Higher risk of viral infections in chronic obstructive pulmonary disease patients recovering from sepsis compared to non-sepsis patients: a propensity score-matched observational study

Abstract

Background:

Patients with chronic obstructive pulmonary disease (COPD) who survive sepsis remain immunocompromised and are at increased risk of subsequent viral infections, including influenza and respiratory syncytial virus (RSV).

Objective:

This study aimed to address the 1-year risk of viral infections after sepsis in patients with COPD through the use of a federated database, TriNetX.

Design:

A propensity score-matched (PSM) retrospective cohort study.

Methods:

We used data of 903,683 COPD patients and identified 113,589 who experienced sepsis. The risk of distinct viral infections, including herpes simplex virus (HSV), varicella-zoster virus (VZV), cytomegalovirus (CMV), RSV and influenza, within 1 year post-sepsis was analyzed, with the employment of PSM to minimize confounding. The effect of vaccination was also assessed to determine its protective efficacy.

Results:

A total of 98,883 COPD patients with sepsis and 1:1 matched COPD without sepsis were eligible for analyses. COPD patients with sepsis had consistently higher risk for viral infections within 1 year after the sepsis compared with COPD patients without sepsis. The hazard ratios (HRs) were as follows: HSV 1.936 (95% CI: 1.775–2.112), VZV 3.050 (2.489–3.737), CMV 2.101 (1.832–2.410), RSV 3.297 (3.158–3.443), and Influenza 3.197 (3.071–3.328). Sensitivity analysis demonstrated the consistently elevated risks across sepsis with varying severities. We further explored the protective effect of vaccinations among patients with COPD and found the significant protective effect of VZV glycoprotein E (HR 0.724, 95% CI: 0.595–0.882), RSV prefusion F protein-based vaccine (HR 0.676, 95% CI: 0.563–0.812) and influenza vaccine (HR 0.709, 95% CI: 0.649–0.776).

Conclusion:

COPD patients recovering from sepsis remain at increased risk of viral infections, highlighting the importance of targeted preventive strategies, including vaccination.

Plain language summary

Recovering from sepsis increases risk of viral infections in people with COPD

People living with chronic obstructive pulmonary disease (COPD) already have a higher risk of lung infections. This study found that if someone with COPD survives a serious illness called sepsis, they are more likely to get viral infections like influenza, RSV (respiratory syncytial virus), or shingles in the following year. Researchers looked at health records from nearly one million patients and compared those who had sepsis with those who didn’t. Even after matching for age and other health problems, people with COPD who had experienced sepsis were two to three times more likely to get certain viral infections. The most common were RSV and influenza. The study also showed that vaccines can help reduce this risk. People who received vaccines for RSV, influenza, or shingles before getting sepsis had a lower chance of getting infected later. Even after sepsis, getting the flu vaccine was still helpful. These findings highlight the need for better protection strategies—especially vaccinations—for people with COPD who have had sepsis. Doctors and caregivers should closely monitor these patients and encourage them to stay up to date with their vaccines.

Keywords

chronic obstructive pulmonary disease real-world data sepsis vaccination viral infection

Introduction

Chronic obstructive pulmonary disease (COPD) is an increasing global health burden due to the aging population and complex environmental exposure, estimated to affect over 480 million people in 2020 worldwide.^1,2 The prevalence of COPD is estimated to increase by 23% from 2020 to 2050, and COPD is expected to cost the global economy approximately INT$4326 trillion between 2020 and 2050.^1,3 Viral infections, particularly respiratory syncytial virus (RSV) and influenza, play a critical role in the disease course of patients with COPD and may lead to disease exacerbation and increased mortality.^4,5 In addition to COPD, sepsis is another leading health threat worldwide, particularly in vulnerable populations, such as the aging population and patients with airway diseases, including COPD.^2,6

The development of sepsis among patients with COPD has been found to be associated with poor prognoses, such as increased ICU admission and prolonged hospital stays.^2,7,8 Furthermore, previous studies have found that COPD patients who survived after sepsis had higher risks of pneumonia, severe acute exacerbations, and mortality compared with COPD patients without sepsis.^7,8 Recent studies, including our studies, have demonstrated that one of the adverse immune consequences of sepsis was immunoparalysis, a state characterized by dysfunctional immune cells, particularly T cells, that may lead to increased risks for further infection.^9
–12 Given that T cell plays a fundamental role in viral infection, post-septic immunoparalysis was found to be associated with increased susceptibility to distinct viral infections, including viral reactivations and new viral infections such as influenza and RSV.^13
–15 However, the majority of studies were conducted to investigate the viral infection during sepsis, and studies focusing on post-septic viral infection in patients with COPD are still lacking. In the present study, we aim to address the development of distinct viral infections after sepsis and the potential protective effect of vaccination in patients with COPD by using a federated data platform, namely TriNetX.

