Abstract
Introduction
Hospital-acquired pneumonia (HAP) is a major type of hospital-acquired infection with substantial morbidity. Large-scale epidemiological data on HAP in China remain limited. This multi-center retrospective study investigates the epidemiology, etiology, and risk factors of HAP to support clinical management.
Methods
From January to December 2019, 543 patients from 37 tertiary hospitals meeting the enrollment criteria (who meet the diagnosis of HAP according to the Guidelines for the Diagnosis and Treatment of HAP in Adults in China and have clear clinical outcomes) were included. Patient data were retrospectively analyzed.
Results
The study recorded 15.8% overall mortality. Multivariate analysis identified independent risk factors including secondary ventilator-associated pneumonia, admission to the intensive care unit, positive airway secretion culture, liver dysfunction, heart failure, cancer, long-term catheter indwelling, sequential organ failure assessment (SOFA) score >2 and quick SOFA (qSOFA) score >1 (p < 0.05). The top four pathogens of HAP were Acinetobacter baumannii (24.89%), Klebsiella pneumoniae (21.57%), Pseudomonas aeruginosa (13.73%), and Staphylococcus aureus (5.88%).
Conclusion
Given the high mortality rate associated with HAP, prevention and early diagnostic assessment of HAP during hospitalization are paramount for these patients. The main pathogens of HAP are Gram-negative bacteria, which frequently exhibit substantial antimicrobial resistance. Therefore, antibiotics should be selected carefully.
Keywords: Hospital-acquired pneumonia, Epidemiology, Risk factors
Introduction
Hospital-acquired pneumonia (HAP) represents a leading type of nosocomial infections, marked by high mortality and considerable management challenges [1]. Globally, accumulating evidence-based data on HAP epidemiology, etiology, diagnosis, and treatment have advanced understanding. Consequently, the “Guidelines for the Diagnosis and Treatment of HAP in Adults in China” were updated in 2018 [1], further standardizing the diagnosis and management protocols of HAP.
HAP demonstrates distinct pathogen distributions and mortality patterns compared to community-acquired and other nosocomial infections, with substantial geographical heterogeneity [2, 3]. Reported mortality ranges from 13% to 30% in Western countries [4, 5], while Chinese single-center studies document rates between 4.5% and 22.6% [6, 7]. Although Gram-negative bacteria predominate among HAP pathogens, regional variations exist. P. aeruginosa is the most frequently isolated pathogen in Western nations, whereas A. baumannii is most common in China [8]. Notably, poor prognosis in HAP is multifactorial, with established factors including advanced age (≥70 years), prolonged hospitalization, and significant comorbidity burden. Evidence suggests that evidence-based interventions targeting modifiable predictors can reduce HAP incidence and improve outcomes [9, 10]. Large-scale epidemiological data on HAP in China are scarce, with most evidence derived from single-center studies. Comprehensive characterization of China’s HAP pathogen profile – including its temporal evolution and risk factors for antimicrobial-resistant and atypical pathogens – are imperative for optimizing prevention, diagnosis, and therapeutic management. This retrospective investigation characterizes the epidemiological and etiological landscape of HAP during the pre-coronavirus disease 2019 (COVID-19) period, while systematically evaluating prognostic determinants. The study ultimately aimed to establish an evidence-informed framework for early preventive strategies and targeted therapeutic interventions to enhance clinical management.
Methods
Study Participants
This multicenter retrospective observational study enrolled 628 HAP patients admitted to 37 tertiary hospitals across China between January and December 2019. A total of 543 participants were ultimately included following application of inclusion criteria and exclusion criteria mentioned below (shown in Fig. 1).
Fig. 1.
STROBE diagram for various phases of the study. Flowchart of the inclusion of patients presenting with hospital acquired pneumonia (HAP). STROBE, strengthening the reporting of observational studies in epidemiology.
HAP diagnosis was established according to the 2018 Chinese HAP Guideline [1]: chest imaging (X-ray or computed tomography) could show a new or progressive lung infiltrate, consolidation or ground-glass pattern, accompanied by at least two of the following clinical manifestations: (a) fever (body temperature > 38°C); (b) purulent respiratory secretions; (c) leukocytosis (>10 × 109/L) or leukopenia (<4 × 109/L). Additionally, all patients provided etiological diagnosis: (a) positive culture from lower respiratory tract secretions (neutrophil count >25 cells/low-power field, epithelial cell count <10 cells/low-power field, or neutrophil/epithelial count ratio >2.5:1), protected specimen brush, bronchoalveolar lavage, lung tissue, or sterile body fluids, with consistency to clinical manifestations; (b) fungi were identified via pathological, cytological, or direct microscopic examination of lung tissue, accompanied by relevant evidence of tissue damage; (c) positive results of serum IgM antibodies against atypical pathogens or viruses, or a fourfold or greater increase in serum IgG antibody titers between acute and convalescent phase; positive results for viral antigens, nucleic acid tests, or viral cultures in respiratory secretions.
Exclusion criteria included (1) patients aged <18 years; (2) cases without clear clinical outcome or with substantial missing data; (3) duplicate strains from the same patient during the same infection episode, derived from the same specimen source. Different episodes of infection in the same patient were considered independent cases. Ventilator-associated pneumonia (VAP) is defined as pneumonia occurring either in patients with tracheal intubation or tracheostomy receiving mechanical ventilation for more than 48 h, or developing within 48 h following the discontinuation of mechanical ventilation or removal of the artificial airway.
