https://orcid.org/0009-0009-3328-156X
Abstract
The purpose of this study was to establish a nomogram for predicting community-acquired pneumonia (CAP) in hospitalized patients with acute exacerbations of chronic obstructive pulmonary disease (AECOPD). The retrospective cohort study included 1249 hospitalized patients with AECOPD between January 2012 and December 2019. The patients were divided into pneumonia-complicating AECOPD (pAECOPD) and non-pneumonic AECOPD (npAECOPD) groups. The least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression were utilized to identify prognostic factors. A prognostic nomogram model was established, and the bootstrap method was used for internal validation. Discrimination and calibration of the nomogram model were evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). Logistic and LASSO regression analysis showed that C-reactive protein (CRP) >10 mg/L, albumin (Alb) <40 g/L, alanine transferase (ALT) >50 U/L, fever, bronchiectasis, asthma, previous hospitalization for pAECOPD in the past year (Pre-H for pAECOPD), and age-adjusted Charlson score (aCCI) ≥6 were independent predictors of pAECOPD. The area under the ROC curve (AUC) of the nomogram model was 0.712 (95% CI: 0.682–0.741). The corrected AUC of internal validation was 0.700. The model had well-fitted calibration curves and good clinical usability DCA curve. A nomogram model was developed to assist clinicians in predicting the risk of pAECOPD.
China Clinical Trials Registry: ChiCTR2000039959
Introduction
Chronic obstructive pulmonary disease (COPD) is a heterogeneous lung condition characterized by chronic respiratory symptoms due to abnormalities of the airways and/or alveoli that cause persistent, often progressive, airflow obstruction [Citation1]. It was noted as the third leading cause of death from disease worldwide in 2020 [Citation2]. Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) lead to an increased rate of lung function decline [Citation3], decline in disease-related quality of life [Citation4], heavy disease burden [Citation5], and even death [Citation6]. Moderate to severe AECOPD often requires emergency medical care or even hospitalization.
Bacterial and viral infections are the main causes of AECOPD [Citation7]. Some cases of infection-induced AECOPD present with community-acquired pneumonia (CAP), which is called pAECOPD [Citation8]. Studies have shown that patients with pAECOPD have a longer hospital stay than those with non-pneumonic AECOPD (npAECOPD), and the proportion of pAECOPD patients who needed to be admitted to the intensive care unit and treated with auxiliary ventilation and dialysis is higher, requiring overall more medical and health resources [Citation9]. Imaging is an indispensable criterion for the diagnosis of CAP [Citation10,Citation11]. However, X-ray chest radiographs are of limited value in the diagnosis of CAP and can be easily missed. A study found that among those with clinically suspected CAP but without infiltrate on chest radiograph, one-third of patients had parenchymal infiltrate on chest computed tomography (CT) consistent with CAP [Citation12]. Another study showed that inpatients who had a chest CT in the emergency department with an early diagnosis of pneumonia resulted in a shortened hospital stay [Citation13]. However, the routine use of CT can lead to increased radiation exposure and medical costs [Citation12]. Moreover, in critically ill patients, chest CT is often not performed in a timely manner because of the severity of the disease. Therefore, there are limitations in using chest CT for the early diagnosis of pAECOPD.
One small study examined multiple biomarkers to predict the diagnosis of pAECOPD and found that only C-reactive protein (CRP) was an ideal predictor [Citation14]. Although highly sensitive, CRP has poor specificity and is elevated in a variety of conditions, including bacterial and viral infections, as well as trauma. Therefore, a more scientific predictive model for pAECOPD is needed in order to help with early identification of pAECOPD in the clinic so that targeted treatment and care measures can be taken. Our study applies logistic regression to develop a nomogram to predict pAECOPD using the routine laboratory examination, comorbidities and other relevant clinical information in patients hospitalized with AECOPD.
