Clinical and etiological characteristics of severe community-acquired pneumonia: A two-center retrospective analysis from Xi’an and Jiaxing in China

Abstract

Objective

This study was conducted to evaluate the baseline clinical profile, causative pathogens, and treatment patterns of severe community-acquired pneumonia across two medical centers (Xi’an and Jiaxing) from January 2023 to August 2025.

Methods

We extracted data regarding medical profile, demographic characteristics, comorbidities, laboratory parameters, treatments during hospitalization, and outcomes for these patients from electronic records. Pathogens were detected using a combination of sputum culture/throat-swab polymerase chain reaction with metagenomic next-generation sequencing or targeted next-generation sequencing.

Results

Among the 204 severe community-acquired pneumonia patients, majority were men (72.55%, 95% confidence interval: 66.05%–78.21%); the age range of the population was 65–79 years (41.18%, 95% confidence interval: 34.65%–48.03%), and chronic obstructive pulmonary disease was the predominant chronic pulmonary condition, present in 28.92% of the study participants. Laboratory evaluations revealed high illness severity scores (Acute Physiology and Chronic Health Evaluation II score: 19.00 (14.00, 24.00) and Sequential Organ Failure Assessment score: 7.00 (6.00, 10.00)). Pathogens were identified using metagenomic next-generation sequencing or targeted next-generation sequencing of blood/sputum/bronchoalveolar lavage fluid and polymerase chain reaction of sputum culture/throat-swab in 117 and 85 cases, respectively. Klebsiella pneumoniae was the most common pathogen, present in 20.10% of the participants (95% confidence interval: 15.18%–26.13%), followed by coronavirus 2019 (16.18%, 95% confidence interval: 11.76%–21.85%), and influenza A virus (12.75%, 95% confidence interval: 8.85%–18.02%). Most frequently administered medicines were β-lactam/β-lactamase-inhibitor combinations (89.22%) and carbapenems (60.78%). All patients received respiratory support, with 15.69% undergoing continuous renal-replacement therapy and 5.39% receiving extracorporeal membrane oxygenation. The in-hospital mortality rate was 38.73% (79/204, 95% confidence interval: 32.31%–45.56%).

Conclusion

Severe community-acquired pneumonia, which is most often caused by K. pneumoniae in our region, regularly complicates chronic obstructive pulmonary disease and other chronic pulmonary diseases, requires broad-spectrum antibiotics and life-support, and is associated with a relatively high mortality rate.

Introduction

Community-acquired pneumonia (CAP) remains the leading infectious cause of adult hospitalization and a major burden on healthcare resources worldwide. ¹ In the general population, the incidence of CAP is 1.5–16 per 1000 adults annually; however, it rises steeply with age, with the incidence being 18–44 per 1000 in those aged ≥65 years and up to 52 per 1000 in persons aged ≥85 years.^2–4 Following the relaxation of coronavirus disease 2019 (COVID-19) public health measures, surveillance programs have reported a 2–3-fold increase in non-COVID-19 CAP, highlighting the post-pandemic resurgence of bacterial and viral respiratory pathogens. ⁵ In addition to immediate financial costs, CAP prolongs hospital stay, accelerates long-term functional decline, and causes more deaths than any other infectious disease, accounting for approximately 2.5 million fatalities worldwide each year. ⁶

Although most cases respond to standard antibiotics, up to 25% patients progress to severe community-acquired pneumonia (SCAP), which is defined by respiratory failure/septic shock or more than two of the Infectious Diseases Society of America/American Thoracic Society (IDSA/ATS) minor criteria, including respiratory rate ≥30 breaths/min and partial pressure of oxygen (PaO₂)/inspired oxygen fraction (FiO₂) ≤250 mmHg.^7,8 In older individuals and those with underlying diseases, this complication is most common. Meanwhile, these patients with SCAP usually harbor a more complex pathogen spectrum, including drug-resistant bacteria infection or mixed viral–bacterial co-infections.^9,10 Therefore, SCAP often incurs higher hospitalization costs than non-severe pneumonia, and its 30-day mortality rate ranges from 17% to 49%, making it the leading contributor to pneumonia-related deaths and long-term functional impairment. ¹¹

Despite its clinical importance, multicenter data depicting the precise demographic profile, pathogen landscape, and real-world management for SCAP have been relatively limited over the past 3 years. In this study, we retrospectively analyzed 204 consecutive SCAP patients from two hospitals (January 2023 to August 2025) in whom pathogens were identified using metagenomic next-generation sequencing (mNGS)/ targeted next-generation sequencing (tNGS) or conventional cultures. The study objective was to outline current risk factors, etiological distribution, and treatment patterns of SCAP in the post-pandemic era.