Setting and Design
Comprehensive patient data collection encompassed demographic characteristics (sex, age), medical history, pre-admission medication use, diagnostic investigations (blood routine examination, infection-related biomarkers, biochemical profiles, arterial blood gases), chest imaging, etiological diagnosis, and therapeutic regimens including antibiotic protocols and respiratory support strategies. Clinical outcomes were categorized into three distinct tiers – “effective,” “ineffective,” and “death” – based on standardized evaluation criteria, as detailed in Table 1.
Table 1.
Clinical outcome definitions
| Clinical outcomes | Definition |
|---|---|
| Effective | Compared with the baseline, clinical symptoms and signs were improved or even completely disappeared. Lung infiltrate, consolidation, or ground-glass opacities showed radiographic resolution. Infection-related laboratory markers (e.g., leukocyte count, neutrophil percentage, CRP, PCT) normalized or trended toward normal |
| Ineffective | Compared with the baseline, there was either persistence or worsening of clinical symptoms and signs. Lung infiltrates, consolidation, or ground-glass opacities showed no improvement or even progressed. There was no improvement or deterioration in infection-related laboratory markers. Lower respiratory tract specimen cultures were still positive, requiring escalation of antibiotic therapy |
| Death | The patient has died |
CRP, C-reactive protein; PCT, procalcitonin.
Clinical Scores
Disease severity was quantified using sequential organ failure assessment (SOFA) and quick SOFA (qSOFA) scoring systems. qSOFA comprises altered mental status, systolic blood pressure ≤100 mm Hg, and respiratory rate ≥22 breaths/min; patients meeting two or more criteria (qSOFA score ≥2) are considered suspected sepsis [11]. The SOFA score is derived from an evaluation of six critical organ systems: respiratory, cardiovascular, hepatic, coagulation, renal, and neurological. For each system, clinical and laboratory parameters are assessed, and the maximum score obtained for each is aggregated to determine the overall SOFA score (missing values were classified as normal per SOFA scoring criteria [12]). This composite score ranges from 0 to 24, with higher scores indicating a greater severity of organ dysfunction (range from 0 to 24, low risk group: 0–1, medium risk group: 2–7, high risk group: ≥8 [13]).
Statistical Analysis
SPSS 27.0 statistical software (IBM SPSS, Chicago, IL) was used for statistical analysis. The categorical variables were presented as the number of cases (n) or/and percentage (%). Independent sample t tests and chi-square tests were used to determine if any demographic and clinical characteristics differed significantly between different groups. Risk factors were analyzed by multivariate logistic regression analysis. A 2-sided p value <0.05 was considered statistically significant.
Results
Study Population
Of the 543 patients, 94 (17.3%) were clinically suspected of having VAP, and 70.2% were male. The mean age was 65.74 ± 16.21 years (range: 19–99 years). The mortality of enrolled patients was 15.8% (n = 86). A total of 150 patients (27.6%) had been hospitalized within 90 days prior to admission, with a mean duration of 12.62 days.
Patients were stratified into two groups according to clinical outcomes: effective (n = 400, 73.7%) and ineffective-death (n = 143, 26.3%). The ineffective-death group exhibited significantly higher rates of VAP, male sex, intensive care unit (ICU) admission, new progressive lung infiltrates on imaging, abnormal temperature (>38°C or <36°C), positive airway secretion cultures, liver dysfunction (alanine aminotransferase [ALT] or aspartate aminotransferase [AST] >40 IU/L), heart failure, cerebral vascular disease, cancer, hospitalization, antibiotic use, and ventilation within 90 days before admission compared to the effective group (all p < 0.05) (shown in Table 2).
Table 2.
Comparison of clinical characteristics between patients with different clinical outcomes
| Variable | Clinical outcomes | p value | |
|---|---|---|---|
| effective, n = 400 | ineffective-death, n = 143 | ||
| Male, n (%) | 271 (67.8) | 110 (76.9) | 0.040 |
| ICU admission, n (%) | 214 (53.5) | 123 (86.0) | <0.001 |
| Age, years | |||
| Mean | 65.55±16.00 | 66.58±16.83 | F = 0.422 |
| p = 0.516 | |||
| 18–45 years | 41 (10.3) | 16 (11.2) | 0.367 |
| 46–65 years | 146 (36.5) | 41 (28.7) | |
| 66–80 years | 143 (35.7) | 55 (38.4) | |
| >80 years | 70 (17.5) | 31 (21.7) | |
| Quarters, n (%) | |||
| Quarters 1st | 94 (23.5) | 29 (20.3) | 0.861 |
| Quarters 2nd | 95 (23.8) | 37 (25.9) | |
| Quarters 3rd | 110 (27.5) | 39 (27.3) | |
| Quarters 4th | 101 (25.2) | 38 (26.5) | |
| Districts, n (%) | |||
| Northern region | 171 (42.8) | 57 (39.9) | 0.123 |
| Eastern region | 174 (43.5) | 56 (39.2) | |