Materials and methods
Population and study design
In this retrospective cohort study, patients with AECOPD hospitalized at the Department of Pulmonary and Critical Care Medicine, Beijing Luhe Hospital, Capital Medical University, from January 2012 to December 2019 were enrolled. Data for patients with repeated hospitalizations was only recorded from the last hospitalization. Patients who were hospitalized for less than 24 h were excluded from the study. This study was performed in accordance with the principles of the Declaration of Helsinki. The study was approved by the Ethics Committee of Beijing Luhe Hospital, Capital Medical University (permit numbers: 2020-LHKY-014-03). The research was a non-interventional observational study, and the data used were anonymous. Therefore, the Ethics Committee of Beijing Luhe Hospital of Capital Medical University issued an informed consent waiver.
Data collection and definitions
The following clinical data were collected: demographic characteristics and general clinical information including, sex, age, smoking history, duration of COPD, long-term oxygen therapy (LTOT), home noninvasive ventilation (NIV), previous hospitalization for AECOPD in the past year (Pre-H for AECOPD), previous hospitalization for pAECOPD in the past year (Pre-H for pAECOPD), age-adjusted Charlson comorbidity index (aCCI) [Citation15]. Respiratory related complications and comorbidities, including hydrothorax, respiratory failure, pulmonary encephalopathy, cor pulmonale, asthma, bronchiectasis, obstructive sleep apnea hypopnea syndrome (OSAHS) were also collected, as well as, routine laboratory test data, days of exacerbation, Anthonisen type, length of stay (LOS), cost of hospitalization.
The diagnosis of COPD was based on the Global Initiative for Chronic Obstructive Lung Disease (GOLD) strategy [Citation16]. The AECOPD defined as an event characterized by increased dyspnea and/or cough and sputum that worsens in < 14 days [Citation16]. The diagnostic criteria for CAP were as follows: A. Onset in community. B. Relevant clinical manifestations of pneumonia: (1) new onset of cough or expectoration, or aggravation of existing symptoms of respiratory tract diseases, with or without purulent sputum, chest pain, dyspnea, or hemoptysis; (2) fever; (3) signs of pulmonary consolidation and/or moist rales; (4) peripheral white blood cell count (WBC) > 10 × 109/L or < 4 × 109/L, with or without a left shift. C. Chest radiograph showing new patchy infiltrates, lobar or segmental consolidation, ground‐glass opacities, or interstitial changes, with or without pleural effusion. Clinical diagnosis can be established if a patient satisfies Criterion A, Criterion C and any one condition of Criterion B and meanwhile, tuberculosis, pulmonary tumor, non‐infectious interstitial lung disease, pulmonary edema, atelectasis, pulmonary embolism, pulmonary eosinophilia and pulmonary vasculitis are all excluded [Citation17].
All reviewed patients underwent chest CT before hospitalization. The chest CT was reviewed independently by a radiologist and a respiratory specialist. The radiologist issued a report on the chest CT and the respiratory physician reviewed the films retrospectively. In case of disagreement, a second respiratory physician reviewed the films to determine whether the imaging was considered characteristic of pneumonia. Echocardiography: left ventricular (LV) enlargement, right ventricular hypertrophy (RVH), right ventricular (RV) enlargement, ejection fractions (EF).
Anthonisen type: The presence of dyspnea, increased sputum and increased purulent sputum was defined as a type I exacerbation. Type II exacerbation was defined as occurring when two of these three symptoms were present. The presence of only one was defined as type III [Citation18].
Statistical analysis
R Studio (version 4.1.2) was used for statistical analysis data . Multiple imputation was performed using the "mice package" for variables with less than 30% missing data. All included patients were divided into two groups: npAECOPD and pAECOPD. Laboratory test data were treated as categorical variables according to hospital laboratory reference values. Age-corrected Charlson comorbidity index was grouped into <6 score and ≥6 score subgroups, according to previous literature [Citation15, Citation19]. Categorical variables were expressed as counts and percentages, and their differences were analyzed using the χ2 test or Fisher’s exact test when necessary. Normally distributed continuous variables were described as mean ± standard deviation (SD), and analyzed using the t-test. Skewed variables were described as median (interquartile ranges, IQR), and analyzed using the Mann-Whitney U test.
