Introduction

Leptospirosis is an infectious disease caused by a spirochete from the Leptospira species after an incubation period of 10 days. It has not been mandatorily notifiable since 1986 in France. The current surveillance system is based on data from the National reference centre (Centre National de Référence, CNR) for leptospires, which receives serological requests from hospital laboratories or laboratories for biological and medical analyses located throughout the whole territory. The yearly incidence estimated by this system in metropolitan France is low and stable, with an average of 290 new cases per year between 1984 and 2000 (268 cases in 2000, incidence of 0.44 cases per 100 000 inhabitants) (1).

A case-control survey carried out in metropolitan France identified the main risk factors in 1999-2000 as the following: presence of wounds, canoeing-kayaking, contact with fresh water for professional reasons, and contact with wild rodents (2). Recently, a working group gathered by the Institut de Veille sanitaire (InVS, National public health institute) aiming at defining priorities in the field of non foodborne zoonoses (2000-01), identified leptospirosis as a priority disease. In this context, reinforced collaboration with the CNR was recommended to undertake epidemilogical surveillance, and to implement an alert system for the human population, as well the publication of a guide book for the investigation of clustered cases.

Alert

On 14 June 2001, the local health authorities of Charente-Maritime (DDASS, Direction départementale des affaires sanitaires et sociales, Departmental directorate of social and health affairs) was informed by the Rochefort sub-prefecture that five teenagers hospitalised in Rochefort presented clinical signs compatible with leptosirosis. For three of them, the search for leptospires by PCR at the Pasteur Cerba laboratory was positive. The patients had swum with thirty other teenagers in the Genouillé canal, in the locality "Les 3 bras" situated two kilometres away from the Libération area, just outside Rochefort (figure 1).

On 26 June, the DDASS and the InVS, in collaboration with the local epidemiological centre, (Cellule inter-régionale d’épidémiologie, CIRE), decided to send a team on site to carry out an epidemiological investigation with the following objectives:

• confirm the common origin of clustered cases,

• search for the origin,

• identify risk factors,

• identify additional potential control and prevention measures,

• strengthen collaboration with the CNR in order to improve the alert,

• design an investigation guide book for clustered cases of leptospirosis.

Materials and methods

An epidemiological investigation (exploratory and case-control) was conducted by the Charente-Maritime DDASS, the CIRE Centre-Ouest, and the InVS. The objectives of the exploratory phase were to describe the cases, to confirm the epidemic nature of the clustered cases, to record other potential cases, and to raise hypotheses to be tested during the analytic phase.

During the first phase, a case was defined as a person living in Rochefort or in the neighbourhood with leptospirosis diagnosed from 25 May 2001.

• Possible case: with a unique serodiagnosis by a micro-agglutination test (MAT) positive with one titre ≥ 1:100 or ELISA ≥ 1:400

• Confirmed case: with positive PCR or culture or positive serology by MAT with two samples taken at a two-week interval with titres ≥ 1:100, and at least a four-fold increase between the two samples.

Cases were investigated from the Pasteur-Cerba laboratory, the CNR for leptospires, the four medical and biological laboratories in Rochefort, the Rochefort Hospital, and the general practitioners of the Libération area.

Cases identified were part of a group of teenagers, the study population, having bathed at least once between 18 May and 10 June in the Genouillé canal, in the locality of "Les 3 bras" (Figure 1). For the case-control study, among these teenagers, a case was defined as any person having bathed at least once during this period in the Genouillé canal, locality "Les 3 bras", presenting a positive biology and an acute infectious syndrome (temperature above or equal to 38.5°C), during the period from 25 May to 17 June 2001. A control was considered as any person who did not present temperature above 38.5°C since 25 May. Moreover, the control should have no history of leptospirosis or vaccination against this disease.

Thanks to the cases’ indications, it was possible to identify and detect among the thirty teenagers exposed, 10 non-sick teenagers recruited as controls. Cases and controls were interviewed by phone, by the same person, and using the same questionnaire. As two controls were absent, a family member was interviewed.

