社区获得性肺炎（Community-acquired Pneumonia, CAP）是威胁人类健康的重要疾病，检测技术的发展大大提高了CAP感染病原体的检出率，但仍约50%的CAP病例无法明确感染病原。病原体检出受多种因素影响，现有研究多集中在检测技术、病原体的复制特点和人体免疫防御等方面，对患者的就医行为、居住环境中污染物情况以及气候等对呼吸道多病原体检出的影响研究十分有限，缺乏基于多病原数据的影响因素系统性研究。本研究通过整合人口学、季节、环境污染物和气候因素、就诊前服用抗生素情况、疾病严重程度及病原体检出数据，分析影响呼吸道感染疾病严重程度以及呼吸道病原体检出的影响因素及关联强度，为完善与CAP疾病严重程度和病原体检出关联因素的认识提供数据参考。

本研究是基于既往多中心的肺炎病原学研究基础的回顾性分析，收集2014年1月1日至2019年12月31日间来自我国8个省份共9家哨点医院的3323例成人CAP住院患者信息，CAP临床诊断参考《美国传染病学会/美国胸科学会社区获得性肺炎指南》，按统一标准入组，入院后48小时内由临床医务工作者采集痰液，标本由本室统一开展病原核酸检测，包括流感病毒（甲、乙、丙型）、1-4型人副流感病毒、四种人冠状病毒、A和B型人偏肺病毒、呼吸道合胞病毒、鼻病毒、腺病毒、肠道病毒、副肠孤病毒、人博卡病毒、巨细胞病毒、肺炎支原体、肺炎衣原体、肺炎链球菌、流感嗜血杆菌、副溶血性嗜血杆菌、金黄色葡萄球菌、卡他莫拉菌、博多氏菌属、肺炎克雷伯杆菌、嗜肺军团菌、沙门氏菌以及耶氏肺孢子虫共33种常见呼吸道病原体。收集患者人口学指标、基础性疾病、临床诊疗和病原学结果，从国家气象局和中国环境监测总站收集各城市环境日值数据，包括温度、相对湿度（Relative Humidity, RH）、颗粒物（Particulate Matter，PM）2.5、PM10、二氧化硫、二氧化氮、一氧化碳以及8小时臭氧（8-hour Ozone，O₃-8h）。用logistics回归分析人口学、季节、环境、就医行为与CAP患者严重程度、病原检出结果间的关联，用SPSS 22.0进行统计分析，用R 4.0.2绘图，P<0.050为有统计学意义。

研究纳入分析的CAP住院患者中位年龄为58岁，四分位间距（Interquartile Range，IQR）为40-69，从症状出现到就医时间平均为7天（IQR 3-10）。其中1936名（58.3%）患者为男性，709名（21.3%）为重症CAP，550名（16.6%）患者超重（身体质量指数BMI≥25 kg/m²），782名（23.5%）入院前自行服用抗生素。2064名（62.1%）患者至少检出一种病原体，其中942名（28.3%）检出细菌（真菌），653名（19.7%）检出病毒，469名（14.1%）有两种及以上病原体混合检出。检出频次依次为肺炎支原体（12.2%）、流感病毒（11.1%）、流感嗜血杆菌（10.5%）、肺炎克雷伯杆菌（10.2%）、鼻病毒（9.9%）、肺炎链球菌（7.6%）、人冠状病毒（4.9%）、金黄色葡萄球菌（4.3%）、人副流感病毒（3.8%）、腺病毒（2.9%）、卡他莫拉菌（2.9%）、呼吸道合胞病毒（2.3%）、耶氏肺孢子虫（2.2%）、人偏肺病毒（2.2%）、嗜肺军团菌（1.1%）和副溶血性嗜血杆菌（1.0%），其他病原体的检出率小于1%。

对人口学指标的分析表明，男性患者病原检出率（60.4%）低于女性（63.9%）（c²=3.9，P=0.048；Adjusted Odds Ratio [aOR]=0.84，95%CI 0.72-0.98）；其中肺炎支原体（aOR=0.61，95%CI 0.48-0.78，P<0.001）与女性呈正向关联，而肺炎克雷伯杆菌（c²=9.9，P=0.002；aOR=1.35，95%CI 1.05-1.75）和肺炎链球菌（c²=11.7，P=0.001；aOR=1.56，95%CI 1.16-2.08）在男性中检出率更高；肺炎支原体在小年龄组高检出（aOR=0.83，95%CI 0.80-0.86，P<0.001），中位年龄为33岁（IQR 26-53），而流感嗜血杆菌、肺炎克雷伯杆菌、肺炎链球菌、流感病毒随着年龄增长检出率增加（P<0.050）；超重患者流感病毒检出率（15.1%）显著高于非超重患者（10.2%）（c²=10.9，P=0.001；aOR=1.40，95%CI 1.06-1.84）。

