Written informed consent was obtained from all patients involved in this study. The study was approved by the Ethics Committee of Nanjing Drum Tower Hospital (EC20130411-1).

During the late March and the early April 2013, several patients with high fever and server pneumonia were admitted to our hospital. On April 2, two of them were confirmed to be infected with avian influenza A(H7N9) virus by the Chinese Center for Disease Control and Prevention in Beijing; one more patient was confirmed the infection by the Jiangsu Provincial Center for Disease Control and Prevention in Nanjing on April 4 [3], [4]. Once these three patients were diagnosed, we wanted to start screening avian influenza A(H7N9) virus in patients with fever and flu-like symptoms who attended to our hospital. However, we did not start the screening until April 15 because of the short of detection agents at that time. We stopped the screening after April 25 because rare cases had been confirmed to be infected with avian influenza A(H7N9) virus within one week, the presumed incubation period, before April 25. With the throat-swabs specimens collectors, we collected the throat swab samples from 200 consecutive febrile (≥38°C) patients with flu-like symptoms when they first went to our hospital from April 15 to April 25.

Each swab was vigorously shaken in 0.5 mL of diethylpyrocarbonate treated water and squeezed on the inside surface of tube to release as much trapped virus as possible before the swab was taken out. Each of the samples was added with 1.0 mL TRIzol (Invitrogen, Life Technologies, USA), vortexed for at least 15 s, and then stored at −70°C.

The total RNA was extracted from samples in TRIzol according to the manufacturer's instructions. To increase the recovery rate of RNA, we added 0.5 µL glycogen (20 mg/mL) in the precipitation of RNA. The final pellet was dissolved in 10 µL diethylpyrocarbonate treated water. Five microliters each was used to detect H7 and N9 genes respectively with the fluorescence reverse transcription (RT) PCR Detection kits (BioPerfectus Technologies, Taizhou, Jiangsu Province, China) on ABI 7500 Real time PCR system (Applied Biosystems). The protocols and primers were prepared in accordance with those provided the WHO Collaborating Centre in Beijing [1].

The virus isolation was performed essentially as described previously [8]. Briefly, throat-swab specimens taken from patients were maintained in viral transport medium on ice fewer than 24 hours, followed by filtration sterilization and inoculation into the allantoic cavities of 10-day-old specific-pathogen-free (SPF) embryonated eggs (Beijing MERIAL Ltd.). After incubation at 35°C for 48–72 h, the allantoic fluid of the inoculated eggs was collected and identified by hemagglutination assay.

The hemagglutinin inhibition (HAI) assay was performed with 0.5% chicken red blood cells as described elsewhere [9] with some modifications. Briefly, the sera from patients were treated with receptor-destroying enzyme (Denke-Seiken, Japan) and inactivated at 56°C for 30 min, followed by incubation with chicken erythrocytes to adsorb nonspecific agglutinins. Then the sera, starting at a 20-fold dilution, were 2-fold serially diluted with PBS, and added with 25 µL/well on a 96-well polystyrene microtiter plate, followed by adding 25 µL of virus suspension containing 4 hemagglutinin units per well. After incubation of the plate at room temperature for 1 hour, 50 µL of 0.5% (v/v) chicken red blood cells was added to each well. After incubation at room temperature for 30 minutes, the HAI titers were determined as the reciprocal of the highest serum dilution that completely inhibited hemagglutination.

Statistical analysis was performed using SPSS software version 13.0 (SPSS, Chicago, USA). We summarized the continuous variables as means and standard deviation. For categorical variables, the percentages of patients in each category were calculated.

From April 15 to April 25, 2013, totally 5,009 patients attended to the outpatient emergency department of our hospital. Of them, 337 with fever ≥38°C were required to attend the fever clinic of our hospital. A total of 200 febrile patients, who also presented flu-like symptoms, were recruited in this study to have their throat-swab specimens tested for the presence of avian influenza A(H7N9) virus with fluorescence RT-PCR. Their general characteristics and results of routine blood tests are shown in Table 1.

One (0.5%) of the 200 patients showed positive in the tests of both avian influenza virus H7 and N9 in RT-PCR and 199 others (99.5%) were negative. Additional throat-swab specimens from the patient with positive avian influenza A(H7N9) virus RNA, which had been collected the next day, were retested with the same kit in two other independent laboratories; both laboratories reported positive. Thus the patient was diagnosed with the infection of avian influenza A(H7N9) virus. All other patients were excluded from the infection of the virus, and recovered after 1–3 weeks treatment.

