that staphylococci of the type responsible forpostoperative infection were rarely found in theair, but this may well reflect the very small airsamples generally examined.

In general, it appears that, in reasonably well-ventilated operating rooms with good staff dis-cipline, the S. aureus count is of the order of 0.01to 0.05 colonies per ft3; in a series of operatingrooms, we have observed a mean settling countof about 0.01 colonies per ft2 per min, while anAmerican cooperative study reported a count aslow as 0.001 colonies per ft2 per min. (33).The operating room ought to be a situation in

which it would be possible to determine the aver-age infecting dose of staphylococci for man. Tak.ing a figure of 0.01 colonies per ft2 per min, andassuming an effective target area of 1 ft2 (to in-clude instruments, etc.) and a duration of opera-tion of 2 hr, a frequency of operating roominfections of 1% would imply that the 1% infec-tive dose is about 1 staphylococcus-carryingparticle. But, to put any real meaning into the

666 WILLIAMS


mbr.asm

.org/D

ownloaded from



figures, we need to measure the air count and thesepsis rate in a far greater number of patients thanhas yet been attempted and to carry out at thesame time elaborate bacteriological cultures onthe patient himself and on the ward to try toassess the relative importance of routes of transferother than air. And we should remember Burke's(12) thesis that sepsis is often determined largelyby the condition of the actual tissue on which thestaphylococcus alights, and on the state of thepatient; if his observations are generally applica-ble, there are usually plenty of staphylococci.

Infection in WardsThere is ample documentation of the rate at

which both newborn infants and adult patientsbecome nasal carriers of the prevalent staphy-lococcus in many hospital wards (56). It wasreasonable to postulate in the first place thatthese staphylococci reached the nose by way ofthe air. Evidence has been sought on this pointin several ways-by examining the order in whichdifferent parts of the body are colonized, byattempting to interfere with transfer by one routeor another, by trying to identify the source of thestaphylococcus more precisely, and by relatingthe acquisition rate to the staphylococci foundin the air.The most precise investigations in this field

concern newborn infants. It was shown, first,that the umbilicus and abdominal skin are gener-ally colonized before the nose (25, 48). Second,Rammelkamp and his collaborators showed thata nurse carrier only conveyed her staphylococcito infants if she handled them (61), and laterthat the colonization of the infants could be de-layed by increasing the precautions against con-tact infection (31, 32). With very strict precau-

TABLE 1. Primary site of colonization in adults

No. of patients who becamensal carriers in wardCarrier sites, etc., positive

before nose for staphylococci ofsame phage type Nose Nosepostwiteve positive Totaltwitcev once

or more only

None............19 25 25 28 44 53Clothing only.... f6 3J Q9Dressing or wound....... 7 5 12Hand or other skin site.... 14 2 16

Total.................. 46 35 81

aPatients swabbed daily; apparent acquisitionson first 3 days of hospital stay, and acquisitions ofuntypable strains, excluded.

tions against cross infection, the rate ofacquisition of staphylococci was reduced from43 to 14%; the latter infections were assumed tobe due to aerial transfer. The relative unimpor-tance of inhalation infection in newborn infantsis perhaps hardly surprising when one considersthat the infant has a minute volume of only about500 ml (about 0.02 ft1), and that he has to be han-dled frequently, usually by nurses who handle agood many other infants. But a 14% acquisitionrate in a 4-day hospital stay is equivalent tosome 3 to 4% per day, which is of the sameorder as observed in adult wards.With the evidence from the newborn infants

in mind, it is pertinent to ask whether the noseor some skin site is the first area to be colonizedin the adults who acquire staphylococci in hos-pitals. It is obviously more difficult to obtain evi-dence on this for the adult than for the infant, butin a study of surgical patients (22) R. A. Hender-son examined swabs daily from the nose, skin ofthe hands, skin near the wound site, wound, bed-clothes, and environment (Table 1). Some 20 7Oof the 81 patients who became nasal carriersyielded staphylococci of the relevant phage typefrom one or other of the two skin sites before itsappearance in the nose, and a further 15% hadyielded the staphylococci from the wound. Inthe remaining 66% of acquisitions, the nose wasthe first site on the patient found to yield thestaphylococcus. Two important provisos have tobe entered here: there was a striking dominanceof staphylococci of one phage type among theacquisitions in the ward, which means that thereis a serious risk of regarding as related two inde-pendent acquisitions; and the area of skin exam-ined was very small and perhaps not representa-tive. Additionally, even skin or clothing con-tamination might result from airborne transfer,which need not operate only to give inhalationinfection. However, the evidence, such as it is,does not contravert the idea that direct inhalationinfection is important in the acquisition of thenasal carrier state in adults.

