Dear Editor,

Microbiological diagnosis of pandemic influenza A (H1N1)v pneumonia is based on rapid and sensitive nucleic acid amplification test of respiratory samples [1], mainly nasopharyngeal aspirate or, alternatively, nasal and pharyngeal swabs. False negatives can be obtained due to errors in collection, inadequate handling or transport, but also because of low viral load in the specimen.

Over 4 months, 18 patients with confirmed influenza virus A (H1N1)v infection were admitted to a medical surgical ICU of a community hospital. Of these, 15 patients presented with severe pneumonia. At ICU admission, respiratory specimens were obtained from the upper respiratory tract. The results of reverse transcriptase polymerase chain reaction (RT-PCR) were positive in 12 cases and negative in 3. Due to high clinical suspicion of the disease, the diagnostic study was extended with other samples.

Based on the hypothesis that more extensive viral replication exists in the lower respiratory tract, bronchoalveolar lavage (BAL) was considered. Fiberoptic bronchoscopy is a safe procedure in critically ill patients without absolute contraindications, but in cases of severe hypoxemia, it is not recommended. One of the three patients presented hypoxemia despite high PEEP, and in this case the diagnosis was achieved by serology. Two cases are presented here in which BAL was performed due to a high clinical suspicion of influenza A (H1N1)v pneumonia despite negative RT-PCR results from upper respiratory samples.

Case 1: Patient with morbid obesity and epilepsy was admitted to the emergency department with fever, dry cough, and myalgia of 5 days of duration, and required ICU admission due to respiratory failure and the presence of bilateral infiltrates on chest radiograph. At admission the patient was intubated for controlled mechanical ventilation. RT-PCR of nasopharyngeal exudate and aspirate performed at ICU admission tested negative for influenza A (H1N1)v. BAL performed 4 days after ICU admission was positive.

Case 2: Young smoker patient was admitted to the pneumology ward with fever, diarrhea, nonproductive cough, and progressive dyspnea. During admission the patient developed respiratory failure and was admitted to the ICU. Chest radiograph showed bilateral alveolar infiltrates, and RT-PCR tests of upper respiratory samples for influenza A (H1N1)v performed at admission and 4 days later were negative. On the 5th day after ICU admission, a BAL was performed, obtaining a positive result.

The incidence of false negatives in critically ill patients was documented in a subgroup of 21 such patients in Australia and New Zealand [2]. BAL samples were positive in all patients, but upper respiratory tract samples tested positive in only 17 of 21 patients. Rello and colleagues documented a 10% false negative rate for nasopharyngeal swabs [3]. Based on these findings and the presented cases, negative results of RT-PCR of upper respiratory samples can not rule out the diagnosis of pandemic influenza A (H1N1)v pneumonia [4].

We did not report viral load, as in cases of influenza A (H1N1)v with severe pneumonia, the natural history of viral shedding from the respiratory tract is uncertain. Well-conducted studies are needed to confirm the hypothesis that viral load is higher in the lower respiratory tract than in the nasopharynx. Clinicians should consider testing lower respiratory tract samples in critically ill patients with negative results for specimens coming from the upper respiratory tract (Fig. 1).

Fig. 1
figure 1

Diagnostic algorithm for influenza A (H1N1)v pneumonia in critically ill patients

Full size image