Tularaemia is a bacterial zoonosis caused by Francisella tularensis .
Humans may become infected through bites from infected ticks or other insects,
contact with infected animals or contaminated animal products, consumption
of vegetables, water or earth contaminated by the faeces or corpses of
infected animals, or inhalation of aerosolised bacteria. The median incubation
period for the disease is 3 to 5 days (range: 1 to 14 days) . The clinical
form of tularaemia depends on the route of entry of the bacterium into
the body: ulceroglandular tularaemia is involved if transmission is transcutaneous;
pulmonary and typhoidal tularaemia if caused by inhalation; oropharyngeal
tularaemia if by ingestion.
On 21 August 2004 a general practitioner informed the local Direction Départementale
des Affaires Sanitaires et Sociales (Departmental Health and Social Services
Division, Ddass) of 15 cases of flu-like infections in patients who had
spent 4 August 2004 at a mill that had been converted into a home in Vendée,
western France. On 8 September, blood tests confirmed the diagnosis of
tularaemia for 3 of the 15 patients.
Because of the similarity of symptoms in the other 12 patients who had
been at the same place on the same date, a diagnosis of tularaemia, and
a common source of contamination were suspected for the whole group. Epidemiological
and environmental investigations were performed to confirm the diagnosis
and identify the source of contamination and the mode(s) of transmission
with a view to taking appropriate outbreak control measures.
This investigation included a visit to the mill, a descriptive investigation
of the cases and a retrospective cohort study of all the subjects who
stayed at the mill from 24 July to 11 August 2004.
A site visit was conducted to describe the house and its surroundings,
and to interview the owners in order to establish a list of animals and
humans who were present during the period under study and retrace their
activities during that time, particularly on 4 August.
A case was defined as any patient with fever and a positive blood test
(agglutination): either a seroconversion, or a significant increase in
antibody titres, or a single titer greater than or equal to 40. The blood
tests for all the patients were performed by the national reference centre
The clinical forms were classified as pulmonary tularaemia (minimum of
one respiratory symptom or abnormalities on the chest x ray picture)
and typhoidal tularaemia based on clinical and biological information
collected from the physician.
The subjects included in the cohort were questioned using a standardised
questionnaire about their symptoms and possible exposure to F. tularensis
(contact with animals, water and soil, possible food exposure and leisure
activities) during the 15 days before onset of symptoms for infected
subjects and from 26 July to 10 August for uninfected subjects.
The strength of the association between the disease and the exposure
was measured by calculating the odds ratio and the 95% confidence interval,
using Epi Info 6fr.
A logistic regression including the variables significantly associated
at a level of p=0.1 in a univariate analysis and those considered the
most biologically plausible was performed using Stata.
The presence of F. tularensis was investigated by culture and
PCR in suspected sources of the contamination, based on samples of the
tank water, mud and fragments of bone from a small mammal collected from
the bottom of the tank, and firewood piled close to the house.
The presence of F. tularensis was investigated by culture and
PCR of cloacal swabs from ducks at the mill and blood samples from the
owners’ dogs. The dogs’ blood specimens were sampled for
The SAGIR network, in charge of French national wildlife health monitoring,
was questioned regarding the possible presence of animal corpses contaminated
by tularaemia in Vendée.
Description of patients
The patients included 10 adults and 5 children, with a median age of
39 years [range: 6 to 49 years]; the male/female ratio was 1,1. The cases
were grouped, and occurred between 9 and 12 August 2004, suggesting a
single point of contamination [FIGURE 1].
The median duration of the symptoms was 6 days [range: 2 to 13 days].
All subjects experienced fever and headaches. Other signs were: asthenia
(93%), myalgia (80%), arthralgia (73%) and respiratory symptoms (73%).
Six of 11 chest x ray pictures were abnormal (four pneumonias, two
pleural effusions). Twelve patients presented with pulmonary clinical
symptoms, and three with typhoidal symptoms. None of the patients was
admitted to hospital. The outcome for all patients was favourable,
whether or not antibiotic treatment was prescribed. Seroconversion
over a minimum of 10 days was documented in three cases, a significant
elevation of antibody titre in 11 cases, and an elevated single titre
in one case. The median duration of incubation was 7 days [range: 5
to 8 days].