In our recent study at St. Mary's Hospital, 53patients were observed to acquire nasal carriageof S. aureus while in the ward. The same phagetype had been recovered from the air prior to itsrecovery from the patient in 64% of cases (Table2). In this ward, there was no marked dominanceof one type, and indeed 25 types are representedamong the 53 acquisitions. Again, this is not for-mal evidence that the nasal carrier state wasacquired by inhalation of cocci, but it is consis-tent with such an explanation.

Further evidence for the importance of aerialtransfer comes from studies of different ward

667VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACTERIOL. REV.

structures. In an open 22-bed ward we found thatseparation of patients by the full length of theward (about 50 feet) only reduced the rate ofacquisition of staphylococci by about one-half,as compared with the acquisition rate for a patientin a neighboring bed (59). At the other extreme,very low nasal acquisition rates have been foundin patients nursed in single rooms opening tofresh air, that is, when the chance of aerial trans-fer from one room to another is very low indeed(40). There also appeared to be very little spreadof tetracycline-resistant strains from patientsnursed in isolation rooms fitted with exhaustventilation, and the acqusition rate for suchstrains was greatly reduced in a ward in whichall patients harboring such strains were isolated(59).

In adult patients, there are technical difficultiesin recognizing the acquisition of nasal carriagethat are not present with infants, since truly per-

TABLE 2. Number of patients showing apparentacquisition of nasal carriage in relation to

previous air exposure (S. M. H.)"

Staphylococci ofNo. of weeks nose acquired type-negative for the

acquired Totalstaphylococcus Found Not found

before acquisition in air in airpreviously previously

1 15 (9)b 11 (8) 26 (17)2 9 (6) 2 (2) 11 (8)3 3 (2) 2 (2) 5 (4)4 3 (1) 2 (2) 5 (3)5+ 4 (3) 2 (1) 6 (4)

Total 34 (21) 19 (15) 53 (36)

a An additional 13 (9) patients were found onadmission to the ward to be carriers of a staphylo-coccus previously found in the air and so may wellhave acquired their nasal carriage in the ward.

bNumbers in parentheses give patients carryingthe acquired strain on one occasion only.

sistent carriers may fail to yield staphylococci onsome occasions. However, since carriage of tet-racycline-resistant staphylococci is even nowrelatively rare (at least in Britain) in people out-side hospitals, such strains form a convenientindicator of hospital acquisition. Some rates ofacquisition of tetracycline-resistant staphylococciin various wards are presented in Table 3. Un-fortunately, data for tetracycline resistance ofstaphylococci isolated from air samples in thesewards are not available, so it is only possible tocompare the ranking of the wards with respect tothe two parameters. The very limited results sug-gest that the acquisition rate was higher in thewards with the higher counts of air staphylococci.For various technical reasons, it has not yet beenpossible to test directly the relation of the acquisi-tion rate to the exposure to particular staphy-lococci, though this clearly needs to be done.Minimal infective dose. If we are to understand

the epidemiology of airborne infection, we mustknow the minimal dose of microbes ordinarilyneeded to effect colonization or infection; thisnumber is not known, but it is so important thatit seems justified to indulge in some extrapolationfrom the few figures available. Shinefield and hiscolleagues, in their investigations of bacterial in-terference, found that they could set up a carrierstate in the nose of 50% of newborn infants bythe inoculation of between 200 and 400 cocci. Asnoted already, most airborne staphylococcus-carrying particles appear to contain no more thanone to six viable cocci.

In experimental infections, it is generally foundthat the relation between dose and attack rate isnot linear, but conforms to an S-shaped curve.For extrapolation to be possible, it is thereforenecessary to apply some transformation to thedata, e.g., to plot the logarithm of the dose inocu-lated against the probit of the percentage attackrate. This has been done in Fig. 6 for the dataobtained by Shinefield and his colleagues (43,supplemented by a personal communication from

TABLE 3. Acquisition of nasal carriage of tetracycline-resistant Staphylococcus aureus in relation to dailyexposure to airborne staphylococci

Median AcquisitionReference Type of ward exposure rateReference ~~~~~~~~~~~~~~~~~(particles/(per cent

24 hr) per day)

Williams et al. (59), Noble (34) 22-24 bed open, S. B. H. 18 0.7Shooter et al. (47) 22-24 bed divided in two parts, S. B H. 9a 0.6

8a 0.3Williams (in preparation) 14 beds in 4 rooms, S. M. H. 4 0.3Lidwell et al. (27a) 30 beds in 9 rooms, S. B. H. 4 0.1

a These values are estimates based on mean counts provided by 0. M. Lidwell, converted to medianson the assumption that the distribution was similar to that in the earlier S. B. H. studies.