CAP病原检出阳性患者发生重症的比率更高（c²=8.1，P=0.004）；年龄每增加5岁、男性、入院前有抗生素服用史与重症CAP呈显著关联（P<0.001）。在重症CAP患者中部分病原体检出率高于非重症患者，包括肺炎克雷伯杆菌（c²=37.3，P<0.001）、流感病毒（c²=8.2，P=0.004）、金黄色葡萄球菌（c²=18.8，P<0.001）、人冠状病毒（c²=5.5，P=0.019）、耶氏肺孢子虫（c²=16.6，P<0.001）和人巨细胞病毒（c²=5.4，P=0.020）。除了肺炎克雷伯杆菌、流感病毒和人冠状病毒，重症患者中还高检出肺炎链球菌、流感嗜血杆菌和鼻病毒，该病原检测组合联合PSI得分（Z=2.4，P=0.017）或者病原检测组合和性别联合CURB-65评分系统（Z=5.5，P<0.001）可以增加CAP患者疾病严重程度的预测效能。

气候因素中，RH每增加10%，与病原检出关联强度增加9%（aOR=1.09，95%CI 1.02-1.15，P=0.010），与病毒检出关联强度增加14%（aOR=1.14，95%CI 1.05-1.23，P=0.002）；对环境污染物的分析表明，PM2.5每升高10μg/m³，与病原检出关联强度增加0.09倍（aOR=1.09，95%CI 1.03-1.15，P=0.005），与病毒检出关联强度增加0.12倍（aOR=1.12，95%CI 1.04-1.20，P=0.002），与流感病毒检出关联显著（aOR=1.17，95%CI 1.08-1.28，P<0.001）；PM10浓度升高与重症CAP关联（P=0.044），与肺炎支原体检出正向关联（aOR=1.08，95%CI 1.00-1.15，P=0.042）；细菌（真菌）月检出率与O₃-8h浓度正相关（P=0.041），O₃-8h浓度升高显著增加肺炎克雷伯杆菌的检出（aOR=1.11，95%CI 1.05-1.18，P=0.001）。

对患者的就医时间进行分析，发现症状出现到就医时间≤2天、3-7天、＞7天病原检出率分别为72.7%、62.3%和56.3%（c²=43.5，P<0.001），就诊的间隔时间每减少1天，与细菌（真菌）（aOR=0.96，95%CI 0.93-0.98，P<0.001）和混合检出的（aOR=0.95，95%CI 0.92-0.97，P<0.001）关联更强，与病毒检出无显著关联，流感病毒除外（aOR=0.96，95%CI 0.93-0.99，P=0.005）。

本研究首次基于CAP病原学数据系统分析了空气污染物（PM2.5、PM10、O₃-8h、二氧化硫、二氧化氮、一氧化碳）、气候因素及入院前抗生素服用史、就诊间隔时间等与病原检出和疾病严重程度的关联。发现基于特定病原组合联合PSI和CURB-65的诊断可以提高CAP患者疾病严重程度的预测效能；此外，环境因素主要与病原检出结果关联，RH的增加与病毒正向关联，PM2.5浓度的增加与病毒尤其是流感病毒检出正向关联，O₃-8h浓度升高与细菌（真菌）检出率正相关，尤其与肺炎克雷伯杆菌的检出显著关联，PM10浓度的升高与肺炎支原体检出正向关联。上述环境变量的升高对增加肺炎链球菌、流感嗜血杆菌和鼻病毒检出的影响均不显著。研究完善了病原检出结果对CAP重症预测以及气候、环境污染物等因素对CAP病原检出影响的认识。

Community-acquired pneumonia (CAP) is an important disease that threatens human health. The development of detection technology has greatly improved the detection rate of infectious pathogens of CAP. But there are still about 50% of CAP cases which cannot be identified the infectious pathogen. The detection of pathogens is affected by varieties of factors. Existing studies mostly focus on detection technology, replication characteristics of pathogens, immune defense in human, and so on. But, the studies focused on the impact of patients' medical behavior, climate and pollutants in the living environment on the detection of multiple pathogens in respiratory tract are very limited. There is a lack of systematic research on the associated factors based on multi-pathogen detected results. By integrating demographic, seasons, environmental pollutants, climatic factors, antibiotics preadmission, disease severity and pathogen detected information, this study analyzed the associated factors and associated intensity for the severity of respiratory-infection disease and the detection results of respiratory pathogens. It’s aimed to provide data reference for improving the understanding of associated factors on CAP severity and pathogen detection.