| Central-Western region | 55 (13.7) | 30 (20.9) | |
| Clinical symptom, n (%) | |||
| New or progressive infiltrates, consolidations, or ground-glass opacities on X-ray/CT | 358 (89.5) | 137 (95.8) | 0.022 |
| Body temperature >38°C or <36°C | 294 (73.5) | 117 (81.8) | 0.047 |
| Leukocyte count >10 × 109/L or <4 × 109/L | 316 (79.0) | 121 (84.6) | 0.146 |
| Purulent discharge in the trachea-bronchi | 298 (74.5) | 118 (82.5) | 0.052 |
| Positive culture of airway secretions | 282 (70.5) | 128 (89.5) | <0.001 |
| Positive smear of airway secretions | 168 (42.0) | 60 (42.0) | 0.993 |
| Evidence of fungal presence and tissue damage on pathology/cytopathology or direct microscopy of lung tissue specimens | 8 (2.0) | 4 (2.8) | 0.578 |
| Low albumin level (<35 g/L) | 221 (55.3) | 93 (65.0) | 0.042 |
| Liver dysfunction (ALT >40 IU/L or AST >40 IU/L) | 113 (28.3) | 69 (48.3) | <0.001 |
| Comorbidities, n (%) | |||
| VAP | 54 (13.5) | 40 (28.0) | <0.001 |
| Diabetes | 79 (19.8) | 28 (19.6) | 0.965 |
| Heart failure | 19 (4.8) | 16 (11.2) | 0.007 |
| Hypertension | 162 (40.5) | 55 (38.5) | 0.669 |
| Coronary heart disease | 58 (14.5) | 25 (17.5) | 0.395 |
| Cerebral vascular disease | 121 (30.3) | 33 (23.1) | 0.102 |
| Cancer | 78 (19.5) | 43 (30.1) | 0.009 |
| Lung cancer | 22 (5.5) | 7 (4.9) | 0.803 |
| Renal insufficiency | 33 (8.3) | 8 (5.6) | 0.302 |
| COPD | 40 (10.0) | 17 (11.9) | 0.688 |
| Chronic interstitial lung disease | 17 (4.3) | 6 (4.2) | 0.978 |
| Pre-infection healthcare interventions (within 90 days before admission), n (%) | |||
| Surgery | 143 (35.8) | 53 (37.1) | 0.779 |
| Hospitalization pre-treatment | 95 (23.8) | 55 (38.5) | <0.001 |
| Antibiotic pre-treatment | 342 (85.5) | 133 (93.0) | 0.020 |
| Steroid pre-treatment | 27 (6.8) | 16 (11.2) | 0.092 |
| Radiotherapy pre-treatment | 20 (5.0) | 6 (4.2) | 0.699 |
| Ventilation pre-treatment | 22 (5.5) | 17 (11.9) | 0.011 |
| qSOFA | |||
| Mean | 0.63±0.73 | 1.01±0.88 | F = 26.970 |
| p < 0.001 | |||
| 0–1 | 356 (89.0) | 101 (70.6) | <0.001 |
| 2–3 | 44 (11.0) | 42 (29.4) | |
| SOFA | |||
| Mean | 3.34±2.47 | 5.30±2.89 | F = 60.517 |
| p < 0.001 | |||
| 0–1 | 100 (25.0) | 11 (7.7) | <0.001 |
| 2–7 | 275 (68.8) | 100 (69.9) | |
| ≥8 | 25 (6.2) | 32 (22.4) | |
| Duration of mechanical ventilation, days | |||
| Mean | 22.94±34.79 | 26.74±34.65 | F = 0.992 |
| p = 0.320 | |||
| 0 days | 149 (37.2) | 19 (13.3) | <0.001 |
| 1–14 days | 132 (33.0) | 54 (37.8) | |
| 15–28 days | 69 (17.3) | 32 (22.3) | |
| 29–56 days | 29 (7.3) | 27 (18.9) | |
| >56 days | 21 (5.2) | 11 (7.7) | |
| Duration of invasive mechanical ventilation, days | |||
| Mean | 17.12±29.86 | 23.25±34.71 | F = 2.665 |
| p = 0.104 | |||
| 0 days | 198 (49.5) | 33 (23.0) | <0.001 |
| 1–14 days | 133 (33.3) | 59 (41.3) | |
| 15–28 days | 39 (9.7) | 29 (20.3) | |
| 29–56 days | 18 (4.5) | 14 (9.8) | |
| >56 | 12 (3.0) | 8 (5.6) | |
| Bacterial infections, n (%) | |||
| Mean | 1.12±0.77 | 1.52±1.03 | F = 23.891 |
| p < 0.001 | |||
| S | 187 (46.7) | 49 (34.3) | <0.001 |
| MDR | 88 (22.0) | 20 (14.0) | |
| XDR | 121 (30.3) | 72 (50.3) | |
| Not sure | 4 (1.0) | 2 (1.4) | |
| Catheterization, n (%) | |||
| Long-term intravenous nutrition | 98 (24.5) | 62 (43.4) | <0.001 |
| Long-term indwelling catheters | 192 (48.0) | 99 (69.2) | <0.001 |
ICU, intensive care unit; CT, computed tomography; VAP, ventilator-associated pneumonia; COPD, chronic obstructive pulmonary disease; SOFA, sequential organ failure assessment; qSOFA, quick SOFA; S, sensitive; MDR, multidrug-resistant; XDR, extensively drug-resistant.
HAP Patient Distribution by Quarters, Districts, and Departments
The distribution of HAP patients was analyzed (Fig. 2). We categorized all centers into three regions based on the geographical location and economic development level: Northern region (n = 228, 42.0%), Eastern region (n = 230, 42.4%), and Central-Western region (n = 85, 15.6%). Distribution across different quarters and districts revealed no significant difference in mortality among different outcome groups (p > 0.05) (shown in Table 2), but it was observed that the Central-Western region exhibited a higher incidence of recurrence and ineffective treatment compared to the eastern region (p < 0.05). Notably, the Northern region exhibited the highest HAP incidence during the fourth quarter (Q4) (30.7%, 70/228), whereas the Eastern region showed peak frequency in Q3 (30.0%, 69/230), and the Central-Western region in Q2 (36.5%, 31/85) (shown in Fig. 2a). The top 3 departments for HAP patients were the Department of Pulmonary (n = 169, 31.1%), Neurosurgery (n = 70, 12.9%), and General Surgery (n = 49, 9.0%), while the three departments with the highest mortality rates were the Department of Cardiology (31.6%, 6/19), Hematology (28.6%, 6/21), and Cardiovascular Surgery (26.9%, 7/26) (shown in Fig. 2b).