Materials and methods

Study design and data source

This study is a multicenter, propensity score-matched (PSM) retrospective cohort study utilizing data from the federated Research Network with 92 healthcare organizations (HCOs) on the TriNetX. TriNetX provides a global federated real-world data and analytics platform, and participating HCOs should sign an official agreement to share de-identified patient data for research purposes on the platform. In brief, the TriNetX platform serves as a federated research tool, enabling the integration and analysis of clinical data from multiple healthcare systems in the common data model. We conducted this study on the TriNetX platform through using the aforementioned existing database. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.¹⁶

Study population

We conducted a retrospective cohort analysis using the TriNetX Research Network, covering patient encounters from 1 January 2000 to 31 December 2022. Eligible individuals were adults who (i) received at least one bronchodilator prescription and (ii) carried ⩾3 International Classification of Diseases, Tenth Revision (ICD-10) codes for chronic obstructive pulmonary disease (COPD: J44.x, J44.9) during the study interval. The exclusion criteria were (i) extreme age (<18 or >90 years) and (ii) septic patients with any viral-infection code recorded ⩽7 days after sepsis onset. Patients who experienced sepsis were defined by the ICD code A41.9. In addition, we used ICD-10, medication, and Current Procedural Terminology (CPT) codes to define varying severities of sepsis by bacteremia (ICD: R78.81), shock defined by the usage of vasopressor (RxNorm: 7512), mechanical ventilation (CPT: 1014859) and hemodialysis (CPT: 90935) in the sensitivity analysis.

Outcomes with distinct viral infections

The outcome of interest in this study was the diagnosis of the viral infection within 1 year following the sepsis episode. The index date of septic patients with COPD was the date of diagnosis with sepsis, and the index date in non-septic COPD patients was defined as the date when patients fulfilled the operational COPD definition in our study: ⩾3 COPD diagnoses (ICD-10: J44/J44.9) and concurrent use of bronchodilators. Given that we focused on the development of viral infection after sepsis, we excluded septic patients who had a diagnosis of viral infection within 7 days of sepsis. We used the ICD codes to define the following viral infections, including HSV (B00.x and A60.x), varicella-zoster virus (VZV) (B01.x and B02.x), CMV (B25.x, B27.1), and RSV (J12.1). With regards to the influenza infection, which was mainly diagnosed by a rapid diagnostic test, we hence use Logical Observation Identifiers Names and Codes (LOINC) of positive influenza A (LONIC: 40982-1) and influenza B (LONIC: 34487-9) to define the diagnosis of influenza. To increase the accuracy of diagnosing viral infections, in addition to using ICD codes, we also used laboratory diagnostic codes from LOINC codes for positive viral detections in sterile sites (e.g., cerebrospinal fluid), as well as positive viral IgM antibody for HSV, VZV, CMV, and RSV. Furthermore, we also attempted to assess the protective effects of vaccinations prior to sepsis in VZV, RSV, and influenza infection. Given that viral vaccines were available in different years, the analysis period for assessing the protective effect of vaccination aligned with the availability period of each vaccine.

Statistical analysis

Baseline characteristics were presented as means (standard deviations) for continuous variables and as counts (percentages) for categorical variables. To minimize the potential confounding effect on the association between sepsis and viral infection, the study employed 1:1 greedy nearest-neighbor PSM (caliper: 0.10) for covariates, including (1) demographics (age, sex, ethnicity), (2) comorbidities (diabetes mellitus, heart failure, ischemic heart disease, cerebrovascular disease peripheral vascular disease, bronchiectasis, liver cirrhosis, chronic kidney disease, presence of neoplasm, metastatic solid tumor, and human immunodeficiency virus disease), and (3) medications related to COPD and viral infections, specifically inhaled corticosteroids (ICS) and systemic corticosteroids. A two-tailed p-value <0.05 was considered statistically significant. Hazard ratios (HRs) and 95% confidence intervals were calculated using proportional hazards models, with time-to-event measured over a 1-year follow-up.

Results

Characteristics of the enrolled patients with COPD before and after the matching

A total of 903,683 adult patients with COPD were eligible for analyses, and 12.6% of them had sepsis (Figure 1). Baseline characteristics analysis between COPD patients with sepsis (n = 113,589) and without sepsis (n = 790,094) revealed significant differences prior to PSM. The sources for sepsis consisted of pneumonia (46%), urinary tract infection (29%), skin soft tissue infection (15%), and intra-abdominal infection (7%; Supplemental Table 1). We employed the use of bronchodilators to increase the diagnostic accuracy of COPD, and the proportions of using inhaled anticholinergics, inhaled sympathomimetics, and oral bronchodilators, including xanthine, were 53%, 67%, and 63%, respectively. The sepsis cohort exhibited higher mean age (69.0 ± 12.6 vs 65.7 ± 13.0 years, p < 0.001) and substantially greater comorbidity burden, particularly in type 2 diabetes mellitus (37.3% vs 15.7%, p < 0.001), heart failure (37.2% vs 9.6%, p < 0.001), and ischemic heart disease (39.3% vs 15.7%, p < 0.001). After 1:1 PSM (n = 98,883 in each group), baseline characteristics were balanced between cohorts, as SMD values were below 0.1 (Table 1).

