Epidemiology and disease burden among hospital-acquired pneumonia/ventilator-associated pneumonia patients in a tertiary care hospital

DOI: https://doi.org/10.21203/rs.3.rs-1809796/v1

Abstract

Background

Hospital-acquired pneumonia (HAP) are one of the major healthcare-associated infection but its burden and financial impact are lesser studied comparing to ventilator-associated pneumonia (VAP). The study aims to investigate the disease burden, etiology and the outcome among HAP/VAP patients in Taiwan.

Methods

This is a retrospective study conducted in a teaching hospital in Taiwan. We enrolled patients with HAP/VAP during 1st Jan. 2018 to 31st Dec. 2018. The definition of HAP/VAP was based on definition proposed by the Taiwan Centers of Disease Control and the hospital-wide data was audited prospectively by Center for Infection Control of the hospital. Patients with HAP were further categorized as non-ventilated HAP (nvHAP) and ventilated HAP (vHAP). The burden of HAP/VAP was estimated and the clinical outcomes were compared between nvHAP and vHAP.

Results

In total, 152 episodes of HAP and 8 episodes of VAP were identified. The incidence of HAP was 0.22 (95% confidence interval [CI], 0.19–0.26) per 1,000 patients-days while that of VAP was 0.27 (95% CI, 0.12–0.53) per 1,000 ventilator-days. The mortality was similar between vHAP and nvHAP in our cohort (55.6% vs. 62.6%, respectively, p = 0.58). However, vHAP was associated with higher length of hospital stay (56 vs. 37 days, p < 0.001) and medical expenses (787,621 vs. 455,347 New Taiwan Dollar, p < 0.001).

Conclusions

The study provides the infection incidence density of HAP and VAP in a tertiary care hospital. Due to the disease burden and the medical cost, prevention and treatment of HAP and VAP are an important element of infection control.

Introduction

Respiratory tract infections are one of the most common site of healthcare-associated infection (HAI) reported worldwide [1]. Depending on the use of mechanical ventilation, pneumonia occurred at in-hospital setting are further classified as hospital-acquired pneumonia (HAP) or ventilator-associated pneumonia (VAP). HAP and VAP are significant public health threat which result in prolong hospitalization, increase medical cost and also an attributable mortality rate as high as 33–50% [2].

The impacts of HAP or VAP vary in different regions and healthcare settings. Recent studies revealed the disease burden of HAP was 3.63 cases per 1000 patient-days and VAP was 1–2.5 cases per 1000 ventilator-days in the United States [3, 4]. A study conducted in 2006 revealed higher infection incidence as 18.3 VAP episodes per 1000 ventilator-days in Europe [5]. The discrepancies between these reports were likely related to the difference in case definition, HAI notification systems and means of diagnosis. The estimated medical cost due to VAP was around $25,000 to $28,000 USD per patient in the United States and £6000 and £22000 per patient in the United Kingdom [6, 7].

However, the disease burden of HAP or VAP in Asia is less reported and mainly available only at institutional level [8]. In a systemic review, we summarized the incidence of HAI from the national surveillance data of Taiwan, South Korea and Japan. The VAP incidence was 1.1 case per 1000 ventilator-days in Taiwan and 1.0 in South Korea according to the national surveillance data in 2015. In Japan, the incidence of VAP was 1.5 cases per 1000 patient-days in 2015 [9]. Staphylococcus aureus, Pseudomonas aeruginosa and Klebsiella pneumoniae were the predominant pathogens among all three countries. In Asia-Pacific region, the morality rate of HAP/VAP ranged from 26–28% reported in Thailand to 58% reported in Pakistan [8]. However, the difference in the data representation and case definition had made the comparison of data challenging.

