Abstract
Background:
How socio-demographic characteristics and comorbidities affect bacterial community-acquired pneumonia (CAP) prognosis during/after hospitalization is important in disease management.
Objectives:
To identify predictors of medical intensive care unit (MICU) admission, length of hospital stay (LOS), in-hospital mortality, and bacterial CAP readmission in patients hospitalized with bacterial CAP.
Methods:
ICD-9/10 codes were used to query electronic medical records to identify a cohort of patients hospitalized for bacterial CAP at a tertiary hospital in Southeastern US between 01/01/2013–12/31/2019. Adjusted accelerated failure time and modified Poisson regression models were used to examine predictors of MICU admission, LOS, in-hospital mortality, and 1-year readmission.
Results:
There were 1,956 adults hospitalized with bacterial CAP. Median (interquartile range) LOS was 11 days (6–23), and there were 26% (513) MICU admission, 14% (266) in-hospital mortality, and 6% (117) 1-year readmission with recurrent CAP. MICU admission was associated with heart failure (RR 1.38;95%CI 1.17–1.62) and obesity (RR 1.26;95%CI 1.04–1.52). Longer LOS was associated with heart failure (adjusted time ratio[TR] 1.27;95%CI 1.12–1.43), stroke (TR 1.90;95%CI 1.54,2.35), type 2 diabetes (TR 1.20;95%CI 1.07–1.36), obesity (TR 1.50;95%CI 1.31–1.72), Black race (TR 1.17;95%CI 1.04–1.31), and males (TR 1.24;95%CI 1.10–1.39). In-hospital mortality was associated with stroke (RR 1.45;95%CI 1.03–2.04) and age≥65 years (RR 1.34;95%CI 1.06–1.68). 1-year readmission was associated with COPD (RR 1.55;95%CI 1.05–2.27) and underweight BMI (RR 1.74;95%CI 1.04–2.90).
Conclusions:
Comorbidities and socio-demographic characteristics have varying impacts on bacterial CAP in-hospital prognosis and readmission. More studies are warranted to confirm these findings to develop comprehensive care plans and inform public health interventions.
Keywords: Community-Acquired Bacterial Pneumonia, Poor prognosis, Mortality, Length of hospital stay, Readmission, Comorbidities
INTRODUCTION
Thirty percent of all adult community-acquired pneumonia (CAP) are hospitalized annually in the US, estimated at over 1.5 million patients and 100,000 deaths annually.(1) Hospitalization for CAP is influenced by chronic comorbidity and socio-demographic risk factors.(1–3) Chronic comorbidities known for the highest incidence of CAP hospitalization include chronic obstructive pulmonary disease (COPD), congestive heart failure, stroke, type 2 diabetes, and obesity.(1) The estimated annual incidence of hospitalization for CAP due to these comorbidities ranges from 674 hospitalizations per 100,000 adults for obesity to 5,832 hospitalizations per 100,000 adults for COPD. Also, socio-demographic factors which include older age and smoking are known to increase the risk of hospitalization for CAP, with age 65 years or older accounting for 7,663 hospitalizations per 100,000 adults and smoking accounting for 822 hospitalizations per 100,000 adults. How these factors affect CAP prognosis during and after CAP hospitalization is a subject of interest and review among experts.
In 2018, the National Heart, Lung, and Blood Institute (NHLBI) identified ‘understanding the disease course during and after pneumonia’ as a high priority initiative in a ‘host-focused’ manner.(4) Characterizing patients who are at risk of poor prognosis during and after CAP will help healthcare policy decision-makers and clinicians prioritize preventive and treatment measures, especially at a regional level. Currently, there is inconclusive evidence about the impacts of certain comorbidities known for high risk of CAP hospitalization – e.g., COPD - on the prognosis of pneumonia.(5–7) Also, limited studies have extensively examined how socio-demographic characteristics affect bacterial CAP prognosis.
The purpose of this study is to determine how comorbidities and socio-demographic characteristics known for high risk of CAP hospitalization affect the length of hospital stay (LOS), in-hospital all-cause mortality, and risk of CAP readmission in a clinical cohort of hospitalized CAP patients from electronic medical records (EMR). We hypothesized that 1) comorbidities including COPD, heart failure, stroke, type 2 diabetes, and obesity will increase MICU admission, LOS, in-hospital all-cause mortality, and CAP readmission; 2) socio-demographic characteristics including smoking, older age, race, sex, and a patient’s admission source will have diverse associations with MICU admission, LOS, in-hospital all-cause mortality, and CAP readmission.