668 WILLIAMS


mbr.asm

.org/D

ownloaded from



0

<12

Dose inoculated

FIG. 6. Relation between dose ofstaphylococci inoc-ulated into infants' noses and acquisition rate for nasalcarriage. Based on figures kindly supplied by Henry R.Shinefield.

Dr. Shinefield), and the points lie very close to astraight line. Extrapolation of the line back wouldsuggest an attack rate of about 0.02% for a doseof five cocci. The observations of Shinefield et al.were made on newborn infants who are presum-ably more susceptible to staphylococcal coloniza-tion than adult subjects, but, in the absence ofany other figures, the calculation may be worthpursuing.The data in Fig. 5 suggest that the median

number of staphylococcus-containing particlesinhaled in the S.B.H. wards may have been about18. Each of these particles probably contained,on the average, about 4 viable cocci, so that thetotal daily dose inhaled could be estimated atabout 70 cocci; if the dose-response relation ob-served by Shinefield were applicable to the adults,this dose might be expected to generate a "take-rate" of just over 10% per day if all the inhaledparticles co-operated to set up the carrier state,or 0.16% if they acted independently. Unfortu-nately, we do not know how many of the airbornestaphylococci were tetracycline-resistant, but theapparent acquisition rate for tetracycline-resist-ant strains was about 0.7% per day.

In the S.M.H. divided ward, the median dose ofsensitive and resistant staphylococci inhaled wasabout 16, which on Shinefield's figures would indi-cate a take-rate of 0.6%, or less than 0.01% if allthe particles acted independently; the actual rate

of acquisition of tetracycline-resistant strains was0.3% per day.

These and some other similar data are presentedin Table 3. Although quite insufficient to indicatea clear relation, they suggest that the staphylococ-cal acquisition rate in different wards may well;be related to the air count. In fact, the acquisitionrates in the wards are, considering the amount ofextrapolation involved, clearly of the same orderas those predicted from Shinefield's figures. Butat least these calculations clearly indicate thatthere is no wild improbability in the idea that theacquisition of the nasal carrier state in surgicalpatients results from the inhalation of such air-borne staphylococci as can be shown to occur inthe wards. The number of complicating factorsin any precise analysis is formidable.

In the first place, as already noted, the figurefor a median bacterial count conceals enormousvariations, and we clearly need to know whethera short exposure to a large number of airbornestaphylococci is equivalent to a more prolongedexposure to smaller numbers. A second complica-tion arises from the fact that staphylococci ap-pear to vary in their ability to colonize the nose(57), so that there is reason to think that inhala-tion of large numbers of cocci of some strains maybe less effective in setting up the carrier state thaninhalation of others. The third complication arisesfrom differences in the recipients. The phenom-enon of bacterial interference, studied in detailby Shinefield et al. (43, 43a) in infants, almostcertainly operates in adults also. Several workershave shown that patients admitted to hospital ascarriers of S. aureus are less liable to acquire hos-pital strains than patients admitted as noncarriers(e.g.,58). The fact that, at least in open wards,patients treated with antibiotics acquire hospitalstaphylococci in the nose more often than thosewho are not (e.g., 39) is presumably another ex-ample of the same phenomenon, which was welldemonstrated experimentally by Boris et al. (9).At the same time, antibiotic treatment probablyprevents nasal acquisition in other patients.

Relevance of Nasal AcquisitionIn the operating room, we must assume that,

whatever the dose-response relation, the aerialtransfer of staphylococci to the wound itself ispotentially important. It may be asked whetherthere is any corresponding relevance in the nasalacquisition of staphylococci in the wards. Thereseem to us to be two ways in which the nasalspread is important.

In the first place, it appears that, at least insome circumstances, nasal carriage of staphy-lococci predisposes to postoperative infection

669,VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACrERIOL. REV.

(58). There has been some discussion on the sig-nificance of these observations (3, 22, 30), butscrutiny of the records of a considerable numberof patients leaves no doubt in my mind that thephenomenon is real, even if not generally so fre-quent as suggested by our original observations.But nasal carriage is also relevant in that it

seems to be the mechanism by which the endemicstaphylococci persist in the hospital. Such per-sistence can often be for a long period. For ex-ample, at Saint Bartholomew's Hospital weobserved the spread of a staphylococcus of phagetype 75/77 which continued from the start of thestudy in one ward in February 1959 until the endof January 1960. During this period of 1 year,there were only 39 days when there was not pres-ent a patient who was either known or reason-ably presumed to be a carrier of the strain. Atotal of 23 patients were infected with the strain,but only 6 of them had any clinically infectedlesion.