In addition, the impact of pneumonia might be different among different subgroups of HAP. In a new definition proposed by Talbot G and colleagues in 2019 [10], they further classified the patients with bacterial HAP into who required subsequent mechanical ventilation as ventilated HAP (ventilated HAP, vHAP), as compared to those who did not need invasive ventilation (non-ventilated HAP, nvHAP). In their study, the all-cause mortality was highest among the patients with vHAP (27.8%), followed by 18% of VAP and 14.5% of nvHAP. As the decision of intubation and ventilation depends on clinical judgement and might vary in different healthcare settings, the definition and prognostic its indication also requires more validation.

Thus, we aimed to conduct the study to address the incidence of HAP (including vHAP and nvHAP) and VAP in a teaching facility in Taiwan and to clarify the pathogen distribution among the three disease categories. The impact of HAP/VAP, including total length of stay (LOS), mortality rate and the total cost per hospitalization were evaluated.

Materials And Methods

Study Setting and Population

This is a retrospective cohort study performed in teaching hospital and tertiary referral center with a capacity of 2,200 acute care bed in northern Taiwan. In 2018, a total of 110,013 patients were discharged from the hospital and accounted for 689,969 patient-days of admission. During 1st January 2018 to 31st December 2018, patients who fulfilled the definition of HAP/VAP would be alerted to the Center for Infection Control (CIC) of the hospital by an automated monitoring system. Each case was reviewed individually, and those who fulfilled the case definition of HAP/VAP were enrolled. The case definitions in our study were published by National Healthcare Safety Network (NHSN) of the United States and adopted by Taiwan Centers for Disease Control (TWCDC) [11, 12]. In the definition, HAP were defined as pneumonias which occurred 48 hours after admission and fulfilled a combination of radiographical and clinical criteria. The details of definition criteria were listed in Table 1. HAP patients were further classified into vHAP (those who required mechanical ventilation later due to HAP) and nvHAP (those who did not need intubation in the episode). VAP was defined as pneumonia onset 48 hours after mechanical ventilation, with same definition as HAP (Table 1).

Table 1

Case definition of HAP/VAP in this study

Radiographical criteria

Laboratory criteria

Clinical criteria

One of the following findings in ≥ 2 consecutive chest radiography:

- New or progressive infiltrates.

- New or progressive consolidations.

- New or progressive cavitation.

One of the following findings:

- Fever above 38℃.

- Leukopenia (WBC < 4,000/mL) or leukocytosis (WBC > 12,000/mL).

Two of the following findings:

- Purulent sputum or increase of sputum production.

- Now cough or dyspnea.

- Rales or bronchial breathing sound on physical examination.

- Impairment of gas exchange (increase oxygen demands or decrease of oxygen saturation).

*A patient must fit radiographical, laboratory and clinical criteria before diagnosed of HAP/VAP.

Causative microorganism

The causative microorganisms were also defined according to the criteria proposed by TWCDC, which including bacteria identified from sterile specimen of blood, pleural fluid or lung tissue; or positive semi-quantitative culture from minimally contaminated lower respiratory tract specimen. Legionella pneumonia was defined as successful detection of urinary Legionella pneumophila serogroup 1 antigen or a > 4-fold increase of serum antibodies to Legionella pneumophila serogroup 1. Pneumonia caused by well-established community-acquired pathogens, including Cryptococcus species, Pneumocystis jirovecii, and Mycobacterium tuberculosis were excluded. Coagulase-negative Staphylococcus species, Enterococcus species and Candida species isolated from blood culture were also excluded, unless the same microorganism were also isolated from lung tissue, pleural fluid or respiratory samples simultaneously.

Estimating the Burden of HAP/VAP

During the study period, all potential episodes of HAP and VAP were identified through the automated notification system and reviewed separately to confirm if they met the defining criteria (Table 1). The incidence and prevalence of HAP and VAP were then calculated. For HAP, the incidence was defined as cases per 1,000 patient-days of admission and the prevalence as cases per 100 discharged patients. In terms of VAP, the incidence was defined as cases per 1,000 ventilator-days and the prevalence being cases per 100 patients discharges from ICU.