Figure 1.

Flowchart of subject enrollment.

Table 1.

Characteristics between patients with COPD categorized by sepsis in the primary cohort and cohort matched with propensity score-matched cohort.

Variables	Before PSM				1:1 PSM
	Sepsis	Non-Sepsis	p-Value	SMD	Sepsis	Non-Sepsis	p-Value	SMD
	n = 113,589	n = 790,094.			n = 98883	n = 98883	p-Value	SMD
Demographics
Age, years	69.0 ± 12.6	65.7 ± 13.0	<0.001	0.256	68.6 ± 12.7	69.3 ± 12.3	<0.001	0.059
Sex (male)	49.6%	45.1%	<0.001	0.090	48.8%	49.6%	<0.001	0.017
Ethnicities
Caucasian	71.8%	72.2%	0.011	0.008	71.8%	71.9%	0.589	0.002
African American	11.5%	11.3%	0.180	0.004	11.3%	10.8%	0.003	0.013
Hispanic/Latino	3.6%	3.1%	<0.001	0.032	3.6%	3.4%	0.046	0.009
Comorbidities
Type 2 Diabetes mellitus	37.3%	15.7%	<0.001	0.505	33.1%	34.9%	<0.001	0.037
Heart failure	37.2%	9.6%	<0.001	0.689	30.5%	31.2%	<0.001	0.017
Ischemic heart disease	39.3%	15.7%	<0.001	0.547	34.3%	35.0%	0.001	0.015
Cerebrovascular disease	15.0%	5.5%	<0.001	0.316	12.7%	13.3%	0.001	0.015
Peripheral vascular disease	12.5%	4.1%	<0.001	0.308	10.2%	10.4%	0.081	0.008
Bronchiectasis	2.8%	0.9%	<0.001	0.140	2.3%	2.3%	0.976	<0.001
Liver cirrhosis	4.5%	1.2%	<0.001	0.204	3.5%	3.6%	0.264	0.005
Chronic kidney disease	27.2%	6.8%	<0.001	0.565	21.7%	22.9%	<0.001	0.037
Neoplasm	29.9%	14.3%	<0.001	0.381	27.2%	27.7%	0.035	0.009
Metastatic solid tumor	8.2%	1.9%	<0.001	0.287	6.6%	6.7%	0.620	0.002
Human immunodeficiency virus disease	1.0%	0.5%	<0.001	0.059	1.0%	0.9%	0.008	0.012
Concomitant medication
Oral corticosteroid	60.8%	23.1%	<0.001	0.059	55.8%	54.5%	<0.001	0.026
Inhaled corticosteroid	41.6%	14.3%	<0.001	0.639	36.1%	32.9%	<0.001	0.067

Data are shown as mean ± standard deviation and percentages.

COPD, chronic obstructive pulmonary disease; PSM, propensity score-matching; SMD, standard mean difference.

The association between sepsis and 1-year risk of distinct viral infections among patients with COPD

We found COPD patients with sepsis had consistently higher risk for viral infections within 1 year after the sepsis compared with COPD patients without sepsis, with the HR were as follows: HSV 1.936 (95% CI: 1.775–2.112), VZV 3.050 (2.489–3.737), CMV 2.101 (1.832–2.410), RSV 3.297 (3.158–3.443), and Influenza 3.197 (3.071–3.328). We also noted that the risk of RSV and influenza was higher than that of HSV, VZV, and CMV. RSV infection demonstrated a 1-year risk of 7.67% in the sepsis group compared to 2.86% in the non-sepsis group. Similarly, influenza infections were substantially higher in the sepsis group (8.62% vs 3.34%). (Table 2). In the sensitivity analysis of COPD patients with varying severities of sepsis, we observed consistently elevated risks of viral infections across all subgroups, including those with bacteremia (n = 17,575), shock (n = 24,754), mechanical ventilation requirement (n = 18,961), and hemodialysis necessity (n = 6122; Table 3).

Table 2.

One-year risk of distinct viral infections between 98,883 COPD patients with sepsis and propensity score-matched COPD patients without sepsis.