CONCLUSIONThe commensal association of staphylococci

with man is universal (56) and to a large degreeharmless. The transfer from one individual toanother must, under ordinary circumstances, veryoften be by direct or indirect contact. But abilityto disperse S. aureus into the air in large numbersis a characteristic-sometimes temporary andsometimes persistent-of a number of healthypeople, and wherever we go indoors there is achance that we shall inhale staphylococci. [A fewobservations in two Post Offices in London havegiven an average sedimentation count of 0.01colonies per ft2 per min, a figure quite similar tothat for hospital wards (J. Corse, personalcommunication)]. But it is only in hospitals thatany detailed study of the processes of transferhas been made.Airborne transfer in hospitals gains its special

significance from the fact that, if this route isactually operative, a single disperser is potentiallyable to infect a considerable number of otherpatients, who need not be confined within thesame room, or even perhaps on the same floor;and the transfer of infection cannot be containedby ordinary methods of asepsis.The evidence that has been reviewed seems to

leave little doubt that airborne transfer can be ofimportance. It suggests that the acquisition ofnasal carriage of S. aureus by patients nursed inhospital wards can be explained if the dose-effectrelationship determined experimentally in infantsis approximately applicable to adults. If the re-sults obtained in the studies reviewed can beconfirmed elsewhere, we should have a rational

basis for assessing one aspect of hospital hygienein relation to the prevention of staphylococcalinfection. We still lack, however, a precise meas-ure of the relative part played by this airbornespread in the etiology of staphylococcal hospital-acquired infection generally.To take surgical wound infection as an ex-

ample, we have to recognize that infection can bederived from: (i) staphylococci carried by thepatient on admission to hospital; (ii) staphy-lococci that the patient has come to carry in thenose and on the skin after admission, which sub-sequently enter the wound; and (iii) staphylococcithat reach the wound directly without the priorintervention of the nose or skin carrier state. Itappears that aerial transfer plays a major partin the second of these categories and a part-sometimes major and sometimes minor-in thethird. But we have insufficient precise evidenceon the relative importance of the three categoriesthemselves. The proportion will clearly differgreatly from one hospital to another, and withinone hospital, from one sort of surgical operationto another, and from time to time.The challenge with which we are faced is to

provide much more firmly based estimates of therelative frequencies in these categories and thefactors that determine them. The practical justi-fication for attempting such an analysis is that itcan provide the only basis for judging how bestto construct and ventilate hospitals. And thefundamental difficulty of performing the analysisis that in any hospital, where the analysis wouldbe practicable, the overall incidence of infectionis probably no more than 1 to 2%, and this smallproportion must be distributed over all thevarious routes and sources.

ACKNOWLEDGMENTSI am grateful to H. R. Shinefield, C. 0. Solberg, and

W. C. Noble for providing me with additional detailsof their experimental results. I am also greatly indebtedto 0. M. Lidwell for allowing me to see, in advance ofpublication, some of the data from a further study atSt. Bartholomew's Hospital, and for numerous valua-ble discussions on the whole topic of the review. Thework at St. Mary's Hospital was supported by a grantfrom the Hospital Endowment Fund.

LITERATURE CITED1. ALDER, V. G., AND W. A. GILLESPIE. 1964. Pres-

sure sores and staphylococcal cross-infection.Detection of sources by means of settle-plates.Lancet 2:1356-1358.

2. BARBER, M., AND S. W. A. KUPER. 1951. Identifi-cation of Staphylococcus pyogenes by the phos-phatase reaction. J. Pathol. Bacteriol. 63:65-68.

3. BAssErr, H. F. M., W. G. FERGUSON, E. HOFFMAN,M. WALTON, R. BLOWERS, AND C. A. CONN.

670 WILLIAMS


mbr.asm

.org/D

ownloaded from



1963. Sources of staphylococcal infection insurgical wound sepsis. J. Hyg. 61:83-94.

4. BERNARD, H. R., R. SPEERS, JR., F. O'GRADY, R.A. SHOOTER. 1965. Reduction of disseminationof skin bacteria by modification of operating-room clothing and by ultraviolet irradiation.Lancet 2:458-461.

5. BETHUNE, D. W., R. BLOWERS, M. PARKER, ANDE. A. PASK. 1965. Dispersal of Staphylococcusaureus by patients and surgical staff. Lancet1:480-483.

6. BLOWERS, R., G. A. MASON, K. R. WALLACE, ANDM. WALTON. 1955. Control of wound infectionin a thoracic surgery unit. Lancet 2:786-794.

