U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter 12 - Pneumonias



The Pneumonias

Section I. Primary Atypical Pneumonia

Norman L. Cressy, M. D.

The occurrence of respiratory illness among troops during periods of mobilization has always been a matter of great importance, and pneumonia as either a primary or secondary disease has usually been a major cause of death. During the winter of 1812 and 1813, there was a high incidence of acute respiratory disease among troops stationed on the northern frontier.1 Measles complicated by pneumonia was epidemic from September to December 1812. During the following winter, although the morbidity and mortality were lower for the Army as a whole, new troops joining the service were as severely affected as the men who were mobilized during the preceding year. In the War Between the States, acute respiratory disease was again an important cause of morbidity and mortality.2During the winter of 1862, there was excessive seasonal variation of respiratory disease rates with a curve for catarrh which was similar to the influenza epidemic of 1918, but there was no epidemic such as the one which occurred in the fall of 1918. In the Mexican-border mobilization of 1916, there was an epidemic of pneumonia with about; 400 cases occurring among 40,000 troops and a 20-percent case mortality.3 During World War I, the great pandemic of influenza swept through troops and civilian populations alike with large numbers of deaths caused in the main by secondary pneumonia. At the same time, measles was widespread, and this disease was also complicated by secondary pneumonia. 4

Knowledge regarding the recognition, epidemiology, and treatment, of pneumonia was more definite at the outbreak of World War II than it had been at the time of any previous mobilization. Primary atypical pneumonia had been recognized for several years. It was first recorded by the Army, in a separate diagnostic category on the individual medical records during 1941 and on the weekly statistical health report in March 1942. Whether this disease or group of diseases was present before or during World War I will

1 Mann. James: Medical Sketches of the Campaigns of 1812, 13, 14. Dedham: H. Mann & Co., 1816, p. 306.

2 The Medical and Surgical History of the War of the Rebellion. Medical History. Washington: Government Printing Office, 1888, vol. 1, pt. 111, p. 719.

3 Nichols, H. J.: The Lobar Pneumonia. Problem in the Army From the Viewpoint of the Recent Differentiation of Types of Pneumococci. Mil. Surgeon 41: 149-161, August 1917.

4 The Medical Department of the United States Army in the World War.Pathology of the Acute Respiratory Diseases, and of Gas Gangrene Following War Wounds. Washington: U. S.,Government Printing Office, 1929, vol. X11, p. 7.


probably never be known, but there is evidence to suggest that it was not a new disease. The following quotation is taken from the history of the Medical Department of the United States Army in World War I: 5

* * * the usual type of pneumonia occurring among young male adults in civil life is of course primary lobar pneumonia * * *. That such cases occurred among the troops is beyond question * * *. However, it was early recognized clinically that in the larger number of cases observed in the camps the pneumonia was of an atypical nature. The onset tended to be slower than that of the lobar pneumonia of civil life; the course more prolonged. Crisis was relatively rare; physical signs were slow of development and of patchy distribution and scattered in several lobes. These facts led careful observers to consider a large proportion of the cases as bronchopneumonia rather than as the usual lobar type. The results of post-mortem study of fatal cases lent confirmation to this distinction: The typical croupons consolidation of lobar pneumonia was relatively rare, patchy consolidation of a suppurative character more frequent. Even when the consolidation involved nearly or quite an entire lobe, careful study often showed evidence of the formation of such lobar consolidation by the confluence of smaller areas, lobular in origin.

The similarity of this description of pneumonia to the picture seen in primary atypical pneumonia during World War II will be obvious to all those familiar with the disease. It would seem quite likely that atypical pneumonia was indeed present in World War I and was classified largely as bronchopneumonia.

The first of the really efficient chemotherapeutic agents, the sulfonamides, had been in use for several years prior to World War II mobilization. These drugs, and later penicillin, proved to be so efficient in the treatment of pneumococcus infections among service personnel that in May 1944 it was recommended that antipneumococcus serum be dropped from the Medical Department supply table. This situation was in sharp contrast to that which existed in World War I when the only specific pneumonia therapy was an antipneumococcus type I serum. The total effect of the use of the sulfonamides and penicillin will probably remain immeasurable. One might expect that, in addition to lowering the mortality case rate of primary bacterial pneumonia and other bacterial infections, it must also have lowered the number of cases occurring as a complication of other diseases.


It was during the late 1930's that primary atypical pneumonia was first recognized as a disease distinct from the bacterial pneumonias. In 1938, Reimann6 published an account of a respiratory illness which he called atypical pneumonia and suggested its probable viral nature. Others had previously described a similar disease which in retrospect might well have been the same thing. Bowen 7 published a radiologic description of an epidemic among

5 The Medical Department of the United States Army in the World War. Communicable and Other Diseases. Washington: U. S. Government Printing Office, 1928, vol. IX, p. 61.