The least absolute shrinkage and selection operator (LASSO) regression analysis was utilized to identify significant prognostic factors for pAECOPD, that were then incorporated into a multivariate logistic regression model to screen for the independent predictors of pAECOPD. The above independent predictors were used to construct the nomogram model. The bootstrap method was used for internal validation. Discrimination and calibration of the nomogram model were evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) curve, with p < 0.05 considered to be statistically significant.
Results
Baseline characteristics of the subjects
A total of 2195 patients with AECOPD hospitalized at the Department of Pulmonary and Critical Care Medicine, Beijing Luhe Hospital, Capital Medical University, from January 2012 to December 2019 were screened. There were 1249 patients who met the inclusion criteria (). Their mean age was 73.6 ± 9.2 years, and there were 880 male and 369 female patients. The pAECOPD group had 387 patients (31.0%), including 271 male patients and 116 female patients, and their mean age was 74.7 ± 9.1 years. There were 862 patients (69.0%) in the npAECOPD group, including 609 male patients and 253 female patients, and their mean age was 73.1 ± 9.2 ().
Figure 1. Study flow diagram.
Table 1. Demographic and baseline clinical characteristics of AECOPD patients.
Compared with the npAECOPD group, the pAECOPD group patients were older (74.7 years vs 73.1 years, p = 0.005), had a longer hospital stay (9 days vs 8 days, p < 0.001) and higher hospital costs (15,542.2 RMB vs 13,631.6 RMB, p < 0.001). No differences were found between the two groups in sex and smoking history ().
The pAECOPD group had higher rates of hydrothorax (21.2% vs. 4.9%, p < 0.001), respiratory failure (40.1% vs. 29.1%, p < 0.001) and bronchiectasis (21.4% vs. 6.7%, p < 0.001) as compared with the npAECOPD group, but also had lower rates of asthma (10.9% vs. 21.0%, p = 0.001). The proportion of patients with aCCI scores ≥6 score was also higher in the pAECOPD patients than in the npAECOPD patients (43.2% vs 33.4%, p = 0.001) ().
Table 2. Complications and comorbidities of AECOPD patients.
In comparison to the npAECOPD group, the pAECOPD group had higher proportions of fever (23.5% vs 15.1%, p < 0.001), CRP > 10 mg/L (68.0% vs 45.1%, p < 0.001), albumin (Alb) < 40 g/L (84.0% vs 71.6%, p < 0.001), PaCO2 > 50 mmHg (35.9% vs 25.4%, p < 0.001) and SaO2 < 95% (39.8% vs 32.9%, p = 0.023), but they also had lower of alanine aminotransferase (ALT) > 50 U/L (3.1% vs 6.1%, p = 0.035), calcium (Ca) < 2.11 mmol/L (50.9% vs 61.0%, p < 0.001), and triglyceride (TG) > 1.70 mmol/L (4.4% vs 8.1%, p = 0.023) blood levels ().
Table 3. Cardiac ultrasound and laboratory findings of AECOPD patients.
Table 4. Arterial blood gas analysis of AECOPD patients.
In the microbial etiology, Acinetobacter baumannii (3.1% vs 1.3%, p = 0.047) was more prevalent in pAECOPD group, as compared with the npAECOPD group. No differences were found between the two groups in other microbial causes ().
Table 5. Microbiologic features of AECOPD patients.
Nomogram development and performance
The variables were initially screened by LASSO regression analysis. The value of λ, which is one standard error away from the minimum mean square error (λ1se), was 0.02668155, and 10 variables were screened (), including elevated CRP (CRP > 10 mg/L), hypoproteinemia (Alb < 40 g/L), elevated ALT (ALT > 50 U/L), elevated PaCO2 (PaCO2≥50 mmHg), fever, bronchial dilatation, asthma, respiratory failure, previous hospitalization for pAECOPD in the past year (Pre-H for pAECOPD), and high aCCI score (aCCI ≥6 score).