Clinical outcome and medical expense of HAP/VAP

For outcome analysis, we excluded recurrent episodes of HAP/VAP in 2018 and only the first episode of HAP/VAP were included. These enrolled patients were analyzed to estimate the clinical outcome and medical expenses related to HAP/VAP. Patients who aged below 20 years of age were also excluded. For each included patient with HAP/VAP, medical record was reviewed to collect demographic and clinical data including age, sex, underlying diseases, laboratory tests and clinically relevant culture results. If the patient was intubated, the duration of mechanical ventilation, duration of intensive care unit (ICU) stay and the length of hospitalization stay (LOS) were recorded. In our study, the outcome of HAP/VAP were categorized as discharged or death. The medical expenditure during hospitalization was also calculated for all enrolled patients. The study was approved by the Research Ethics Committee of National Taiwan University Hospital (registration number: 202009032RIPD) and informed consent was waived.

Statistical analyses

The overall incidence and prevalence of HAP/VAP, with a 95% confidence interval (CI) were calculated as the aforementioned definitions. After excluding repeated HAP/VAP episodes, the demographic and clinical characteristics of included patients were summarized. Non-categorical variables were compared using Student’s t-test or Mann–Whitney U-test, and categorical variables were compared using chi-square test or Fisher’s exact test. For non-categorical data with more than 2 subgroups, one-way ANOVA test or Kruskal-Wallis test was used to detect differences between groups and pairwise comparisons or Dunn’s test with Bonferroni correction would be applied if post-hoc analysis was needed [13]. All statistical analyses were performed using STATA software v.14.0 S/E (StataCorp LP, College Station, TX). All p values were two-sided, with a p value < 0.05 is defined as statistically significant.

Results

During the 1-year study period, 110,013 patients was discharged from the hospital and contributed for 689,969 patient-day of admission; and 9,238 patients were admitted in the ICU and contributed for 29,865 ventilator-day of use. In total, 152 episodes of HAP was identified throughout the study period, with 70 episodes being nvHAP and 74 episodes being vHAP and 8 episodes of VAP. This gave a HAP prevalence of 0.13% with an incidence of 0.22 (95% confidence interval [CI], 0.19–0.26) per 1,000 patient-days of admission. In terms of VAP, the prevalence among all ICU patients was 0.09% with an incidence of 0.27 (95% CI, 0.12–0.53) per 1,000 ventilator-days of use.

After excluding repeated events of HAP/VAP and patients who aged below 20 years, there were 134 episodes of HAP included in further analysis, including 63 episodes nvHAP (63/134, 47.0%) and 71 vHAP (71/134, 53.0%), as well as 8 episodes of VAP (Table 2). For patients who were included for HAP, the median age was 69 (IQR 60–79) years old with a male predominance of 61.2% (87/134). A large proportion of patients with HAP presented with underlying medical conditions, with solid organ malignancies (71/134, 53.0%), hypertension (58/134, 43.3%) diabetes mellitus (42/134, 31.3%) and hematological malignancy (35/134, 26.1%) being the most commonly associated diagnoses. 37.3% (50/134) of HAP patients reported to ever smoked, with 13.4% (18/134) being active smoker before their admission in 2018.

Table 2

The epidemiology character of patients with HAP and VAP in a tertially care hospital in 2018

 

HAP (n = 134)

nvHAP (n = 63)

vHAP (n = 71)

VAP (n = 8)

Age, median (IQR), years

69 (60–79)

73 (62–82)

65 (58–78)

64.5 (62-72.5)

Male, n (%)

82 (61.2%)

38 (60.3%)

44 (62.0%)

5 (62.5%)

HTN, n (%)

58 (43.3%)

20 (31.8%)

38 (53.5%)

6 (75.0%)

DM, n (%)

42 (31.3%)

15 (23.8%)

27 (38.0%)

5 (62.5%)

ESRD, n (%)

9 (6.7%)

0 (0%)

9 (12.7%)

0 (0%)

Autoimmune disease, n (%)

9 (6.7%)

2 (3.2%)

7 (9.9%)

1 (12.5%)

Solid organ cancer, n (%)

71 (53.6%)

36 (57.1%)