This study is based on the retrospective analysis of previous multi-center pneumonia etiology research, collecting the information of 3323 adult inpatients with CAP from a total of 9 sentinel hospitals in 8 provinces in China from January 1, 2014 to December 31, 2019. The clinical diagnosis of CAP referred to the 2007 Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. All patients were enrolled according to the unified standard and sputum was collected by clinical medical staff within 48 hours after admission. Sputum were uniformly tested by pathogenic nucleic acid by our laboratory. 33 common respiratory pathogens included influenza viruses (A, B, C), human parainfluenzae viruses types 1-4, four human coronaviruses, human metapneumovirus types A and B, respiratory syncytial virus, rhinovirus, adenovirus, enterovirus, parechovirus, human bocavirus, cytomegalovirus, mycoplasma pneumoniae, chlamydia pneumoniae, streptococcus pneumoniae, haemophilus influenzae, haemophilus parahaemolyticus, staphylococcus aureus, moraxella catarrhalis, bordetella, klebsiella pneumoniae, legionella pneumophila, salmonella and pneumocystis jirovecii. Patients’ demographic indicators, underlying diseases, clinical diagnosis and treatment and etiological results were collected. Daily environmental data of each city were collected from the National Meteorological Administration and China National Environmental Monitoring Centre, including temperature, relative humidity (RH), particulate matter (PM) 2.5, PM10, sulfur dioxide, nitrogen dioxide, carbon monoxide, and 8-hour ozone (O₃-8h). The associations of factors including demography, seasons, environment factors, medical behaviors with severity and pathogen detection results of CAP patients were analyzed by logistics regression. Statistical analysis was performed with SPSS 22.0, and R 4.0.2 was mainly used for drawing. P<0.050 was considered statistically significant.

The median age of hospitalized patients with CAP included in the study was 58 years, and the Interquartile Range (IQR) was 40 to 69. The mean time-period from symptoms onset to admission was 7 days (IQR 3 to 10). Among them, 1936 (58.3%) patients were male, 709 (21.3%) patients were diagnosed as severe CAP, 550 (16.6%) patients were overweight (Body mass index, BMI≥25 kg/m²), and 782 (23.5%) were taking antibiotics preadmission. There were one or more pathogens detected in 2064 (62.1%) patients, including 942 (28.3%) patients with bacteria (fungi), 653 (19.7%) patients with virus, and 469 (14.1%) patients with mixed detection. The frequency of detection was sorted by mycoplasma pneumoniae (12.2%), influenza viruses (11.1%), haemophilus influenzae (10.5%), klebsiella pneumoniae (10.2%), rhinovirus (9.9%), streptococcus pneumoniae (7.6%), human coronavirus (4.9%), staphylococcus aureus (4.3%), human parainfluenza virus (3.8%), adenovirus (2.9%), moraxella catarrhalis (2.9%), respiratory syncytial virus (2.3%), pneumocystis jirovecii (2.2%), human metapneumovirus (2.2%), legionella pneumophila (1.1%) and haemophilus parahaemolyticus (1.0%). And detection rates of other pathogens less than 1%. 

The analysis of demographic indicators showed that the pathogen detection rate of male patients (60.4%) was lower than that of female patients (63.9%) (c²=3.9, P=0.048; Adjusted Odds Ratio [aOR]=0.84, 95%CI 0.72-0.98). Mycoplasma pneumoniae (aOR=0.61, 95%CI 0.48-0.78, P<0.001) was positively associated with female, while klebsiella pneumoniae(c²=9.9, P=0.002; aOR=1.35, 95%CI 1.05-1.75) and streptococcus pneumoniae (c²=11.7, P=0.001; aOR=1.56, 95%CI 1.16-2.08) had higher detection rates in male. Mycoplasma pneumoniae was detected in smaller age groups (aOR=0.83, 95%CI 0.80-0.86, P<0.001), with a median age of 33 years (IQR 26-53). While detection of haemophilus influenzae, klebsiella pneumoniae, streptococcus pneumoniae, and influenza viruses increased with age (P<0.050). The detection rate of influenza virus in overweight patients (15.1%) was significantly higher than that in non-overweight patients (10.2%) (c²=10.9, P=0.001; aOR=1.40, 95%CI 1.06-1.84). 