Groups	Risk	HR (95%CI)
HSV
Non-sepsis	0.83%	Reference
Sepsis	1.36%	1.936 (1.775–2.112)
VZV
Non-sepsis	0.13%	Reference
Sepsis	0.34%	3.050 (2.489–3.737)
CMV
Non-sepsis	0.32%	Reference
Sepsis	0.58%	2.101 (1.832–2.410)
RSV
Non-sepsis	2.86%	Reference
Sepsis	7.67%	3.297 (3.158–3.443)
Influenza
Non-sepsis	3.34%	Reference
Sepsis	8.62%	3.197 (3.071–3.328)

CMV, cytomegalovirus; COPD, chronic obstructive pulmonary disease; HSV, herpes simplex virus; RSV, respiratory syncytial virus; VZV, varicella-zoster virus.

Table 3.

Sensitivity analysis of the estimated risk of post-septic viral infections in COPD patients with varying severities of sepsis.

Groups	HR (95%CI)
HSV
Sepsis + bacteremia (n = 17,575)	2.284 (1.909–2.733)
Sepsis + shock (n = 24,754)	2.137 (1.819–2.510)
Sepsis + mechanical ventilation (n = 18,961)	1.863 (1.546–2.245)
Sepsis + hemodialysis (n = 6122)	2.697 (2.007–3.624)
VZV
Sepsis + bacteremia (n = 17,575)	4.081 (2.661–6.260)
Sepsis + shock (n = 24,754)	4.342 (2.996–6.292)
Sepsis + mechanical ventilation (n = 18,961)	3.149 (2.069–4.793)
Sepsis + hemodialysis (n = 6122)	3.727 (1.937–7.171)
CMV
Sepsis + bacteremia (n = 17,575)	2.429 (1.917–3.077)
Sepsis + shock (n = 24,754)	2.900 (2.347–3.583)
Sepsis + mechanical ventilation (n = 18,961)	2.857 (2.267–3.601)
Sepsis + hemodialysis (n = 6122)	1.915 (1.424–2.575)
RSV
Sepsis + bacteremia (n = 17,575)	3.088 (2.816–3.387)
Sepsis + shock (n = 24,754)	3.012 (2.775–3.269)
Sepsis + mechanical ventilation (n = 18,961)	2.556 (2.329–2.806)
Sepsis + hemodialysis (n = 6122)	2.416 (2.094–2.788)
Influenza
Sepsis + bacteremia (n = 17,575)	2.906 (2.667–3.167)
Sepsis + shock (n = 24,754)	3.354 (3.111–3.616)
Sepsis + mechanical ventilation (n = 18,961)	2.404 (2.202–2.625)
Sepsis + hemodialysis (n = 6122)	2.576 (2.259–2.939)

CMV, cytomegalovirus; COPD, chronic obstructive pulmonary disease; HSV, herpes simplex virus; RSV, respiratory syncytial virus; VZV, varicella-zoster virus.

Protective effect of vaccination in distinct viral infections among patients with COPD and septic patients with COPD

We further addressed the protective effects of vaccinations for VZV, RSV, and influenza. We explored the protective effect of vaccinations among patients with COPD and found the significant protective effect of VZV glycoprotein E (HR 0.724, 95% CI: 0.595–0.882), RSV prefusion F protein-based vaccine (HR 0.676, 95% CI: 0.563–0.812) and influenza vaccine (HR 0.709, 95% CI: 0.649–0.776), whereas the protective effect of Zoster virus live vaccine tended to be neutral (HR 1.095, 95% CI: 0.741–1.619; Table 4). We further aimed to explore the protective effect of vaccination after sepsis, although the number of patients with COPD who received vaccinations after sepsis was apparently quite low. We identified a significant protective effect of influenza vaccination after sepsis (HR 0.414, 95% CI 0.333–0.515) and the potential protective trends of vaccination after sepsis in VZV and RSV.

Table 4.

Protective effect of vaccination in distinct viral infections among COPD patients with and without sepsis.

Groups	Vaccinated COPD patients versus non-vaccinated COPD patients			Vaccinated after sepsis among COPD patients versus non-vaccinated septic patients with COPD
	Number	Risk	HR (95%CI)	Number	Risk	HR (95%CI)
VZV^a
Vaccination (–)	39,843	0.6%	Reference	6315	1.4%	Reference
Varicella-zoster virus glycoprotein E (Shingrix)	39,843	0.4%	0.724 (0.595–0.882)	6315	1.1%	0.799 (0.580–1.101)
VZV^b
Vaccination (–)	11,133	0.4%	Reference	3299	0.8%	Reference
Zoster virus vaccine live (Oka/Merck) (Zostavax)	11,133	0.5%	1.095 (0.741–1.619)	3299	1.1%	0.676 (0.406–1.126)
RSV^c
Vaccination (–)	9964	3.8%	Reference	311	11.6%	Reference
RSV prefusion F protein-based vaccine	9964	2.1%	0.676 (0.563–0.812)	311	6.8%	0.910 (0.518–1.599)
Influenza^d
Vaccination (–)	44,803	3.4%	Reference	7869	4.8%	Reference
Influenza vaccine (+)	44,803	2.0%	0.709 (0.649–0.776)	7869	2.0%	0.414 (0.333–0.515)

50–90-year-old, 2017–2022, 5-year risk.