Figure 2. Outcome of LASSO regression (A, B). Outcome of logistic regression (C).
Multivariate logistic regression was performed for the primary screening variables, of which, eight variables were identified as independent predictors: elevated CRP, hypoproteinemia, elevated ALT, fever, bronchiectasis, asthma, Pre-H for pAECOPD, and high aCCI score. Among these eight variables, ALT elevation and asthma were found as independent protective factors, and the rest were independent risk factors (). The nomogram was drawn based on the regression results (), and each variable was assigned a corresponding score based on the regression coefficients. The corresponding to the score on the uppermost scoring axis, and the scores of the eight predictor variables were summed to give a total score, and the value corresponding to the bottom axis was the predictive value for the occurrence of pAECOPD.
Figure 3. A nomogram prediction model for diagnosis of pAECOPD.
The ROC curves were plotted for the individual predictor variables and the combination model, and the area under the curve (AUC) was calculated to evaluate the predictive efficiency of the model. The AUC value of the combination model was 0.712 (95% CI: 0.682–0.741), which was superior to each individual variable (). The Barier score of the combination model was 0.187, and the Hosmer-Lemeshow good of fit test suggesting a good fit (X-squared = 13.783, p = 0.08759). DCA and calibration curves indicated that the cohort had good agreement (). The combination model had a sensitivity of 0.665, specificity of 0.672, and Youden’s index of 0.326. The corrected AUC of internal validation was 0.700, and the corrected Barier score was 0.193, suggesting a good internal validation discriminant performance.
Figure 4. The ROC curves of the combination model and variables.
Figure 5. Decision curve analysis of the combination model (A). Clinical impact curve of the combination model (B). Calibration plots of the combination model (C).
Discussion
pAECOPD is a type of AECOPD, and the limitations of imaging lead to difficulties in definitively diagnosing pAECOPD [Citation12]. We developed a clinical prediction model for early identification of pAECOPD using comprehensive clinical data related to routine laboratory tests, complications and comorbidities. After screening, the model finally included eight variables, including elevated CRP, hypoproteinemia, elevated ALT, fever, bronchiectasis, asthma, Pre-H for pAECOPD, and high aCCI score. Moreover, the model was internally validated for good discriminatory performance.
Studies describing pAECOPD predictive models are scarce. In a recent study that included 145 patients with AECOPD, CRP was found to be an independent predictor of pAECOPD by logistic regression of multiple biomarkers in patients [Citation14]. However, the sample size of this study was small, and CRP as a single indicator has limited value in predicting the diagnosis of pneumonia as it is a nonspecific inflammatory marker [Citation20]. Our study included not only commonly used biomarkers, but also other relevant clinical data such as complications and comorbidities, and ultimately identified eight independent predictors of pAECOPD. We then constructed a combination model using these eight variables which produced better AUC values than each individual variable alone.
In our model, CRP was an independent risk factor for predicting pAECOPD with an AUC value of 0.614, which is consistent with two other studies [Citation14, Citation21]. However, in both studies, the cutoff values of CRP were 158 mg/L and 25.7 mg/L, respectively, which were higher values than in our study. We speculate that this may be due to the influence of inconsistent inclusion and exclusion criteria in the studies. For example, Halıcı and his colleagues [Citation21], excluded all patients treated with prehospital antibiotics, thus possibly affecting CRP levels. Beside in our study model, hypoproteinemia was an independent risk factor for the development of pAECOPD, suggesting that hypoproteinemia is both a consequence of and a risk factor for infection [Citation22].
Studies have shown that, if AECOPD is caused by a bacterial infection, the bacterial acute exacerbation (AE) were more likely to be repeated in subsequent exacerbations within a subject, that may be related to imbalance and colonization of the lung microbiome [Citation23]. Another study found that in patients with COPD infection with respiratory viruses increases the microbial load of the lower respiratory tract, which predisposes to secondary respiratory bacterial infections [Citation24], and increases the risk of the next exacerbation [Citation25]. In our study, the proportion of Pre-H for pAECOPD patients was found to be higher and an important predictor in the model (OR = 2.53). Thus, our research suggests that whether pAECOPD being caused by bacterial or viral, the risk of the next AE for pAECOPD is increased. Once secondary bronchiectasis occurs in COPD, bacterial colonization will occur that will easily lead to bacterial caused AE [Citation26]. This was also confirmed by our model.