35 (49.3%)

3 (37.5)

Hematology malignancy, n (%)

35 (26.1%)

18 (28.6%)

17 (23.9%)

0 (%)

COPD, n (%)

5 (3.7%)

4 (6.4%)

1 (1.4%)

0 (0%)

Asthma, n (%)

1 (0.8%)

0 (0%)

1 (1.4%)

0 (0%)

Liver cirrhosis, n (%)

8 (6.0%)

3 (4.8%)

5 (7.0%)

0 (0%)

Current smoker, n (%)

18 (13.4%)

4 (6.4%)

14 (19.7%)

2(25.0%)

Quitted smoker, n (%)

32 (23.9%)

22 (34.9%)

10 (14.1%)

1 (12.5%)

Abbreviation: HAP, healthcare-associated pneumonia; nvHAP, non-ventilated healthcare-associated pneumonia; vHAP, ventilated healthcare-associated pneumonia; VAP, ventilator-associated pneumonia; IQR, interquartile range; HTN, hypertension; DM, diabetes mellitus; ESRD, end-stage renal diseases; COPD, chronic obstructive pulmonary diseases.

Clinical outcome and medical expanses of HAP/VAP

The outcome and medical expenses of included patients were shown in Table 3. The overall mortality rate among patients with HAV were 59.0% (79/134); with 55.6% (35/63) and 62.0% (44/71) patients of nvHAP and vHAP died, respectively (p = 0.58). Additionally, the mortality rate among patients with VAP were 37.5% (3/8), which was not statistically different from those with nvHAP or vHAP in our cohort. Among those with HAP, the median duration between admission to development of HAP was 14 (interquartile range [IQR], 8–30) days, with the median LOS being 43 (IQR, 29–68) days. For those with nvHAP, the median LOS was 37 (IQR, 26–63) days, while the LOS among vHAP and VAP was 56 (IQR, 32–74) days and 30 (25–69) days, respectively (p = 0.049). In the post-hoc analysis, the LOS was significantly longer among patients with vHAP as compared with those who had nvHAP, but not with thosw who had VAP.

Table 3

Outcome and medical expenditure of the HAP/VAP patients

 

HAP (n = 134)

nvHAP (n = 63)

vHAP (n = 71)

VAP (n = 8)

LOS, median (IQR), days

43 (29–68)

37 (26–63)

56 (32–74)

29.5 (25–69)

Duration of mechanical ventilation, median (IQR), days

3.5 (0–13)

0 (0–0)

12 (7–21)

15.5 (9-46.5)

ICU stay, median (IQR), days

7.5 (0–20)

0 (0–4)

17 (10–30)

18.5 (10.5–47)

Mortality, n (%)*

79 (59.0%)

35 (55.6%)

44 (62.0%)

3 (37.5%)

Discharged, n (%)

55 (41.0%)

28 (44.4%)

27 (38.0%)

5 (62.5%)

Medical Cost median (IQR), NTD

675,016 (393,621-1,212,018)

455,347 (207,371–816,296)

787,621 (584,468-1,390,192)

522,078 (347,421-1,156,087)

*Included those who were discharged under critical condition and were considered improbable to survive without life support.

Abbreviation: HAP, healthcare-associated pneumonia; nvHAP, non-ventilated healthcare-associated pneumonia; vHAP, ventilated healthcare-associated pneumonia; VAP, ventilator-associated pneumonia; LOS, length of hospital stays; IQR, interquartile range; ICU, intensive care units; NTD, New Taiwan Dollar.

In terms of the medical expenses, the median medical expenditure during hospitalization for those with HAP in our study was 675,016 (IQR, 393,621-1,212,018) New Taiwan Dollar (NTD), which was significantly higher among those with vHAP as compared with those with nvHAP (787,621 vs. 455,347 NTD, p < 0.001).

Causative microorganisms for HAP/VAP

Among the 142 patients with HAP/VAP, 60 patients (42.3%) had culture available culture results with 83 pathogens identified (see supplement Table S1). The most common pathogens for nvHAP and vHAP were listed in Table 4.