Fig. 2.
HAP patient distribution by districts, quarters, and departments. a Distribution of HAP patients by districts and quarters. b Distribution of HAP patients by departments.
Severity Assessment of HAP Patients
The severity in HAP patients was assessed using the qSOFA score and the SOFA score. Among the patients, 86 cases (15.8%) had a qSOFA score ≥2, and 57 cases (10.5%) had a SOFA score ≥8. The effective group demonstrated significantly lower mean qSOFA (0.63 ± 0.73 vs. 1.01 ± 0.88) and SOFA scores (3.34 ± 2.47 vs. 5.30 ± 2.89) compared to the ineffective-death group (both p < 0.001) (shown in Table 2).
Respiratory Support and Catheterization
A total of 375 (69.1%) patients received respiratory support (e.g., mechanical ventilation, tracheostomy), with a mean duration of 24.20 ± 34.74 days and a median duration of 15.00 days. Among these, 312 (57.5%) patients received invasive ventilation. Although no statistical difference was found in the duration of respiratory support among different outcome groups (p > 0.05), the effective group had a shorter mean ventilation duration (19.28 ± 31.73 vs. 23.25 ± 34.71 days) and a higher proportion of patients requiring ≤14 days of support (37.2% vs. 13.3%, p < 0.001).
Additionally, 160 (29.5%) patients received long-term intravenous nutrition, and 291 (53.6%) patients had long-term indwelling catheters. Significant differences were observed in the use of intravenous nutrition and indwelling catheters among the clinical evaluation groups (p < 0.001) (shown in Table 2).
Pathogen Isolation and Antimicrobial Resistance
A total of 663 pathogenic strains were isolated from the specimens submitted by the enrolled patients, comprising 511 (77.1%) Gram-negative bacteria, 81 (12.2%) Gram-positive bacteria, and 71 (10.7%) fungi. We observed a marked increase in the mean amounts of strains among patients in the ineffective-death group compared to those in the effective group (F = 23.891, p < 0.001). Based on the resistance to common antibiotics of collected strains (intermediate resistance was considered resistant), they were categorized into 3 groups: (1) sensitive (S); (2) multidrug-resistant (MDR): strains resistant to three or more other antibiotics; (3) extensively drug-resistant (XDR): strains resistant to almost all commonly used antibiotics, especially carbapenems (for S. aureus, XDR strain is defined as methicillin-resistant S. aureus [MRSA], resistant to oxacillin, β-lactams, and cephalosporins). Patients of the effective group are less likely to acquire the infection due to MDR/XDR organisms (p < 0.05) (shown in Table 2).
The most common organism was A. baumannii (n = 171, 25.8%), followed by K. pneumoniae (n = 143, 21.6%), P. aeruginosa (n = 92, 13.9%), and S. aureus (n = 39, 5.9%) (shown in Fig. 3). Among A. baumannii, there were 128 XDR strains (74.9%) and 18 MDR strains (10.5%), demonstrating >50% resistance to most antibiotics, with imipenem resistance as high as 80.1%. For K. pneumoniae, a total of 85 MDR and XDR strains (59.4%) were identified, with a resistance of 30%–40% to commonly used antibiotics (39.2% resistance to imipenem). For P. aeruginosa, there were 51 MDR and XDR strains (55.4%), exhibiting 30%–40% resistance to first-line antibiotics (47.8% resistance to imipenem). Notably, most carbapenem-resistant P. aeruginosa isolates retained susceptibility to non-carbapenem antibiotics. Among S. aureus, there were 4 MRSA strains (10.3%). All MRSA isolates except one pan-resistant strain were susceptible to glycopeptides and had a 100% sensitivity rate to linezolid (shown in Table 3).
Fig. 3.
Microbiology results of HAP patients. The top 4 pathogens of HAP were A. baumannii (25.8%), K. pneumoniae (21.6%), P. aeruginosa (13.9%), and S. aureus (5.9%).
Table 3.
Resistance of 4 pathogens with high detection rates (A. baumannii, K. pneumoniae, P. aeruginosa and S. aureus)
| Antibiotics | Resistance rate, n (%) | |||
|---|---|---|---|---|
| A. baumannii, n = 171 | K. pneumoniae, n = 143 | P. aeruginosa, n = 92 | S. aureus, n = 39 | |
| Amikacin | 95 (55.6) | 41 (28.7) | 5 (5.4) | 3 (7.7) |
| Gentamicin | 100 (58.5) | 52 (36.4) | 23 (25.0) | 10 (25.6) |
| Oxacillin | 46 (26.9) | 20 (14.0) | 16 (17.4) | 19 (48.7) |
| Ampicillin | 80 (46.8) | 68 (47.6) | 20 (21.7) | 9 (23.1) |
| Sulbactam/ampicillin | 113 (66.1) | 64 (44.8) | 22 (23.9) | 5 (12.8) |
| Ureidopenicillin | 114 (66.7) | 68 (47.6) | 24 (26.1) | 5 (12.8) |
| Cefazolin | 54 (31.6) | 63 (44.1) | 20 (21.7) | 5 (12.8) |
| Cefaclor | 42 (24.6) | 17 (11.9) | 18 (19.6) | 3 (7.7) |
| Cefuroxime | 49 (28.7) | 49 (34.3) | 19 (20.7) | 5 (12.8) |
| Cefotaxime | 73 (42.7) | 37 (25.9) | 22 (23.9) | 3 (7.7) |
| Cefoperazone | 115 (67.3) | 49 (34.3) | 20 (21.7) | 3 (7.7) |
| Ceftazidime | 133 (77.8) | 71 (49.7) | 20 (21.7) | 3 (7.7) |
| Ticarcillin | 48 (28.1) | 23 (16.1) | 15 (16.3) | 4 (10.3) |
| Imipenem | 137 (80.1) | 56 (39.2) | 44 (47.8) | 2 (5.1) |
| Ciprofloxacin | 134 (78.4) | 72 (50.3) | 33 (35.9) | 15 (38.5) |
| Erythromycin | 37 (21.6) | 2 (1.4) | 12 (13.0) | 13 (33.3) |
| Vancomycin | 35 (20.5) | 14 (9.8) | 12 (13.0) | 1 (2.6) |
| Teicoplanin | 33 (19.3) | 9 (6.3) | 11 (12.0) | 1 (2.6) |
| Linezolid | 32 (18.7) | 12 (8.4) | 11 (12.0) | 0 |
Besides, among the 71 fungi isolates identified, all Candida albicans except one (4.0%) were sensitive to antifungal drugs, with the exception of intermediate resistance to amphotericin B in one isolate. Among the 9 Aspergillus fumigatus isolates, one was resistant to amphotericin B, and another was resistant to caspofungin, fluconazole, and flucytosine. 2 of the 16 Candida tropicalis isolates and 1 of the 4 Candida glabrata isolates showed intermediate resistance to all common antifungal drugs.