ALT is reduced in the elderly with combined frailty, disability and muscle loss [Citation27] , and there is an increased risk of CAP [Citation28]. In our study, we found that those with low ALT were more likely to develop pAECOPD. Since ALT measurements were qualitative variables in our study, at less than 50 U/L, we did not explore the cutoff value of low ALT to predict pAECOPD, and future studies will be needed to clarify this. COPD predominates in the elderly, and comorbidities are not only the risk factors for AECOPD, but also affect patient prognosis [Citation29]. A study has shown that COPD patients with comorbidities are more likely to experience deterioration induced by infection, and the more comorbidities, the greater the risk of infection [Citation30]. The aCCI score is a standard scale based on comorbidity and age that is widely used to predict patient prognosis [Citation15]. Our study found that a high aCCI score is a risk factor for pAECOPD. These results indicate that aCCI score not only predicts patient prognosis [Citation15], but also predicts the risk of pAECOPD.
Our study showed a higher proportion of patients with asthma in the npAECOPD group, which is consistent with the findings of Søgaard and colleagues [Citation9]. Moreover, in our model, asthma was a protective factor for pAECOPD. It has been shown that ICS increases the incidence of pneumonia in COPD patients [Citation31,Citation32], but in COPD patients with asthma features or COPD-asthma overlap (ACO), the use of inhaled corticosteroids (ICS) instead decreases the rate of pneumonia hospitalization [Citation33]. We speculate that this may be related to the fact that ACO patients induce a higher proportion of eosinophils in sputum [Citation34], and respond better to ICS treatment [Citation34], resulting in lower rate of AE [Citation35].
There are some limitations in this study. (1) This study is a single-center study, and although internal validation was performed using Bootstrap method, no external validation was performed. (2) This study is a retrospective study, which may produce selective bias and missing data. To address this limitation, we tried to include all patients who met the criteria and performed multiple imputation for data with less than 30% missingness. (3) Lung function data were not included in our study due to the high amount of missing data. However, arterial blood gas analysis indexes that respond to the severity of the disease were included which, to some extent, compensated for the impact of missing pulmonary function on the model. (4) In the present study, procalcitonin (PCT) was not included as a potential predictor also due to missing PCT data. However, this omission seems reasonable given that previous studies have shown that the diagnostic value of PCT is not superior to that of CRP in predicting pAECOPD [Citation14, Citation36].
Conclusion
We developed a clinical prediction model for the early identification of pAECOPD based on clinical data related to routine laboratory tests, complications and comorbidities. Eight variables were screened, including elevated CRP, hypoproteinemia, elevated ALT, fever, bronchiectasis, asthma, Pre-H for pAECOPD, and high aCCI score. Application of this model can help clinicians in the early identification of factors associated with pAECOPD, leading to timely and targeted treatment and care measures ultimately resulting in better outcomes for the patient.
Data availability statement
The datasets used in the current study are available from the corresponding author on reasonable request.
Disclosure statement
The author reports no conflicts of interest in this work.
Additional information
Funding
References
- Labaki WW, Rosenberg SR. Chronic obstructive pulmonary disease screening. Ann Intern Med. 2020;173(3):ITC17–ITC32. doi: 10.7326/AITC202008040.
- Soriano JB, Kendrick PJ, Paulson KR, et al. Prevalence and attributable health burden of chronic respiratory diseases, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet Respirat Med. 2020;8(6):585–596. doi: 10.1016/S2213-2600(20)30105-3.
- Donaldson GC, Seemungal TAR, Bhowmik A, et al. Relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax. 2002;57(10):847–852. doi: 10.1136/thorax.57.10.847.
