Grabenstein JD, Winkenwerder, Jr W. US Military Smallpox Vaccination Program Experience. JAMA. 2003;289(24):3278-3282. doi:10.1001/jama.289.24.3278
Author Affiliations: Military Vaccine Agency, US Army Medical Command, Falls Church, Va (Dr Grabenstein); Assistant Secretary of Defense (Health Affairs), Pentagon, Washington, DC (Dr Winkenwerder).
Context The United States recently implemented smallpox vaccination of selected
military personnel in a national program of preparedness against use of smallpox
as a biological weapon. The resumption of smallpox vaccinations raises important
questions regarding implementation and safety.
Objective To describe the US military smallpox vaccination program.
Design Descriptive study of the vaccination program from its inception on December
13, 2002, through May 28, 2003.
Setting US Department of Defense (DoD) fixed and field medical treatment facilities
on multiple continents and ships at sea.
Subjects US service members and DoD civilian workers eligible for smallpox vaccination.
Main Outcome Measures Numbers of vaccinations and rates of vaccination exemptions, symptoms,
and adverse events. Data were collected via reports to headquarters and rigorous
surveillance for sentinel events.
Results In 5.5 months, the DoD administered 450 293 smallpox vaccinations
(70.5% primary vaccinees and 29.5% revaccinees). In 2 settings, 0.5% and 3.0%
of vaccine recipients needed short-term sick leave. Most adverse events occurred
at rates below historical rates. One case of encephalitis and 37 cases of
acute myopericarditis developed after vaccination; all cases recovered. Among
19 461 worker-months of clinical contact, there were no cases of transmission
of vaccinia from worker to patient, no cases of eczema vaccinatum or progressive
vaccinia, and no attributed deaths.
Conclusions Mass smallpox vaccinations can be conducted safely with very low rates
of serious adverse events. Program implementation emphasized human factors:
careful staff training, contraindication screening, recipient education, and
attention to bandaging. Our experience suggests that broad smallpox vaccination
programs may be implemented with fewer serious adverse events than previously
In December 2002, the United States implemented a program of smallpox
vaccinations for approximately 500 000 military personnel. The directive
came as part of a national program of preparedness against biological attack.1 Preattack vaccination was determined to be the best
way to personally protect troops so they can continue their missions.2,3 The program was therefore mandatory
for designated service members and employees except those with contraindications.
This article describes the first 450 293 vaccinations.
To develop vaccination policy, the US Department of Defense (DoD) drew
from its own physicians, scientists, and administrators3 as
well as colleagues in government and academia. The military vaccination program
included vaccination for smallpox epidemic response teams (2000-5000 people)
to assist with epidemic control and contact tracing in an outbreak, medical
teams for hospitals and clinics (10 000-25 000 people) to care for
smallpox cases, and operational forces (up to 500 000 people) to preserve
The US Army conducted a medical training conference in October 2002.
This conference was videotaped and posted at http://www.smallpox.army.mil. Medical directors completed 8 hours of training; physicians and supervising
nurses completed 6 hours; vaccinating nurses and medics completed 3 hours.
Leaders educated vaccine candidates and reinforced messages after vaccination.
We provided brochures and curricula as well as customized information services
via telephone (877-GET-VACC) or e-mail (email@example.com).
To identify people with contraindications to vaccination,4- 8 clinicians
used a standardized recording form to screen for skin conditions, immunosuppression,
pregnancy or breastfeeding, and allergies (http://www.smallpox.army.mil/resource/forms.asp). Completed forms were reviewed by medics, nurses, nurse practitioners,
physician assistants, and physicians in a triaged manner before vaccination.
The DoD used its periodic testing program for human immunodeficiency virus
(HIV) infection to identify that contraindication. Women were advised that
pregnancy is a contraindication, asked about pregnancy status, and provided
pregnancy tests, if requested.
Each hospital and clinic established bandage evaluation stations.6,7 For non–health care workers,
infection control practices focused on bandages (eg, simple Band-Aid type
bandages with 20 × 25-mm pads), sleeves, and hand washing. To calculate
time at risk, each vaccinee was assumed capable of spreading vaccinia for
21 days after vaccination.