Materials and methods

Study participants

We retrospectively screened all adult patients (≥18 years) with a primary diagnosis of SCAP at two tertiary hospitals (Xi’an People's Hospital (Xi’an Fourth Hospital) and First Hospital of Pinghu) from January 2023 to August 2025. The inclusion criteria were as follows: (a) diagnosis of CAP, with onset in the community setting, new pulmonary infiltrates visible on chest imaging, and at least two out of the following clinical features, including fever (axillary temperature ≥38°C) or hypothermia (<36°C), cough with or without sputum, pleuritic chest pain, dyspnea, and altered breath sounds on auscultation; (b) fulfillment of the SCAP criteria within 24 h of hospital admission, defined based on the presence of ≥1 major IDSA/ATS criterion (respiratory failure necessitating mechanical ventilation (invasive or non-invasive) or septic shock with systolic blood pressure (BP) <90 mmHg despite adequate fluid resuscitation plus need of vasopressors) or ≥2 minor IDSA/ATS criteria (respiratory rate ≥30 breaths/min, PaO₂/FiO₂ ≤250 mmHg, multilobar infiltrates on chest imaging, systolic BP <90 mmHg requiring fluid resuscitation, core temperature <36°C, blood urea nitrogen (BUN) ≥ 20 mg/dL, white blood cell (WBC) count <4 × 10⁹/L, platelet count (PLT) <100 × 10⁹/L, and altered mental status); and (c) complete electronic medical records and written informed consent provided by the patient or his/her surrogate. We excluded patients with hospital-acquired or healthcare-associated pneumonia; those whose primary diagnosis could not be clearly differentiated from heart failure, aspiration, or non-infectious pulmonary conditions; and those who were unwilling to participate. Patients were selected consecutively.

Data collection

We extracted data regarding demographic characteristics, comorbidities, vital signs, and laboratory values from the electronic medical record system for all 204 enrolled patients within the first 24 h of admission. Disease severity was evaluated using the Acute Physiology and Chronic Health Evaluation II (APACHE II) and Sequential Organ Failure Assessment (SOFA) scores, which were calculated at 24 h and 96 h post-enrollment. Microbiological documentation combined conventional sputum culture/throat-swab polymerase chain reaction (PCR) with m/tNGS of the blood, sputum, or bronchoalveolar lavage fluid (BALF). Information regarding the use of antimicrobial agents, respiratory support modalities (invasive mechanical ventilation (IMV), non-invasive mechanical ventilation (NIMV), and high-flow nasal cannula (HFNC)), renal-replacement therapy (CRRT), respiratory and circulatory replacement therapy (extracorporeal membrane oxygenation (ECMO)), and outcomes of all the patients were recorded until discharge or death. All patient data were anonymized and deidentified prior to analysis.

Pathogen identification

We adopted the following criteria based on clinical guidelines and previous literature to distinguish true pathogens from colonization or contamination.

Bacteria and fungi. An organism was considered a causative pathogen if it was isolated from a sterile site (e.g. blood) or when isolated from respiratory specimens (sputum, BALF, or tracheal aspirate), the patient presented with compatible clinical signs of pneumonia, and the treating physician judged the organism to be clinically significant and initiated pathogen-directed antimicrobial therapy. However, organisms isolated from upper respiratory tract solely without corresponding lower respiratory tract symptoms were not considered causative.

For viruses. A respiratory virus (including COVID-19, influenza A, adenovirus, and respiratory syncytial virus) was considered a true pathogen if it was detected using PCR or next-generation sequencing (NGS) from respiratory specimens in a patient with compatible clinical features (fever, cough, and dyspnea) and no alternative etiology adequately explained the clinical presentation.

For mixed infections. Patients in whom two or more pathogens were identified (e.g. bacterial and viral) were classified as having polymicrobial infection. All detected pathogens that met the above clinical criteria were included in the pathogen spectrum analysis.