6 Reimann, H. A.: An Acute Infection of the Respiratory Tract With Atypical Pneumonia. A Disease Entity Probably Caused by a Filtrable Virus. J. A. M. A. 111: 2377-2384, December 1938.

7 Bowen, A.: Acute Influenza Pneumonitis. Am. J. Roentgenol. 34: 168-174, August 1935.


troops in Hawaii which he called acute influenza pneumonitis. Gallagher 8 described bronchial pneumonia and acute pneumonitis in adolescents. In 1916, Clough and Richter 9 published an account of a patient with a respiratory illness in whom autohemagglutinins were demonstrated. The failure of some cases of pneumonia to respond promptly to the sulfonamides undoubtedly influenced the thinking of investigators and helped to bring about the concept of this disease as a separate entity of virus etiology.

Since many of the, early reports dealt with epidemics in young adults in camps and boarding schools, it, came as no great surprise when cases began to be recognized in the Armed Forces.


Statistics for pneumonia in the Armed Forces in World War II are neither accurate nor complete. Reasons for this vary all the way from the intrinsic difficulties of gathering statistics in wartime to the ability of the medical personnel to make accurate differential diagnoses. The disease became officially reportable on the weekly statistical health report as Primary Atypical Pneumonia, Etiology Unknown, by direction of Circular Letter No. 19, Office of the Surgeon General, United States Army, 2 March 1942. It was probably some months following this before all medical personnel became sufficiently familiar with this new classification to make the differential diagnosis regularly. Officers concerned almost exclusively with the respiratory disease problem often had trouble ill making accurate distinction between the various types of pneumonia even while working under the best conditions. It would be reasonable to expect that officers responsible for all the medical problems of entire units working under less favorable and often hazardous conditions would have even greater difficulty. The fact that there was some difficulty ill making accurate differential diagnoses was emphasized in a report from ETOUSA (European Theater of Operations, United States Army) for April 1944 by Maj. Charles D. May, MC.10 He stated

It was possible to substantiate the diagnosis of atypical pneumonia in 72% of the patients so diagnosed. But in only 35% of the patients diagnosed by the hospital as primary [i. e., bacterial or lobar] pneumonia was the evidence considered adequate to justify the diagnosis.

One hypothesis to account for the apparent increase in the incidence of both primary and atypical pneumonia without a corresponding rise in the incidence of common respiratory disease is that there was an actual increase in atypical pneumonia with a confusion in diagnosis leading to many of the cases being reported as primary pneumonia.

Available data show that primary atypical pneumonia was present in all theaters and that its clinical characteristics with but very few exceptions were similar wherever it was reported.

Gallagher, J. R.: Bronchopneumonia in Adolescence. Yale J. Biol. & Med. 7: 23-40, October 1934.

9Clough, M. C., and Richter, I. M.: A Study of an Autoagglutinin Occurring in a Human Serum. Bull. Johns Hopkins Hosp. 29: 86-93, April 1918.

10Memorandum, Maj. C. D. May, MC, for Lt. Col. J. E. Gordon, MC, 2 May 1944, subject: Pneumonia Study, ETOUSA, 1944--Considerations for Further Study.


During the summer of 1941, an unexpectedly high incidence of pneumonia was observed in the soldiers at Camp Claiborne, La. The unusual character of this disease was noted by the commanding officer and by the chief of the medical service at the station hospital who recognized its similarity to the clinical syndrome of atypical pneumonia which had recently been described. The outbreak was reported to The Surgeon General and to the Surgeon, Fourth Corps Area, which resulted in the institution of preliminary surveys at Camp Claiborne by Drs. A. R. Dochez, Yale Kneeland, Colin M. MacLeod, and Kenneth Goodner. These workers felt that further investigation was warranted, and, accordingly in December 1941, Drs. John H. Dingle and W. Barry Wood took up residence at Camp Claiborne and remained until 1 May 1942 to direct a group study of the problem. The results of the work led to the establishment of a laboratory for the Commission on Acute Respiratory Diseases at Fort Bragg, N. C., on 19 October 1942.

It is probable that the figures of morbidity and mortality obtained at Camp Claiborne during 1941 and 1942 and at Fort Bragg during the remainder of the war are fairly accurate because of the special studies which were conducted at these two stations. The early work at Camp Claiborne showed that the disease represented a real problem to the Army with an average attack rate of 28 per 100,000 per week and a recorded peak incidence of 88 per 100,000 per week during an epidemic. It was shown further that the average hospitalization period for patients with atypical pneumonia was 32 days.11 This clearly presented a threat to the well-being of troops in training and potentially to those in combat.