Patients with positive CAP pathogen detection had a higher rate of severe CAP (c²=8.1, P=0.004). Each 5-year increase in age, male, and history of antibiotics preadmission were significantly associated with severe CAP (P<0.001). The detection rate of some pathogens in patients with severe CAP was higher than that in non-severe patients, including klebsiella pneumoniae (c²=37.3, P<0.001), influenza viruses (c²=8.2, P=0.004), staphylococcus aureus (c²=18.8, P<0.001), human coronaviruses (c²=5.5, P=0.019), pneumocystis jirovecii (c²=16.6, P<0.001) and cytomegalovirus (c²=5.4, P=0.020). In addition to klebsiella pneumoniae, influenza viruses, and human coronaviruses, streptococcus pneumoniae, haemophilus influenzae, and rhinovirus were also highly detected in severe patients. And detection of the grouped pathogens combined with PSI score (Z=2.4, P=0.017), or detection of the grouped pathogens and gender combined with CURB-65 score (Z=5.5, P<0.001) could increase the prediction of disease severity in CAP patients.

Among the climatic factors, for each 10% increment in RH, the associated strength with pathogen detection increased by 9% (aOR=1.09, 95%CI 1.02-1.15, P=0.010), and with virus detection increased by 14% (aOR=1.14, 95%CI 1.05-1.23, P=0.002). The analysis of environmental pollutants showed that for every 10 μg/m³ increment in PM2.5, the associated strength with pathogen detection increased by 0.09 times (aOR=1.09, 95%CI 1.03-1.15, P=0.005), and with virus detection increased by 0.12 times (aOR=1.12, 95%CI 1.04-1.20, P=0.002), significantly associated with influenza virus detection (aOR=1.17, 95%CI 1.08-1.28, P<0.001). Increment in PM10 concentration was associated with severe CAP (P=0.044), and positively associated with mycoplasma pneumoniae (aOR=1.08, 95%CI 1.00-1.15, P=0.042). The monthly detection rate of bacteria (fungi) was positively correlated with O₃-8h concentration (P=0.041), and the detection of klebsiella pneumoniae increased significantly for increment in O₃-8h concentration (aOR=1.11, 95%CI 1.05-1.18, P=0.001). 

In the perspective of analyzing medical behavior of patients, the pathogen detected rates of time-period from symptoms onset to admission with ≤2 days, 3-7 days and >7 days were 72.7%, 62.3% and 56.3%, respectively (c²=43.5, P<0.001). For each decrease of 1 day in time-period from symptoms onset to admission, there were more strongly associated with bacteria (fungi) (aOR=0.96, 95%CI 0.93-0.98, P<0.001) and mixed detection (aOR=0.95, 95%CI 0.92-0.97, P<0.001), not significantly associated with virus detection, except for influenza viruses (aOR=0.96, 95%CI 0.93-0.99, P=0.005).

Firstly based on CAP etiological data, this study systematically analyzed the association between air pollutants (PM2.5, PM10, O₃-8h, sulfur dioxide, nitrogen dioxide, carbon monoxide), climatic factors, history of antibiotics preadmission, time-period from symptoms onset to admission, and pathogen detection results, CAP severity. This study found that the combination of PSI and CURB-65 based on the combination of detecting specific pathogens could improve the prediction of disease severity in CAP patients. In addition, environmental factors were mainly associated with pathogen detection results. The increase of RH was positively associated with the detection of viruses. The increase of PM2.5 concentration was positively associated with the detection of viruses, especially influenza viruses. The increase in O₃-8h concentration was positively correlated with the detection rate of bacteria (fungi), especially with the detection of klebsiella pneumoniae. And the increase in PM10 concentration was positively associated with the detection of mycoplasma pneumoniae. But there was no significant association between increment of these environmental variables and increased detection of streptococcus pneumoniae, haemophilus influenzae, and rhinovirus. This study improved the understanding of the effect of detected pathogens on predicting disease severity in CAP patients, and the learning of association of climate, environmental pollutants and other factors on the detected results of CAP pathogens.