50–90 year-old, 2007–2022, 5-year risk.

20–90-year-old, 2020–2023, 1-year risk.

20–90-year-old, 2000 and 2022, 1-year risk.

Discussion

Patients with COPD, particularly those with sepsis, are vulnerable to viral infections, but the risk of viral infection among COPD patients after sepsis has remained unclear. Our study, utilizing a federated database and PSM approach, demonstrated that COPD patients with sepsis had a significantly higher risk of distinct viral infections compared to COPD patients without sepsis. We validated the protective effect of vaccinations, including VZV, RSV, and influenza, among patients with COPD. Furthermore, we found that vaccinations tended to protect COPD patients from viral infections, particularly influenza, after sepsis. These findings provide clinical evidence with regard to the need for vigilance of viral infections in COPD patients after sepsis.

Sepsis can induce immune suppression, including endotoxin tolerance and impaired energy metabolism of immune cells, including lymphocytes, and lead to increased susceptibility to secondary infections with associated late mortality.^17,18 Previous studies, including our study, have shown the critical role of post-septic T cell dysfunction, which may account for the increased risk of viral infection, given that T cell immunity plays a substantial role in suppressing viral infection and reducing disease severity of viral infection.^9,10,12,19 A number of studies have found the dynamic change of T cell subtypes after sepsis, with an increased proportion of regulatory T cells and decreased numbers of effector T cells in sepsis.^20,21 Moreover, recent studies have demonstrated that the post-septic impaired production of inflammatory cytokines and cellular metabolism of T cells may further lead to newly or reactivated viral infection.^13,18,22 Notably, viral infection may exacerbate the pathogenesis of COPD by causing immune dysregulation and damaging airway structures, worsening symptoms, and increasing bacterial coinfections.^4,23 In patients with COPD, the immune dysfunction may be further exacerbated by an imbalance of regulatory T cells, which suppress effector T cell responses, and by impaired T cell immunity during acute exacerbations.^24,25 In addition, chronic inflammation and oxidative stress in patients with COPD may further lead to immune senescence, reducing the ability to clear pathogens effectively and increasing susceptibility to infections.²⁶ The aforementioned evidence of sepsis-associated impaired immunity, particularly in patients with COPD, indicates the need to address the previously ignored increased risk of viral infection after sepsis, particularly in patients with COPD, as we have shown in this study.

To improve the reliability of viral infection diagnoses, we combined clinical diagnoses and laboratory evidence in this study. In line with our data that the 1-year risk for RSV was 2.86% in non-septic patients with COPD and 7.67% in COPD patients with sepsis, Falsey et al., conducting a 4-year prospective cohort study, found that RSV infection was observed annually in 3%—7% of healthy older adults and in 4%—10% of adults with chronic cardiopulmonary disease.²⁷ With regards to influenza in patients with COPD, one Taiwanese population-based 6-year study found that 9.3% (10,872/117,125) of patients with COPD had influenza.²⁸ In this study, the 1-year risk of influenza among sepsis-COPD and non-sepsis-COPD was 8.62% and 3.34%, respectively. The impact of sepsis on the risk of influenza in patients with COPD was consistent in the sensitivity analyses using distinct severities of sepsis (Table 3). Notably, the protection effects of influenza vaccination were both significant in patients with COPD (HR 0.709, 95% CI 0.649—0.776) and sepsis-COPD patients (HR 0.414, 95% CI 0.333—0.515; Table 4). These real-world data highlight the essential need for vigilance of influenza as well as vaccination in patients with COPD. In the present study, HSV and VZV infections tended to be low, and we postulated that the relatively less severe symptoms of HSV/VZV might lead to underreporting, whereas CMV disease is a rare, severe CMV infection. Importantly, there should be no differential error regarding viral infections between septic patients with COPD and non-sepsis COPD; therefore, the association between HSV/VZV/CMV infection and sepsis in patients with COPD should be reliable. Although the severity of COPD cannot be ascertained in this study, we further conducted the subgroup analyses based on the recent use of systemic and ICS, which should be a surrogate of exacerbation of COPD (Supplemental Tables 2 and 3). Intriguingly, we found that the strength of association between sepsis and viral infections, particularly RSV and influenza, tended to be slightly stronger in COPD patients receiving systemic and ICS. We believe more evidence are warranted to validate this finding. Furthermore, we conducted analysis using distinct washout periods, including 1, 7, 14, 21, and 28 days, and the results were consistent (Supplemental Table 4). Moreover, we performed analysis through using patients hospitalized for etiologies other than sepsis and patients hospitalized for exacerbation of COPD to identify the consistent association between sepsis and increased risk of post-septic viral infections (Supplemental Tables 5 and 6). In brief, we used distinct sensitivity analyses, including stratification by sepsis severity (bacteremia, shock, mechanical ventilation, and hemodialysis), corticosteroid exposure subgroups, alternative washout windows, and comparator cohorts (non-sepsis hospitalizations and COPD exacerbations) to demonstrate the consistent the association of sepsis among patients with COPD with higher 1-year risks of HSV, VZV, CMV, RSV, and influenza.