7. BLOWERS, R., AND B. CREW. 1960. Ventilation ofoperating-theatres. J. Hyg. 58:427-448.

8. BLOWERS, R., AND M. MCCLUSKEY. 1965. Designof operating-room dress for surgeons. Lancet2:681-683.

9. BORIS, M., T. F. SELLERS, JR., H. F. EICHENWALD,J. C. RIBBLE, AND H. R. SHINEFIELD. 1964. Bac-terial interference. Protection of adults againstnasal Staphylococcus aureus infection aftercolonisation with a heterologous S. aureusstrain. Am. J. Diseases Children 108:252-261.

10. BOURDILLON, R. B., AND L. COLEBROOK. 1946.Air hygiene in dressing-rooms for burns ormajor wounds. Lancet 1:561-565.

11. BURKE, J. F. 1963. Identification of the sources ofstaphylococci contaminating the surgical woundduring operation. Ann. Surg. 158:898-904.

12. BURKE, J. F. 1965. Factors predisposing to infec-tion in the surgical patient, p. 143-155. In H. I.Maibach & G. Hildick-Smith [ed.], Skin bac-teria and their role in infection. McGraw-HillBook Co., Inc., New York.

13. BURKE, J. F., AND E. A. CORRIGAN. 1961. Staphy-lococcal epidemiology on a surgical ward.Fluctuations in ward staphylococcal content,its effect on hospitalized patients and the extentof endemic hospital strains. New Engl. J. Med.264:321-326.

14. DAVIES, R. R., AND W. C. NOBLE. 1962. Dispersalof bacteria on desquamated skin. Lancet 2:1295-1297.

15. DAVIES, R. R., AND W. C. NOBLE. 1963. Dispersalof staphylococci on desquamated skin. Lancet1:1111. v1

16. EICHENWALD, H. F., 0. KoTsEvALOv, AND L. A.FAsso. 1960. The "cloud baby": an example ofbacterial-viral interaction. Am. J. DiseasesChildren 100:161-173.

17. EHRENKRANZ, N. J. 1964. Person-to-person trans-mission of Staphylococcus aureus. New Engl. J.Med. 271:225-230.

18. FREINKEL, R. K., J. S. STRAUSS, S. Y. YIP, AND P.E. POCHI. 1965. Effect of tetracycline on thecomposition of sebum in acne vulgaris. NewEngi. J. Med. 273:850-854.

19. HARE, R. 1964. The transmission of respiratoryinfections. Proc. Roy. Soc. Med. 57:221-230.

20. HARE, R., AND E. M. COOKE. 1961. Self-contami-nation of patients with staphylococcal infec-tions. Brit. Med. J. 2:333-336.

21. HARE, R., AND M. RIDLEY. 1958. Further studieson the transmission of Staphylococcus aureus.Brit. Med. J. 1:69-73.

22. HENDERSON, R. J., AND R. E. 0. WILLIAMS. 1963.Nasal carriage of staphylococci and post-operative staphylococcal wound infection. J.Clin. Pathol. 16:452-456.

23. HINTON, N. A., J. R. MALTMAN, AND J. H. ORR.1960. The effect of dessication on the ability ofStaphylococcus pyogenes to produce disease inmice. Am. J. Hyg. 72:343-350.

24. HOwE, C. W., AND A. T. MARSTON. 1962. A studyon sources of post-operative staphylococcal in-fection. Surg. Gynecol. Obstet. 115:266-275.

25. HuRST, V. 1960. Transmission of hospital staphy-lococci among newborn infants. II. Coloniza-tion of the skin and mucous membranes of theinfants. Pediatrics 25:204-214.

26. LIDWELL, 0. M. 1959. Impaction sampler for sizegrading airborne bacteria-carrying particles. J.Sci. Instr. 36:3-8.

27. LIDWELL, 0. M., W. C. NOBLE, AND G. W. DOL-PHIN. 1959. The use of radiation to estimate thenumbers of micro-organisms in airborne parti-cles. J. Hyg. 57:299-308.

27a. LIDWELL, 0. M., S. POLAKOFF, M. P. JEVONS,M. T. PARKER, R. A. SHOOTER, V. I. FRENCH,AND D. R. DUNKERLEY. 1966. Staphylococcalinfection in thoracic surgery: experience in asubdivided ward. J. Hyg. 64:1-17.

28. McDADE, J. J., AND L. B. HALL. 1963. An experi-mental method to measure the influence ofenvironmental factors on the viability andpathogenicity of Staphylococcus aureus. Am. J.Hyg. 77:98-108.

29. MAY, K. R. 1964. Calibration of a modified An-dersen bacterial aerosol sampler. Appl. Mi-crobiol. 12:37-43.

30. MOORE, B., AND A. M. N. GARDNER. 1963. Astudy of post-operative wound infection in aprovincial general hospital. J. Hyg. 61:95-113.

