From May 25 through October 31, 2007, USAFSAM received 959 respiratory specimens from the 12 secondary training sites that initially participated in enhanced surveillance and from 9 additional sites (Table 1). Adenovirus accounted for 413 (89.8%) of the 460 specimens with known etiologic agents; the other viruses identified included parainfluenza (31 [6.7%]), influenza type A (5 [1.1%]), respiratory syncytial virus (2 [0.4%]), and enterovirus (1 [0.2%]). Among the specimens that were culture positive for adenovirus, 358 (86.7%) were tested for Ad14, of which 341 (95.3%) were positive. Ad14 was identified at 8 secondary sites located in California, Florida, Mississippi, Texas, and South Korea; collection dates of the first Ad14-positive specimen at each site ranged from May 30 through October 30. Most patients (331 [97.1%]) with confirmed Ad14 infection were advanced training students, while 9 (2.6%) infections occurred in active duty members outside the training population, and 1 (0.03%) occurred in a dependent child.

Patient survey data were available for 538 of the 959 (56.1%) patients from whom specimens were collected; of these, 220 (40.9%) were Ad14 positive. The following results are only for those 220 patients with confirmed Ad14 infection and available patient survey data (Table 2). Patient ages spanned 17–29 years, though most (183 [84.7%]) patients were 18–22 years of age; the median age was 19 years. In addition, most (197 [89.5%]) patients were male. Regarding patient symptoms, the median temperature recorded was 101.0°F. The most common signs and symptoms reported by patients were sore throat (90.9%), chills (83.2%), fatigue (78.6%), cough (78.2%), headache (75.9%), body aches (70.0%), and nasal congestion (61.4%). One patient was hospitalized with pneumonia and recovered fully without complications. A total of 191 (86.8%) patients were placed on quarters. Of the 125 patients for whom length of quarters information was available, most (108 [86.4%]) were placed on quarters for 24 hours. In addition, 147 (66.8%) patients had recently traveled; of these, most (143 [97.3%]) had recently traveled from Lackland AFB.

At the 3 secondary sites receiving the most BMT graduates, Sheppard AFB, Goodfellow AFB, and Keesler AFB, AdARD incidence rates among active duty personnel were tracked and compared with concurrent rates calculated at Lackland AFB (Figure 3). AdARD rates at Lackland AFB ranged from 0.1–2.0 cases per 100 personnel, with 2 peaks in June and September 2007. AdARD activity at Sheppard AFB waxed and waned throughout the surveillance period, ranging from 0.2–0.8 ARD cases per 100 personnel. The largest peak of activity occurred on September 22, 2007, 2 weeks following the onset of Lackland AFB’s second wave of activity. However, this AdARD activity was short lived, decreasing over a course of 4 weeks to 0.2 cases per 100 personnel. Activity at Goodfellow AFB and Keesler AFB was highest following the initial peak at Lackland AFB, and then tapered off. All 3 sites placed ill students on quarters, which resulted in the short-term removal of >600 students from training activities. Only 1 person required hospitalization for adenovirus-associated pneumonia, at Sheppard AFB, during this time (0.01/100 trainees for this 23-week time period). As of October 31, 2007, prevention and control efforts were terminated at Goodfellow AFB and Keesler AFB but continued at Sheppard AFB.

Ad14 spread readily to secondary training sites because of the rapid mobility of BMTs following their graduation from basic training. For the most part, the onset of Ad14-related illness occurred first at sites that received the most BMT graduates. Although Lackland AFB made a concerted effort to identify and segregate outgoing febrile BMTs, more than one quarter of the trainees were likely shedding virus or recovering from illness. In addition, many were possibly preclinical and incubating adenovirus as they departed Lackland AFB, with illness developing shortly after arrival at secondary sites.

Although Ad14 was exported continuously to the secondary sites, neither the ARD rates nor the severity of illness at those sites reached the levels seen at Lackland AFB. Control efforts by Lackland AFB that placed febrile BMTs on medical hold and prevented them from leaving the base seemed to affect severity of illness at Lackland AFB and likely reduced the number of ill persons arriving at secondary sites. The lower rates of illness at secondary sites may also have been due to a decreased number of susceptible persons in the secondary training population, berthing differences that resulted in less contact between trainees at secondary sites, and decreased stress levels among trainees. Additionally, decreased illness severity at the secondary training sites may have resulted from the early identification of patients with suspected cases and their placement on quarters, allowing for rest and recovery. Illness trends at Sheppard AFB tracked Lackland AFB ARD activity most closely, possibly because Sheppard AFB received the largest proportion of BMT graduates compared with other secondary training sites and because healthcare personnel at that base more aggressively identified cases and submitted specimens. Keesler AFB and Goodfellow AFB experienced an earlier decline of infection rates due partially to a lower influx of BMTs from Lackland and partially due to lessening participation in enhanced surveillance efforts after the first 3–4 weeks of surveillance. Adenovirus rates mirrored the overall ARD rates, which suggests that adenovirus accounted for most of the ARD cases during the entire period. As of October 31, 2007, ARD rates had subsided somewhat but not sufficiently to cease surveillance and control efforts.

Surveillance findings indicated that spread of Ad14 to active duty members and dependents outside the training population was minimal, although surveillance efforts were not as robust in these populations. Because secondary training students are segregated from the base population at installations, limited contact takes place between trainees and other military members and their dependents. Still, the lack of spread of this readily transmissible pathogen to persons outside the training population was remarkable.

Eliminating the spread from Lackland AFB to secondary training sites was difficult, because any control measures deployed had to operate within the constraints of the recruit training system. Particularly in wartime, military operational requirements do not permit slowing or canceling the training of new military recruits. Even within this limited setting, segregating ill patients and checking students for fever both before leaving Lackland AFB and upon arrival at their next duty station seemed to affect transmission, because ARD rates remained relatively low at secondary training sites and peak ARD activity did not persist.

A new vaccine for Ad4 and Ad7 is currently under development; phase III has been completed (14). The degree of cross-protection this vaccine will offer against Ad14 is unknown, although evidence suggests some protection can be expected. Findings from the Lackland AFB outbreak indicated that Ad7 serum neutralizing antibody was protective against Ad14 illness, mitigating the severity of symptoms (12). Previous studies have indicated that Ad4 and Ad7 vaccination was also protective against Ad3 (8,15) and Ad14 (8,15–18). Thus, implementing the Ad4 and Ad7 vaccine may affect AdARD rates in general in future military trainees.

Following the loss of the Ad4 and Ad7 vaccines, the spread of Ad4 from an Army basic training site to a secondary training installation was associated with a large Ad4 outbreak at a secondary site (19). In this report we describe the spread of Ad14 to Air Force secondary training sites by recently graduated basic trainees who moved quickly from Lackland AFB to their next assignment. Military planners must focus on how best to control the spread of infectious respiratory diseases in highly mobile military populations that travel between geographically dispersed locations. Additionally, these planners must consider that rigid public health interventions may be unacceptable because of interference with critical operations. In this instance, we found that interventions could not interfere with the flow of programmed training operations for a military at war. Additionally, detailed studies aimed at better describing the transmission of adenoviruses may result in better focused and more effective control measures. Our results show that public health leaders in both the military and civilian communities should be concerned about the geographic spread of respiratory disease agents by highly mobile populations.