Intriguingly, a number of studies have found an association between sepsis and severe viral infections during hospitalization, but the long-term post-septic risk of viral infection is underexplored. Cabler et al. conducted a multicenter study among 401 pediatric septic patients and found that 47.6% (191/401) of them had viral DNAemia with at least one virus.²⁹ Ong et al. conducted a study to address the prevalence of previously immunocompetent patients with septic shock in the Netherlands and found that 68% (223/329) of patients had viremia. Furthermore, 34% (112/329) of previously immunocompetent patients with septic shock had multiple concurrent viremia events.³⁰ Recently, Charalampous et al. used a routine metagenomic approach among critically ill patients with a respiratory infection and revealed the previously ignored colonization/infection of viruses, such as HSV, Coronavirus (COVID-19)-19, and CMV.³¹ The majority of aforementioned studies aim to investigate viral colonization/infection during sepsis; therefore, studies focusing on post-septic viral infection are required. These evidence indicated the increased risk of viral infection in patients with sepsis, and our study, focusing on patients with COPD, further extends the follow-up period to 1 year and provides clinical evidence of the need for studies with regard to long-term viral infection prevention strategies, including vaccination among post-septic COPD patients.

Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines advocate the administration of vaccines against influenza and RSV, as well as prophylactic measures in patients with stable COPD.² Cherukuri et al. identified the immunosenescence of RSV with reduced RSV-specific T cell responses in adults 65 years old and older.³² In this study, we focused on patients with COPD, with the average age being 68–69. Therefore, the 1-year risk of RSV was relatively high, particularly for those who had sepsis (7.67%). In addition to the immunosenescence, the increased risk of RSV among COPD patients with sepsis may also result from sepsis-associated T cell immunity impairment. In line with the evidence of the protective effect of RSV vaccination in older adults, we found a significant protective effect of vaccination of RSV in patients with COPD (HR 0.676, 95% CI 0.563–0.812).^27,33 Our study focuses on the increased risk of post-sepsis viral infections in COPD patients, emphasizing the importance of preventive measures such as vaccination. While the protective effects of vaccinations, particularly for influenza and RSV, prior to sepsis were observed in this study, the optimal timing of vaccinations post-sepsis remains beyond the scope of this study, as vaccinations were administered prior to sepsis among COPD patients in this study. However, we note that novel vaccine modalities, such as self-amplifying RNA (saRNA) for RSV, may offer potential benefits in boosting immune responses among septic patients with impaired T cell immunity in the future.³⁴ Further research is needed to explore the efficacy and optimal timing of distinct vaccinations among COPD patients who experienced sepsis.

While randomized controlled trials (RCTs) are the gold standard for establishing causal relationships through randomization, their controlled environments and patient selection may limit generalizability.³⁵ In contrast, real-world data (RWD), as demonstrated using the TriNetX platform in this study, provides insights into diverse, heterogeneous patient populations reflective of actual clinical settings.³⁶ However, the observational nature of RWD may introduce potential confounders, and advanced statistical techniques like PSM, which was used in this study, reduced standardized mean differences, ensuring minimal bias and validating the association between sepsis and subsequent viral infections. Collectively, RWD may complement RCT and provide real-world evidence for guiding clinical practice and improving patient outcomes.

In this study, we utilized ICD codes to identify sepsis, COPD, and viral infections. While ICD coding has known limitations, one recent study reported moderate accuracy for sepsis diagnosis, with sensitivity and specificity around 75% and 85%, respectively.³⁷ For COPD, validation studies have shown that ICD codes can be less reliable, with approximately 64% of cases confirmed by physician verification.³⁸ However, we employed stringent criteria using three inpatient diagnoses with COPD and bronchodilators. In addition, to enhance the accuracy of viral infection diagnoses, we supplemented ICD codes with laboratory data, including positive viral detections and IgM antibody tests. These combined approaches aimed to improve the robustness of our findings despite the inherent challenges with ICD coding. Our study exhibited large effect sizes for viral infections following sepsis (RSV HR 3.297; influenza HR 3.197), and we demonstrated the consistent finding across sepsis severity strata and corticosteroid‑treated COPD subgroups. Concordant results from additional sensitivity analyses employing alternative washout periods further reinforce the robustness of the observed association.