31. MORTIMER, E. A., JR., P. J. LIPSITZ, E. WOLINSKY,A. J. GONZAGA, AND C. H. RAMMELKAMP, JR.1962. Transmission of staphylococci betweennewborns. Importance of the hands of person-nel. Am. J. Diseases Children 104:289-295.

32. MORTIMER, E. A., JR., E. WOLINSKY, A. J. GON-ZADA, AND C. H. RAMMELKAMP JR. 1966. Roleof airborne transmission in staphylococcal in-fections. Brit. Med. J. 1:319-322.

32a. MURLEY, R. S. 1965. Infection in the theatresuite. Bull. Soc. Intern. Chir. No. 6, p. 636-643.

33. NATIONAL RESEARCH COUNCIL. 1964. Post-opera-tive wound infections: the influence of ultra-violet irradiation on the operating room and ofvarious other factors. Ann. Surg. Vol. 160,Suppl. to No. 2.

34. NOBLE, W. C. 1962. The dispersal of staphylococciin hospital wards. J. Clin. Pathol. 15:552-558.

35. NOBLE, W. C. 1964. The production of sub-cutaneous staphylococcal skin lesions in mice.Brit. J. Exptl. Pathol. 46:254-262.

36. NOBLE, W. C. 1964. Staphylococcal infections in

671VOL. 30, 1966


mbr.asm

.org/D

ownloaded from


BACrERIOL. REV.

man and animals. Ph.D. Thesis, Univ. London,London, England.

37. NOBLE, W. C., AND R. R. DAVIES. 1965. Studies onthe dispersal of staphylococci. J. Clin. Pathol.18:16-19.

38. NOBLE, W. C., 0. M. LIDWELL, AND D. KINGSTON.1963. The size distribution of airborne particlescarrying micro-organisms. J. Hyg. 61:385-391.

39. NOBLE, W. C., R. E. 0. WILLIAMS, M. P. JEVONS,AND R. A. SHOOTER. 1964. Some aspects ofnasal carriage of staphylococci. J. Clin. Pathol.17:79-83.

40. PARKER, M. T., M. JOHN, R. T. D. EMOND, ANDK. A. MACHACEK. 1965. Acquisition of Staphy-lococcus aureus by patients in cubicles. Brit.Med. J. 1:1101-1105.

41. RIDLEY, M. 1959. Perineal carriage of Staph.aureus. Brit. Med. J. 1:270-273.

42. ROUNTREE, P. M. 1963. The effect of desiccation onthe viability of Staphylococcus aureus. J. Hyg.61:265-272.

43. SHINEFIELD, H. R., J. C. RIBBLE, M. BORIS, ANDH. F. EICHENWALD. 1963. Bacterial interference:its effect on nursery-acquired infection withStaphylococcus aureus. 1. Preliminary observa-tions on artificial colonization of newborns.Am. J. Diseases Children 105:646-654.

43a. SHINEFIELD, H. R., J. D. WILSEY, J. C. RIBBLE, M.BORIS, H. F. EICHENWALD, AND C. I. DITrMAR.1966. Interactions of staphylococcal coloniza-tion. Am. J. Diseases Children 111:11-21.

44. SHOOTER, R. A., J. D. GRIFFITHS, J. COOK, AND R.E. 0. WILLIAMS. 1957. Outbreak of staphy-lococcal infection in a surgical ward. Brit. Med.J. 1:433-436.

45. SHOOTER, R. A., M. A. SMITH, J. D. GRIFFITHS, M.E. A. BROWN, R. E. 0. WILLIAMS, J. E. RIPPON,AND M. P. JEVONS. 1958. Spread of staphy-lococci in a surgical ward. Brit. Med. J. 1:607-613.

46. SHOOTER, R. A., G. W. TAYLOR, G. ELLIS, ANDJ. P. RoSS. 1956. Postoperative wound infection.Surgery Gynecol. Obstet. 103:257-262.

47. SHOOTER, R. A.) B. T. THOM, D. R. DUNKERLEY,G. W. TAYLOR, M. T. PARKER, M. JOHN, ANDI. D. G. RICHARDS. 1963. Pre-operative segre-gation of patients in a surgical ward. Brit. Med.J. 2:1567-1569.

48. SIMPSON, K., C. TOZER, AND W. A. GILLESPIE.1960. Prevention of staphylococcal sepsis in amaternity hospital by means of hexachloro-phane. Brit. Med. J. 1:315-317.

49. SOLBERG, C. 0. 1965. A study of carriers of Staph-ylococcusaureus. Acta Med. Scand. 178(suppl.):436.

50. SPEERS, R. JR., H. BERNARD, F. O'GRADY, AND R.A. SHOOTER. 1965. Increased dispersal of skinbacteria into the air after shower-baths. Lancet1:478-480.

