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

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

Effect of Route of Inoculation on ExperimentalRespiratory Viral Disease in Volunteers and

Evidence for Airborne TransmissionROBERT B. COUCH,1 THOMAS R. CATE,2 R. GORDON DOUGLAS, JR.,1 PETER J. GERONE, AND

VERNON KNIGHT1Laboratory of Clinical Investigations, National Institute of Allergy and Infectious Diseases, U.S. Public Health

Service, Bethesda, Maryland, and U.S. Army Biological Laboratories, Fort Detrick,Frederick, Maryland

INTRODUCTION................................................................ 517

MATERIALS AND METHODS................... 518

Volunteers...................... 518Inocula .................................................................... 518Inoculation Procedures............... 518Collection of Cough, Sneeze, Talking, and Room Air Samples...................... 518

Virus Isolation and Identification Procedures.................................... 519

Serological Tests ................................................. 519RESULTS..................................................................... 519

Response to Inoculation with Aerosol and Nasal Drops 519

Coxsackievirus A type 21: 50% human infectious doses (HIDIo) .... .............. 519Rhinovirus NIH 1734: HID6o ............................................... 521Adenovirus type 4: HID5t.................................................. 522

Evidence for Airborne Transmission ............................................. 525Detection of virus in particles produced by coughing, sneezing, and normal expiration. 525Virus in room air................................................. 525Preliminary report on a transmission experiment 526

DISCUSSION................................................................... 526

SUMMARY.................................................................... 528

LITERATURE CITED................................................... 528

INTRODUCTIONInitiation of respiratory viral infection, with

some possible exceptions, appears to dependupon deposition of infectious virus at some pointon the respiratory tract. There appear to be twopossible mechanisms of transmission, contact orairborne. The former term is meant to refer totransfer of virus by physical contact between aninfected and a susceptible subject, or indirectlythrough personal articles or fomites. Trans-mission by this route would result in depositionof virus predominantly in the nasopharynx.

Airborne transmission is intended to meantransfer of infection by means of small-particleaerosols (11, 16). These particles are the evap-orated residues of infected respiratory secre-tions which are of such small size (mostly lessthan 5 .t in diameter) that they will remain air-borne for long periods of time. As a function oftheir small size, such droplets, when inhaled,

1 Present address: Baylor University College ofMedicine, Houston, Tex.

2Present address: Washington University Collegeof Medicine, St. Louis, Mo.

deposit predominantly in the lower respiratorytract. Particles between 5 and 15 1A to 20 1i indiameter represent an intermediate stage, andmost particles in this size range will be trapped inthe nose, although some will penetrate to belowthe larynx. (Lower respiratory tract will refer tothat portion of the respiratory tract below thelarynx.) Still larger particles may be producedby coughing and sneezing, etc., but since, be-cause of their large size, they do not producestable aerosols, transmission will ordinarilyoccur only by direct impaction on the naso-pharynx of persons in the immediate vicinity ofan infected case. Such transmission would bedifficult to distinguish from that resulting fromcontact, and is best considered under this cate-gory.

This report will describe studies of the trans-mission of respiratory viral diseases which werea joint undertaking of the U.S. Army BiologicalLaboratories, Fort Detrick, Md., and theLaboratory of Clinical Investigations, NationalInstitute of Allergy and Infectious Diseases,Bethesda, Md.

517

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

5COUCH Er AL.

The first part of the report will describe aninvestigation of the infectiousness of respiratoryviruses given by methods which attempt tosimulate natural contact and airborne trans-mission, namely, nasal drops and aerosolscontaining virus. Coxsackievirus A type 21, astrain of rhinovirus, and adenovirus type 4 wereused in these studies.The second part will describe recovery of

virus from natural aerosols produced by cough-ing and sneezing and from air of rooms con-taminated by such discharges. In addition,preliminary results of an experimental attempt totransmit respiratory viral infection in volunteersby the airborne route will be presented.

MATERIALS AND METMODSVolunteers

Subjects were healthy adult male inmates fromseveral federal correctional institutions andwere selected on the basis of serum antibodydeterminations, willingness to participate, anddemonstration of good health. For studies per-formed at the Clinical Center, National Insti-tutes of Health, volunteers were isolated two orthree to a room for 3 to 4 days prior to inocula-tion and 10 to 14 days after inoculation. Examina-tions were performed daily by physicians havingno knowledge of which of several respiratoryagents was administered to a particular volun-teer.An experimental transmission experiment was

performed at the Federal Prison Camp, EglinAir Force Base, Fla. Volunteers were housed ina converted barracks building, and were evaluatedbefore inoculation and twice daily after inocula-tion by physicians who knew which volunteer wasinoculated and which was an exposed susceptible.Complete separation of the two groups, asdescribed in the text, was carefully maintained;however, only partial separation from theremaining camp population was maintained.

InoculaVirus strains used in these studies were ob-

tained from Marines with acute respiratorydisease at Parris Island, S.C., or Camp Lejeune,N.C. (through the courtesy of K. M. Johnson,H. H. Bloom, and R. M. Chanock). Eachinoculum had been passaged once or twice (seeResults) in either human embryonic kidney(HEK) or human embryonic fibroblast (HEF,strain WI-26) tissue cultures (17). The harvests ineach case were frozen and thawed, pooled,centrifuged at 1,000 X g for 20 min, and filteredthrough 800-mu membrane filters (Millipore).

The filtrates were stored at -60 C until used.Each inoculum was safety-tested for adventitiousagents in a manner previously described (19). Inaddition to the above described procedures, thecoxsackievirus A type 21 strain 48654 HEF2 wassubmitted to vacuum concentration and tri-fluortrichlorethane (Gelman) treatment. Furtherdetails of these procedures have been described(6, 8, 9).

Inoculation ProceduresVolunteers received aerosol inoculation by

means of a molded rubber face mask attached toa cylindrical chamber containing a continuousflow of aerosol generated by a Collison ato-mizer. Virus was approximately 10 sec old atthe time of inoculation. This equipment andother necessary auxiliary components were con-tained in a mobile truck and semitrailer and havebeen previously described (15). Each man in-haled 10 liters (±5%) through the nose, andexhaled by mouth into a discharge bag. Eachinoculation required 30 to 60 sec and usuallyfollowed a training period on a previous day withuse of the same equipment. The size of particlesin the aerosol ranged from 0.2 to 3.0 ,u in diam-eter. Particles 1 to 2 ,u in diameter comprised54% of the total particle volume and contained68% of recoverable virus. Further details of theaerosol will be described in a subsequent reportin this symposium (14). Aerosol inoculationswith particles 15 A in diameter were performedwith the same equipment, except that the vibratingreed method of Wolf was used to generate theaerosol (25). Volunteer doses for both aerosolswere calculated from virus assays in simul-taneously collected Shipe impinger samples of theaerosol.