Data collected were the following:

• Demographic data;

• Presence, type of symptoms, and date of onset;

• Exposure during the three weeks before onset of symptoms: location and frequency of contacts with the canal water;

• Estimation of the total duration of the contact with water, use of protection, presence of wounds and their type during swimming, knowledge of vaccination against leptospirosis, presence of any chronic disease and its potential treatment, knowledge of an antimicrobial ongoing treatment, during or a month before swimming;

• Hospitalisation and hospital clinical data;

• Biological data and leptospirosis diagnosis (PCR, serology).

The case-control study’s objectives were to analyse the potential risk factors, namely the presence of wounds during swimming, the use of protection, the duration of contacts with the canal water, the presence of chronic diseases, of a chronic or antimicrobial treatment during the exposure period. The data analysis was carried out using Epi Info, version 6.04.

A bacteriological analysis (search for E.coli and coliforms) was carried out on 15 June on the canal waters by the Departmental direction of water and forests (Direction Départementale des Eaux et Forêts, DDAF).

Between August and September, the veterinary school in Nantes (ENV, Laboratoire B2ML) and the National federation of departmental groups for the protection of cultures (Fédération Nationale des Groupes Départementaux de la Protection des Cultures, FNGDPC) placed traps near the locality of "Les 3 bras" to estimate the presence of rodents potentially excreting of leptospires. Three rodent species were trapped: coypus, muskrats and field rats, and serologies were undertaken.

Results

Characteristics of the cases

In total, 5 cases (3 confirmed and 2 possible) were identified by the Hospital Centre of Rochefort. All cases occurred in a clustered way between 4 and 12 June (date of onset of symptoms). This suggests an exposure period limited in time. Depending on the dates of occurrence of the cases, and incubation periods (between 4 and 19 days), the exposure period was estimated to be between 18 May and 10 June (figure 2).

Four out of five cases were male (sex ratio M/F = 4). The cases were aged between 10 and 19 years (average 13.2 years, median 11 years). All lived in the Libération area located in the north east suburb of Rochefort, two kilometres away from the suspected exposure area.

Clinically, all cases presented fever with headaches, abdominal pain and vomiting, and in four cases, chills and myalgia. Three cases presented conjunctivitis, one meningeal symptoms, and one a pulmonary attack. No jaundice was observed (table 1). Four cases were hospitalised in the Rochefort Hospital, the fifth consulted in the same hospital. The duration of hospitalisation varied from two to eight days (median 6 days).

Tableau 1 / Table 1. Fréquence des signes cliniques (n=5), cas groupés de leptospirose, Charente Maritime, juin 2001 /  Frequency of clinical signs (n=5), clustered cases of leptospirosis, Charente Maritime, June 2001

For three cases, positive PCR and seroconversion were observed. The DNA sequences were similar to each other and to the sequence of a strain from icterohaemorrhagiae serovar. For the other two, only an IgM initial titre of 1/400th was shown by ELISA.

Potential risk factors

In the case-control study, the cases included are the five confirmed and probable cases identified during the exploratory survey. One case and three controls suffered from a chronic disease (asthma for one case and two controls, and "convulsions" for another control), and were taking a long term treatment (p=0.6). No case or control used protections (goggles, shoes, other) while swimming. Only one control had wounds at the time of swimming (p=0.7).

The average cumulated duration of swims was significantly higher for the cases (median 23.8 hours) than for controls (median 14.4 hours), p=0.035.

Environmental survey and hydrologic data

The analyses of the canal water testified a moderate quality (E.coli between 100 and 2000/100 ml), according to swimming water criteria.

At the locality "Les 3 bras", three canals join and pour their fresh waters in the Charente. Two locks regulate the flow between the canals fresh waters and brackish waters. Between 28 May and 11 June, the locks had not been opened.

The rodent traps made by the ENV and the FNGDPC has permitted the analysis of serological status of 130 rodents. The prevalence of seropositive rodents, all antigenes considered, was 30.8%, 23.1% for L. icterohaemorrhagiae, 4.6% for L. saxkoebing , 3.8% for L. australis. The proportion of rodents with positive serology for serovar icterohaemorrhagiae was higher in coypus (30%) than in muskrats (11%) (p=0.014).

