Remembering Jonathan M. Mann in a World Ajar

1181 LETTERS Her hemoglobin level had dropped further to 8.4 g/dL. Sputum culture yielded substantial growth of methi-cillin-sensitive Staphylococcus aureus and Pseudomonas aeruginosa. RT-PCR results of throat and nasal swabs were positive twice for coronavirus, but coronavirus cultures from these areas were negative. One month after onset, her coronavirus antibody titer was 200. In view of possible relapse of SARS, she was treated with oral ribavirin (1,200 mg/day) and lopinavir (133.3 mg/day)/ritonavir (33.3 mg) combination (3 capsules twice a day) in addition to intravenous piperacillin/tazobactam combination. The patient was afebrile, and symptoms improved 3 days after admission. Serial chest radiograph showed gradual resolution of shadowing. Subsequent RT-PCR and sputum culture were negative. This case illustrates several important issues regarding problems of infection control, diagnosis, and management of SARS. As the definition of SARS is nonspecific, patients with upper respiratory infection or community acquired pneumonia could be mislabeled as having SARS. Accommodating confirmed SARS patients and patients mislabeled as having SARS in the same facility may be disastrous. Unfortunately, isolating every single case is impossible, particularly when a large number of patients are admitted. Our patient may have acquired the disease after admission since she was placed in the same ward with other patients confirmed to have SARS. For this reason, special cohorting of SARS patients with closely related signs and symptoms should be strictly implemented at admission. Since fever is the most common feature of SARS, isolating febrile cases with respiratory or gas-trointestinal symptoms may be appropriate. Even patients with fever alone should be quarantined since the other symptoms of SARS may not be clinically obvious. Secondly, the sensitivity of diagnosing a coronavirus infection on admission is only 32% to 50% by nasopharyngeal RT-PCR test (3,4). Many infected cases will be missed as a result. Our patient may have had a relapse of disease during her second admission, although she had positive RT-PCR and antibody surge only 1 month after onset. However, we could not conclude whether the first RT-PCR on admission was a false negative or whether the patient acquired coronavirus infection in the hospital. Our study showed that sensitivity for diagnosing coronavirus infection could be increased by performing RT-PCR on samples from different parts of the body (4). Unfortunately, these samples were not taken from our patient. Furthermore, the chest infection with organisms recovered from her sputum could be the sole reason for her second admission, especially when …