Nasopharyngeal inoculations were performedby the instillation of 0.25 ml of virus inoculuminto each nostril of the volunteer while he wasprone. This inoculation was accompanied by asensation of liquid in the nose but not by adesire to expectorate or swallow. In addition,some volunteers received 0.5 ml of inoculuminto each nostril as well as 0.5 ml sprayed intoeach nostril by a no. 127 DeVilbiss (12) handatomizer. Studies on the aerosol produced bythis atomizer have shown that 99.95% of theinoculum is contained in particles greater than5 IA in diameter and most would be deposited inthe nasopharynx (unpublished data).Collection of Cough, Sneeze, Talking, and Room

Air SamplesParticles produced in selected expiratory

events were collected for size analyses and virus

518 BAcrERIOL. REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

assay. In addition, room air samples werecollected in a large-volume air sampler. Descrip-tion and analysis of the methods used will alsobe described in a subsequent report in thissymposium (14).

Virus Jolations and Identification Procedures

Specimens obtained varied with the virus beingstudied but included nose, throat, and analswabs, nasal washes, and expectoration speci-mens. Specimens were collected prior to andsubsequent to inoculation. Expectoration speci-mens were stored in that form until tested. Nasalwashes were performed with 10 ml of VealInfusion Broth (Difco) containing 0.5% bovinealbumin with antibiotics; swabs were agitated in2 ml of this medium and then discarded. Allspecimens were stored at -20 C until tested.Testing for virus was performed by inoculating0.4 ml of specimen fluid into one HEK and HEFtissue culture tube that contained 1.5 ml of equalparts of medium 199 and Eagle's MEM, 2%inactivated calf or chicken serum, and anti-biotics. The cultures were incubated in a rollerdrum turning at 12 rev/min at 33 to 34 C andwere observed for cytopathic effect (CPE). Thisobservation period was 14 days for coxsackie-virus A type 21 and rhinovirus NIH 1734, but 60days for adenovirus type 4. All the latter studieswere performed in HEK cultures. Tissue culturefluid and cells were harvested when CPE in-volved 75 to 100% of the cell sheet. For cox-sackievirus A type 21 and adenovirus type 4, thefirst and last isolates, as well as interveningisolates, when indicated, were identified byhemagglutination-inhibition (HI) with 20 anti-body units of specific hyperimmune guinea pigserum or rabbit serum. HEF cultures were usedfor identification of comparable specimens ofrhinovirus NIH 1734 by neutralization of 32 to100 TCID5o of virus with specific hyperimmuneguinea pig serum. Further details of theseprocedures have been reported (6, 8, 9).

Serological Tests

Serial fourfold dilutions of inactivated serumwere tested for neutralizing antibody for eachvirus by mixing equal volumes of the serumdilution with a test dose of virus, incubating atroom temperature, inoculating each of two tissueculture tubes with 0.2 ml of the mixture, andobserving thereafter for CPE.

All neutralizing antibody titers, calculated bythe method of Karber, are expressed as the initialdilution of serum completely inhibiting CPE of32 to 100 TcID5o of coxsackievirus A type 21 andadenovirus type 4, but 10 to 16 TCID5o of rhino-

virus NIH 1734. Further details of the proce-dures have been reported (6, 8, 9, 13).

RESULTSResponse to Inoculation with Aerosol and Nasal

DropsCoxsackievirus A type 21: 50%, human in-

fectious doses (HID50). Volunteers free of detect-able antibody were inoculated with a range ofdoses of coxsackievirus A type 21 by small-particle aerosol (diameter of particles, 0.3 to 2.5ju), large-particle aerosol (diameter of particles,15 ju), and nasal drops (0.25 ml in each nostril).An example of the type of response obtained isshown in Table 1. Twenty-eight volunteersreceived strain 49889 HEK1 in a small-particleaerosol, and 18 became infected. The doses,number of volunteers who received each dose,and the number who became infected, as deter-mined by virus isolation and antibody rise, areshown. Based on these findings, the HiD50 for thisinoculum administered in this way corresponds to28 TCID5 (Spearman-Karber method; 13). Onlytwo of the infected volunteers failed to developillness, indicating that the 50% infectious doseand 50% illness dose are nearly the same.

In this experiment, three volunteers developedunexplained mild cases of rhinitis. Experiencewith over 300 volunteer inoculations indicatesthat such an illness is recorded in about 15% ofuninfected individuals. The phenomenon occurseven though virus is inactivated with specifichyperimmune serum, in men with all levels ofserum antibody, and irrespective of virus type ormaterials and methods used for inoculum prep-aration (8). Attempts to isolate a causativeagent in HEK, HEF, and rhesus monkey kidneytissue cultures have been unsuccessful.The HID5o for strain 49889 HEK1 and another

inoculum (strain 48654 HEF2) of coxsackievirus

TABLE 1. Response of antibody-free volunteersinoculated with 0.3 to 2.5-M particle aerosol ofcoxsackievirus A type 21 (strain 49889 HEK,)a

Inhaled dose No. of volun- No. infected No. ill(TCIM60) teers

832 1 1 1676 3 3 2316 3 3 383 2 2 271 5 5 447 4 3 318 4 1 2b6 6 0 2b

a HID50 = 28 TCID50.b Three cases of afebrile URI without infection.

519VOL. 30, 1966

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TABLE 2. HID5o for coxsackievirus A type 21

Inoculum Inoculation method No. of No.9rID695 Confi-volunteers infected Hs denc e limits

Strain 49889 HEK a.Aerosol, 0.3 to 2.5-tL 28 18 28 TCIDs5 15-49particles

Strain 48654 HEF2b.......... Aerosol, 0.3 to 2.5-,u 14 8 34 TCID.s 22-52particles

Aerosol, 15-/ particles 29 12 32 TCID50 13-78Nose drops 14 7 6 TCID5 3-13

a One passage in human embryonic kidney tissue cultures.b Two passages in human embryonic fibroblast tissue cultures.

TABLE 3. Clinical response of antibody-free volunteers to coxsackievirus A type 21

Predominant illnessInoculurn ~Inoculation No. of No.

NoilInocmethod volunteers infected No. illAfebrile Febrile FebrileURIa URI LRIb

Strain 49889 HEKjc........ Aerosol, 0.3 to 28 18 16 1 3 122.5-u particles

Coarse spray and 13 13 8 2 6nose drops

Strain 48654 HEF2d........ Aerosol, 0.3 to 14 8 8 1 72.5-,e particles

Aerosol, 15-,. 29 12 11 1 8 2particles

Nose drops 14 7 5 2 3

a Upper respiratory tract illness.b Lower respiratory tract illness.c One passage in human embryonic kidney tissue cultures.d Two passages in human embryonic fibroblast tissue cultures.