This study has limitations. First, being similar to studies using the claim database, ICD codes might be inaccurate, although we have used laboratory diagnostic codes to increase the accuracy of viral infection. Second, the TriNetX network mainly consists of academic medical centers or research-focused HCOs in the United States, and more studies are warranted for generalizability to other populations, including distinct centers and non-U.S. countries.³⁹ Third, we enrolled patients with COPD diagnosed between 2000 and 2022; therefore, COVID-19 was not analyzed in this study.

Conclusion

In conclusion, we used the federated database to identify the increased one-year risk of viral infections, including HSV, VZV, CMV, RSV, and influenza, after sepsis among patients with COPD. Furthermore, we demonstrate the protective effect of vaccinations for influenza and RSV. Our data indicate the need for vigilance of viral infections in COPD patients who had sepsis and support prioritizing vaccination in COPD management strategies, particularly the need for vaccinations among patients with stable COPD. Further prospective studies are needed to validate our findings and elucidate the post-septic immunological status in patients with COPD.

Supplemental Material

sj-docx-1-tar-10.1177_17534666251385679 – Supplemental material for Higher risk of viral infections in chronic obstructive pulmonary disease patients recovering from sepsis compared to non-sepsis patients: a propensity score-matched observational study

Supplemental material, sj-docx-1-tar-10.1177_17534666251385679 for Higher risk of viral infections in chronic obstructive pulmonary disease patients recovering from sepsis compared to non-sepsis patients: a propensity score-matched observational study by Tzu-I Chuang and Wen-Cheng Chao in Therapeutic Advances in Respiratory Disease

Supplemental Material

sj-pdf-2-tar-10.1177_17534666251385679 – Supplemental material for Higher risk of viral infections in chronic obstructive pulmonary disease patients recovering from sepsis compared to non-sepsis patients: a propensity score-matched observational study

Supplemental material, sj-pdf-2-tar-10.1177_17534666251385679 for Higher risk of viral infections in chronic obstructive pulmonary disease patients recovering from sepsis compared to non-sepsis patients: a propensity score-matched observational study by Tzu-I Chuang and Wen-Cheng Chao in Therapeutic Advances in Respiratory Disease

Footnotes

Acknowledgements

Portions of these findings were previously disseminated in abstract form at the 2025 American Thoracic Society International Conference (https://www.atsjournals.org/doi/abs/10.1164/ajrccm.2025.211.Abstracts.A2093) and Review of ATS 2025 International Conference at the American Medical Journal Respiratory (Respir AMJ. 2025;3 [1]:43-44. ).

Declarations

ORCID iD

Wen-Cheng Chao

Supplemental material

Supplemental material for this article is available online.

References

Boers

Barrett

, et al. Global burden of chronic obstructive pulmonary disease through 2050. JAMA Network Open 2023; 6(12): e2346598.

Venkatesan

GOLD COPD report: 2024 update. Lancet Respir Med 2024; 12(1): 15–16.

Chen

Kuhn

Prettner

, et al. The global economic burden of chronic obstructive pulmonary disease for 204 countries and territories in 2020-50: a health-augmented macroeconomic modelling study. Lancet Glob Health 2023; 11(8): e1183–e1193.

Guo-Parke

Linden

Weldon

, et al. Mechanisms of virus-induced airway immunity dysfunction in the pathogenesis of COPD disease, progression, and exacerbation. Front Immunol 2020; 11: 1205.

Ackerson

Tseng

, et al. Severe morbidity and mortality associated with respiratory syncytial virus versus influenza infection in hospitalized older adults. Clin Infect Dis 2019; 69(2): 197–203.

Goodacre

Fuller

Conroy

, et al. Diagnosis and management of sepsis in the older adult. BMJ 2023; 382: e075585.

Chen

, et al. Association between chronic obstructive pulmonary disease and 28-day mortality in patients with sepsis: a retrospective study based on the MIMIC-III database. BMC Pulm Med 2023; 23(1): 435.

Chen

Lai

Wang

, et al. The Impact of Sepsis on the Outcomes of COPD Patients: a Population-Based Cohort Study. J Clin Med 2018; 7(11): 393.

Yende

Kellum

Talisa

, et al. Long-term host immune response trajectories among hospitalized patients with sepsis. JAMA Network Open 2019; 2(8): e198686.

10.

Martin

Badovinac

Griffith

TS.

CD4 T cell responses and the sepsis-induced immunoparalysis state. Front Immunol 2020; 11: 1364.

11.

van der Slikke

Hancock

REW

, et al. Exploring the pathophysiology of post-sepsis syndrome to identify therapeutic opportunities. EBioMedicine 2020; 61: 103044.

12.

Chen

Jiang

, et al. Whole transcriptome analysis to explore the impaired immunological features in critically ill elderly patients with sepsis. J Transl Med 2023; 21(1): 141.

13.