Available figures suggest that, excluding the common respiratory diseases, atypical pneumonia represented the major respiratory disease problem for the Army as a whole. In the 4-year period from 1942-45, total Army admissions for atypical pneumonia were 160,940 with an annual admission rate of 6.32 per 1,000 (table 34). Comparable figures for all other pneumonias were 109,882 and 4.31 (table 35). This general relationship was true both in the United States and in overseas areas taken as a whole. It is of interest that in 1942, the year in which atypical pneumonia was first accepted as an official diagnosis for the statistical health reports, the incidence of reported atypical pneumonia in all areas except the Central and South Pacific was less than that for all other pneumonias. In 1943, the relationship of the two groups, for the total Army and for the United States, was reversed, in that atypical pneumonia admissions exceeded admissions for other pneumonias; for total overseas admissions, however, other pneumonia still exceeded atypical pneumonia, although not as markedly as in 1942. In 1944 and 1945, so far as the figures are available, the diagnosis of atypical pneumonia far exceeded the total for all other pneumonias. This was true in all areas except in 1944 in China-Burma-India, the Middle Fast, and North America, exclusive of the United States. Whether this represented in part a growing awareness among medical personnel of the

11 Dingle, J. H., Abernethy, T. J., Badger, G. F., Buddingh, G. J., Feller, A. E., Langmuir, A. D., Ruegsegger, J. M., and Wood, W. B.: Primary Atypical Pneumonia, Etiology Unknown.War Med. 3:223-248, March 1943.


TABLE 34.-Admissions for primary atypical pneumonia in the U. S. Army, by area and year, 1942-45

presence of atypical pneumonia or whether it represented a true, increase in the incidence of this disease cannot be determined.

The peak incidence of atypical pneumonia for the entire Army was reached in 1943, when the rate per 1,000 per year reached 7.45 (table 34). This was largely a reflection of the rate for troops in the United States where the rate reached 8.95 in that year. The peak incidence in Europe, however, occurred in 1942, when it reached 9.23 per 1,000 per annum. Thereafter it declined to a low of 4.80 in 1944 but rose slightly to 5.35 in 1945. The Middle East had its greatest incidence in 1942. All other areas had a peak rate in 1945, when the incidence in the Mediterranean (North African) theater reached 14.13 per 1,000 per annum, the highest rate recorded anywhere for an entire area.

Generally, there seems to have been a somewhat higher incidence of atypical pneumonia during colder months when all respiratory disease was more prevalent, but it is obvious that sharp outbreaks did occur during the warmer months as well. From the date on which the disease first became reportable, it was present in all areas almost constantly. Comparative figures for all other pneumonias indicate that atypical pneumonia was the major respiratory illness.


TABLE 35.-Admissions for pneumonia, other than primary atypical type, in the U. S. Army, by area and year, 1942-45

In spite of the widespread morbidity due to atypical pneumonia, the mortality was fortunately low in all areas. There were 170 deaths attributed to atypical pneumonia in the entire Army from 1942 through 1945. This is in contrast to 1,041 deaths caused by other pneumonias during the same period. It is of interest that the case fatality rates were higher overseas than in the United States for all types of the disease; however, the death rate (number of deaths due to pneumonia per 100,000 average, strength per year) was higher for both atypical and other pneumonias among those troops stationed in the United States than it was for those stationed overseas. The over-all case fatality rate (number of deaths per 100 admissions for pneumonia) for atypical pneumonia was 0.12 per 100 admissions for the 4 -year period 1942 through 1945 as compared with a rate of 0.88 per 100 admissions for all other pneumonias.

What conclusions can be drawn from the foregoing discussion? As stated previously, the figures are incomplete and are based on sample tabulations. It can be said, nevertheless, that, apart from the common respiratory diseases, atypical pneumonia was the major respiratory disease problem during World War II. Atypical pneumonia was present in all areas in significant


numbers and accounted for a large share of the total morbidity from respiratory disease.

Since atypical pneumonia was not recognized as an entity in 1918, no direct comparison of figures for the two World Wars for this disease can be made. Study of the figures for all pneumonia for the two periods is very interesting. The admission rate for all pneumonia per 1,000 average strength per year for the total Army from April 1917 to December 1919 was 18.98, and the approximate case fatality rate per 100 admissions was 24.46. The admission rate for all pneumonia in the entire Army from 1942 to 1945 was 10.63 per 1,000 average strength per year, and the average case mortality rate (deaths per 100 admissions) for the same period was approximately 0.42. This great reduction of morbidity and mortality figures could have been caused by many factors, ranging from the type of warfare carried on to the many aspects of personal hygiene. Undoubtedly, the sulfonamides and penicillin were a great influence and may well have been the most important. The absence of pandemic influenza played an undetermined but probably important role.