A type 21 administered by each of the describedmethods in shown in Table 2. As can be seen, theHID5o is virtually identical for the three aerosoltitrations; however, for virus administered bynasal drops, it is about fivefold less. Naturalvirus (virus recovered from naturally infectedindividuals, but not cultivated in vitro) ad-ministered by small-particle aerosol (not shown)produced infection in one of two volunteers at a

dose of 28 TCID5o, and in none of six who received7 TCID50, suggesting a similar degree of in-fectivity (8).The HID50 for each aerosol inoculum is based

on inhaled virus. Available information indicatesthat only 50 to 75% of particles of the size rangein the small-particle aerosol would be retainedand that the majority of these would deposit inthe lower respiratory tract (11, 16). This indi-cates that the true HID50 for the inocula ad-ministered in this way is appreciably less thanthat indicated in Table 2. All of the nasal dropinoculation was retained, and therefore theHID5o for this method corresponds to the HID5ogiven in the table. Since virtually all 15-,u particles

would be retained, and the majority would betrapped in the nose, one would expect the HID50by this route of inoculation to be similar to thatobtained by nasal drops. No explanation ispresently available for the observed difference.The clinical responses that correspond to the

strains and inoculation methods in Table 2 areshown in Table 3. In addition, the responses to3,000 TCIDm0 of strain 49889 HEK1 administeredto the nasopharynx by coarse spray and drops areincluded (22). The frequencies of occurrence ofillness in each of the five categories were notsignificantly different. As can be seen, the pre-dominant clinical response to strain 49889inoculated by small-particle aerosol was febrilelower respiratory tract illness. All 12 volunteerswith this response were clinically diagnosed ashaving acute tracheobronchitis. The pertinentdata obtained on a volunteer from a more recentexperiment, but typical of the syndrome, areshown in Fig. 1. Characteristic of this syndromewas the occurrence of pain in the neck (tracheal)and chest, the latter usually being substernal.Cough, often paroxysmal, was usually non-

520 COUCH ET AL. BAcrERIOL. REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

J.J. d 37 Yr, Coxsockievirus A-21, 95 TCID50, Aerosol

ys *-2, -I, O .I 2, 3, 4, 5, 6, 7, 8, 9, 10,

Illness

Virus Recovery(Nosophoryngeol

wosh)Neut. Antibody(reciprocali

0

128

FIG. 1. Case report of an antibody-free volunteerinoculated with coxsackievirus A type 21 by small-particle aerosol.

productive, although auscultation of the chestoccasionally revealed scattered rhonchi, and, intwo cases, there was X-ray evidence of pneu-monia. These lower respiratory tract symptomswere accompanied by malaise, myalgias, chillysensations, sweats, headache, and anorexia.Illness was not limited to the lower respiratorytract, however, since 9 of the 12 volunteers withtracheobronchitis also had upper respiratorytract illness that was characterized by rhinorrheaand nasal obstruction. Four of the remaining sixinfected volunteers had upper respiratory tractillness only, and the other two had infectionwithout apparent illness.

In contrast to the small-particle aerosolresponse, 8 of 13 volunteers who received naso-pharyngeal inoculation developed upper respira-tory tract (nasopharyngeal) illness only. The factthat virus was deposited in the nasopharynx inthis case and predominantly in the lower respira-tory tract in the former suggested that virusdeposition site accounted for this difference andthat it might be the factor that determines theclinical response. However, when strain 48654was administered by small-particle aerosol, thelower respiratory tract illness, which was char-acteristic of strain 49889 given in this way, wasnot seen. The predominant clinical response tostrain 48654 in a small-particle aerosol wasfebrile upper respiratory tract illness (Table 3).Thus, virus deposition site and inoculum dif-ferences both appeared as important factors indetermining the type of clinical response.

Febrile upper respiratory tract illness was alsothe predominant clinical response for strain48654 administered by 15-A particle aerosol andby nasal drops (Table 3). Not shown are clinicalresponses to natural virus and to still anotherstrain of virus administered by small-particle

aerosol (8, 20). For these inocula, febrile upperrespiratory tract illness also predominated. Thiscombined experience suggests that virus deposi-tion site may be an important factor in deter-mining the type of clinical response that occurs.However, for coxsackievirus A type 21, moststrains appear to lack the capability of producinglower respiratory tract illness when presentedsuch an opportunity by virus deposition at thissite.

In all other aspects, the clinical responses weresimilar for each inoculum and inoculationmethod. The incubation period was 2 to 5 days,illness usually lasted 2 to 3 days, fever rarelyexceeded 38.5 C, and fever usually persisted lessthan 1 day.The effect of pre-existing serum neutralizing

antibody on the responses following inoculationof volunteers with coxsackievirus A type 21(strain 49889) has not been completely evaluated,but the data available are shown in Table 4. Ascan be seen, all individuals with intermediatetiters of antibody were infected after naso-pharyngeal inoculation, but infection occurred inonly 5 out of 11 with high titers. A similarsuggestion of reduction in infection also occurredin the small-particle aerosol groups.

Rhinovirus NIH 1734: HID50. Volunteers freeof detectable antibody to this virus were inocu-lated with a range of doses of rhinovirus NIH1734 by small-particle aerosol and by nasal drops.The HID50 for each inoculation method is shownin Table 5. Nasal drop doses of 1 TcID50 and lesswere extrapolated values based on dilutions of apool with known virus concentration, andaerosol doses of 2 and less were extrapolatedfrom measured concentrations of virus in aerosolsproduced, during the inoculation period, byhigher concentrations of virus. Repeated tests ofseveral dilutions of virus run in sequence havebeen shown to produce proportionate changes inaerosol virus concentration. Assays for virus wereperformed in HEF (WI-38) tissue cultures, in a

TABLE 4. Response of volunteers with pre-existingantibody to inoculation with coxsackievirus A type 21

Nasopharyngeal Aerosol, 0.3 to 2.5 isinoculation particles

Level ofantibody No. of No. N No. of No .

voln 0nN. voluon- in NO.teers fected ill teers fected ill

Intermediate 6 6 4 5 3 2(1:4-1:128)High I1 5 0 3 0 3(1:256 or greater) 11 S 0 3 0 3a

a Each illness was mild rhinitis.

0 000 0 + + + + + + + +

<2

VOL. 30, 1966 521

21 28F=,:.

+

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

COUCH ET AL.

