Abstract
OBJECTIVE: To analyze the clinical characteristics and risk factors for death of severe pneumonia (SP) in adults and explore the application value of metagenomic next-generation sequencing in the detection of pathogens. METHODS: A total of 132 adult patients with SP admitted from May 2021 to October 2023 were selected. Data on gender, age, smoking, underlying diseases, laboratory tests and prognosis were collected. BALF samples were sent for mNGS, smear-stained microscopy and culture. Meanwhile, conventional methods were used for pathogen detection of blood, urine and throat swab specimens. The detection efficiencies of different methods were compared and the associated pathogen profiles were analyzed. RESULTS: Among the 132 patients, there were 92 males and 40 females, with a total of 52 deaths. Age≥65 years, heart failure, renal insufficiency, positive of COVID-19, use of vasoactive drugs, use of mechanical ventilation and use of CRRT were statistically different between the survivors and non-survivors. Heart failure (OR=4.751) and use of mechanical ventilation (OR=11.914) were risk factors of SP mortality. The bacteria detected were mainly Klebsiella pneumoniae, Acinetobacter baumannii and Pseudomonas aeruginosa. The fungi detected were mainly Candida and Aspergillus. The viruses detected were mainly COVID-19 and influenza virus. The positive rate of mNGS was higher than conventional methods in both bacteria, fungus and virus (82.58% vs 63.64%, 50.76% vs 37.88% and 67.42% vs 37.88%, respectively) (P<0.05). The sensitivity and accuracy of mNGS in bacterial detection were significantly higher than traditional methods (P<0.05). Compared to culture, mNGS detected more Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus influenzae and Escherichia coli, and had a significant advantage in the detection of Mycobacterium tuberculosis complex, Nontuberculous mycobacterial, Legionella pneumophila, Chlamydia psittaci, Pneumocystis jirovecii and Aspergillus. Moreover, mNGS can better indicate mixed infections of bacteria, viruses, or fungi. CONCLUSION: Elderly people with chronic diseases were the main group of severe pneumonia in adults. The pathogenic microorganisms that caused SP are complex, and mixed infection is common. mNGS enhanced the effectiveness of pathogen detection, makes up for the shortcomings of conventional methods, especially in identifying unexpected pathogens, and provides a new means for early targeted anti-infection treatment.