The results of the early investigations at Camp Claiborne, in 1941 and 1942, showed the need for continued study of the problem of atypical pneumonia. To this end, the Commission on Acute Respiratory Diseases was founded, and a laboratory was later equipped at Fort Bragg. This Commission functioned as an active investigating unit from 1 August 1942 throughout the duration of the war. The scope of its activities included not only primary atypical pneumonia but also influenza and other respiratory diseases. The complete work is fully reported in numerous articles under the authorship of the Commission on Acute Respiratory Diseases which were published from 1943 through 1946. Only the most important aspects of the work will be set forth here. The most significant work of the Commission concerned the study of atypical pneumonia in human volunteer subjects who were drawn from the ranks of conscientious objectors.12 These studies were carried out over a period of 3 years. The disease was successfully transmitted to humans by inoculation with bacteria-free filtrates of respiratory secretions which had been collected from patients with atypical pneumonia. This accomplishment alone lends strong support to the widely held theory that atypical pneumonia is a virus disease. These carefully controlled cases furnished a unique opportunity to study the clinical, roentgenographic, and laboratory characteristics of atypical pneumonia.

Clinical Aspects

Onset.-The Commission's studies of these cases confirmed and extended previous descriptions of the clinical picture of atypical pneumonia. The exact


12 Commission on Acute Respiratory Diseases: The Transmission of Primary Atypical Pneumonia to Human Volunteers. Bull. Johns Hopkins Hosp. 79: 97-167, August 1946.


time of onset proved to be almost as difficult to determine in these cases as in the naturally occurring disease. Often the earliest symptoms were mild and inconstant with no objective evidence of illness. This was quite in keeping with the previous observations of the disease. The incubation period varied from 7 to 14 days and was in general shorter for those who received untreated inoculation and longer for those who received filtered material. The reason for this variation is not clear but could represent a difference in the amount of infectious material present since presumably some of the agent was adsorbed by the filter in processing.

The character of onset varied considerably among the 16 patients studied. In five, the simultaneous occurrence of fever and constitutional and local symptoms marked a rather sudden onset of illness. In the remaining 11, the onset was gradual and marked by varying local and constitutional symptoms. The latter type of onset was in accord with that described by numerous authors in the naturally occurring disease. Early symptoms included dry or sore throat, nasal stuffiness, and headache. Feverishness and headache developed early, and chilliness was common and most prevalent on the second day following the onset. Malaise occurred in more than half the patients and was also an early symptom. Anorexia was present at some time in all patients. Local symptoms included nasal stuffiness, mild sore throat, and hoarseness. Cough was a conspicuous feature and was usually dry at first but later became paroxysmal and productive. All patients developed coughs between the first and fourth day. Sputum was mucoid at first, later becoming purulent. No patient developed grossly bloody or rusty sputum, but two of the most severely ill produced a slight blood streaking. Fourteen of the sixteen patients with pneumonia experienced chest discomfort, usually described as a sense of pressure or substernal soreness. Only one developed sharp pleuritic pain.

Physical findings.-Fine and coarse rales were present in 15 of the 16 patients. Only one had no rales at any time. Rales first appeared from the first to the ninth day after onset. They developed on the third and fourth day in the greatest number of patients. Slight dullness to percussion was present in 10 patients, changes in tactile and vocal fremitus in 5, and alterations of the breath sounds in 7. Only one patient developed pleural fluid. Two patients showed signs of central nervous system disturbance. One developed partial loss of bladder and rectal function with motor impairment of the lower limbs. In the other, there was transient areflexia of the lower limbs. Both eventually recovered completely.

Roentgenographic findings.-Roentgenographic findings conformed generally to those seen in the naturally occurring cases. The earliest findings were peribronchial infiltration, most commonly seen at the lung bases.The appearance was usually that of soft patchy densities of irregular size and shape. In some cases, the densities were small, discrete, and nodular, tending to become larger and confluent as the disease process continued. There was much variation in the development of these lesions. In four patients, it was limited to one or both lower lobes. In six patients, it spread toward the hilar region.


The distribution involved the lower lobes solely in 75 percent of the patients. In the others, the lower lobes were involved with concomitant lesions in other lobes. Average duration of roentgen-ray lesions was about 10 days with a few being present for only 1 to 3 days. It was not uncommon to find roentgenographic evidence of pneumonia before physical signs developed.

Fever.-All of the patients had fever. It began as early as the first day of illness in some and reached its peak incidence of onset between the second and fifth days. The maximum temperature observed was 104.8o F., and the average maximum was 102.8° F. By the ninth day following onset, most temperatures had returned to normal. The pulse and respiratory rates were not strikingly elevated except in a few patients who were severely ill with extreme pneumonic infiltration. The cases varied considerably in severity from very mild with minimal infiltration and fever for only 3 days to rather severe with extensive involvement of all lobes and fever for 15 days.

Complications.-Complications were observed in only four patients. One had maxillary sinusitis during the recovery period; one had a toxic psychosis associated with fever. Pleural effusion and encephalomyelitis each occurred in one patient.

Minor respiratory illness. It is of interest that many of the subjects who were inoculated but did not develop atypical pneumonia did, however, develop evidence of a minor respiratory illness. Whether these represented mild infections with atypical pneumonia or infection with other agents present in the inoculum could not be determined.