TABLE 5. HID5o for rhinovirus NIH 1734

a ~~~~95% Con-Inoculation method , c HID5so fidence

*0 limits> 6

Nasal drops....... 17 11 0.032 TCID50 0aAerosol, 0.3 to 2.5 IA 26 20 0.68 TCID6s 0.2-2.0

particles ..... _

a Indicates no intermediate response between 100 and 0%infection.

manner described previously (6). Other types oftissue culture [HEK and HEF (WI-26)] andtissue culture assay [HEF (WI-38) plaque assay]were tested and found to be equal to or less sensi-tive than the cultures and methods used.As can be seen in Table 5, the HID50 for both

inoculation methods was below the practicallimits of detection. Failure to infect all volunteerswith small-particle aerosol inoculation firstoccurred at an inhaled dose of 2 TCID50, and noneof three who inhaled 0.06 TCID50 became in-fected. The HIDI5 for this inoculation method was0.68 TCID50 (Spearman-Karber; 13). In contrast,all volunteers who received 0.1 TCID5o by nasaldrops became infected, although none becameinfected at two lower doses. The HID50 for thismethod corresponded to 0.032 TCID5o. Theseresults indicate an approximately 20-fold dis-parity between infectivity for the virus given bythe two methods. The disparity could be ac-counted for by assuming that the 10 to 20% ofsmall-particle aerosol particles that deposit inthe nasopharynx are responsible for all infectionin volunteers inoculated in this way. However,the fact that this is not the case is suggested bythe occurrence of lower respiratory tract illnessin some of these volunteers. In any event, thedata suggest that the nasal mucosa is somewhatmore susceptible to rhinovirus NIH 1734 than isthe lower respiratory tract. Although the dif-ference was less for coxsackievirus A type 21,it was similar in direction.The clinical responses of all volunteers who

have received either nasal or small-particleaerosol inoculation with rhinovirus NIH 1734are shown in Table 6. As can be seen, the char-acteristic response to either method of inocula-tion is an upper respiratory tract illness which inall respects is a common cold. The pertinentdata obtained from one of the volunteers inocu-lated by nasal drops are shown in Fig. 2. Hisresponse consisted of a common cold syndromecharacterized by nasal obstruction and discharge,and was accompanied by throat irritation andsystemic symptoms. The extent of the rhinorrhea

TABLE 6. Clinical response of volunteers toinoculation with rhinovirus NIH 1734

No. of IllnessInoculation method infected No. illvolun-

UIteers URI LRja LRI

Coarse sprayand nosedrops ........ 48 43 41 2 0

Aerosol, 0.3 to2.5 , particles 41 33 23 5 5

a Upper and lower respiratory tract illness.

is shown in the figure. Fever was absent in thisvolunteer and occurred in less than 10% of thevolunteers, regardless of method of inoculation.As can be seen in Table 6, lower respiratory

tract illness (acute tracheobronchitis) was pre-dominant in five volunteers who received small-particle aerosol inoculation, and diffuse respira-tory tract disease without a predominant localiza-tion was seen in five others. Predominant lowerrespiratory tract illness was not seen in meninoculated by nasal drops, although two volun-teers exhibited a combination of upper and lowerrespiratory tract illness. These findings suggestthat aerosol inoculation may produce lowerrespiratory tract involvement, but the char-acteristic response to infection produced byeither method is an upper respiratory tract ill-ness.The incubation period of the illnesses produced

by both inoculation methods was 2 to 4 days, theillness usually lasted 2 to 3 days, and fever, whenit occurred, was usually 1 day in duration.The effect of pre-existing serum neutralizing

antibody on responses to inoculation withrhinovirus NIH 1734 is shown in Table 7. Ascan be seen, no significant reduction in frequencyof infection occurred unless high levels of serumantibody were present. This reduction in fre-quency of infections occurred for both methodsof inoculation and was accompanied by a similarreduction in illnesses. [Data are grouped forconvenience. Individual values were tested inSpearman's rank correlation or Yates meanscore tests (13). Reduction in infection and ill-ness with increasing serum antibody was statis-tically significant (P < 0.05) for both inocula-tion methods.]

Adenovirus type 4: HID50. Nine volunteersfree of detectable antibody to adenovirus type 4received small doses of this virus by small-particle aerosol. Six volunteers received the virusby 15-,u particle aerosol. The results of these

522 BACTERIOL. REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

RHINOVIRUS COMMON COLD - W.W.A., 38 Yr.d9 VOLUNTEERDay After Inoculation -2 -I 0 2 3 4 5 6 7 8 9 10g13 21

20DA I LYNASAL

SECRETION(GMS) 10 _

Illness L Rhinitis, Malaise, Throat Irritation

NIH 1734±Nasopharynx 0 + + + + + + + + + *+Neut. Antib. <1:2 1:128

FIG. 2. Case report ofan antibody-free volunteer inoculated with rhinovirus NIH 1734 by coarse spray and nasaldrops. (Reproduced with the permission of the Journal of Clinical Investigation.)

studies are shown in Table 8. As can be seen,all volunteers who received doses of 11 and5 TCID50 by small-particle aerosol became in-fected, but only one out of three became infectedat a dose of 1 TCID5o. Other volunteers wereinoculated in this way, and, although the data areincomplete, the studies indicate that the HID5ofor small-particle aerosol inoculation is about 1TCLD50. It should be stated that these virus assayswere performed in HEK tissue cultures, the mostsensitive tissue available for adenovirus, and thecultures were observed for 60 days for CPE withsubpassage as needed. This time period wasshown to provide maximal detection of adeno-virus (9).Only one dose level of adenovirus type 4 has

been administered by 15-As particle aerosol, andthis was 1,000 TCID5o. All six volunteers whoreceived this dose became infected. Preliminaryresults on inoculation of volunteers by nasaldrops indicate that the HID50 by this method isabout 20 TCID5o. This combined experience withadenovirus type 4 suggests that a greater dose ofthis virus is required to initiate infection in thenasopharynx than in the lower respiratory tract.