The patient diagnosed with the infection of avian influenza A(H7N9) virus was a 36-year-old man who lived in Xinyi city, which is located in north part of Jiangsu Province, 310 km away from Nanjing. He was a living poultry retailer and had transported living poultry from Xinyi city to Hangzhou city, the capital of Zhejiang Province, where sporadic avian influenza A(H7N9) occurred from February to April, 2013 [2], [4]. He had a known exposure history to live chicken within a week before the onset. He developed fever (38.4°C), non-productive cough, malaise, and myalgia on April 18, and his condition was deteriorated with the highest temperature 40.5°C and dyspnea on the subsequent days. The patient was referred to our hospital on April 22, 2013.

On the physical examination in our hospital, the patient had high fever (40.2°C), dyspnea and lip cyanosis. Blood tests revealed that leukocytes were 3.3×10⁹/L with neutrophils of 2.6×10⁹/L, lymphocytes of 0.5×10⁹/L, platelet of 93×10⁹/L, and red blood cells 4.38×10¹²/L, and the liver and kidney functions were generally normal. The sputum cultures for bacteria in the first three days were all negative. Arterial blood gas analysis presented the results of pH 7.441, PO₂ 64 mmHg, PCO₂ 30.5 mmHg, and SaO₂ 93%, under nose tubular oxygen supplement (5 L/min). Chest x-ray and computed tomographic (CT) scanning revealed consolidation and diffuse infiltrates in the right lower lobe and patchy infiltrates in the left lower lobe (Figure 1A and 1C). He was diagnosed with severe pneumonia (suspected H7N9 influenza) and respiratory failure. He was admitted to the Respiratory Department, and given the oseltamivir (150 mg, twice daily) and antibiotics as well as other symptomatic treatment. Meanwhile, throat-swab samples were collected for detecting avian influenza A(H7N9) virus, which turned out to be positive on the test of April 23. On the next day, throat-swab samples were collected again for detecting avian influenza A(H7N9) virus; the viral RNA was positive in the two independent laboratories reported on April 24.

Once the diagnosis was established on April 24, the patient was quarantined in the isolation room to prevent the potential transmission and treated with the combination therapy against influenza A, including oseltamivir and piperacillin/tazobactam (Tazocin). The main clinical features, laboratory findings, and treatment during the disease course are presented in Figure 2. The patient received continuous nasal oxygen supplementation (2–5 L/min) in the period of April 23–April 25 and intermittent nasal oxygen supplementation (2 L/min) during April 26 and 27; he never used mask oxygen inhalation or mechanical ventilation. The patient's condition was substantially improved within three days after admission; the dyspnea disappeared and hypoxemia was corrected. The isolation of the virus from the throat-swab sample collected on April 25 was failed even after the sample was propagated in the allantoic sac and amniotic cavity of embryonated chicken eggs. Avian influenza A(H7N9) virus RNA in the throat-swab samples was still positive on April 26 but became undetectable on April 28 and remained negative thereafter, while the viral RNA was undetectable in the serum samples collected on the April 23, 26, and 28 respectively. The chest CT-scanning on May 9 showed that the patchy infiltrates in the left lower lobe disappeared and the diffuse infiltrates in the right lower lobe significantly improved (Figure 1D). In agreement with this, no consolidation was observed on the chest x-ray on May 13 (Figure 1B). On May 15, the patient was discharged after he had a good recovery. The follow-up on May 31 showed that the physical examinations were generally well and hematological and biochemical investigations did not reveal abnormal results, but his chest CT-scanning still had some abnormal in the right lower lobe (Figure 1E). A further follow-up conducted on July 2 revealed no abnormal findings in physical examinations, blood tests and chest CT-scanning (Figure 1F).

The hemagglutinin inhibition test showed that the serum at the early phase of the illness, on day 6 after the onset, had the titers lower than 1∶20 against avian influenza A H7N9, H5N1, and 2009H1N1, respectively, but had a titer of 1∶80 against seasonal influenza A H1N1. However, the serum of convalescent phase, on day 44 after the onset, had titers of 1∶80 against avian influenza A(H7N9) virus (Table 2).

The patient had not been quarantined until his diagnosis was established on April 24, 2013. Some persons closely contacted the patient without strict preventive measures. These included his parents, his wife and two children, and 18 physicians and nurses. We followed these persons to investigate whether avian influenza A(H7N9) virus was transmitted from human to human. None of the 23 close contacts developed the fever and flu-like symptoms within 14 days after contact.