The licensed full-strength smallpox vaccine (Dryvax, Wyeth Laboratories,
Marietta, Pa) containing the New York City Board of Health strain of vaccinia
was used in this program.4,8 First-time
vaccination entailed 3 punctures with a bifurcated needle. Previous vaccinees
received 15 punctures. Those who did not respond with a major reaction as
defined by the World Health Organization (WHO)5,8 were
Smallpox vaccinations began at 4 pilot sites: Walter Reed Army Medical
Center, Washington, DC; Aberdeen Proving Ground, Md; Wilford Hall Air Force
Medical Center, Lackland Air Force Base, San Antonio, Tex; and the National
Naval Medical Center, Bethesda, Md. For quality control, clinics tracked the
vaccination response rates of the first 25 people for each vaccinator.
Postvaccination surveillance assesses the screening process, health
events shortly after vaccination, and events months or years after vaccination.
Vaccine Healthcare Centers evaluate complex cases. A registry was established
for women vaccinated before recognition of pregnancy, a joint project with
the Centers for Disease Control and Prevention (CDC).9 An
independent safety monitoring panel, coordinated by the Armed Forces Epidemiological
Board and the Advisory Committee on Immunization Practices, reviews weekly
updates.10 We shared experience weekly with
CDC and state health departments.
Adverse events reported here were collected via symptom surveys from
each vaccinated member of the first 7 operational units vaccinated (n = 526),
plus reports to headquarters and rigorous surveillance for sentinel events.
Frequencies are reported for the units described. The number of adverse events
identified was divided by the size of the vaccinated cohort. Exact confidence
intervals were calculated for Poisson distribution of rare events.
Definitions of adverse events were derived from expert consultants.
Generalized vaccinia involved vesicles or pustules in several locations distant
from the vaccination site that evolved like vaccinial vesicles and pustules.
Encephalitis involved altered consciousness for at least 24 hours with more
than 5 white blood cells/µL in cerebrospinal fluid. Myopericarditis
involved chest pain and elevated cardiac enzymes (eg, troponin, creatine kinase–MB),
as reported by Halsell et al11 elsewhere in
this issue of THE JOURNAL.
From December 13, 2002, through May 28, 2003, a total of 450 293
personnel were vaccinated against smallpox with the New York City Board of
Health strain (Table 1). Vaccinations
occurred in fixed and field medical facilities in North America, Europe, Asia,
Africa, and Oceania and on ships at sea. At peak, more than 50 000 vaccinations
were administered per week.
The median age of vaccinees was 26 years and 87% were men; 70.5% were
receiving their primary (first) vaccination. Contraindications due to either
personal conditions or conditions among household contacts varied by military
unit, temporarily deferring vaccination in 11% to 34% of eligible personnel.
Skin conditions predominated, followed by pregnancy and immune conditions.
Exemptions ranged from 4.9% to 7.8% among deployed units, where household
contacts were not present.
Military immunization clinics reported high rates of successful vaccination
(major reactions, response rates).5,8 Some
clinicians sought advice to confirm that small dermal responses fit the WHO
definition. At the 4 pilot clinics, 971 of 1017 primary vaccinees (95.5%)
were successfully vaccinated with a single vaccination. Among 975 people who
had been vaccinated decades earlier, 934 (95.8%) were successfully vaccinated
with a single vaccination. Around the world, 2 vaccine recipients were treated
successfully with epinephrine for acute allergic events.
Vaccinated health care workers remained on the job (including surgery
and labor and delivery), using bandages, sleeves, hand washing, and evaluation
stations. Vaccinated workers were not prohibited from caring for immunocompromised
patients. No transmission of vaccinia from a health care worker to a patient
was identified during an estimated 19 461 worker-months of interaction
with patients. Conversely, no vaccinee later treated as a patient transmitted
vaccinia to a health care worker.
Expected temporary symptoms occurred after vaccination, notably itching,
muscle ache, "feeling lousy," headache, swollen lymph nodes, irritation from
bandages, subjectively reported fever, and local rash (Table 2). Approximately 1% of vaccine recipients developed cutaneous
eruptions beyond the vaccination site.