Also shown in Table 8 are the clinical responsesseen in the volunteers inoculated by aerosol. Ascan be seen, all volunteers infected by means ofsmall-particle aerosol inoculation became ill,and the illness was usually febrile. Three volun-teers had predominantly upper respiratory tractillness, and, in three others, illness was pre-dominantly in the lower respiratory tract. Thelatter included one instance of mild pneumonia.Only three of the six volunteers infected by 15-,uparticle aerosol inoculation became ill, two withfebrile upper respiratory tract illness and onewith pneumonia. The incubation period for these

TABLE 7. Response of volunteers with pre-existingantibody to inoculation with rhinovirus NIH 1734

Nasopharyngeal Aerosol, 0.3 to 2.5 juinoculation particles

Level ofanioy No. of No. No No. of No. in- No.

volun- in- ill volun- fecte lteenr fected teern ted ill

Low 3 3 2 5 4 4(1:2-1:8)

Intermediate 9 7 8 6 4(1:16-1:64) 9 8 7 8 6 4

Hi~gh 1(1:128 or greaser) 13 4 4 1

illnesses varied between 6 and 13 days, durationof illness varied between 2 and 10 days, and feverbetween 1 and 8 days. In addition, the severity ofillness, as manifested in respiratory tract involve-ment and constitutional symptoms, also wasquite variable. Upper respiratory tract findingsoccurred in all men in the 15-,u particle aerosolgroup, whereas this finding was variable in thesmall-particle aerosol group. The pertinent dataobtained on one of the volunteers who exhibitedthe syndrome described as acute respiratorydisease (ARD) of military recruits are shown inFig. 3. Bacteriological cultures were negative forpathogens, and spontaneous recovery occurredwithout antibiotic therapy.

It is notable that the syndromes of febrilerespiratory tract illness that occurred after aerosolinoculation resemble the naturally occurring type4 adenovirus diseases of military recruits (3, 7,21). Previous studies by others, in which volun-teers were inoculated in the nasopharynx, usuallyresulted in asymptomatic infection or mildafebrile upper respiratory illness (1). Inocula-

523VOL. 30, 1966

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TABLE 8. Response of antibody-free volunteers to adenovirus type 4

Illness

Inoculation method Dose' No. of No. infected No. illvolunteers Afebrile Febrile FebrileURIb URI LRI"

Aerosol, 0.3-2.5-pA 11 3 3 3 1 1 1particles

5 3 3 3 1 21 3 1 1 1

Aerosol, 15-p& par- 1,000 6 6 3 2 1tides

a Expressed as TCID50.b Upper respiratory tract illness.¢ Lower respiratory tract illness.

2/Mr d' Adenovirus type 4, /000 TC/D50 -15a aerosol partic/es.

39 I

MAX. TEMP. 00C 38-

37

Illness

inoculation

White blood cell countl 10.7 9.2 8.9

Sed. rate mm/hrl 5Nos

Virus Recovery ThrocAnu

Neut. Antibody (reciprocc

Pharyngitis, rhinitis \///"~ prostration

10.0 9.2 6.2 4.0

9 25 39 27ie 0 0 0 0 0 + 0 + + + +itlO 0 0 0 + + + + + + +As o 0 0 0 0 0 0 + + + +

1)<2|| . " ' , , , , ~(DAY 20,>1024)01 <2. A I .- I DAY O _?_O )0 2 4 6 8 10 12 14 e 16

DAYS AFTER INOCULATIONFIG. 3. Case report of an antibody-free volunteer inoculated with adenovirus type 4 by 15-p particle aerosol.

White blood cell counts are times 103 per cm. (Reproduced with the permission of the American Review of Respir-atory Diseases.)

tions into the conjunctival sac resulted in oc-currence of conjunctivitis only or pharyngo-conjunctival fever, illnesses which rarely occurnaturally in type 4 infection, and which were notseen in the present studies (1). These findingssuggest that the unique feature of the presentinoculations, deposition of virus in the lowerrespiratory tract, was the major factor accountingfor the recruit-type illnesses. This is supportedby the fact that small doses of virus given bysmall-particle aerosol produced illness in allvolunteers infected by this method of inoculation,

whereas the large dose given by 15-,u particleaerosol caused illness in only three of six infectedmen. Evidence indicates that most of the 15-ptparticles were deposited in the upper respiratorytract, but the possibility exists of deposition in thelower respiratory tract either by direct inhalationor as a result of particle fragmentation (16). It issuggested that in three men this occurred andcaused febrile illness.Three volunteers with pre-existing antibody

titers of 1:32 to 1:64 received 6 TCVo by small-particle aerosol, and none became infected or ill.

524 COUCH ET AL. BACTERIOL.' REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

Inoculations of volunteers with pre-existing anti-body into the nasopharynx or conjunctival sacby others have also demonstrated the protectiveeffect of serum antibody (1).

Evidence for Airborne TransmissionDetection of virus in particles produced by

coughing, sneezing, and normal expiration. By useof methods (14) for recovery of virus from par-ticles produced by coughs and sneezes, virustitration was carried out on 61 cough collectionsand 58 sneeze collections from volunteers in-fected with coxsackievirus A type 21 (Table 9).The collection method involved coughing orsneezing into a collapsed weather balloon througha tight-fitting face mask. The air in the balloonwas evacuated through a Shipe impinger toremove airborne particles, and material impactedon the wall of the balloon was collected by rins-ing with sterile tissue culture fluid. When theresults of both samples were combined, 39% ofcough specimens and 50% of sneeze specimenswere positive for virus. Thirty per cent of airsamples were positive for both events, and themean quantity present was 30 TcDw0 and 60TCID5o for Cough and sneeze samples, respectively.This close similarity in results is of interest inview of the approximately 20-fold greater numberof particles and particle volumes produced bysneezing (14). This finding suggests that the con-centration of virus in secretions released in smallparticles produced by coughing is greater thanthat produced by sneezing.

Analysis of balloon wall samples revealed adisparity between the two events. Wall samples ofsneezes were more frequently positive than thewall samples of coughs, but the mean quantitypresent was only twofold greater. However, themean quantity of virus present in the wall samplesfrom sneezes does not include four sneezes inwhich gross contamination with large quantities

TABLE 9. Virus recovery from particles in coughsand sneezes produced by volunteers infected with

coxsackievirus A type 21

1 ~~~~~MeanPhenomenon No. Per cent Source Per cent quan-tested positive positive tity

(TCIDso)

Sneeze.... 58 52 Aira 30 60Wallb 45 100

Cough .... 61 39 Air" 30 30Wallb 20 50

a Assay of Shipe impinger collection of particlessuspended in air in balloon.bAssay of 10-ml liquid rinse of balloon wall.

of nasal secretion occurred. The wall samples ofthese sneezes contained 30,000 to 500,000 TCID50of virus. The reasons for the disparity in fre-quency of detection of virus on the balloon wallfor the two events are not known at the presenttime, since studies have revealed similar particlesize distributions for both events (12, 14).