Laboratory studies.-Total and differential leukocyte counts were in general within normal limits. Slight elevations were observed in a few of the more severely ill patients. Sedimentation rates showed no constant variation and, although they rose above normal in some individuals, the average for the group showed no striking increase.

Bacterial studies indicated that none of the common organisms found in the respiratory tract seemed to play any role in the infection. There were no concentrations of any one organism that suggested bacterial influence. Special efforts were made to recover the streptococcus MG described by Mirick and others.13 The Commission's studies failed to relate these organisms causally to atypical pneumonia. Cold autohemagglutinins were found in 13 of the 16 cases of pneumonia in significant titers. The significance of this test in its relationship to the etiology of atypical pneumonia is not known. It is apparently of some diagnostic value in those cases in which it is present.

Etiologic Studies

From the beginning of the first work of the Commission group at Camp Claiborne until the end of the war, research was in progress to uncover the

13 Mirick, Cx. S., Thomas, L., Curnen, E. C., and Horsfall, F. L., Jr.: Studies on a Non-Hemolytic Streptococcus Isolated From the Respiratory Tract of Human Beings. J. Exper. Med. 80: 391-440, November 1944.


agent of atypical pneumonia. Commission studies involved the use of chick embryos, mice, rats, cotton rats, hamsters, guinea pigs, cats, and monkeys. A few members of the Commission were sent to Puerto Rico to work with mongooses. In no instance was it possible to reproduce the disease until human volunteers were used. The electronic microscope and ultracentrifuge at Duke University, Durham, N. C., were used in an attempt to find virus particles but without success. Acute and convalescent sera from patients with atypical pneumonia failed to show antibodies against any of the known viruses or rickettsiae. Much of this work was done in the laboratory of Dr. Thomas Francis, Jr., at the University of Michigan at Ann Arbor. Extensive bacterial studies over a period of 5 years failed to reveal any bacterial agent responsible for the disease.

Epidemiologic Studies

The early work of the investigators at Camp Claiborne showed that a moderately severe epidemic of atypical pneumonia occurred during the summer of 1941. The peak epidemic rate was about three times as high as the average endemic rate prevalent at that camp. Studies failed to reveal any possibility of contamination of water, milk, or food supplies as a transmitting agent. At the same time, the disease was too widely spread to be easily charged to direct person-to-person contact of overt cases except for a small number which occurred among medical personnel. It was noted that many cases were mild and indistinguishable from common respiratory disease infections except by roentgenogram. It was concluded that these cases probably formed an inapparent reservoir which spread the disease from person to person. This was substantiated by subsequent work. It will be recalled that many of the inoculated human volunteers described earlier developed minor respiratory illness without evidence of pneumonia. In addition, previously unknown cases of atypical pneumonia, without symptoms, were found during roentgenographic surveys of entire units.

Later studies at Fort Bragg showed that new recruits experienced high rates of respiratory illness during the first 4 weeks after their arrival at camp.14 The peak incidence of atypical pneumonia was likewise greatest during this period. The attack rate at Fort Bragg for respiratory diseases, in general, and for atypical pneumonia, in particular, followed a more or less constant ratio of 10:1. This led to some speculation as to the possibility of a common etiology. However, this possibility was not supported by results from subsequent studies in human volunteers who failed to develop pneumonia following inoculation with material recovered from patients with common respiratory disease. Reports from England specifically note the increased incidence of atypical pneumonia in the presence of normal figures for common respiratory

14 Commission on Acute Respiratory Diseases: Acute Respiratory Disease Among New Recruits. Am. J. Pub. Health 36: 439-450, May 1946.


diseases.15 A report from India takes note of the sharp increase in all respiratory diseases, including atypical pneumonia, during the hottest months of the year.16 A similar experience was reported from Camp Claiborne during the summer of 1941, at which time the incidence of common respiratory diseases was at its usual seasonal level.17

The evidence indicates that atypical pneumonia is an infectious disease of virus etiology. It is generally, but not invariably, more prevalent at the time of greatest incidence of other respiratory diseases. It can be spread from person to person by infected respiratory secretions, and this is probably its natural mode of spread. The frequent finding of inapparent cases suggests the probability that such cases form a reservoir of infection from which clinical cases may arise.


Atypical pneumonia, which first became recognized as a clinical entity in the late 1930's, made its appearance in the Army soon after large-scale mobilization began. It became clear after preliminary studies at Camp Claiborne that the disease was of major importance to the Army, and it eventually became an outstanding respiratory problem. It appeared in significant numbers in all theaters.

Special research in this field was carried on by the Army throughout the war. These studies showed that atypical pneumonia is an infectious disease which can be transmitted to human volunteers by the inhalation of infected bacteria-free filtrates. The specific etiologic agent was not determined, and extensive serologic studies failed to suggest a relationship to any known virus or rickettsia.