Risk Factors for Ineffective Treatment and Mortality
Variables significantly associated with poor outcomes in univariate analysis (p < 0.05) were included in a multivariable logistic regression analysis: secondary VAP, male sex, admission to ICU, new progressive lung infiltrates on imaging, abnormal temperature, positive airway secretion cultures, liver dysfunction, heart failure, cerebral vascular disease, cancer, hospitalization, antibiotic use, and ventilation within 90 days before admission. The results showed that for HAP patients, secondary VAP, ICU admission, positive airway secretion culture, liver dysfunction, heart failure, cancer, long-term catheter indwelling, qSOFA > 1 and SOFA > 2 were risk factors of ineffective-death outcomes (p < 0.05, variance inflation factor <2) (shown in Table 4). In the final logistic regression model (Nagelkerke R2 = 0.353), the overall classification accuracy of the model was 78.6%.
Table 4.
Multivariate analysis of risk factors for ineffective-death outcomes
| | β | p value | OR | 95% CI | VIF |
|---|---|---|---|---|---|
| VAP | 0.763 | 0.006 | 2.144 | 1.247–3.687 | 1.181 |
| ICU admission | 1.296 | <0.001 | 3.655 | 2.028–6.588 | 1.370 |
| Positive airway secretion culture | 1.039 | <0.001 | 2.828 | 1.450–5.514 | 1.182 |
| Liver dysfunction | 0.770 | 0.001 | 2.161 | 1.362–3.427 | 1.091 |
| History of heart failure | 0.826 | 0.043 | 2.285 | 1.026–5.090 | 1.062 |
| History of cancer | 1.202 | <0.001 | 3.327 | 1.923–5.755 | 1.102 |
| Long-term catheter indwelling | 0.578 | 0.021 | 1.782 | 1.089–2.916 | 1.706 |
| qSOFA score <1 | −0.745 | 0.011 | 0.475 | 0.267–0.843 | 1.249 |
| SOFA score <2 | −2.843 | <0.001 | 0.029 | 0.006–0.141 | 1.128 |
VAP, ventilator-associated pneumonia; ICU, intensive care unit; SOFA, sequential organ failure assessment; qSOFA, quick SOFA; VIF, variance inflation factor.
Discussion
This multicenter retrospective study, encompassing 543 patients from 37 tertiary hospitals across China, provides a comprehensive epidemiological profile of HAP and reveals critical insights into its clinical outcomes. The overall mortality rate of 15.8% underscores the substantial burden of HAP. Through multivariate logistic regression analysis, we identified several independent risk factors significantly associated with poor prognosis, including secondary VAP, admission to ICU, positive airway secretion culture, liver dysfunction, heart failure, cancer, long-term catheter indwelling, qSOFA > 1 and SOFA > 2. Microbiologically, the top four pathogens isolated were A. baumannii, K. pneumoniae, P. aeruginosa, and S. aureus, highlighting the dominant role of Gram-negative bacteria.
A key finding of our study is the regional disparity in both the incidence and outcomes of HAP. Seasonal peaks differed across geographical regions – Q4 in the North, Q3 in the East, and Q2 in the Central West – potentially linked to extreme local climatic conditions driving hospitalizations for respiratory diseases [14]. Also, patients in the Central-Western regions demonstrated a tendency towards poorer prognosis, potentially reflecting socioeconomic and healthcare disparities. This aligns with global data showing higher HAP-associated mortality in resource-limited settings [15–17] and highlights the need for targeted interventions, such as strengthening infection control and training in these regions.
Multivariate analysis identified several independent risk factors for poor prognosis, consistent with established literature. This study identified 94 patients with secondary VAP, demonstrating significantly elevated mortality (25.5%; p < 0.05) compared to non-ventilated HAP cases. This mortality discrepancy may derive from increased utilization of non-invasive ventilation and high-flow oxygen therapy, which reduces immediate intubation needs but expands the population of critically ill patients at high risk for severe requiring subsequent mechanical ventilation [18]. In ICU settings, prolonged immobilization contributes to impaired airway clearance, aspiration risk, and recurrent pneumonia [19], while confined environments facilitate transmission of nosocomial pathogens [20]. Patients with heart failure or cancer may experience weakened immunity or prolonged ICU stays, contributing to HAP severity [21]. Liver dysfunction may signal sepsis, working as a poor prognostic factor and a leading cause of death [22]. Additionally, invasive devices such as indwelling catheters and nasogastric tubes – more frequent in patients with unfavorable outcomes (p < 0.05) – elevate the risk of aspiration and Gram-negative pneumonia [23].