Breathing samples were tested by collectingthe entire amount of expired air in Shipe im-pingers through a closed system for 30-minperiods. This testing constituted sampling ofair expired for 2 hr per day from four infectedvolunteers during the period that included oc-currence of illness and maximal virus shedding. Avolume equivalent to 12 hr of expired air wastested in this way, and all samples were negativefor virus.A number of factors were evaluated to deter-

mine the cause for virus release in the process ofcoughing and sneezing. These evaluationssuggested that the presence of nasal obstructionand discharge was the most important deter-minant for release of virus when infected personssneeze [with nasal obstruction and discharge, 19of 24 sneeze samples were positive; without nasalobstruction and discharge, 11 of 34 samples werepositive (P < 0.001)]. In contrast, positive coughspecimens bore a relationship only to the quantityof virus present in respiratory secretions, andthis relationship occurred for air samples only[combined nasal and oral secretions, Yates meanscore, test, P = 0.05 (13)]. Since cough particleswould presumably be derived from pharyngealand lower respiratory secretions, it is suggestedthat the concentration of virus in these secretionsvaried proportionately with the secretions tested.

Virus in room air. The contribution to roomair contamination by coughing, sneezing, andpossibly by other expiratory phenomena of maninvolves frequency and occurrence of the phe-momenon, inactivation of virus, and physical lossof aerosol particles, in addition to quantity ofvirusreleased. The significance of these factors indetermining environmental contamination wastested by collecting particles present in the air ofrooms occupied by volunteers infected withcoxsackievirus A type 21 and then assaying thecollections for virus.The large-volume air sampler was used to

collect particles from approximately 70% of roomair after a period of 2 hr with no ventilation.Shown in Table 10 are the results of testing 30such samples collected during the acute phase ofillness and maximal virus shedding. Of the 30samples, 14 were positive, and, as cao be seen, thefrequency of positive samples increased withincreasing quantity of virus present in respiratory

525VOL. 30, 1966

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

COUCH ET AL.

TABLE 10. Relation of virus quantity in respiratorysecretions to virus in room air samples

Air sample

Mean (3 vol.) virus quantityin secretions No. of No. Mean

tests positive virus.quantitity

10-30a 5 1 5b30-100 11 2 160100-300 5 4 250300-1,000 6 4 50

1,000->1,000 3 3 500

a Expressed as TCID50 per milliliter of secretion.b Expressed as TCID50.

secretions [Smirnoff test, P < 0.01 (13)]. Themean virus quantity in positive samples is shownin the last column and was sufficiently variableso that no quantitative relationship to virus inrespiratory secretions was detected, although it isof interest that the largest quantity present inroom air, 28,000 TCID50, was in the room withthe highest virus concentration in secretions.

Since both positive cough and room airsamples were related to quantity of virus inrespiratory secretions of infected volunteers, itwas suggested that coughing was responsible forcontamination of room air with virus. When theresults were analyzed by room, it was found thatthe presence of virus in cough air samples fromvolunteers occupying a room was significantlyrelated to the recovery of virus from the air of thatsame room on the same day. [Positive room airsamples, 10 of 11 rooms with positive cough airsamples; negative room air samples, 2 of 7 roomswith positive cough air sample (P = 0.03)]. Thisfurther suggested that cough is the importantintermediary between virus in secretions andvirus in room air. No such relationship wasdetected for sneezing. These findings are notsurprising, since cough as a symptom wasrecorded as being frequently present in thesesame volunteers at this time, whereas sneezingwas not.

Preliminary report on a transmission experi-ment. An experiment designed to test whetherthe occurrence of air contamination with virus issufficient to produce airborne transmission hasbeen performed (unpublished data). Nineteenplacebo-inoculated volunteers were exposed toair surrounding infected volunteers by housingthe two groups in a converted barracks andseparating them with a double-wire barrier.Even distribution of air on both sides of the testbuilding was accomplished by means of largefloor fans and was proved by generating an

aerosol containing a fluorescein dye on one sideand then collecting and analyzing air samplesfrom different locations throughout the building.Coxsackievirus A type 21 infection was producedin 10 volunteers with aerosol inoculation, andall exposed individuals became infected with thisvirus during the course of the study. A specificseparation of results in terms of contact and air-borne-acquired infection is not completed, but itis possible to state that airborne transmission un-questionably occurred.

DIscussIoNThe theory that respiratory viruses are trans-

mitted by the airborne route has been popularin the past, primarily because it seemed reason-able to assume that coughing and sneezing,common symptoms of viral respiratory disease,produce aerosols that would accomplish suchtransmission. Despite this assumption, proof thatman produces aerosols that contain virus andthat sufficient viral contamination of air occursto result in this type of transmission, bothessential requirements for airborne transmission,has not been obtained (24). The results presentedin this report provide this important informa-tion. It was shown that individuals infected withrespiratory viruses, in this case coxsackievirus Atype 21, produce airborne virus in quantitiessufficient to infect susceptible individuals. Thecapacity to produce viral aerosols was tested forthree expiratory events. Breathing samples wereuniformly negative for virus, whereas cough andsneeze samples were frequently positive. Thus,whereas in man the former event is probably in-significant in producing transmission of therespiratory viruses, it seems likely that it isimportant in the mouse-influenza system ofShulman and Kilbourne (23) in which airbornetransmission has also been conclusively demon-strated. For man, coughing and sneezing appearto be the significant events for producing viralaerosols.

Studies in which virus released by coughingand sneezing was collected in a balloon andseparated into an air phase and a wall phaseprovided quantitative results that correspondroughly to virus involved in airborne transmissionand contact transmission, respectively. Virus wasrecovered more often from the air sample fromcoughs than from the wall samples, although thewall samples of sneezes were more commonlypositive than the air samples. This would suggestthat sneezing may be of some significance for thatform of transmission involving direct impactionof large particles in the nasopharynx, whereascough contributes primarily to small-particle

526 BAcrERioL. REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

aerosol transmission. Despite the differences infrequency of recovery, the difference in meanquantity of virus in each phase was small andquite similar for each event. These findings are incontrast to the findings of Buckland et al. (5),in which the vast majority of virus released insneezing was found on the sides of a large sam-pling bag. However, the different collection meth-ods involved may account for this disparity.