Morbidity and mortality rates for all pneumonia taken as a group were much lower in World War II than in World War I. There are probably many reasons for these differences, but unquestionably the use of chemothera peutic agents and the absence of pandemic influenza were important factors. Extensive clinical, laboratory, and epidemiologic investigations were pursued which confirmed and extended the findings of previous workers. Available evidence suggests that the natural mode of spread is by person-to-person contact. Inapparent cases are known to exist which probably furnish a reservoir of infection. In general, atypical pneumonia was most prevalent during the colder months when there was an increase in all respiratory diseases. There were, however, some notable exceptions to this rule, and several epidemics were reported during the warm months.

15 Gordon, J. E.: A History of Preventive Medicine in the European Theater of Operations, United States Army. 1941-194.5. Pt. 111, Epidemiology, sec. 3, Acute Respiratory Infections, No. 3, The Pneumonias, pp. 1-14. [Official record.]

16 Blumgart, H. L., and Pike, G. M.: History of Internal Medicine in India-Burma Theater. Pt. II, Internal Medicine in the India-Burma Theater. Respiratory Diseases, pp. 116-123.[In preparation.]

17 See footnote 11, p. 214.


Section II. Bacterial Pneumonia

Richard G. Hodges, M. D.

In great contrast to World War I and probably to all previous wartime mobilizations, bacterial pneumonia during World War II did not present a major problem. The reasons for this were several. Although influenza, both A and B, involved the military population, nothing resembling the pandemic of 1918-19 occurred with its wake of pneumococcal, streptococcal, and influenzal pneumonia. Furthermore, there were no epidemics of measles to introduce pneumonia as a complication. This may be due to the automobile and the motion picture, both of which brought about an earlier and more general exposure of the rural population to measles. Finally, the widespread use of the sulfonamide drugs in the early treatment of febrile respiratory infections probably resulted in the prevention or abortion of many cases and certainly reduced the mortality to an extremely low level.

No meaningful figures can be given as to the incidence of bacterial pneumonia during the war years. The general decline of accurate bacteriologic diagnosis made it impossible to distinguish between bacterial and primary atypical pneumonia. During the early years of the war, it is probable that many cases of nonbacterial pneumonia were diagnosed as bacterial; when the diagnosis of primary atypical pneumonia had become popularized, it is probable that the error was in the opposite direction. The best available data on the comparative incidence of the two conditions is given in the preceding section dealing with primary atypical pneumonia.

Historically, the most important aspect of bacterial pneumonia was the information gathered about the spread and particularly the prevention of pneumococcal pneumonia.18 The studies were carried out at a single installation, the Army Air Force Technical School at Sioux Falls, S. Dak. This was the only large military establishment that suffered severely from pneumococcal pneumonia. The investigation represented a joint project of the Commission on Pneumonia, Army Epidemiological Board, Office of the Surgeon General, and the Army Air Force Rheumatic Fever Control Program, Office of the Air Surgeon.


Studies on the epidemiology of pneumococcal pneumonia covered a 3-year period at the Army Air Force Technical School at Sioux Falls, S. Dak. During this period of observation, more than 1,600 cases of pneumonia occurred. On several occasions, the attack rate exceeded 150 cases per annum per 1,000 ,average strength. The experience of the first year was studied in retrospect from hospital records; the data for the second year were gathered directly by an epidemiologist; during the third year, extensive bacteriologic and statistical

18 Hodges, R. G., and MacLeod, C. M.: Epidemic Pneumococcal Pneumonia, pts. I-V. Am. J. Hyg. 44: 183-243, September 1946.


facilities were available. From this large experience, certain factors which contributed to the high pneumonia rates could be detected.

Influence of Population Characteristics and Environment

The two important population characteristics which influenced the pneumonia attack rates appeared to be length of service and duration of stay on the post. During the first months after the establishment of the post, it was impossible to distinguish between these two factors since men were brought to the school direct from basic training. Later, when the population was a mixture of new recruits and of seasoned troops, it was possible so to distinguish. There was a small but definite excess incidence in men newly introduced into military service. Regardless of length of service, the majority of pneumonia cases occurred in men spending their first 8 weeks at the post. Thus, it appeared that the troops became twice seasoned, first to military life in general and secondly to the particular environment of this particular technical school.Of these factors, the latter was by far the more important.

No way was devised of subjecting the environmental factors to controlled study. However, there was much to indicate that environment was important. In many ways, the operations of the technical school were admirably devised to promote the spread of respiratory disease. The barracks were of the theater of operations type, ill-suited to the climate of South Dakota. No efforts were made to promote ventilation or to control dust. The school buildings were no better than the barracks in these respects. The exact role of dust could not be determined, but pathogenic pneumococci were cultured from 29 percent of 147 dust samples. Moreover, there was a thorough mixing of the school population. Each class was composed of men from several different squadrons and from many different barracks. These arrangements facilitated cross-infection. It was shown that common respiratory diseases, pneumonia, streptococcal sore throat, epidemic influenza, and even specific serologic types of pneumococci spread rapidly and evenly throughout the school population.