Currently, accurate diagnosis and prognosis assessment for HAP remain clinically challenging. Current guidelines recommend integrating clinical manifestations, laboratory tests, and chest imaging findings [1], yet these parameters often lack specificity for reliable prediction. Clinically, the qSOFA and SOFA scores are commonly used for mortality risk stratification. Our study showed that the SOFA score differed significantly among clinical evaluation groups (p < 0.05), indicating its potential in assessing disease severity and predicting prognosis. For patients with suspected HAP, early SOFA assessment is essential. A score exceeding 2 indicates an elevated risk of poor outcomes and warrants urgent intervention. Continuous monitoring using SOFA throughout the clinical course enables dynamic evaluation and tailored management.
Respiratory support, particularly mechanical ventilation, carries an inherent risk of introducing pathogens into the lower respiratory tract and is a well-established risk factor for HAP [24, 25]. Our analysis revealed that patients in the effective group were significantly more likely to have received ≤14 days of respiratory support (including no ventilation) compared to those in the ineffective-death group (p < 0.05). This implies that critically ill patients often require prolonged respiratory support, face greater weaning difficulties, and have an increased risk of secondary VAP. Therefore, implementing prevention strategies during mechanical ventilation and striving to shorten ventilation time can effectively improve patient outcomes.
In China, HAP is predominantly caused by Gram-negative bacteria (e.g., A. baumannii, K. pneumoniae, and P. aeruginosa) and S. aureus. This pattern aligns with the global trend of increasing Gram-negative bacterial prevalence in HAP, likely driven by escalating antimicrobial resistance due to antibiotic overuse [26, 27]. Notably, our cohort demonstrated carbapenem resistance rates exceeding the 2019 CHINET national surveillance data, highlighting particularly severe resistance profiles among HAP patients in China. Although the Chinese Guidelines for HAP (2018 edition) [1] has provided detailed recommendations for the types and combination regimens of empirical antimicrobial therapy, acute-phase treatment decisions remain largely empirical in practice. Challenges including regional susceptibility variations, delayed laboratory reporting, and heterogeneous prescribing practices complicate appropriate antibiotic use. These findings underscore the urgent need for implementing region-specific antimicrobial stewardship programs guided by local resistance patterns. Additionally, fungal co-detection was frequently observed, especially in immunocompromised hosts, though the clinical significance remains uncertain due to the lack of definitive diagnostic criteria for fungal HAP [28, 29].
This investigation has several limitations that warrant consideration. First, the retrospective design relying exclusively on medical records – without prospective physician assessments or standardized evaluations – may introduce documentation inaccuracies and limit analysis of dynamic clinical indicators during HAP progression. Second, generalizability may be limited by the inclusion of only tertiary hospitals and confirmed HAP cases, which also precluded assessment of overall incidence and predisposing factors. Third, inter-institutional variations in treatment protocols impeded systematic evaluation of therapeutic impacts. Fourth, analysis of risk factors – particularly within the VAP subgroup – was constrained by limited sample size, missing data, and potential selection bias. Additionally, heterogeneity in antimicrobial susceptibility reporting – especially for newer agents like polymyxin, tigecycline, and ceftazidime-avibactam – complicated resistance pattern interpretation. Although metagenomic next-generation sequencing shows diagnostic promise [30], its limited adoption due to cost prevented assessment of its clinical utility in HAP diagnosis. Diagnostic challenges further restricted evaluation of fungal and viral pathogens, the latter being more commonly linked to community-acquired pneumonia [31]. Finally, it is important to note that this study reflects the HAP landscape in 2019, prior to the COVID-19 pandemic. While this limits generalizability to contemporary practice, it simultaneously establishes a valuable pre-pandemic baseline. This reference point will facilitate future comparisons and help elucidate how the epidemiology, management, and outcomes of HAP may have evolved in the post-COVID era.
Conclusion
Our multicenter study delineates the epidemiology, etiology, and risk factors of HAP in Chinese adults before COVID-19. Independent risk factors for poor outcomes included secondary VAP, ICU admission, positive airway secretion culture, liver dysfunction, heart failure, cancer, long-term catheter indwelling, qSOFA > 1 and SOFA > 2. High-risk populations necessitate prioritized surveillance and early interventions targeting modifiable predictors (e.g., catheter indwelling duration). Given the predominance of MDR Gram-negative pathogens in HAP, antimicrobial selection should align with local pathogen distribution and resistance patterns.
Acknowledgments
We thank all the members in the CHAPTER Research Group for making this study possible. The member list for the CHAPTER Research Group includes Qu Jieming, Zhu Yinggang, Xu Yanping, Chen Yanxin, Zhao Hongyu, Tian Xinlun, Pan Siqi, Sun Wenkui, Liu Xuedong, Wei Fengqin, Huang Yi, Li Yi, Wang Hongmin, Yang Kangli, Chen Yu, Lu Ye, Shen Ning, Liu Beibei, Zhang Jianquan, Pan Mianluan, Zhang Wei, Zhao Hongwen, Tan Wei, Chang de, Jiang Shujuan, Su Xin, Gu Yu, Li Yuping, Wang Lingwei, Xian Zhihong, Zhou Qiong, He Xinliang, Zhang Qiao, Liu lin, Wu Yaoyao, Su Zhenzhong, Wang qi, She Danyang, Wang Kaifei, Cao Bin, Lai Guoxiang, Wen Wen, Sun He, Li Tian, Zhang Tiantuo, Feng Dingyun, Fan Hong, Tong Xiang, Li Haichao, Wang Xi, Li Huayin, Hou Dongni, Xu Jinfu, Yang Jiawei, Li Yuanyuan, Jiang Juan, Zhao Haiyan, Hai Mo, Liu Yi, Wang Xiang, Jie Zhijun, Ding Yi, Zhou Hua, Sheng Lingyan, Chen Yiqiang, Huang Haizhu, Xu Feng, and Zhou Hui.