Table 4

Comparison of culture results among the HAP patients

 

HAP

nvHAP

vHAP

Culture-negative, n (%)

76/134 (56.7)

40/63 (63.5)

35/71 (49.3)

Most common pathogens (n)

Rank 1

Acinetobacter baumannii complex (14)

Pseudomonas aeruginosa (5)

Acinetobacter baumannii complex (10)

Rank 2

Pseudomonas aeruginosa (13)

Acinetobacter baumannii complex (4)

Klebsiella pneumoniae (9)

Rank 2

Klebsiella pneumoniae (13)

Klebsiella pneumoniae (4)

Pseudomonas aeruginosa (8)

Rank 4

Escherichia coli (5)

Escherichia coli (3)

Stenotrophomonas maltophilia (4)

Rank 5

Stenotrophomonas maltophilia (5)

Elizabethkingia species (2)

Escherichia coli (2)

Abbreviation: HAP, healthcare-associated pneumonia; nvHAP, non-ventilated healthcare-associated pneumonia; vHAP, ventilated healthcare-associated pneumonia.

Discussion

In this retrospective cohort study involving patients with HAP/VAP in a tertiary teaching hospital in Taiwan, the incidence of HAP was incidence of 0.22 (95% CI, 0.19–0.26) per 1,000 patient-days and that of VAP was 0.27 (95% CI, 0.12–0.53) per 1,000 ventilator-days of use. Overall, our study reported a high all-cause mortality rate and high medical expenditure associated with patients who were complicated with HAP/VAP.

According to the Taiwan nosocomial infection surveillance system (TNIS) system, the burden of VAP was 0.6 per 1,000 ventilator-days in the medical center and 0.9 per 1,000 ventilator-days in the regional hospitals in 2018 [14]. The incidence rate reported in our study was slightly lower than the national surveillance data. Since all cases reported from the automated system was reviewed individually to fit the diagnostic criteria, it is possible that our system might underestimate the incidence HAP/VAP by only selecting those with the most obvious clinical findings and severities. However, the low incidence rate in our study as well as the national surveillance data were still comparable with our prior systemic review, we had demonstrated a statistically significant reduction of HAI incidence from 2008 to 2015 in Taiwan, South Korea and Japan [9]. In Taiwan, the improvement may attributable to the development of TNIS system (available since 2000 and online since 2006), hand hygiene program since 2007, and care bundle program since 2011. The incidence of VAP had 57.7% reduction during the study period in Taiwan. But the incidence of HAP is not available in Taiwan’s national surveillance data. Chawla R, et al had tried to provide the comparison of HAP incidence among 10 Asia countries [8], and they had reported an HAP incidence between 0.51–0.85 per 1000 patient-days in Taiwan in 2008, which was possibly generated through personal communication. To our best knowledge, this is the first article in English provide the disease burden of HAP in Taiwan.

The reported incidence of HAP/VAP also varied greatly in different regions. The NHSN reported an VAP incidence of 1–2.5 cases per 1000 ventilator-days in 2011 in the United States [4]. However, Koulenti and colleagues reported a much higher incidence of 18.3 episodes per 1000 ventilator-day among 27 European ICUs in 2006 [5]. The huge differences among incidence rate might partially explained by the difference of case definition. The case definition used in our hospital surveillance system (and hence in our study) was in line with the definition proposed by the US CDC, while the study by Koulenti et al. was based on clinical judgment and interpretation of culture results. This highlighted the importance of a globally unified definition in order to compare the burden of HAP/VAP in different healthcare settings.