No evidence was obtained that implicated previous geographic environment, age, chilling, or fatigue as being important factors in the production of pneumococcal pneumonia in the population as a whole.

Influence of Pneumococcal Carrier State

During the third year of the study, extensive carrier surveys were made in an attempt to relate the behavior of the causative organism to the incidence of pneumococcal pneumonia. Serologic typing was carried out with great care. Three methods of survey were used. A single squadron was sampled three times a week throughout the year. Cultures were made from all men admitted to hospital for respiratory diseases. Cultures were also taken from the occupants of a single barracks three times a week for 9 consecutive weeks. The carrier rates for the single squadron and for the hospital admissions, which


came from all squadrons, were identical and were combined to represent the population as a whole.

The total carrier rate was affected by season. Starting with a rate of approximately 40 percent in September, there was a sharp rise to 60 percent in November, and this high level was maintained throughout the winter season. Within the total carrier rate, the individual serologic types behaved rattier independently, each attaining its own peak of incidence. Surprisingly, neither the total nor the specific carrier rates appeared to be affected by the incidence of common respiratory diseases.

Men newly arrived at the post were relatively free of pneumococci and were almost entirely free of the types which were known to produce pneumonia frequently. However, the new arrivals rapidly acquired pneumococci. After 4 weeks in the environment, their carrier rate was equal to that of the total population. Moreover, the new men became rapidly infected with the pathogenic types. This was demonstrated in the single barracks study. Only 7 percent of new men carried either type IV or XII, the types that were currently the leading cause of pneumonia. After 7 weeks in the barracks, 40 percent of the new men had become carriers of one or both of these types.

The single barracks study also demonstrated that the carrier state was dynamic in character. Men rapidly acquired and lost several different serologic types of pneumococci during the period of observation.

The incidence of pneumonia did not correspond to the total carrier rate nor did it correlate closely with the carrier rates for the highly infective types of pneumococci. At the times when the pneumonia rate was high, it was usual for the carrier rate to be high also, but there were several periods when the carrier rate for infective types was high and the pneumonia rate low. This indicated that some other factor besides the presence of the infective agent was necessary to produce high pneumonia rates.

Influence of Nonbacterial Respiratory Disease

During the entire 3 years of observation, there was a close relationship between the incidence of nonbacterial respiratory disease and that of pneumococcal pneumonia. On the average, 1 case of pneumonia was admitted to the hospital for every 10 patients admitted with nonbacterial respiratory disease. The seasonal occurrences of the two conditions paralleled each other closely, there being only two periods when the 1:10 ratio was not closely approximated. The first of these was in the 4 months after the. post was opened. At that time, the incidence of nonbacterial respiratory disease was high but that for pneumococcal pneumonia low. Presumably, this was before the population had become thoroughly seeded with pneumococci. The other period was during the second winter of the study when, for a time, the ratio was more nearly 1:5 than 1:10. This corresponded to a period when the incidence of type II pneumonia was very high, approximately 60 percent of the cases of pneumonia. Later, it was possible to show that there were considerable differ-


ences in the "infectivity" of the various types of pneumococci. Infectivity was expressed as the number of men admitted with pneumonia due to type X divided by the number of respiratory admissions carrying type X. The value for type II was 0.52; whereas for type IV, it was 0.17; for type XII, 0.26; for type III, 0.10; and for type VIII, 0.09. Thus, when type II was prevalent, more cases of pneumonia per case of nonbacterial respiratory disease would be expected to occur.

Two epidemics, one of influenza A and one of influenza B, were identified. Both resulted in a sharp rise in the pneumococcal pneumonia rates, but during both the usual 1:10 ratio held good.

It was concluded that the incidence of pneumonia was governed by the prevalence of pneumococci, by the infectivity of the serologic types which were present, and by the incidence of nonbacterial respiratory, disease.


The circumstances in the Army Air Force Technical School, Sioux Falls, S. Dak., appeared to be ideal to test the efficacy of immunization against pneumococcal pneumonia. Of the recorded attempts to conduct such im munization,19 each had been handicapped by one or more of the following difficulties: (1) Differences in the composition of the immunized and control groups, (2) uncertainty as to whether the specific pneumococcal types included in the immunizing preparation were the same as those currently causing pneumonia, (3) failure to determine whether the observed decline in cases in the immunized group was due to a decrease in cases caused by the pneumococcal types in the vaccine, and (4) inadequate control of the antigenicity of the preparation used.