For atypical pathogens. Detection of organisms such as Chlamydia psittaci or Mycoplasma pneumoniae using mNGS/tNGS or PCR was considered diagnostic when accompanied with a compatible clinical presentation, irrespective of culture results.

Exclusion of colonization/contamination. We excluded the following from our analysis: (a) isolation of coagulase-negative staphylococci, diphtheroids, or Bacillus species from respiratory specimens without clinical infection evidence and (b) Candida species detected in respiratory specimens (except Candida cultured from blood or sterile sites) as these are generally considered colonization in the absence of histopathologic evidence for invasive disease.

According to these criteria, the final proportion of each pathogen was calculated as (number of patients in whom the pathogen was identified as clinically relevant)/(204 cases) × 100%. Polymicrobial infections are common in SCAP; therefore, the sum of these individual proportions exceeds 100%.

Statistical analyses

All statistical analyses were performed using Microsoft Excel and the Statistical Package for the Social Sciences (SPSS) software (version 26.0). Categorical data, such as the number of patients with underlying diseases or medication usage distribution, are presented as frequencies and percentages (%). Normally distributed continuous data, including laboratory indicators, are expressed as mean ± SD. Non-normally distributed continuous data are expressed as median and interquartile range (IQR) values. The pie chart illustrating pathogen distribution was generated using Microsoft Excel. For between-group comparisons, continuous variables were analyzed using unpaired Student's t-test or Mann–Whitney U test, as appropriate, and categorical variables were compared using chi-square test or Fisher's exact test. A two-sided p-value <0.05 was considered statistically significant. Since no variable had >5% missing values, missing data were not imputed. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Results

Patient enrollment and exclusion

An initial screening identified 286 patients with a SCAP diagnosis across the two participating medical centers. After we excluded cases in which the primary etiology could not be reliably distinguished from heart failure, aspiration, or non-infectious pulmonary conditions, 227 patients remained eligible. A further 23 cases were subsequently removed from the analysis due to loss to follow-up, absence of written informed consent, or refusal to provide biological samples. The final analytic cohort included 204 patients (Figure 1). Among the 204 patients included in the final cohort, 189 (92.65%) were enrolled from Xi’an People's Hospital (Xi’an Fourth Hospital) and 15 (7.35%) from First Hospital of Pinghu. Given the small number of patients enrolled from the Pinghu center, between-center comparative analysis was not performed.

Figure 1.

Flowchart of patient screening and enrollment. In total, 286 patients with SCAP were initially screened across two tertiary hospitals. After excluding 59 patients whose primary etiology could not be clearly distinguished from heart failure, aspiration, or non-infectious pulmonary disorders, 227 patients remained eligible. A further 23 patients were excluded due to loss to follow-up (n = 3), lack of written informed consent (n = 13), or refusal to provide samples (n = 7), yielding a final cohort of 204 patients.

SCAP: severe community-acquired pneumonia.

Baseline clinical characteristics

Table 1 summarizes the baseline demographic and clinical features of the 204 enrolled patients. The majority were men (72.55%), and the largest age group was 65–79 years (41.18%). The most common chronic pulmonary comorbidity was chronic obstructive pulmonary disease (COPD, 28.92%), followed by bronchiectasis (5.88%) and lung cancer (2.94%). Turning to systemic conditions, hypertension ranked first (42.16%), with diabetes (21.08%), cerebrovascular disease (20.59%), chronic renal failure (18.63%), other systemic malignancies (9.31%), and immune system diseases (6.37%) also present. Notably, 11.27% of patients were on long-term, high-dose glucocorticoids or immunosuppressive agents for underlying conditions. Regarding the complications, acute myocardial injury and acute kidney injury (AKI) were the most common complications, affecting 50.98% and 35.29% of patients, respectively. Other complications included pneumothorax (2.94%) and pulmonary embolism (PE, 0.98%). The overall in-hospital mortality rate in our cohort was 38.73% (95% confidence interval (CI): 32.31%–45.56%). In univariate comparison between survivors and non-survivors, significant differences were observed in terms of male sex, COPD, invasive pulmonary mold disease, use of glycopeptide/oxazolidinone antibiotics, glucocorticoid use, vasoactive agent use, parenteral nutrition, and ECMO support (Supplementary Table 1).

Table 1.

Baseline characteristics of 204 patients with SCAP.