- Seemungal TA, Donaldson GC, Paul EA, et al. Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 1998;157(5 Pt 1):1418–1422. doi: 10.1164/ajrccm.157.5.9709032.
- Hilleman DE, Dewan N, Malesker M, et al. Pharmacoeconomic evaluation of COPD. Chest. 2000;118(5):1278–1285. doi: 10.1378/chest.118.5.1278.
- Soler-Cataluna JJ. Severe acute exacerbations and mortality in patients with chronic obstructive pulmonary disease. Thorax. 2005;60(11):925–931. doi: 10.1136/thx.2005.040527.
- Bafadhel M, McKenna S, Terry S, et al. Acute exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2011;184(6):662–671. doi: 10.1164/rccm.201104-0597OC.
- Steer J, Norman EM, Afolabi OA, et al. Dyspnoea severity and pneumonia as predictors of in-hospital mortality and early readmission in acute exacerbations of COPD. Thorax. 2012;67(2):117–121. doi: 10.1136/thoraxjnl-2011-200332.
- Søgaard M, Madsen M, Løkke A, et al. Incidence and outcomes of patients hospitalized with COPD exacerbation with and without pneumonia. COPD. 2016;11:455–465. doi: 10.2147/COPD.S96179.
- Mandell LA, Wunderink RG, Anzueto A, et al. Infectious diseases society of america/American thoracic society consensus guidelines on the management of Community-Acquired pneumonia in adults. Clin Infect Dis. 2007;44(Supplement_2):S27–S72. doi: 10.1086/511159.
- Lim WS, Baudouin SV, George RC, et al. BTS guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax. 2009;64(Suppl 3):iii1–iii55. doi: 10.1136/thx.2009.121434.
- Claessens Y, Debray M, Tubach F, et al. Early chest computed tomography scan to assist diagnosis and guide treatment decision for suspected community-acquired pneumonia. Am J Respir Crit Care Med. 2015;192(8):974–982. doi: 10.1164/rccm.201501-0017OC.
- Claessens Y, Berthier F, Baqué-Juston M, et al. Early chest CT-scan in emergency patients affected by community-acquired pneumonia is associated with improved diagnosis consistency. Eur J Emerg Med. 2022;29(6):417–420. doi: 10.1097/MEJ.0000000000000955.
- Mou S, Zhang W, Deng Y, et al. Comparison of CRP, procalcitonin, neutrophil counts, eosinophil counts, sTREM-1, and OPN between pneumonic and nonpneumonic exacerbations in COPD patients. Can Respir J. 2022;2022:1–9. doi: 10.1155/2022/7609083.
- C M, S TP, P J, et al. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47(11):1245–1251.
- Global Initiative for Chronic Obstructive Lung Disease (GOLD). Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease; 2020. www.goldcopd.org.
- Cao B, Huang Y, She DY, et al. Diagnosis and treatment of community-acquired pneumonia in adults: 2016 clinical practice guidelines by the chinese thoracic society, chinese medical association. Clin Respir J. 2018;12(4):1320–1360. doi: 10.1111/crj.12674.
- Anthonisen NR, Manfreda J, Warren CP, et al. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med. 1987;106(2):196–204. doi: 10.7326/0003-4819-106-2-196.
- Dessai SB, Fasal R, Dipin J, et al. Age-adjusted charlson comorbidity index and 30-day morbidity in pelvic surgeries. South Asian J. Cancer. 2020;7(4):240–243.
- Hausfater P. Biomarkers and infection in the emergency unit. Med Mal Infect. 2014;44(4):139–145. doi: 10.1016/j.medmal.2014.01.002.
- Halıcı A, Hür 0, Abatay K, et al. The role of presepsin in the diagnosis of chronic obstructive pulmonary disease acute exacerbation with pneumonia. Biomark Med. 2020;14(1):31–41. doi: 10.2217/bmm-2019-0183.
- Wiedermann CJ. Hypoalbuminemia as surrogate and culprit of infections. IJMS. 2021;22(9):4496. doi: 10.3390/ijms22094496.