Among carefully observed workers at Walter Reed Army Medical Center,
the frequency of vaccinated workers taking sick leave was 3.0% (16/530), typically
8 to 12 days after vaccination. The most common duration was 1 day (mean,
1.5 days; maximum, 4 days). Eleven of 198 primary vaccinees (5.5%) took sick
leave, as did 5 of 332 revaccinees (1.5%). Among 16 deployed US Air Force
clinics serving 20 272 vaccinees, 0.5% took 1 or 2 days of sick leave
after vaccination. Other sites reported similar results.
Skin eruptions or rashes caused concern among clinicians. These rashes
were generally mild and treated symptomatically. In 16 cases with maculopapular
rashes, vaccine recipients were hospitalized, in part because of uncertainty
about contagiousness or clinical course. This was especially true early in
the program. In each case, patients returned to duty promptly. Subsequent
patients of comparable acuity were typically not hospitalized. Six service
members developed cellulitis as diagnosed by physicians, were treated with
intravenous antibiotics, and returned to duty.
Noteworthy adverse reactions are summarized in Table 3. Thirty-six vaccine recipients may qualify as mild cases
of generalized vaccinia (based on reports to the Vaccine Adverse Event Reporting
System), all among primary vaccinees (35 men and 1 woman). Each was treated
symptomatically (eg, with antihistamines, analgesics, antipyretics) and returned
to duty. Nine were hospitalized for observation, notably by clinicians seeing
their first such patient. Erythema multiforme major confirmed by skin biopsy
developed in a 20-year-old service member 19 days after receiving smallpox
and several other vaccines. This case was deemed to possibly be associated
with smallpox vaccination, given multiple possible causes.
Thirty-eight reports of inadvertent implantation ("autoinoculation")
of the skin were submitted, with varying certainty of association to vaccinia.
Affected sites included chest, cheek, arm, finger, wrist, genitals, groin,
and mouth. Ten others involved possible or probable ocular vaccinia, typically
treated with trifluridine eye drops. Four (skin = 3, eye = 1) were confirmed
by culture and polymerase chain reaction (PCR) test.
We identified 21 instances of contact transfer of vaccinia, all involving
primary vaccinees, in 306 673 person-months of potential contact with
vaccinia. These included 10 spouses, 2 children, 5 intimate contacts, and
4 friends. No cases of contact vaccinia were observed in patients. The most
commonly named mechanism involved failure to observe bandaging discipline
(eg, touching vaccination sites, touching bandages). Most cases were clinically
mild, manifesting as pustules on arms, shoulders, or hips. In 2 cases, conjunctivitis
or blepharitis developed; each was treated with trifluridine. One was treated
with vaccinia immune globulin (VIG) and discharged from the hospital the following
day.12 Hundreds of thousands of vaccinees flew
from North America to southwest Asia before their vaccinial scabs had fallen
off, yet no spread to passengers or flight attendants was reported.
Two cases of encephalitis as defined by attending physicians presented
7 and 9 days after vaccination, the first in a 23-year-old after primary vaccination
and the second in a 38-year-old after revaccination. In both cases, the patients
recovered and returned to duty. The first case met objective criteria for
encephalitis but the second did not. Diagnostic studies in the first case
by viral culture and PCR did not establish vaccinia or 9 other viruses as
a cause. Twenty-three other neurologic events occurred, 1 fatal. Their association
to vaccination is unclear, given multiple possible causes (neuromuscular weakness
or palsy = 6, seizures = 6, meningitis = 4, optic neuritis = 2, demyelinating
conditions [eg, Guillain-Barré syndrome] = 2, altered cognition = 2,
transient visual defect = 1).
Thirty-seven suspected, probable, or confirmed cases diagnosed as acute
myopericarditis developed in male primary vaccinees aged 21 to 33 years at
7 to 19 days after vaccination. In each of the initial 18 cases described
by Halsell et al,11 the patient presented with
chest pain, and elevated cardiac enzymes were found; electrocardiogram (ECG)
and echocardiogram findings varied. Clinical conditions ranged from mild (no
ECG or echocardiogram changes) to severe (1 case of transient lowered ejection
fraction).13 All survived and either have or
are expected to return to duty.
