Distinction of clinical features and microbiological methods between and pneumonia confirmed by metagenomic next-generation sequencing.

Objectives: Detection and diagnosis of () pneumonia is often overlooked due to conventional methods limitations and similarity to other atypical community acquired pneumonia (CAP). Using mNGS, we aimed to distinguish psittacosis from legionellosis for early pneumonia diagnosis and better prognosis.

Methods: Thirty-seven patients diagnosed with atypical CAP were enrolled in this retrospective study, including 14  pneumonia and 23 () pneumonia. We collected and compared baseline, lab results, radiology imaging, conventional microbiological methods and more importantly, mNGS results of clinical samples, as well as the treatments and prognosis between psittacosis and legionellosis.

Results: Patients with and had similar symptoms and were presented with high levels of inflammatory markers. However, patients with pneumonia were more likely to have exposure to birds or parrots [11 (78.6%) vs. 2 (8.7%),  < 0.001], had higher proportions of fever and chill ( = 0.015 and 0.035), higher levels of hemoglobin and albumin ( = 0.002 and 0.018) compared with those with . Of 14  patients, only one had positive IgM antibody, with no positive cultures. Early identification of pathogens by mNGS method contributed to timely antibiotics' adjustment and better outcomes then, yet with similar hospital mortality between two groups [7.1% (1/14) vs. 34.8% (8/23),  = 0.112].

Conclusion: Early mNGS detection of atypical pathogens in multiple samples improves on traditional methods, promptly adjust empirical antimicrobial treatment to pathogen-targeted antibiotics, further improve prognosis.