In the investigation conducted at Sioux Falls, the situation was such that each of these obstacles could be eliminated.20 In the first place in the 2 preceding years, the population had been subjected to a thorough epidemiologic study of respiratory disease. It was known that the population was very uniform in respect to such epidemiologic characteristics as age, length of service, and duration of stay on the post, and that the environment of the troops was admirably devised to facilitate a rapid and uniform dissemination of respiratory disease throughout the entire population. This had been proven for streptococcal sore throat, pneumonia, influenza A, and the common respiratory diseases. Second, for the 2 preceding years, the pneumococcal pneumonia rates had been extremely high, and, equally important, the distribution of pneumococcus types causing pneumonia appeared to be uniform. For each of the 2 preceding years, the approximate distribution was: Type II, 34 percent; types I, V, and VII, 9 percent each; types XII and IV, 7 and 5 percent, respectively. Third, statistical machinery was established whereby the

19 Heffron, R.: Pneumonia With Special Reference to Pneumococcus Lobar Pneumonia.New York: The Commonwealth Fund, 1939, pp. 446-483.

20MacLeod, C. M., IIodges, It. G., Heidelberger, AT., and Bernhard, W. G.: Prevention of Pneumococcal Pneumonia by Immunization With Specific Capsular Polysaccharides. J. Exper. Med. 82: 445-465, December 1945.


population could be carefully followed, and a laboratory capable of doing extensive pneumococcal typing was organized. Finally, through the efforts of Dr. Michael Heidelberger, highly purified capsular polysaccharides of known antigenicity were available.

Accordingly, in September 1944, the entire personnel of the technical school was subjected to this test. A barracks was roped off longitudinally, and as the men were marched through they were given a random choice of which side of the rope they preferred. Those passing down one side received 1.0 ml. of saline containing 0.03 to 0.06 mg. of types I, II, V, and VII type-specific polysaccharide; those choosing the opposite side were injected with 1.0 ml. of saline containing 0.5 percent phenol. Subsequently, when a new troop shipment arrived at the post, alternate men received the polysaccharide solution or the saline, respectively. In all, 8,586 men were injected with the polysaccharide solution and 8,449 with saline. In terms of man-days exposure, the experience was 745,997 days for the immunized and 772,898 days for the nonimmunized. Many samplings of the population were taken to test for random distribution, and in each instance immunized and nommmunized men were found to be present in equal numbers.

The effect of the immunization on the development of clinical pneumonia is shown in table 36. Pneumonia due to the types against which immunization was not practiced was equally divided between the treated and control groups. Pneumonia due to types I, II, V, and VII occurred 4 times in the treated group and 26 times in the control group. Moreover, each of the four cases in the immunized group occurred within 2 weeks after the individual was injected, whereas the cases in the control group were distributed at random over the period of observation (table 37). The number of type II cases was large enough to afford sound evidence of the protective value of the type-specific polysaccharide. There was no reason to believe that the specific protection against the other types was not equally good.

TABLE 36.-Incidence of pneumonia in immunized and nonimmunized groups


TABLE 37.-Interval between injection and the development of the several types of pneumonia in immunized and nonimmunized subjects

The incidence for pneumonia due to types IV and XII, against which immunization was not practiced, was approximately that expected from the experience of the preceding 2 years.For types I, II, V, and VII, even in the nonimmunized group, the incidence of pneumonia was far below that expected. This strongly suggested that immunization of one-half of the population conferred a real protection on the nonimmunized subjects. There were plausible explanations for this. Reduction of the number of cases would comparably reduce the number of case contacts and might thus inhibit the spread of the organism. A more potent reason stemmed from the results of the carrier study which was being carried out simultaneously (table 38).The carrier rates for the types against which immunization was not practiced were almost equal in the immunized and nonimmunized groups. However, the rates for types I, II, V, and VII were significantly lower in the immunized group as compared with the control. Thus, the immunization of a given individual appeared to render him relatively resistant to becoming a carrier of the specific types contained in the vaccine. Consequently, in a population consisting of intimately mixed immunes and nonimmunes, every second transfer of a pneumococcus would result in the organisms falling on relatively infertile ground. The consistent behavior of types IV and XII in each of the 3 years of observation provided a means of calculating the amount of reduction in the incidence of type I, II, V, and VII pneumonia among the nonimmunes which was achieved by immunizing one-half the population. Only 17.6 percent of the expected cases were observed.


TABLE 38.-Distribution of individual types of pneumococci between immunized and nonimmunized groups (excluding cases of pneumococcal pneumonia)


Bacterial pneumonia was not a major problem in World War II. However, studies conducted under the auspices of the Armed Forces did demonstrate the steps that should be taken to control pneumococcal pneumonia during future periods of mobilization. A high carrier rate for pneumococci and a high incidence of nonbacterial respiratory disease were shown to stimulate the occurrence of pneumonia. Although neither of these factors could be attacked directly, the implication was clear that more care in choosing the geographic location of military schools and more attention to the sanitation of installations in unfavorable climates would serve to lessen the spread of both bacterial and of nonbacterial respiratory disease.

Finally, it was possible to prove that a high degree of protection against pneumococcal pneumonia can be given by immunization with type-specific capsular polysaccharides.