Despite a larger number of particles in sneezesthan in coughs, the quantity of virus expelled inthe two events was remarkably similar, suggest-ing that, in these volunteers, the concentration ofvirus in secretions atomized in coughing wasrelatively greater than that in secretions atomizedin sneezing (12, 14). Inoculation of these volun-teers was performed by small-particle aerosol,and, although lower respiratory tract secretionswere not quantitated, virus is known to have beendeposited at this site and probably inducedinfection there. Thus, it is possible that themethod by which infection was induced mayhave contributed to the virus recovery resultsfrom coughing.The fact that infected persons are capable of

producing airborne virus does not necessarilyindicate that virus can be transmitted in this way.Viral aerosols produced by infected persons aresubject to dilution in room air, biological decay,and sedimentation. Nevertheless, assuming nor-mal breathing by susceptible volunteers and an

infectious dose of about 6 to 30 TCID5o, assay ofair samples from rooms occupied by infectedvolunteers indicated that transmission would beaccomplished in from 5 min to 24 hr. Further-more, in view of the observed efficiency (11%)of the air-sampling equipment, larger thanmeasured doses of virus were actually availablefor inhalation (14). In addition, the present datasuggest that cough is a most important event inproducing viral contamination of air.The findings described above stimulated the

performance of an experiment to test the assump-tion that airborne transmission is possible, and

preliminary results revealed that airborne trans-mission occurred from infected cases to suscepti-bles across a wire barrier.

Airborne and contact transmisison was simu-lated in volunteers by aerosol and nose dropinoculation, respectively. Studies with three dif-ferent strains of coxsackievirus A type 21 indi-cated a similar HID5o of about 30 TCID5o for thisvirus, when predominant deposition was in thelower respiratory tract (small-particle aerosol),and a lower value when nasal drops were used.Since the latter inoculation method provideddeposition only in the nasopharynx, it is suggestedthat the nasal mucosa exhibited a greater sus-ceptibility to infection with this virus than didthe lower respiratory tract. Another picorna-virus, rhinovirus NIH 1734, exhibited an evengreater difference between the HID50 for nasal dropinoculation and for small-particle aerosolinoculation. Thus, the data suggest that, for bothof these viruses, the nasal mucosa is the pre-ferred site for infection. Although definitivecomparisons are incomplete, present evidencesuggests a disparity in infectivity in the oppositedirection for adenovirus type 4. This virus exhibitsa high degree of infectivity for the lower respira-tory tract, but the nasopharynx appears to lackthis degree of susceptibility.The most common illness response to each

virus that followed inoculation by nasal dropsand small-particle aerosol is shown in Table 11.For comparative purposes, the most commonnaturally occurring illness response to each virusis also listed. As can be seen for coxsackievirus Atype 21, regardless of method of inoculation aswell as dose, febrile upper respiratory illnessusually results in volunteers, whereas naturallyoccurring illness is reported to be usually afebrile(2, 18). This disparity may well be explained bythe fact that feveL in volunteers is usually sobrief in duration that, without 24-hr observation,the majority of volunteers would have beendesignated afebrile. The predominant lowerrespiratory tract illness that was seen with one

TABLE 11. Characteristic natural and experimentally induced clinical responses torespiratory viruses

Experimental inoculationVirus Natural inoculation

Nasopharyngeal Aerosol, 0.3 to 2.5 ,u particles

Coxsackievirus A type 21 .......... Febrile URIa Febrile URI Afebrile URIRhinovirus NIH 1734............... Afebrile URI Afebrile URI Afebrile URIAdenovirus type 4.................. Afebrile URI Febrile URI or LRI, Febrile URI or LRI,

or both or both

Upper respiratory tract illness.

527VOL. 30, 1966

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

COUCH ET AL.

inoculum administered by small-particle aerosolappears to have been relatively unique, and wasdue to properties of the virus in that inoculumthat are not usually exhibited by strains of thisvirus.For rhinovirus NIH 1734, afebrile upper

respiratory tract illness occurs in volunteersregardless of inoculation method and is also thecharacteristic natural clinical response to thisand other rhinoviruses (4, 10). Data thus faravailable indicate that naturally occurringadenovirus type 4 disease can regularly be re-produced in volunteers only by aerosol inocula-tion. Nasal inoculation, throat swabbing, andconjunctival inoculation have all failed to re-produce naturally occurring type 4 adenovirusdisease (1).

It is therefore suggested that adenovirus type 4disease is transmitted in natural circumstancesprimarily by the airborne route. The informationavailable on coxsackievirus A type 21 and rhino-virus NIH 1734 indicates that either airborne orcontact transmission would result in the upperrespiratory tract illness characteristic of naturallyoccurring illness. However, the small-particleaerosol inoculation results suggest that airbornetransmission would produce a more variedresponse and account for the lower respiratorytract illness which is sometimes associated withnaturally occurring upper respiratory tractdisease (4, 10, 18).Thus, the data presented on production of air-

borne virus, environmental air contaminationwith virus, and the demonstration of airbornetransmission summarized in the present reportindicate that airborne transmission probablyoccurs naturally. Present information, however,does not indicate whether airborne transmissionis the predominant mechanism of natural trans-mission. At the present time, it seems mostreasonable to suggest that both contact and air-borne transmission occur in natural circum-stances, and that the predominant method oftransmission varies with the virus and the op-portunity presented in a particular situation.For those viruses and situations in which air-borne transmission predominates, it may bepossible to devise suitable methods of control ofrespiratory viral infection.

SUMMARY

Volunteers were inoculated with respiratoryviruses by means of nasal instillations andinhalation of aerosols. The former method wasused to simulate contact transmission, and thelatter to simulate airborne transmission. TheHID50 for coxsackievirus A type 21 was about 30

TCID50 by aerosol and 6 TCID50 by nose drops.Similar determinations for rhinovirus NIH 1734revealed HID50 of 0.68 TCID50 by aerosol and 0.032TCID50 by nasal drops. The clinical response wascharacteristically an upper respiratory tractillness for both viruses by both inoculationmethods, although coxsackievirus A type 21illness was usually febrile, and rhinovirus illnessusually was not. Incomplete infectivity studieswith adenovirus type 4 suggest a disparity in theopposite direction for this infection. Aerosolinoculation revealed an HID50 of about 1 TCID50and thus far is the only inoculation method whichregularly reproduced naturally occurring ARD.The suggestion that airborne transmission

accounted for some naturally occurring acuterespiratory disease was further evaluated bystudying the production of airborne virus bycoughs and sneezes and the contamination ofroom air with virus. Coughing and sneezingregularly produced quantities of virus sufficientto infect, whereas breathing did not. Room airsamples revealed contamination probably suf-ficient to infect susceptibles. In addition, pre-liminary results of a transmission experimentwith coxsackievirus A type 21 indicate that air-borne transmission unquestionably occurred. Itwas concluded that both contact and airbornetransmission of the respiratory viruses probablyoccur in natural circumstances, and that thepredominant method of transmission may varywith the virus and with the particular environ-mental situation.