Statement of Ethics
This study was approved by the Ruijin Hospital Ethics Committee, Shanghai Jiao Tong University School of Medicine (No. [2021]305). Due to the nature of the retrospective study and the anonymous processing of data prior to analysis, the Ethics Committee approved the waiver of informed consent. The study we carried out strictly complies with the Declaration of Helsinki.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This work was supported by grants from the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2023ZD0506200), National Natural Science Foundation of China (No. 82300008), Shanghai Key Laboratory of Emergency Prevention, Diagnosis and Treatment of Respiratory Infectious Diseases (No. 20dz2261100), and Cultivation Project of Shanghai Major Infectious Disease Research Base (No. 20dz2210500). The grant agency did not involve in the design of the study, collection, analysis and interpretation of data, and writing or revising the manuscript.
Author Contributions
J.Q. conceived this study, supervised all the project, participated in the study design, and revised the manuscript. Y.Z. and Y.X. participated in the study design and revised the manuscript. Ya.C. collected clinical data, interpreted the results, prepared all the figures and tables, and wrote and revised the manuscript. H.Z., X.T., S.P., W.S., X.L., F.W., Y.H., Yi.L., H.W., K.Y., Yu.C., Ye.L., N.S., B.L., J.Z., and M.P. participated in collecting data. All authors read and approved the final manuscript.
Funding Statement
This work was supported by grants from the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2023ZD0506200), National Natural Science Foundation of China (No. 82300008), Shanghai Key Laboratory of Emergency Prevention, Diagnosis and Treatment of Respiratory Infectious Diseases (No. 20dz2261100), and Cultivation Project of Shanghai Major Infectious Disease Research Base (No. 20dz2210500). The grant agency did not involve in the design of the study, collection, analysis and interpretation of data, and writing or revising the manuscript.
Data Availability Statement
The data that support the findings of this study are not publicly available due to participant privacy and institutional data use agreements but are available from the corresponding author upon request.
References
- 1. Subspecialty Group of Infectious Diseases; Chinese Society of Respiratory Diseases . Guidelines for the diagnosis and treatment of hospital-acquired pneumonia and ventilator-associated pneumonia in adults in China (2018 edition). Chin J Tuberc Respir Dis. 2018;41:255–80. [Google Scholar]
- 2. Liu YN, Cao B, Wang H, Chen L, She D, Zhao T, et al. Adult hospital acquired pneumonia: a multicenter study on microbiology and clinical characteristics of patients from 9 Chinese cities. Zhonghua Jie He He Hu Xi Za Zhi. 2012;35(10):739–46. [PubMed] [Google Scholar]
- 3. Chung DR, Song JH, Kim SH, Thamlikitkul V, Huang SG, Wang H, et al. High prevalence of multidrug-resistant nonfermenters in hospital-acquired pneumonia in Asia. Am J Resp Crit Care. 2011;184(12):1409–17. [DOI] [PubMed] [Google Scholar]
- 4. Ghadimi M, Siemieniuk RAC, Guyatt G, Loeb M, Hazzan AA, Aminaei D, et al. Empiric antibiotic regimens in adults with non-ventilator-associated hospital-acquired pneumonia: a systematic review and network meta-analysis of randomized controlled trials. Clin Microbiol Infec. 2024;30(11):1351–63. [DOI] [PubMed] [Google Scholar]
- 5. Jones BE, Sarvet AL, Ying J, Jin R, Nevers MR, Stern SE, et al. Incidence and outcomes of non–ventilator-associated hospital-acquired pneumonia in 284 US hospitals using electronic surveillance criteria. JAMA Netw Open. 2023;6(5):e2314185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Fang WF, Yang KY, Wu CL, Yu CJ, Chen CW, Tu CY, et al. Application and comparison of scoring indices to predict outcomes in patients with healthcare-associated pneumonia. Crit Care. 2011;15(1):R32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. An YB, Tian YL. Clinical analysis of healthcare associated pneumonia in emergency department. J Clin Pulm Med. 2015;20:1842–4. [Google Scholar]
- 8. Weiner-Lastinger LM, Abner S, Edwards JR, Kallen AJ, Karlsson M, Magill SS, et al. Antimicrobial-resistant pathogens associated with adult healthcare-associated infections: summary of data reported to the National Healthcare Safety Network, 2015–2017. Infect Cont Hosp Epidemiol. 2020;41(1):1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Mitchell BG, Russo PL, Cheng AC, Stewardson AJ, Rosebrock H, Curtis SJ, et al. Strategies to reduce non-ventilator-associated hospital-acquired pneumonia: a systematic review. Infect Dis Health. 2019;24(4):229–39. [DOI] [PubMed] [Google Scholar]
- 10. Lyons PG, Kollef MH. Prevention of hospital-acquired pneumonia. Curr Opin Crit Care. 2018;24(5):370–8. [DOI] [PubMed] [Google Scholar]
- 11. Singer AJ, Ng J, Thode HC, Spiegel R, Weingart S. Quick SOFA scores predict mortality in adult emergency department patients with and without suspected infection. Ann Emerg Med. 2017;69(4):475–9. [DOI] [PubMed] [Google Scholar]