Description on August 4, 2004
The mill had been renovated into a comfortable home and was supplied
with water from a tank. On 4 August, 19 subjects, including the 15
patients, and domestic animals (five ducks, one donkey, one sheep,
eight cats and three dogs) were present [FIGURE 2]. The four subjects
who did not become ill had spent the day upstairs in the house and
left before 7 pm. Four of the patients had handled firewood in the
mill’s woodpile, carrying it through the ground floor of the
mill. Fifteen people attended dinner between 8 pm and midnight in a
room on the ground floor where dogs were also present.
Thirty nine subjects, including 24 asymptomatic people, were included
in the cohort. The incidence rate (IR) was 38%. Being at the mill during
dinner on 4 August 2004 was strongly associated with contracting the
disease (IR = 100%; p <10-8). Patients who developed disease were
all exposed to bread or pizza cooked in the bread oven at the mill
(IR = 60%; p=4x10-4) and to water from the tank at the mill
(IR = 52% ; p=6x10-3).
Environmental and veterinary investigation
The analyses performed on environmental specimens and on domestic animals
were negative for F. tularensis. One dog tested positive (titre
1:160). During the study, the SAGIR network did not identify any contaminated
wild animal corpses.
These investigations confirmed the occurrence of 15 grouped cases of
tularaemia. It was shown that the contamination took place in a mill
in Vendée on the evening of 4 August 2004 .
The pulmonary clinical form suggests contamination via aerosolised bacteria.
This could be explained by dust particles suspended in the air while
firewood was carried through the ground floor in the afternoon, or by
contaminated particles present in dog fur. Dogs may carry bacteria in
their fur after contact with an infected animal or contaminated environment
[4,5,6]; bacteria are then disseminated when the dog shakes itself. The
mill is located in an area where tularaemia is endemic ; one dog tested
positive for previous contact with F. tularensis. This type
of contamination has been described during a similar episode in the United
States in 1978 [5,6].
All samples were analysed by PCR. The negative PCR test results can be
explained by late performance of testing, since F. tularensis does
not survive more than several days in animal bodies .
In France cases of tularaemia are notified sporadically, and modes of
transmission include contact with wild game and tick bites . Two thirds
of the cases are ulceroglandular; the pulmonary form is unusual .
Additionally, the pulmonary form, as described in other countries, is
usually severe [2,9,10].
This incident has demonstrated that a pulmonary form of tularaemia exists
in France; given the uncharacteristically mild form of the disease, and
in the absence of specific clinical symptoms, the diagnosis may often
Tularaemia should be considered in cases of pneumonia of unexplained
origin, especially if risk from exposure has been reported.
The mandatory reports required when a unusual phenomenon such as tularaemia
cases is observed must be submitted to health authorities urgently to
facilitate investigation and expedite rapid action. The transmission
of F. tularensis by inhalation may be prevented by wearing protective
equipment (goggles, gloves, masks), mainly used by professionals (gardener,
Basic hygiene measures can help prevent transmission of the disease from
pets, for example, washing pets (avoid splashing) before they enter the
house if they have been rolling in mud or have been in contact with dead
animals. Thorough handwashing is recommended for all people after contact
with any animals, including pets.
We would like to express our appreciation to all participants in this
study, with special thanks to:
the Mérieux (Dr S Trombert) and Cerba (Dr G Desnoyel) laboratories,
biological and medical laboratories, general practitioners, especially
the notifying physician (Dr C Ennaert) and assisting veterinarians, departmental
veterinary laboratories and the Office national de la chasse et de la
faune sauvage (National office of hunting and wildlife), the Institut
départemental d’analyses et de conseils de Loire-Atlantique
(medical laboratory and advisory institute in the Loire-Atlantique region),
and the DDASS in the Vendée region.