Jensen

Sjaastad

Griffith

, et al. Sepsis-induced T cell immunoparalysis: the ins and outs of impaired T cell immunity. J Immunol 2018; 200(5): 1543–1553.

14.

Cheng

Chen

Jiang

, et al. Using RNA-Seq to investigate immune-metabolism features in immunocompromised patients with sepsis. Front Med 2021; 8: 747263.

15.

Zhang

Yuan

, et al. The potential immunological mechanisms of sepsis. Front Immunol 2024; 15: 1434688.

16.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 2007; 4(10): e296.

17.

van Vught

Klein Klouwenberg

Spitoni

, et al. Incidence, risk factors, and attributable mortality of secondary infections in the intensive care unit after admission for sepsis. JAMA 2016; 315(14): 1469–1479.

18.

Cheng

Scicluna

Arts

, et al. Broad defects in the energy metabolism of leukocytes underlie immunoparalysis in sepsis. Nat Immunol 2016; 17(4): 406–413.

19.

Torres

Pickkers

van der Poll

Sepsis-induced immunosuppression. Annu Rev Physiol 2022; 84: 157–181.

20.

Sharma

Yang

Matsuo

, et al. Differential alterations of tissue T-cell subsets after sepsis. Immunol Lett 2015; 168(1): 41–50.

21.

Jiang

Chen

Sun

, et al. Unveiling dynamic changes and regulatory mechanisms of T cell subsets in sepsis pathogenesis. Immunotargets Ther 2024; 13: 29–44.

22.

McBride

Owen

Stothers

, et al. The metabolic basis of immune dysfunction following sepsis and trauma. Front Immunol 2020; 11: 1043.

23.

Planer

Morrisey

EE.

After the storm: regeneration, repair, and reestablishment of homeostasis between the alveolar epithelium and innate immune system following viral lung injury. Annu Rev Pathol 2023; 18: 337–359.

24.

Hou

Sun

Role of regulatory T cells in disturbed immune homeostasis in patients with chronic obstructive pulmonary disease. Front Immunol 2020; 11: 723.

25.

Xiong

Zhu

, et al. T-cell immune status in patients with acute exacerbation of chronic obstructive pulmonary disease: a case-control study. Front Med 2025; 12: 1433844.

26.

Bhat

Panzica

Kalathil

, et al. Immune dysfunction in patients with chronic obstructive pulmonary disease. Ann Am Thorac Soc 2015; 12(Suppl 2): S169–S175.

27.

Falsey

Hennessey

Formica

, et al. Respiratory syncytial virus infection in elderly and high-risk adults. N Engl J Med 2005; 352(17): 1749–1759.

28.

Liao

Chen

Shen

, et al. The influence of influenza virus infections in patients with chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis 2022; 17: 2253–2261.

29.

Cabler

Storch

Weinberg

, et al. Viral DNAemia and DNA virus seropositivity and mortality in pediatric sepsis. JAMA Netw Open 2024; 7(2): e240383.

30.

Ong

DSY

MJM

Bonten

Spitoni

, et al. Epidemiology of multiple herpes viremia in previously immunocompetent patients with septic shock. Clin Infect Dis 2017; 64(9): 1204–1210.

31.

Charalampous

Alcolea-Medina

Snell

, et al. Routine metagenomics service for ICU patients with respiratory infection. Am J Respir Crit Care Med 2024; 209(2): 164–174.

32.

Cherukuri

Patton

Gasser

, et al. Adults 65 years old and older have reduced numbers of functional memory T cells to respiratory syncytial virus fusion protein. Clin Vaccine Immunol 2013; 20(2): 239–247.

33.

Wildenbeest

Lowe

Standing

, et al. Respiratory syncytial virus infections in adults: a narrative review. Lancet Respir Med 2024; 12(10): 822–836.

34.

Vijayan

Vogels

Groppo

, et al. A self-amplifying RNA RSV prefusion-F vaccine elicits potent immunity in pre-exposed and naive non-human primates. Nat Commun 2024; 15(1): 9884.

35.

Concato

Corrigan-Curay

Real-world evidence – where are we now?

N Engl J Med 2022; 386(18): 1680–1682.

36.

Hong

JC.

Strategies to turn real-world data into real-world knowledge. JAMA Netw Open 2021; 4(10): e2128045.

37.

Kumar

Hammond

Grattan

, et al. Accuracy of international classification of disease coding methods to estimate sepsis epidemiology: a scoping review. J Intensive Care Med 2024; 39(1): 3–11.

38.

Ruan

Huang

, et al. Validity of ICD9-CM codes to diagnose chronic obstructive pulmonary disease from National Health Insurance claim data in Taiwan. Int J Chron Obstruct Pulmon Dis 2018; 13: 3055–3063.

39.

Palchuk

London

Perez-Rey

, et al. A global federated real-world data and analytics platform for research. JAMIA Open 2023; 6(2): ooad035.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.15 MB