Eight additional cardiac events occurred 2 to 12 days after vaccination
(myocardial infarction = 4, angina = 2, coronary artery spasm = 1, atrial
fibrillation = 1). A 55-year-old male service member died after a myocardial
infarction 5 days after vaccination. Autopsy findings showed 3-vessel coronary
occlusion, left ventricular hypertrophy, and cardiomegaly.13
Despite extensive efforts to avoid vaccinating women who were pregnant,
85 of 62 622 vaccine-eligible women (0.14%) were vaccinated before they
knew they were pregnant. Once found to be pregnant, these women were offered
clinical support and medical counseling, and enrolled in a prospective registry.9,14
Ten male service members aged 23 to 37 years were vaccinated before
recognition that they were infected with HIV. Each responded successfully
to vaccination and their sites healed at the expected pace. Later studies
showed that their CD4 cell counts ranged from 303 to 751 cells/µL shortly
One service member received extensive burns 5 days after vaccination.
To preclude vaccinial complications of his recovery, this man received VIG
prophylactically. He developed new vesicles 13 days after receiving VIG, but
the vesicles tested negative for vaccinia by PCR and culture. Another burned
service member whose vaccination site had scabbed over was not treated with
VIG. He did not develop other vaccinial lesions.
No cases of eczema vaccinatum or progressive vaccinia occurred. We attribute
no deaths to smallpox vaccination to date.13
Smallpox affected soldiers and sailors for centuries.4,8,15,16 United
States service members received smallpox vaccination from the War of 1812
until 1990.17,18 A 1977 report
notes 54 hospitalizations per 1 million vaccinations (mainly secondary infections).19 Between 1965 and 1975, the US Air Force recorded
6 encephalitis cases after vaccination.19 Service
members spread vaccinia to several civilians in the 1980s.20- 23 After
extensive searching, we could not confirm any deaths due to smallpox vaccination
of military personnel since 1943.17- 19,24- 26
Surveillance of the current US military smallpox vaccination program
identified an increased risk of myopericarditis after smallpox vaccination.11 We believe myopericarditis resulted from smallpox
vaccination in the 1960s but was underrecognized because cardiac enzyme and
imaging technologies were still being developed. We observed ischemic events
after vaccination but at a rate that was less than the age-adjusted rate expected
among unvaccinated military personnel. Allowing for underreporting, we found
no causal association between vaccination and ischemic events.
The incidence of adverse events in Table 3 reflects rates largely below those reported in the 1960s
and 1970 (although such comparisons must be interpreted with caution given
differences in age, prior vaccination, population immunity, and other variables).27- 29 In addition, underreporting
may affect our rates and historical rates,27- 31 but
our global communications and surveillance would tend to reduce underreporting.
Military commanders were vigorous in identifying troops needing treatment.
Teleconsultation identified noteworthy cases that may have been missed in
earlier eras. Historical rates are not directly applicable to contact transfer
because of differences in population immunity and underlying disease.
Our individualized screening for contraindications may explain the lack
of eczema vaccinatum or progressive vaccinia.6,7 Extensive
education may have contributed to low rates of inadvertent inoculation. Interestingly,
reports of inadvertent inoculation from the United States did not differ substantially
from rates reported in Asia, where sanitary conditions are rudimentary. Medical
officers attributed the low rates overseas to disciplined wearing of long
sleeves, attention to personal hygiene, and training. We observed 1 probable
case of autoinoculation among 27 801 military health care workers, comparable
with the 3 confirmed cases reported among 34 541 civilian health care
Successful implementation of bandage evaluation stations in our health
care settings validates this approach.6,7 Our
experience suggests that no disruption of routine clinical care need occur
after vaccination given vigorous infection control.
Similarly, we observed a low rate of contact transfer of vaccinia, despite
the close living conditions associated with mobilization and deployment. These
conditions are known risk factors for other contagious diseases. Most vaccine
recipients used simple bandages, not semipermeable dressings, suggesting the
primacy of human factors and postvaccination discipline. Health care workers
in both military and civilian settings have been comparably successful in
not spreading vaccinia to their contacts.32
This descriptive report cannot establish the efficacy of the screening,
educational, and surveillance methods used because it lacks an explicit control
group. Additional efforts will be needed to evaluate the relative efficiency
of alternative approaches.
The DoD is responsible for protecting and preserving the health of US
service members. We vaccinated with this responsibility in mind. Because smallpox
vaccine is associated with temporary unpleasant symptoms and carries a small
risk of serious adverse events, we sought to implement our program with due
attention to safety. We offer our experience to help the civilian community.