Introduction of SARS in France, March-April, 2003 - SARS Epidemiology

Emerging Infectious Diseases, Feb, 2004 by Jean-Claude Desenclos, Sylvie van der Werf, Isabelle Bonmarin, Daniel Levy-Bruhl, Yazdan Yazdanpanah, Bruno Hoen, Julien Emmanuelli, Olivier Lesens, Michel Dupon, Francois Natali, Christian Michelet, Jacques Reynes, Benoit Guery, Christine Larsen, Caroline Semaille, Yves Mouton, Daniel Christmann, Michel Andre, Nicolas Escriou, Anna Burguiere, Jean-Claude Manuguerra, Bruno Coignard, Agnes Lepoutre, Christine Meffre, Dounia Bitar, Benedicte Decludt, Isabelle Capek, Denise Antona, Didier Che, Magid Herida, Andrea Infuso, Christine Saura, Gilles Brucker, Bruno Hubert, Dominique LeGoff, Suzanne Scheidegger

We describe severe acute respiratory syndrome (SARS) in France. Patients meeting the World Health Organization definition of a suspected case underwent a clinical, radiologic, and biologic assessment at the closest university-affiliated infectious disease ward. Suspected cases were immediately reported to the Institut de Veille Sanitaire. Probable case-patients were isolated, their contacts quarantined at home, and were followed for 10 days after exposure. Five probable cases occurred from March through April 2003; four were confirmed as SARS coronavirus by reverse transcription-polymerase chain reaction, serologic testing, or both. The index case-patient (patient A), who had worked in the French hospital of Hanoi, Vietnam, was the most probable source of transmission for the three other confirmed cases; two had been exposed to patient A while on the Hanoi-Paris flight of March 22-23. Timely detection, isolation of probable case-patients, and quarantine of their contacts appear to have been effective in preventing the secondary spread of SARS in France.

**********

Severe acute respiratory syndrome (SARS) was recently identified as a new clinical entity (1). SARS likely originated in the Guangdong Province of People's Republic of China (2) and subsequently spread worldwide as infected persons traveled. During the 2003 outbreak, SARS was primarily transmitted by person-to-person contact between healthcare workers or household members and ill patients (2). Community transmission also occurred in several of the most affected areas, and an explosive outbreak from a common source occurred in Amoy Garden in Hong-Kong (3). As of June 2003, a total of 8,477 probable cases and 811 deaths had been reported from 32 countries (4). A novel coronavirus has been identified as the cause of SARS (5-7). Based on current knowledge, SARS is transmitted from symptomatic patients by close direct or indirect contacts through respiratory droplet secretions (2). In specific situations, other modes of transmission, such as airborne spread, may be possible (8). The incubation period ranges from 2 to 10 days, allowing SARS to spread over long distances by infected persons who travel (8,9).

We describe how SARS was introduced in France through a single patient who returned from Vietnam on March 23 and present data that suggest transmission from this patient to other passengers may have occurred during his flight back from Hanoi to Paris.

Materials and Methods

After the World Health Organization (WHO) alert on March 12, 2003, a centralized surveillance system was set up for SARS in France (10). All persons who returned from an area affected by recent transmission, had been in contact with a probable case during the previous 10 days, and in whom fever was >38[degrees]C, with cough or difficult breathing, were advised to call the emergency service. These persons were transported to the closest university-affiliated infectious disease ward or one of the nine infectious disease wards designated as a regional reference center in the French plan of action against bioterrorism, using masks for droplet protection. After performing clinical and biologic evaluation and chest x-ray, the attending clinician notified the Institut de Veille Sanitaire through a unique telephone number. On the basis of the results of the initial and subsequent evaluations, each notified case was either discharged, kept as a suspect case, or classified as a probable case using the WHO SARS case definition (10,11). Probable and suspected case-patients were kept in isolation until recovery or until the diagnosis was changed, respectively. For this investigation, a probable case of SARS was defined as previously described (12).

For patients who fulfilled the definition of a probable case, respiratory secretion specimens were taken from the nose, throat, or sputum to detect for SARS--associated coronovirus (CoV) by reverse transcription--polymerase chain reaction (RT-PCR) (7) at the National Reference Center for Influenza (Northern France), Institut Pasteur, Paris. RNA extraction and RT-PCR mixes were prepared in designated rooms. RT-PCR procedures included appropriate negative and positive controls in each run: two negative controls for the extraction procedure and one water control and one positive control for each PCR run. Two RT-PCR, either both nested or one nested and one real-time, were performed for each sample. Real-time RT-PCR, using the SARS-CoV detection kit from Artus (Germany), included an internal control that detected PCR inhibitory substances. One-step nested RT-PCR targeting either the Bernhard Nocht Institute (BNI) or the Centers for Disease Control and Prevention (CDC) fragment of the polymerase gene was used (7,13). When real-time RT-PCR was performed, which targets the BNI fragment, the other RT-PCR was the nested RT-PCR targeting the CDC fragment of the polymerase gene. The real-time and nested RT-PCR, which targeted the BNI fragment reliably, detected 10 copies of RNA in the assay corresponding to 800 RNA molecules per milliliter of specimen.

Acute and convalescent serum samples were also obtained from probable cases. They were tested for immunoglobulin (Ig) G antibodies against the SARS-CoV using indirect immunofluorescence with Vero E6 cells infected by the SARS-CoV, negative control Vero E6 cells and fluorescein-labeled goat antihuman IgG. Results of serologic testing were considered positive either in case of seroconversion or a fourfold increase of observed titers, or if the serum exhibited a titer >160. The detection limit of our indirect immunofluorescence assay corresponded to the first dilution used: 1/40.

For each probable and confirmed case, information was collected on clinical symptoms, chest x-ray findings, leukocyte counts, illness onset date, demography, all possible contacts with a probable case, and exposures when traveling to affected area (contact with any hospital or place of potential transmission). Persons who did not use masks for droplet protection and had contact with a symptomatic probable or confirmed case of SARS were quarantined at home for 10 days after exposure and contacted daily by telephone. As recommended by WHO, this follow-up included the passengers who sat within two rows of a SARS case-patient on the Air France Hanoi-Paris flight of March 22 and 23, 2003 (14). The crew of the Air France flight was also followed for 10 days by the Air France medical service. During follow-up interviews with the passengers seated close to the index patient (patient A), we obtained a detailed description of his clinical condition, his movements in the aircraft, the contacts he may have had with other persons on board, and the timing of his boarding and deplaning in relation to other passengers, including the stopover in Bangkok. Passengers on a flight in which a person with a symptomatic probable case had traveled were informed publicly through the media and mail of the potential exposure and advised to call the emergency service phone number to be evaluated and admitted to the closest university-affiliated infectious disease ward if a fever of >38[degrees]C developed within 10 days of the flight.

We estimated the incidence density of SARS among passengers who sat within two rows of a case of SARS in the AF171 flight of March 22-23 by using the total number of person-hours as the denominator. Ninety-five percent confidence intervals (95% CI) were calculated by using the exact binomial method (15).