- 12. Moreno R, Rhodes A, Piquilloud L, Hernandez G, Takala J, Gershengorn HB, et al. The sequential organ failure assessment (SOFA) score: has the time come for an update? Crit Care. 2023;27(1):15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Wolfensberger A, Meier AH, Kuster SP, Mehra T, Meier MT, Sax H. Should international classification of diseases codes be used to survey hospital-acquired pneumonia? J Hosp Infect. 2018;99(1):81–4. [DOI] [PubMed] [Google Scholar]
- 14. He Q, Liu Y, Yin P, Gao Y, Kan H, Zhou M, et al. Differentiating the impacts of ambient temperature on pneumonia mortality of various infectious causes: a nationwide, individual-level, case-crossover study. EBioMedicine. 2023;98:104854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Aloush SM, Al-Rawajfa OM. Prevention of ventilator-associated pneumonia in intensive care units: barriers and compliance. Int J Nurs Pract. 2020;26(5):e12838. [DOI] [PubMed] [Google Scholar]
- 16. Bonell A, Azarrafiy R, Huong VTL, Viet TL, Phu VD, Dat VQ, et al. A systematic review and meta-analysis of ventilator-associated pneumonia in adults in Asia: an analysis of national income level on incidence and etiology. Clin Infect Dis. 2019;68(3):511–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Zhao JH, Qin B, Yan RN, Ma J, Hu LL, Liu YL. Status and trend of the main healthcare-associated infection indicators in tertiary public hospitals in China (2018-2020). Chin J Infect Control. 2022;21:524–31. [Google Scholar]
- 18. Vallecoccia MS, Dominedò C, Cutuli SL, Martin-Loeches I, Torres A, De Pascale G. Is ventilated hospital-acquired pneumonia a worse entity than ventilator-associated pneumonia? Eur Respir Rev. 2020;29(157):200023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Duncan S, Mcauley DF, Walshe M, McGaughey J, Anand R, Fallis R, et al. Interventions for oropharyngeal dysphagia in acute and critical care: a systematic review and meta-analysis. Intens Care Med. 2020;46(7):1326–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Chen J, Ma H, Huang X, Cui Y, Peng W, Zhu F, et al. Risk factors and mortality of carbapenem-resistant Klebsiella pneumoniae bloodstream infection in a tertiary-care hospital in China: an eight-year retrospective study. Antimicrob Resist Infect Control. 2022;11(1):161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kreitmann L, Helms J, Martin-Loeches I, Salluh J, Poulakou G, Pène F, et al. ICU-acquired infections in immunocompromised patients. Intens Care Med. 2024;50(3):332–49. [DOI] [PubMed] [Google Scholar]
- 22. Strnad P, Tacke F, Koch A, Trautwein C. Liver guardian, modifier and target of sepsis. Nat Rev Gastroenterol Hepatol. 2016;14(1):55–66. [DOI] [PubMed] [Google Scholar]
- 23. Guzmán-Herrador B, Molina CD, Allam MF, Navajas RFC. Independent risk factors associated with hospital-acquired pneumonia in an adult ICU: 4-year prospective cohort study in a university reference hospital. J Public Health. 2016;38(2):378–83. [DOI] [PubMed] [Google Scholar]
- 24. Muscedere J, Rewa O, Mckechnie K, Jiang X, Laporta D, Heyland DK. Subglottic secretion drainage for the prevention of ventilator-associated pneumonia: a systematic review and meta-analysis. Crit Care Med. 2011;39(8):1985–91. [DOI] [PubMed] [Google Scholar]
- 25. Stoclin A, Rotolo F, Hicheri Y, Mons M, Chachaty E, Gachot B, et al. Ventilator-associated pneumonia and bloodstream infections in intensive care unit cancer patients: a retrospective 12-year study on 3388 prospectively monitored patients. Support Care Cancer. 2020;28(1):193–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Rachina SА, Fedina LV, Sukhorukova MV, Sychev IN, Larin ES, Alkhlavov A. Diagnosis and antibiotic therapy of nosocomial pneumonia in adults: from recommendations to real practice. a review. Ter Arkh. 2023;95(11):996–1003. [DOI] [PubMed] [Google Scholar]
- 27. Rhodes NJ, Cruce CE, O’Donelle JN, Wunderink RG, Hauser AR. Resistance trends and treatment options in Gram-negative ventilator-associated pneumonia. Curr Infect Dis Rep. 2018;20(2):3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Hage CA, Carmona EM, Epelbaum O, Evans SE, Gabe LM, Haydour Q, et al. Microbiological laboratory testing in the diagnosis of fungal infections in pulmonary and critical care practice. An official American Thoracic Society Clinical Practice Guideline. Am J Resp Crit Care. 2019;200(5):535–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Donnelly JP, Chen SC, Kauffman CA, Steinbach WJ, Baddley JW, Verweij PE, et al. Revision and update of the consensus definitions of invasive fungal disease from the European organization for research and treatment of cancer and the Mycoses study group education and research consortium. Clin Infect Dis. 2020;71(6):1367–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ruppé E, Greub G, Schrenzel J. Messages from the first international conference on clinical metagenomics (ICCMg). Microbes Infect. 2017;19(4–5):223–8. [DOI] [PubMed] [Google Scholar]
- 31. Zilberbeg MD, Khan I, Shorr AF. Respiratory viruses in nosocomial pneumonia: an evolving paradigm. Viruses. 2023;15(8):1676. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are not publicly available due to participant privacy and institutional data use agreements but are available from the corresponding author upon request.