Discussion

The exploratory phase showed that all the recorded cases had bathed in the Genouillé canal, in the locality called "Les 3 bras". As it was not possible to reconstitute the cohort of all teenagers who swam at this site, we have not been able to estimate the attack rate. This led us to conduct a case-control study for the analytic phase of the investigation.

Swimming in this canal from 18 May to 10 June was probably the origin of all leptospirosis cases. The total duration of swims in this canal was associated with acquiring leptospirosis. However, an information bias is possible due to the tendency for cases to increase the length of their exposure, or inversely, for controls to decrease it.

No other risk factor (pre-existing conditions, wounds, ongoing treatments) was found. However, the strength of this study remained weak due to the small study population. Moreover, in the absence of serology for controls, asymptomatic persons may have been misclassified as controls.

The occurrence of clustered cases suggests recent contamination of the waters in " Les 3 bras " that can be explained by one or several of the following reasons:

• The presence of rodents excreting of leptospires was reported by residents around the swimming area. A study on coypus (July 2000–May 2001) trapped in marshes or rivers has shown that in Charente-Maritime, 52% of animals (63/120) were seropositive for MAT, the icterohaemorrhagiae (61/63) being the predominant serogroup; (3)

• An increase or a shift of rodents’ population excreting of leptospires towards the swimming zone may have interfered recently (because of heavy rains in other areas more visited by rodents);

• The trapping results carried out in August showed a predominance of the icterohaemorrhagiae serogroup, although not specifically in the swimming area of cases, especially in coypus, suggesting the potential role of these rodents in the occurrence of clustered cases in Rochefort ;

• Between 28 May and 11 June, the lock was not opened, and that could have facilitated the concentration of leptospires in the swimming area due to water stagnation.

The occurrence of clustered cases of leptospirosis raises the question of using doxycycline as a secondary preventive measure. The use of this antibiotic in endemic areas has shown a protective effect on morbidity and mortality during epidemics, without however reducing the leptospires infection. As soon as the first leptospirosis cases are detected, a chemoprophylaxis with 200 mg of doxycycline per week for individuals exposed to leptospires could prevent the occurrence of clinical signs.

The investigation carried out on site has not only allowed us to describe clustered cases, but also to sensitise clinicians and private laboratories to the existence of the disease, and the notification of suspected cases, and to relaunch discussions on the CNR’s role, the alert mechanism, and the potential use of antibioprophylaxis for the persons exposed. It has also supplied indications for the writing of an investigation guide book.

Control measures and prevention

Upon the diagnosis confirmation, the Rochefort hospital issued information on the risk linked to swimming for the relatives of the teenagers hospitalised. Press releases about the disease, its symptoms and the swimming area were published by the DDASS through the media.

This information however seemed to have a limited impact: during the investigation, most teenagers interviewed and their families ignored the transmission routes of the disease, although these were indicated in the press releases.

Road signs around the locality of "Les 3 bras" reminding the ban on swimming was reinforced by the town hall. Surveillance measures were set up in the community laboratories and the Rochefort hospital (reporting to the DDASS of any new leptospirosis case, dispatch of strains to the CNR for confirmation).

The occurrence of clustered cases and the environmental conditions of the suburban area of Rochefort justify the reinforcement of surveillance of human cases of leptospirosis, during risk periods (swimmimg months). It would be useful that all information from general practitioners, laboratories and hospitals, could result in immediate reporting of any suspected case (without waiting for biological confirmation), even if notification is not mandatory, to local health authorities (DDASS), together with the dispatch of samples to the CNR. This procedure could be set up in the city of Rochefort, and applied every year.

Information to the public need to be maintained in parallel (information on the disease, banned swimming areas). In the context of developing investigation guidelines for clustered cases of leptospirosis, it will be useful to study the procedures to carry out serology in controls, in order to reduce the classification bias in analytical studies.