Characteristics on admission	Patients with SCAP (n %)
Age (years)
18–64	70; 34.31%
65–79	84; 41.18%
≥ 80	50; 24.51%
Male	148; 72.55%
Current smoker	53; 25.98%
Current alcoholic	30; 14.71%
Mode of admission
Emergency department	123; 60.29%
Internal medicine department	58; 28.43%
Surgical department	3; 1.47%
Outside hospital ICU	20; 9.80%
Underlying disease
COPD	59; 28.92%
Asthma	3; 1.47%
Lung cancer	6; 2.94%
Bronchiectasis	12; 5.88%
Hypertension	86; 42.16%
Cerebrovascular disease	42; 20.59%
Diabetes	43; 21.08%
Immune system diseases	13; 6.37%
Chronic renal failure	38; 18.63%
Chronic liver dysfunction	11; 5.39%
Other malignant conditions	19; 9.31%
Treatment with immunosuppressive agents	23; 11.27%
Complications
Acute myocardial injury	104; 50.98%
AKI	72; 35.29%
Pneumothorax	6; 2.94%
PE	2; 0.98%
Invasive pulmonary mold disease (Aspergillus or Mucorales)	28; 13.73%
In-hospital mortality	79; 38.73%

Data are presented as n (%); 95% confidence intervals for key proportions are provided in the main text.

ICU: intensive care unit; COPD: chronic obstructive pulmonary disease; AKI: acute kidney injury; PE: pulmonary embolism; SCAP: severe community-acquired pneumonia.

Laboratory findings

Table 2 presents key laboratory parameters of the cohort at admission. Patients exhibited elevated levels of inflammatory markers, including WBCs (WBC count: 10.46 (7.00, 15.46) × 10⁹/L, indicting leukocytosis), procalcitonin (PCT) (1.02 (0.30, 7.90) ng/mL), and C-reactive protein (CRP) (94.61 (27.00, 188.66) mg/L). Even with respiratory support, impaired oxygenation existed, reflected by a mean PaO₂ of 72.75 (59.00, 90.05) mmHg. Prolonged activated partial thromboplastin time (APTT, 32.90 (28.75, 38.15) s) and markedly elevated D-dimer levels (2550.0 (1470.0, 5420.0) μg/L) confirmed coagulation abnormalities. Additionally, liver and renal function indices indicated significant dysfunction: serum alanine aminotransferase (ALT) levels, 36.0 (15.0, 55.5) U/L; aspartate aminotransferase (AST) level, 38.0 (24.0, 70.75) U/L, albumin (ALB) level, 30.7 (26.9, 34.7) g/L, lactate dehydrogenase (LDH) level, 327.0 (232.0, 489.0) U/L; creatinine (Cr) level, 84.0 (61.95, 127.5) μmol/L; and BUN level, 8.92 (6.31, 14.13) mmol/L. Disease severity scores were high at admission (APACHE II 19.00 (14.00, 24.00) and SOFA 7.00 (6.00, 10.00)), with some improvement by 96 h (SOFA 6.00 (4.00, 8.00)). To address the clinical utility of biomarkers in guiding antibiotic therapy, we analyzed CRP and PCT levels stratified by etiological category (Supplementary Table 2). Among the 91 patients with confirmed bacterial infection, 30 (32.97%) had a PCT level ≤0.5 ng/mL, and 9 (9.89%) had a CRP level ≤10 mg/L. Conversely, among the 40 patients with isolated viral infection, 17 (42.50%) had a PCT level >0.5 ng/mL, and 34 (85.0%) had a CRP level >10 mg/L. These findings demonstrate substantial overlap between groups and suggest that neither biomarker alone can reliably distinguish bacterial from viral infection at the individual patient level.

Table 2.

Laboratory parameters of 204 patients with SCAP.

