In late August 2007, an outbreak of highly pathogenic avian influenza A virus (HPAIV) of subtype H5N1 was reported in a large duck farm in Wachenroth, in the county Erlangen-H?chstadt, state of Bavaria, in Southern Germany (index farm). Daily mortality among ducklings was approximately 2% and the presence of HPAIV H5N1 was confirmed in several samples (although clinical signs were not typical for HPAIV H5N1). Protection and observation zones were installed according to German legislation, and over 165,000 animals were subsequently culled. The epidemiological contact investigation of the flocks (forward and backward tracing) led to the identification of H5 infection of ducks in two other farms in the nearby county of Schwandorf (contact farms 1 and 2). As in the previous farm, clinical signs were not typical for HPAIV H5N1. The Friedrich-Loeffler-Institute, the Federal Research Institute for Animal Health in Germany, confirmed the presence of the Asian lineage of HPAIV H5N1. Culling of another 205,000 ducks was started at these two sites on 7 September, exactly two weeks after the first outbreak. Two days later, low pathogenic avian influenza A virus (LPAIV) of subtype H5N8 was detected in two farms (contact farms 3 and 4) in the two adjacent counties of Rottal-Inn and Dingolfing in Bavaria (approximately 250 kilometres away from the index farm). On these five farms, 438,000 animals were culled. A specialised company from the Netherlands was commissioned to perform the culling. We currently have no information about how many foreign nationals took part in the operation, or whether other foreign companies were also involved. To our knowledge. the involvement companies and the hiring of foreign nationals is not unusual.

On 22 September, the National Institute for Public Health and the Environment; (RIVM) of the Netherlands informed the Robert Koch Institute, its German counterpart, about a Dutch poultry worker hospitalised with high fever and radiologically confirmed pneumonia who was reported to have been involved in the culling in contact farms 1 and 4.

Case report
The onset of the disease with flu-like symptoms, including light fever (38o C), headache, myalgia, a slight cough and general malaise, was reported by the Dutch poultry worker to have been on 18 September, just inside the upper limit of the potential incubation period of avian influenza (eight days) [1]. Prophylactic oseltamivir was changed into therapeutic dosage (75 mg twice daily), but hospitalisation was not initially necessary. In accordance with the Dutch preparedness protocol for exposure to avian influenza virus [2], nose and throat swabs of the patient were taken on 18 September at home by the Municipal Health Services (the guidelines recommend isolation of cohorting when the patient is hospitalised to avoid contact with other patients; analogous mixing with other outpatients should be avoided by taking the samples at home). The swabs were analysed simultaneously by PCR at the RIVM and the Erasmus Medical Centre (Erasmus MC), Rotterdam, for general influenza virus A, specific influenza virus A (H1, H3, H5, H7), influenza virus B, other respiratory viruses (respiratory syncytial virus, human metapneumovirus, rhinovirus, parainfluenza virus 1-4, adenovirus, bocavirus and coronaviruses) and Mycoplasma pneumoniae. All result, including those for influenza virus, were negative. However, the patient reported to have taken oseltamivir (irregularly) as prophylaxis, which could have caused a negative influenza result.

On 21 September, the condition of the patient worsened, and he was admitted in isolation in the hospital. The patient then started with antibiotic treatment for atypical pneumonia. The patient’s condition did not initially improve. As a result of this, and because lower respiratory tract specimens have a higher positive rate for H5N1 [3-5], it was decided to redo the laboratory diagnosis with material from a bronchoalveolar lavage.

After information about the case from RIVM, the data in the German national surveillance system were screened for cases of ornithosis because of the known risk to poultry workers. Results showed five human infections with Chlamydophila psittaci in the county Erlangen-H?chstadt, where the index farm was located, between January 2001 and September 2007. In four of these – two cases from 2001 and one each in 2004 and 2005 – work in a poultry abattoir was reported as the suspected source of infection. Screening the national database for all cases of ornithosis with additional information about the source disclosed two further cases in the states of Thuringia from 2001 and 2004 that were potentially linked to professional slaughtering of poultry in the county of the index farm.

Provided with the information on possible exposure of the patient to the ornithosis agent, the doctor in charge of the patient, in consultation with the RIVM, expanded the differential laboratory diagnosis by PCR testing for the pathogens as indicated above by including testing for bacteria in the Order Chlamydiales [6]. The only positive finding for the bronchoalveolar lavage material was in the Chlamydiales PCR of which the product was identified as C. psittaci by sequencing as the causative agent of the pneumonia in this patient. The clinical conditions improved rapidly and the patient was released from the hospital after seven days.

Discussion and conclusion
This case showed the usefulness of national and international surveillance and communication for the rapid diagnosis of human illness following exposure to potential zoonotic micro-organisms. The rapid exchange of relevant information on the circulation of ornithosis was crucial in directing the diagnostic procedure in a situation were primarily exposure to HPAI H5N1 was the most likely cause of the clinical symptoms.

The risk of transmission of C. psittaci from ducks is well known, and high seropositivity rates have been reported previously in farm workers [7]. Since protective measures for HPAI would have been sufficient for C. psittaci infection, it is not clear how the patient contracted the infection and other unidentified exposure situations can not be excluded.

This case is also a reminder that culling of large numbers of animals, especially in farms suspected to be infected with avian influenza virus, often involves large numbers of workers coming from several countries, where recommendations for worker protection might differ. Therefore, when involved in mass culling, local authorities should assure the following:

1. Informing cullers about recommended protective measures according to local recommendations and the current risk assessment (if final confirmation of the causative agent is pending, strict adherence to protective measures against potential animal to human transmission is necessary);
2. Supervising the adherence to these recommendations;
3. Assuring the possibility of tracing and contacting exposed workers, in case increased risk for exposure to HPAIV is suspected;
4. Ensuring early close collaboration between veterinary and public health authorities [8].

Furthermore, pre- and post-exposure serum samples of exposed workers involved in culling might be considered in order to detect clinically unapparent transmission, enhance knowledge on the risk for infection and improve rapid differential diagnosis should workers develop symptoms after exposure.

A lack of personnel in local authorities responsible and a fear of economic losses among farming and culling businesses may hamper these activities. However, the potential risk to public health demands alertness and close cooperation between the authorities in charge and the businesses involved, locally and even internationally, as this case report has illustrated.