- Mayhew D, Devos N, Lambert C, et al. Longitudinal profiling of the lung microbiome in the AERIS study demonstrates repeatability of bacterial and eosinophilic COPD exacerbations. Thorax. 2018;73(5):422–430. doi: 10.1136/thoraxjnl-2017-210408.
- Mallia P, Footitt J, Sotero R, et al. Rhinovirus infection induces degradation of antimicrobial peptides and secondary bacterial infection in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2012;186(11):1117–1124. doi: 10.1164/rccm.201205-0806OC.
- Wark PAB, Tooze M, Powell H, et al. Viral and bacterial infection in acute asthma and chronic obstructive pulmonary disease increases the risk of readmission. Respirology. 2013;18(6):996–1002. doi: 10.1111/resp.12099.
- Martínez-García MÁ, de la Rosa-Carrillo D, Soler-Cataluña JJ, et al. Bronchial infection and temporal evolution of bronchiectasis in patients with chronic obstructive pulmonary disease. Clin Infect Dis. 2021;72(3):403–410. doi: 10.1093/cid/ciaa069.
- Vespasiani-Gentilucci U, De Vincentis A, Ferrucci L, et al. Low alanine aminotransferase levels in the elderly population: frailty, disability, sarcopenia, and reduced survival. J Gerontol A Biol Sci Med Sci. 2018;73(7):925–930. doi: 10.1093/gerona/glx126.
- Cillóniz C, Rodríguez-Hurtado D, Torres A. Characteristics and management of Community-Acquired pneumonia in the era of global aging. Med Sci. 2018;6(2):35. doi: 10.3390/medsci6020035.
- Almagro P, Cabrera FJ, Diez J, et al. Comorbidities and short-term prognosis in patients hospitalized for acute exacerbation of COPD: the EPOC en servicios de medicina interna (ESMI) study. Chest. 2012;142(5):1126–1133. doi: 10.1378/chest.11-2413.
- Seo H, Sim YS, Min KH, et al. The relationship between comorbidities and microbiologic findings in patients with acute exacerbation of chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis. 2022;17:855–867. doi: 10.2147/COPD.S360222.
- Wedzicha JA, Calverley PMA, Seemungal TA, et al. The prevention of chronic obstructive pulmonary disease exacerbations by salmeterol/fluticasone propionate or tiotropium bromide. Am J Respir Crit Care Med. 2008;177(1):19–26. doi: 10.1164/rccm.200707-973OC.
- Halpin DMG, Gray J, Edwards SJ, et al. Budesonide/formoterol vs. salmeterol/fluticasone in COPD: a systematic review and adjusted indirect comparison of pneumonia in randomised controlled trials. Int J Clin Pract. 2011;65(7):764–774. doi: 10.1111/j.1742-1241.2011.02685.x.
- Kendzerska T, Aaron SD, To T, et al. Effectiveness and safety of inhaled corticosteroids in older individuals with chronic obstructive pulmonary disease and/or asthma. A population study. Ann Am Thorac Soc. 2019;16(10):1252–1262. doi: 10.1513/AnnalsATS.201902-126OC.
- Lee S, Lee H, Park H. Transcriptome analysis of sputum cells reveals two distinct molecular phenotypes of "asthma and chronic obstructive pulmonary disease overlap" in the elderly. Eur J Med Res. 2022;27(1):215. doi: 10.1186/s40001-022-00861-2.
- Leigh R, Pizzichini MMM, Morris MM, et al. Stable COPD: predicting benefit from high-dose inhaled corticosteroid treatment. Eur Respir J. 2006;27(5):964–971. doi: 10.1183/09031936.06.00072105.
- Titova E, Christensen A, Henriksen AH, et al. Comparison of procalcitonin, C-reactive protein, white blood cell count and clinical status in diagnosing pneumonia in patients hospitalized with acute exacerbations of COPD: a prospective observational study. Chron Respir Dis. 2019;16:147997231876976. doi: 10.1177/1479972318769762.