Laboratory parameters	Reference range	Patients with SCAP, median (IQR)
Vital signs
MAP (mmHg)	70–100	76.0 (65.0, 92.75)
Temperature (°C)	36.0–0.5	36.9 (36.5, 38.45)
HR (beats/min)	60–100	103.0 (80.0, 124.5)
RR (breaths/min)	12–20	21.0 (15.25, 29.0)
Routine blood test
WBC count (×109/L)	3.5–10.0	10.46 (7.00, 15.46)
Neutrophil count (×109/L)	1.8–7.5	9.28 (5.83, 13.61)
Lymphocyte count (×109/L)	1.0–4.0	0.55 (0.28, 0.98)
Monocyte count (×109/L)	0.2–0.8	0.33 (0.17, 0.65)
Eosinophil count (×109/L)	0.1–0.4	0.00 (0.00, 0.03)
Basophil count (×109/L)	0–0.1	0.01 (0.00, 0.03)
HGB level (g/L)	115–170	112.0 (94.0, 128.5)
PLT (×109/L)	100–300	174.5 (105.5, 233.0)
PCT level (ng/mL)	<0.5	1.02 (0.30, 7.90)
CRP level (mg/L)	<10	94.61 (27.00, 188.66)
Blood gas analysis (within 24 h after respiratory support)
pH	7.35–7.45	7.36 (7.29, 7.45)
PaO2 (mmHg)	80–100	72.75 (59.0, 90.05)
Coagulation index
APTT (s)	22–30	32.9 (28.75, 38.15)
PT (s)	10–14	13.30 (12.05, 14.70)
D-dimer level (ug/L)	<500	2550.0 (1470.0, 5420.0)
Liver function indices
Total bile acids level (µmol/L)	0–10	5.00 (3.00, 7.00)
Total bilirubin level (µmol/L)	3–21	16.00 (12.25, 20.00)
ALT level (U/L)	≤45	36.0 (15.0, 55.5)
AST level (U/L)	≤40	38.0 (24.0, 70.75)
ALB level (g/L)	35–50	30.7 (26.9, 34.7)
LDH level (U/L)	100–205	327.0 (232.0, 489.0)
ALP level (U/L)	30–120	84.0 (63.5, 124.0)
Renal function indices
Cr level (µmol/L)	62–115	84.0 (61.95, 127.5)
BUN level (mmol/L)	2.5–7.1	8.92 (6.31, 14.13)
UA level (µmol/L)	150–420	428.0 (258.0, 553.8)
SOFA score (within 24 h)	<2	7.00 (6.00, 10.00)
SOFA score (within 96 h)	<2	6.00 (4.00, 8.00)
SOFA-non-neuro score (within 24 h)	<2	6.00 (4.00, 8.00)
SOFA-non-neuro score (within 96 h)	<2	5.00 (3.00, 7.00)
APACHE II score (within 24 h)	<5	19.00 (14.00, 24.00)
APACHE II-non-neuro score (within 24 h)	<5	16.00 (12.50, 21.00)

Data are presented as median (IQR) values.

MAP: mean arterial pressure; HR: heart rate; RR: respiratory rate; WBC = White blood cell count; HGB: hemoglobin; PLT: platelet count; PCT: procalcitonin; CRP: C-reactive protein; APTT: activated partial thromboplastin time; PT: prothrombin time; ALT = alanine aminotransferase; AST = aspartate aminotransferase; ALB: albumin; LDH: lactate dehydrogenase; Cr: creatinine; BUN: blood urea nitrogen; UA: uric acid; SOFA: Sequential Organ Failure Assessment; APACHE II: Acute Physiology and Chronic Health Evaluation II; SCAP: severe community-acquired pneumonia; IQR: interquartile range; PAO₂: partial pressure of oxygen.

Etiological distribution

Pathogen identification was achieved via m/tNGS of blood/sputum/BALF (117 cases) or conventional culture/PCR (85 cases). In total, 117 patients (57.35%) were diagnosed based on the evaluation of lower respiratory tract specimens (76 BALF mNGS, 38 BALF tNGS, 2 blood mNGS, and 1 blood + BALF mNGS), whereas 85 (41.67%) were identified based on upper respiratory tract specimen evaluation; no pathogens were identified in the remaining 2 patients (0.98%). As illustrated in Figure 2, the top 10 pathogens were Klebsiella pneumoniae (41/204, 20.10%), COVID-19 (33/204, 16.18%), influenza A virus (26/204, 12.75%), Acinetobacter baumannii (24/204, 11.76%), Aspergillus fumigatus (20/204, 9.80%), Pseudomonas aeruginosa (19/204, 9.31%), Streptococcus pneumoniae (19/204, 9.31%), Escherichia coli (11/204, 5.39%), Hemophilus influenzae (11/204, 5.39%), and Staphylococcus aureus (11/204, 5.39%). Among the top 10 pathogens, A. fumigatus (80.0%, 16/20) and A. baumannii (75.0%, 18/24) exhibited the highest co-infection rates, whereas COVID-19 (45.5%, 15/33) and S. aureus (45.5%, 5/11) had the lowest rates (Figure 2(a)). Regarding infection patterns, bacterial mono-infection was the most common type (45.7%, 91/199), followed by bacterial–viral co-infection (23.1%, 46/199) and viral mono-infection (20.1%, 40/199); fungal or atypical pathogen mono-infections accounted for 11.0% (22/199) of all infections (Figure 2(b)).