ACKNOWLEDGMENTS

We thank Holly A. Smith, Carol Uhlendorf, JoanC. Enterline, James Turner, and Leonard P. Durocherfor their technical work; Edward B. Derrenbacher andCharles 0. Masemore, U.S. Army Biological Labora-tories, Fort Detrick, Md., for assistance with theaerosol inoculations; Mollie M. Fletcher for prepara-tion of the manuscript; and the following who mate-rially assisted in the program: James Bennett, formerDirector, and Charles E. Smith, Chief Medical Officer,Bureau of Prisons, U.S. Department of Justice; andFranklyn Gray, Assistant Chief, Normal VolunteerProgram, Clinical Center, National Institutes ofHealth. David Alling kindly performed the statisticalanalyses. The volunteers are commended for theirexcellent cooperation.

LITERATURE CITED

1. BELL, J. A., T. G. WARD, R. J. HUEBNER, W. P.ROWE, R. G. SUSKIND, AND R. S. PAFFEN-BARGER, JR. 1956. Studies of adenoviruses(APC) in volunteers. Am. J. Public Health46:1130-1146.

2. BLOOM, H. H., K. M. JOHNSON, M. A. MUFSON,AND R. M. CHANOCK. 1962. Acute respiratory

528 BACTERIOL. REV.

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

TRANSMISSION OF RESPIRATORY VIRUS DISEASES

disease associated with Coxsackie A-21 infec-tion. II. Incidence in military personnel: ob-servations in a non-recruit population. J. Am.Med. Assoc. 179:120-125.

3. BLOOM, H. H., B. R. FORSYTH, K. M. JOHNSON,M. A. MUFSON, H. C. TURNER, M. B. DAVISON,AND R. M. CHANOCK. 1964. Patterns of adeno-virus infections in Marine Corps personnel. I.A 42-month survey in recruit and non-recruitpopulations. Am. J. Hyg. 80:328-342.

4. BLOOM, H. H., B. R. FORSYTH, K. M. JOHNSON,AND R. M. CHANOCK. 1963. Relationship ofrhinovirus infection to mild upper respiratorydisease. I. Results of a survey in young adultsand children. J. Am. Med. Assoc. 186:38-45.

5. BUCKLAND, F. E., M. L. BYNOE, AND D. A. J.TYRRELL. 1965. Experiments on the spread ofcolds. II. Studies in volunteers with cox-sackievirus A21. J. Hyg. 63:327-343.

6. CATE, T. R., R. B. COUCH, AND K. M. JOHNSON.1964. Studies with rhinoviruses in volunteers:Production of illness, effect of naturally ac-quired antibody, and demonstration of a pro-tective effect not associated with serum anti-body. J. Clin. Invest. 43:56-67.

7. COMMISSION ON ACUTE RESPIRATORY DISEASES.1947. Clinical patterns of undifferentiated andother acute respiratory diseases in Army re-cruits. Medicine 26:441-464.

8. COUCH, R. B., T. R. CATE, P. J. GERONE, W. F.FLEET, D. J. LANG, W. R. GRIFFITH, AND V.KNIGHT. 1965. Production of illness with asmall-particle aerosol of Coxsackie A21. J.Clin. Invest. 44:535-542.

9. COUCH, R. B., T. R. CATE, W. F. FLEET, P. J.GERONE, AND V. KNIGHT. 1966. Aerosol-induced adenovirus illness resembling thenaturally-occurring illness in military recruits.Am. Rev. Respirat. Diseases 93:529-535.

10. FORSYTH, B. R., H. H. BLOOM, K. M. JOHNSON,AND R. M. CHANOCK. 1963. Patterns of illnessin rhinovirus infections of military personnel.New Engl. J. Med. 269:602-606.

11. DAVIES, C. N. 1960. Deposition of dust in thelungs; a physical process, p. 44-58. In E. J.King and C. M. Fletcher [ed.], Industrial pul-monary diseases. Little Brown & Co., Boston.

12. DUGUID, J. P. 1946. The size and duration of air-carriage of respiratory droplets and droplet-nuclei. J. Hyg. 44:471-479.

13. FINNEY, D. J. 1964. Statistical method in bio-logical assay. Hafner Publishing Co., NewYork.

14. GERONE, P. J., R. B. COUCH, G. V. KEEFER, R. G.

DOUGLAS, JR., E. B. DERRENBACHER, AND V.KNIGHT. Assessment of experimental andnatural viral aerosols. Bacteriol Rev. 30:576-584.

15. GRIFFITH, W. R. 1964. A mobile laboratory unitfor exposure of animals and human volunteersto bacterial and viral aerosols. Am. Rev.Respirat. Diseases 89:240-249.

16. HATCH, T. R., AND P. GROSS. 1964. Pulmonarydeposition and retention of inhaled aerosols.Academic Press, Inc., New York.

17. HAYFLICK, L., AND P. S. MOORHEAD. 1961. Theserial cultivation of human diploid cell strains.Exptl. Cell Res. 25:585-621.

18. JOHNSON, K. M., H. H. BLOOM, M. A. MUFSON,AND R. M. CHANOCK. 1962. Acute respiratorydisease associated with Coxsackie A-21 virusinfection. I. Incidence in military personnel:observations in a recruit population. J. Am.Med. Assoc. 179:112-119.

19. KNIGHT, V. 1964. The use of volunteers in medicalvirology, p. 1-26. In E. Barger and J. L.Melnick [ed.], Progress in medical virology.S. Karger, Basel.

20. LANG, D. J., T. R. CATE, R. B. COUCH, V. KNIGHT,AND K. M. JOHNSON. 1965. Response of volun-teers to inoculation with hemagglutinin-posi-tive and hemagglutinin-negative variants ofCoxsackie A21 virus. J. Clin. Invest. 44:1125-1131.

21. RowE, W. P., J. R. SEAL, R. J. HUEBNER, J. E.WHITESIDE, R. L. WOOLRIDGE, AND H. C.TURNER. 1956. A study of the role of adeno-viruses in acute respiratory infections in aNavy recruit population. Am. J. Hyg. 64:211-219.

22. SPICKARD, A., H. EVANS, V. KNIGHT, AND K.JOHNSON. Acute respiratory disease in normalvolunteers associated with Coxsackie A-21viral infection. III. Response to nasopharyngealand enteric inoculation. J. Clin. Invest. 42:840-852.

23. SCHULMAN, J. L., AND E. D. KILBOURNE. 1963.Experimental transmission of influenza virusinfection in mice. I. The period of transmissi-bility. J. Exptl. Med. 118:257-266.

24. TYRRELL, D. A. J. 1965. Common colds andrelated diseases. The Williams & Wilkins Co.,Baltimore.

25. WOLF, W. R. 1961. Study of the vibrating reed inthe production of small droplets and solidparticles of uniform size. Rev. Sci. Instr. 10:1124-1129.

529VOL. 30, 1966

on May 21, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from