One of the challenges in HAP/VAP is the identification of causative microorganism. In past literatures, clinical cultures, including expectorated sputum, tracheal aspirate and blood, could identify around 50% of the causative pathogen [15]. Since the selection of antibiotics might be impacted by the difficulty in identifying pathogens and might resulted in increased risks of mortality or morbidity, the 2016 treatment guideline published by Infectious Diseases Society of America (IDSA) and the American Thoracic Society (ATS) had suggest each hospital should build-up its own antibiogram to facilitate the optimal choice of antibiotics [16, 17]. In our study, the top five common pathogens in nvHAP and vHAP were Gram-negative bacteria, including Acinecobacter baumanii complex, Pseudomonas aeruginosa, Klbsiella pneumoniae, Escherichia coli, Stenotrophomonas maltophilia and Elizabethkingia species, which was also consistent with the finding in the national surveillance data of VAP in Taiwan [14]. Our results might offer an evidence in choosing optimal empiric antibiotics, and de-escalating should be considered while culture result available.

The all-cause mortality among patients with HAP and VAP in our report was 59.0% and 37.5%, respectively. Furthermore, there was no statistically significant difference in the mortality rate among those with vHAP and nvHAP in our study. However, unlike our observation, Talbot G. et al reported a lower much overall HAP mortality and had observed a slightly lower risk of mortality among those with nvHAP, when compared with patients of vHAP [10]. This difference might be partially explained by the inclusion criteria of different studies. The study from Talbot et al. included data from several drug trials and might had excluded those who were ineligible for clinical studies. Also, as aforementioned, our study might have underestimated the incidence of HAP/VAP and overlooked less severe patients.

In the addition, HAP/VAP also lead to substantial increase of medical expenditure. In a systemic review, Zimlichman E, et al estimated that an episode of VAP would associated with 13.1 additional days of hospital stay (8.4 days in ICU) and surplus expenses of 40,144 USD [18]. According to the NHSN, VAP ranked 2nd for the most overall attributable cost among the major HAI in the United States, following only by surgical site infection (SSI). Wagh H, et al also reported the estimated cost due to VAP was £6000 and £22000 per patient in the United Kingdom (7). According a study using administrative data from Malaysia, Indonesia and Philippines, Azmi S et al. reported the LOS due to HAP ranged from 6.9–10.2 in these countries, and the estimated cost per patient was 784,300-1,882,140 USD [19]. Our study added pieces to the evidence which estimating the cost of HAP/VAP, although the difference related socioeconomic infrastructure had made the comparison difficult.

There are several limitations of this study. Since the data is obtained from a tertiary care hospital, thus the incidence of HAP/VAP and outcome may not be generalized to other hospitals setting with patients with fewer comorbidity and lesser disease severity. Second, our study likely underestimated the incidence of HAP/VAP and warrant careful interpretation when applied. However, our results might still help guiding clinicians dealing with most severe HAP or VAPs.

Conclusion

In summary, our study provides an updated information of disease burden of HAP and VAP in a tertiary care hospital in Taiwan and reported the causative microorganisms identified. For the financial impact and the mortality caused by the HAP/VAP, further efforts should be put on aggressive prevention and proper treatment of these important HAIs.

Declarations

Acknowledgement

We thank the support from the staffs of the Centers for Infection Control, National Taiwan University Hospital.

Ethical Approval

The study was approved by the Research Ethics Committee of National Taiwan University Hospital (registration number: 202009032RIPD) and informed consent was waived due to the retrospective design.

Consent for publication

Not applicable.

Availability of Data and Materials

Aggregated data without personal information is available upon request.

Funding

This study was supported by Merck Sharp & Dohme, Taiwan. The authors declared that the manuscript was based on the research result and not been influenced by the sponsorship.

Conflict of Interest

All authors report no potential conflict of interest.

Authors’ contributions

S-C. P. and Y-C. C. conceptualized the study. S-H. C and K-L. T. retried data from hospital registry. G-J. C. and S-C. P performed data cleaning and statistical analysis. G-J. C prepared the first draft of manuscript. All authors reviewed the final manuscript.