Figure 2.

Pathogen distribution and co-infection patterns in 204 patients with SCAP. (a) Mono-infection versus co-infection proportions for the ten most frequently detected pathogens. For each pathogen, the orange segment represents single-infection, and the yellow segment represents co-infection. (b) Pie chart of overall infection categories. Patients were classified into four groups based on pathogen types detected: bacterial mono-infection (91, 45.7%), bacterial–viral co-infection (46, 23.1%), viral mono-infection (40, 20.1%), and fungal or atypical pathogen mono-infection (22, 11.0%). Percentages were calculated using the 199 pathogen-positive cases as the denominator.

COVID-19: coronavirus disease 2019; SCAP: severe community-acquired pneumonia.

Treatment modalities

Table 3 details the treatment strategies employed for all 204 patients with SCAP. β-lactam/β-lactamase inhibitor combinations (89.22%) and carbapenems (60.78%) were the most widely used antibiotic classes. Among antiviral agents, drugs for influenza (14.71%) and COVID-19 (11.76%) were used, along with broad-spectrum antiviral drugs (5.39%). Antifungal treatment included triazole (30.39%), echinocandin (10.29%), and polyene (3.92%) agents. Glucocorticoids were used in 73.04% of the patients. All patients required respiratory support, with IMV being the most common modality (85.78%), followed by HFNC (12.75%), NIMV (0.98%), and mask oxygen inhalation (0.49%). A subset of patients underwent continuous renal-replacement therapy (CRRT; 15.69%) or extracorporeal membrane oxygenation (ECMO; 5.39%).

Table 3.

Treatment methods employed for 204 patients with SCAP.

Treatment methods	Patients with SCAP (n %)
Antibiotics
Carbapenem antibiotics	124; 60.78%
Quinolone antibiotics	36; 17.65%
Glycopeptide antibiotic/Oxazolidinones	67; 32.84%
Aminoglycoside antibiotics	4; 1.96%
Cephalosporin	16; 7.84%
β-lactam/β-lactamase inhibitor combinations	182; 89.22%
Macrolides antibiotics	5; 2.45%
Polypeptide antibiotics	20; 9.80%
Glycylcycline antibiotics	46; 22.55%
Sulfonamide antimicrobials	22; 10.78%
Antiviral drugs
Antiviral drugs for influenza	30; 14.71%
Antiviral drugs for COVID-19	24; 11.76%
Broad-spectrum antiviral drugs	11; 5.39%
Antifungal agents
Echinocandin antifungals	21; 10.29%
Triazole antifungals	62; 30.39%
Polyene antifungal	8; 3.92%
Glucocorticoids	149; 73.04%
Immunosuppressants	7; 3.43%
Gastric mucosal protective agents	181; 88.73%
Vasoactive agents	122; 59.80%
Enteral nutrition	192; 94.12%
Parenteral nutrition	42; 20.59%
Respiratory support methods
IMV	175; 85.78%
NIMV	2; 0.98%
HFNC	26; 12.75%
Mask oxygen inhalation	1; 0.49%
CRRT	32; 15.69%
ECMO	11; 5.39%

IMV: invasive mechanical ventilation; NIMV: non-invasive mechanical ventilation; HFNC: high-flow nasal cannula; CRRT: continuous renal-replacement therapy; ECMO: extracorporeal membrane oxygenation; SCAP: severe community-acquired pneumonia; COVID-19: coronavirus disease 2019.

Discussion

In this two-center retrospective investigation, we characterized the clinical features, etiological profile, and therapeutic approaches employed in 204 patients with SCAP during the post-pandemic period. The cohort was predominantly composed of elderly men, with COPD emerging as the most frequently observed comorbidity. An in-hospital mortality of 38.73% was recorded, underscoring the considerable disease severity in this population.