51. TAYLOR, G. W., R. A. SHOOTER, P. H. FRANDSEN,W. R. FIELDER, AND W. J. KERTH. 1962. Staphy-lococcal wound infection. An experimentalstudy in guinea-pigs. Brit. J. Surg. 49:569-571.

52. THOM, B. T., AND R. G. WHITE. 1962. The dis-persal of organisms from minor septic lesions.J. Clin. Pathol. 15:559-562.

53. WALTER, C. W., R. B. KUNDSIN, AND M. M.BRUBAKER. 1963. The incidence of airbornewound infection during operation. J. Am. Med.Assoc. 186:908-913.

54. WHITE, A. 1961. Relation between quantitativenasal cultures and dissemination of staphy-lococci. J. Lab. Clin. Med. 58:273-277.

55. WHITE, A., J. SMITH, AND D. T. VARGA. 1964. Dis-semination of staphylococci. Arch. InternalMed. 114:651-656.

56. WILLIAMS, R. E. 0. 1963. Healthy carriage ofStaphylococcus aureus: its prevalence and im-portance. Bacteriol. Rev. 27:56-71.

57. WILLIAMS, R. E. 0. 1963. Ecological evidence forthe existence of virulent strains of staphylococci.Recent Progr. Microbiol. 8:551-560.

58. WILLIAMS, R. E. 0., M. P. JEVONS, R. A. SHOOTER.C. J. W. HUNTER, J. A. GIRLING, J. D. GRIF-FITHS, AND G. W. TAYLOR. 1959. Nasal staphy-lococci and sepsis in hospital patients. Brit.Med. J. 2:658-662.

59. WILLIAMS, R. E. 0., W. C. NOBLE, M. P. JEVONS,0. M. L1DWELL, R. A. SHOOTER, R. G. WHITE,B. T. THOM, AND G. W. TAYLOR. 1962. Isolationfor the control of staphylococcal infection insurgical wards. Brit. Med. J. 2:275-282.

60. WOLF, H. W., M. M. HARRIS, AND W. R. DYER.1959. Staphylococcus aureus in air of an operat-ing room. J. Am. Med. Assoc. 169:1983-1987.

61. WOLINSKY, E., P. J. LIPSITZ, E. A. MORTIMER, JR.,AND C. H. RAMMELKAMP JR. 1960. Acquisitionof staphylococci by newborns. Direct versusindirect transmission. Lancet 2:620-622.

62. ULRICH, J. A. 1965. Dynamics of bacterial skinpopulations, p. 219-234. H. I. Maibach and G.Hildick-Smith [ed.], Skin bacteria and theirrole in infection. McGraw-Hill Book Co., Inc.,New York.

DiscussionALEXANDER D. LANGMUIR

Epidemiology Branch, Communicable Disease Center, U.S. Public Health Service, Atlanta, Georgia

Dr. Williams has presented a perceptive reviewof our knowledge of the occurrence of Staphy-lococcus aureus in the air of hospital wards and

surgical operating rooms and of its spread topatients. It is impressive how much work hasbeen reported during the past decade and what a

672 WILLIAMS


mbr.asm

.org/D

ownloaded from


EXTRACELLULAR LIPIDS OF YEASTS

128. SVENNERHOLM, E., AND L. SVENNERHOLM. 1963.The separation of neutral blood-serum glyco-lipids by thin-layer chromatography. Bio-chim. Biophys. Acta 70:432-441.

129. SWEELEY, C. C., AND B. KLIONSKY. 1963. Fabry'sdisease: classification as a sphingolipidosis andpartial characterization of a novel glycolipid.J. Biol. Chem. 238:PC3148-3150.

130. SWEELEY, C. C., AND B. KLIONSKY. 1965. Glyco-lipid lipidosis: Fabry's disease, p. 618-632. InJ. B. Stanbury, J. B. Wyngaarden, and D. S.Fredrickson [ed.], The metabolic basis ofinherited disease, 2nd ed. McGraw-Hill BookCo., New York.

131. SWEELEY, C. C., AND E. A. MOSCATELLI. 1959.Qualitative microanalysis and estimation ofsphingolipid bases. J. Lipid Res. 1:40-47.

132. TAKAHASHI, H. 1948. Bacterial components ofCorynebacterium diphtheriae. V. Phospho-lipides. 2. Structure of Chargaff's corynin. J.Pharm. Soc. Japan 68:292-296.

133. TEETS, D. W., P. JOHNSON, C. L. TIPTON, AND H.E. CARTER. 1963. Isolation of a new type ofsphingolipid. Federation Proc. 22:414.