References

  1. World Health Organization. Prevention of hospital-acquired infections: a practical guide. 2002. Available at: https://apps.who.int/iris/handle/10665/67350. Assessed 10 Jun 2022.
  2. Mandell LA, Campbell GDJ. Nosocomial pneumonia guidelines: an international perspective. Chest 1998; 113:188S-193S.
  3. Giuliano KK, Baker D, Quinn B. The epidemiology of nonventilator hospital-acquired pneumonia in the United States. Am J Infect Control 2018; 46:322–327.
  4. Dudeck MA, Horan TC, Peterson KD, et al. National Healthcare Safety Network report, data summary for 2011, device-associated module. Am J Infect Control 2013; 41:286–300.
  5. Koulenti D, Tsigou E, Rello J. Nosocomial pneumonia in 27 ICUs in Europe: perspectives from the EU-VAP/CAP study. Eur J Clin Microbiol Infect Dis 2017; 36:1999–2006.
  6. Centers for Disease Control and Prevention, United States. The direct medical costs of healthcare-associated infections in U.S hospitals and the benefit of prevention. 2009. Available at: https://www.cdc.gov/hai/pdfs/hai/scott_costpaper.pdf. Assessed 10 Jun 2022.
  7. Wagh H, Acharya D. Ventilator Associated Pneumonia – an Overview. BJMP 2009; 2:16–19.
  8. Chawla R. Epidemiology, etiology, and diagnosis of hospital-acquired pneumonia and ventilator-associated pneumonia in Asian countries. Am J Infect Control 2008; 36:S93-100.
  9. Chiang C-H, Pan S-C, Yang T-S, et al. Healthcare-associated infections in intensive care units in Taiwan, South Korea, and Japan: recent trends based on national surveillance reports. Antimicrob Resist Infect Control 2018; 7:129.
  10. Talbot GH, Das A, Cush S, et al. Evidence-Based Study Design for Hospital-Acquired Bacterial Pneumonia and Ventilator-Associated Bacterial Pneumonia. J Infect Dis 2019; 219:1536–1544.
  11. National Healthcare Safety Network. Pneumonia (Ventilator-associated [VAP] and non-ventilatorassociated Pneumonia [PNEU]) Event. 2022. Available at: https://www.cdc.gov/nhsn/pdfs/pscmanual/6pscvapcurrent.pdf. Assessed 10 Jun 2022.
  12. Taiwan Centers for Disease Control. Surveillance definition for healthcare associated infection. 2018. Available at: https://www.cdc.gov.tw/File/Get/vaCzISfjhsSg_7siqy44CQ. Assess 10 Jun 2022.
  13. Dinno A. Nonparametric pairwise multiple comparisons in independent groups using Dunn’s test. Stata J 2015; 15:292–300. Available at: http://www.stata-journal.com/article.html?article=st0381. Assessed 10 Jun 2022.
  14. Taiwan Centers for Disease Control. Statistics of Communicable Diseases and Surveillance Report. 2018. Available at: https://www.cdc.gov.tw/En/InfectionReport/Info/A1HdZwK8fUzN1IiBKDPkVQ?infoId=Oxeyq9kDOBy43jTcCjzWMw. Assessed 10 Jun 2022.
  15. Welte T, Torres A, Nathwani D. Clinical and economic burden of community-acquired pneumonia among adults in Europe. Thorax 2012; 67:71–79.
  16. Kalil AC, Metersky ML, Klompas M, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis 2016; 63:e61–e111.
  17. Vacas-Córdoba M, Cardozo-Espinola C, Puerta-Alcalde P, Cilloniz C, Torres A, García-Vidal C. Empirical treatment of adults with hospital-acquired pneumonia: lights and shadows of the 2016 Clinical Practice ATS/IDSA Guidelines. Rev Esp Quimioter 2017; 30 Suppl 1:30–33.
  18. Zimlichman E, Henderson D, Tamir O, et al. Health care-associated infections: a meta-analysis of costs and financial impact on the US health care system. JAMA Intern Med 2013; 173:2039–2046.
  19. Azmi S, Aljunid SM, Maimaiti N, et al. Assessing the burden of pneumonia using administrative data from Malaysia, Indonesia, and the Philippines. Int J Infect Dis 2016; 49:87–93.