The clinical profile of SCAP patients in this study, particularly the age distribution, male predominance, and high prevalence of COPD, are consistent with those reported in earlier investigations.^12,13 However, the mortality rate observed here (38.73%) is at the higher end of the range (17%–49%) reported in other studies.^7,11 This may be attributed to the advanced age of our cohort as well as the high incidence of complications such as acute myocardial injury (50.98%) and AKI (35.29%), factors known to worsen prognosis in SCAP.^9,10 Laboratory findings revealed marked inflammatory responses (elevated WBC count and increased PCT and CRP levels), impaired oxygenation, coagulation abnormalities (elevated D-dimer levels and prolonged APTT), and significant organ dysfunction, aligning with the systemic nature of severe infection described in prior research.^6,8 The high APACHE II and SOFA scores at admission further underscore the severity of illness in these patients, although some improvement was observed by 96 h, suggesting a response to initial therapeutic interventions.

K. pneumoniae was detected as the most prevalent pathogen (20.10%); thus, the etiological spectrum identified in our study contrasts with the epidemiological patterns described in many Western cohorts, wherein Streptococcus pneumoniae has been consistently identified as the predominant pathogen, especially in studies that have employed conventional culture methods.^3,14,15 However, similar findings have been reported in several studies from tropical and subtropical regions, including Southeast Asia and South Africa, wherein K. pneumoniae has emerged as a leading pathogen in CAP patients.^16,17 The high rate of viral pathogens (COVID-19 and influenza A virus) likely reflects the post-pandemic resurgence of respiratory viruses. ⁵ Our comprehensive approach, combining m/tNGS of the BALF/sputum/blood with conventional methods, likely improved detection yield compared with studies relying solely on culture or upper respiratory tract PCR. ² The high detection rate of K. pneumoniae and other Gram-negative pathogens supports the widespread broad-spectrum antibiotics use in our cohort, particularly β-lactam/β-lactamase inhibitor combinations (89.22%) and carbapenems (60.78%). These choices are generally appropriate and in line with guidelines for managing severe infections with risk factors for multidrug-resistant organisms.^18–20 The targeted use of antiviral and antifungal agents also appears justified. However, the high mortality rate suggests that initial empirical therapy requires optimization. The prominent role of K. pneumoniae might necessitate even more rapid deployment of combination therapy or newer antimicrobial agents (e.g. ceftazidime–avibactam) in high-risk patients, pending local antibiogram data. ² Furthermore, the high rate of viral co-infections underscores the importance of rapid multiplex PCR or NGS testing upon admission to guide appropriate and timely antiviral therapy. ²¹ In addition to antimicrobial optimization, improvements in supportive care remain central to better outcomes in SCAP. For instance, early referral to specialized centers equipped to provide venovenous ECMO may serve as an effective strategy for the most critically ill patients who fail to respond to conventional support measures. ²² Notably, several factors may explain why K. pneumoniae outnumbered S. pneumoniae in our cohort, including both a genuine regional shift and improved detection via molecular techniques. Carbapenems were used in 60.78% of the patients; this is justified by the high severity of illness and local resistance patterns, but also calls for antimicrobial stewardship. For empirical therapy, our data support initial coverage of K. pneumoniae combined with early molecular diagnostics to enable timely de-escalation.

Certain study limitations warrant careful consideration. The retrospective nature of the investigation introduces the potential for selection bias. Although several methods were used to enhance pathogen detection, the sensitivity and specificity variation across different testing platforms may have introduced some heterogeneity. Furthermore, as this study was conducted during a defined post-pandemic period, the generalizability of the findings to other populations may be limited. Additionally, vaccination data for influenza and pneumococcus were not available for our cohort, and the absence of this information limited our ability to assess its full impact on the pathogen distribution observed. Future prospective studies should incorporate systematic documentation of vaccination history to better address this confounder. Although we performed univariate comparisons, the relatively modest sample size and descriptive nature of this study precluded a robust multivariable analysis. We acknowledge this as a study limitation, and future larger-scale studies are needed to identify independent risk factors for mortality.

In conclusion, this study provides a contemporary overview of SCAP in the post-pandemic era. The high prevalence of Klebsiella pneumoniae and viral pathogens, along with the substantial disease burden and mortality, highlights the need of rapid diagnostic techniques and timely treatments. In future, prospective studies incorporating antimicrobial resistance data will be essential for further optimize management.