134. THANNHAUSER, S. J., AND E. FRANKEL. 1931.Uber das Lignocerylsphingosine. Z. Physiol.Chem. 203:183-188.

135. THIERFELDER, H., AND E. KLENK. 1930. DieChemie der Cerebroside und Phosphatide. J.Springer, Berlin.

136. THORPE, S. R., AND C. C. SWEELEY. 1967.Chemistry and metabolism of sphingolipids.On the biosynthesis of phytosphingosine byyeast. Biochemistry (in press).

137. THUDICHUM, J. L. W. 1874. Researches on thechemical constitution of the brain. Report ofthe Medical Officer of Privy Council a. LocalGovernment Board. New series no. III, 113,London.

138. THUDICHUM, J. L. W. 1880. Further researches onthe chemical constitution of the brain, and ofthe organoplastic substances. 9th A nnualReport of the Local Government Board, 1879-1880. Supplement B, no. 3, Report of theMedical Officer for 1879, London, p. 143-188.

139. THUDICHUM, J. L. W. 1884. A treatise on thechemical constitution of the brain. Bailliere,Tindall and Cox, London, p. 105.

140. TULLOCH, A. P., AND J. F. T. SPENCER. 1964.

Extracellular glycolipids of Rhodotorulaspecies. The isolation and synthesis of 3-D-hydroxypalmitic and 3-D-hydroxystearic acids.Can. J. Chem. 42:830-835.

141. TULLOCH, A. P., J. F. T. SPENCER, AND P. A. J.GORIN. 1962. The fermentation of long-chaincompounds by Torulopsis magnoliae. I.Structures of the hydroxy fatty acids obtainedby the fermentation of fatty acids and hydro-carbons. Can. J. Chem. 40:1326-1338.

142. VAN AMMERS, M., M. H. DEINEMA, C. A.LANDHEER, AND M. H. M. VAN ROOYEN. 1964.Note on the isolation of B-hydroxypalmiticacid from the extracellular lipids of Rhodo-torula glutinis. Rec. Trav. Chim. 83:708-710.

143. WAGNER, H., AND W. SOFCSIK. 1966. U0ber neueSphingolipide der Hefe. Biochem. Z. 344:314-316.

144. WICKERHAM, L. J., AND F. H. STODOLA. 1960.Formation of extracellular sphingolipids bymicroorganisms. I. Tetraacetylphytosphingo-sine from Hansenula ciferri. J. Bacteriol. 80:484-491.

145. WOODBINE, M. 1959. Microbial fat: micro-organisms as potential fat producers. Prog.Ind. Microbiol. 1:181-245.

146. YABUTA, T., Y. SUMIKI, AND K. TAMARI. 1941.Chemical constituents of inekoji. VIII.Arabityl margarate. J. Agr. Chem. Soc. Japan17:307-310.

147. YAMAKAWA, T., S. YOKOYAMA, AND N. Kiso.1962. Structure of main globoside of humanerythrocytes. J. Biochem. (Tokyo) 52:228-229.

148. YAMAKAWA, T., AND S. SUZUKI. 1951. Thechemistry of the lipids of post-hemolyticresidue or stroma of erythrocytes. I. Con-cerning the ether-soluble lipids of lyophilizedhorse blood stroma. J. Biochem. (Tokyo)38:199-212.

149. ZABIN, I. 1957. Biosynthesis of ceramide by ratbrain homogenates. J. Am. Chem. Soc. 79:5834-5835.

150. ZABIN, I., AND J. F. MEAD. 1953. The bio-synthesis of sphingosine. I. The utilization ofcarboxyl-labeled acetate. J. Biol. Chem. 205:271-277.

151. ZABIN, I., AND J. F. MEAD. 1954. The biosynthe-sis of sphingosine. II. The utilization ofmethyl-labeled acetate, formate, and ethanol-amine. J. Biol. Chem. 211:87-93.

ERRATUM

Epidemiology of Airborne Staphylococcal InfectionsR. E. 0. WILLIAMS

Wright-Fleming Institute of Microbiology, St. Mary's Hospital Medical School1 London, England

Volume 30, no. 3, p. 664, column 2: on Fig. 4, the left-hand scale should read "cols. per 0.2 sq.ft. per 24 hour" and the figures on the right-hand scale should read, from top to bottom, "0.05,0.025, and 0.005."

213VOL. 3 1, 1967

Epidemiology of Airborne Staphylococcal Infection · cially true of the forms of staphylococcal...

Documents

Transcript of Epidemiology of Airborne Staphylococcal Infection · cially true of the forms of staphylococcal...