METHODS
Study Design/Population/Setting/Approval
This was a retrospective cohort study which utilized secondary data from EMR. The study population consisted of a clinical cohort of patients admitted to the University of Alabama at Birmingham (UAB) Healthcare System between 01/01/2013 – 12/31/2019 with bacterial pneumonia diagnosis. The UAB Healthcare System is a level 1 trauma tertiary hospital with over 1,200 bed capacity. The UAB Institutional Review Board (IRB) approved this study (IRB number: IRB-300002043–007).
Inclusion/Exclusion
The criteria for inclusion were age ≥ 18 years; admission from a clinician’s office or a non-healthcare facility (home); bacterial pneumonia diagnosis with International Classification of Diseases (ICD) codes listed in Supplement table 1; bacterial pneumonia diagnosis reported in the EMR as present on admission. Where there was no information about pneumonia present on admission, we used collection of microbiology culture within 48 hours of admission as a proxy. Patients with multiple admissions with bacterial pneumonia diagnosis were only included once, using only their first encounter. We excluded patients with cystic fibrosis, organ transplant, and those who were hospitalized for more than 100 days.
Data source
The EMR data used for the study were obtained through the UAB Informatics for Integrating Biology and the Bedside (i2b2) program which is an NIH-funded National Center for Biomedical Computing based at Partners HealthCare System. Data obtained included socio-demographic characteristics; microbial culture and susceptibility; hospitalization and clinical records.
Outcomes
Medical intensive care unit admission, length of hospital stay (in days), in-hospital all-cause mortality, and readmission for bacterial CAP within one year of the first case were the primary outcomes.
General and clinical characteristics
The comorbidities of interest were those known as the major comorbid risk factors for CAP hospitalization.(8) These included COPD, heart failure, stroke, type 2 diabetes, and obesity. The ICD codes for these comorbidities are listed in Supplement table 1. Charlson comorbidity index (CCI) was calculated using comorbidities and weights previously validated.(9, 10) Patients were further grouped into two categories (high CCI and low CCI) based on the CCI median value of 4. Body mass index (BMI) of patients was classified into underweight (BMI < 18.5), normal (BMI 18.5 to < 25.0), overweight (BMI 25.0 to < 30.0), and obese (BMI 30.0 or higher). Admission to medical intensive care unit (MICU) was used as a proxy for severe CAP. Information was also obtained on socio-demographic characteristics, smoking, and other clinical data, including admission source, presence of Pseudomonas, multi-drug resistant (MDR) Pseudomonas, or methicillin-resistant Staphylococcus aureus (MRSA) isolates in respiratory samples. Respiratory samples that were considered included sputum, bronchoalveolar lavage, bronchial wash, or tracheal aspirate. MDR Pseudomonas was defined based on criteria used in previous literature.(11)
Statistical analysis
Descriptive statistics were done with chi-square test, t-statistics, and non-parametric test where appropriate. Accelerated failure time (AFT) models were used to examine associations of LOS with comorbidities and socio-demographic characteristics. The choice of AFT was made because the proportionality of hazard assumption was not met for some of the risk factors of interest (e.g. COPD) when Cox proportional hazard model was considered; AFT models have been recommended in situations like this (See Supplement).(12) Time from admission to discharge was used as the time variable. Death during hospitalization and LOS greater than 30 days were censored. Estimates of LOS ratio, called time ratio (TR), and corresponding adjusted TR were calculated; TR > 1 indicates that the LOS of a covariate is greater than the LOS of its reference category, and TR < 1 indicates that the LOS of a covariate is lower than the LOS of its reference category. Modified Poisson regression models with robust error variance and unstructured correlation matrix were used to examine the associations of MICU admission, in-hospital all-cause mortality, and 1-year bacterial CAP readmission risk with comorbidities and socio-demographic characteristics. Risk ratio (RR) and adjusted risk ratio (aRR), 95% confidence interval, and p-value were reported. MICU admission was used as a proxy for pneumonia severity when analyzing LOS and in-hospital all-cause mortality outcomes. All adjusted models adjusted for age, sex, race, smoking, and Charlson comorbidity index. Adjusted models for LOS, in-hospital all-cause mortality, and 1-year bacterial CAP readmission also included MICU admission. We used an alpha of 0.05 for significance testing. SAS version 9.4 software (SAS Institute, Cary, NC) was used for statistical analyses.
RESULTS
General Characteristics
The study cohort consisted of 1,956 patients. Their median age (interquartile range [IQR]) was 58 years (46 – 69), and were mostly males (1,178, 60.2%), white (1,149, 58.8%), non-Hispanic (1,838, 94.1%), and current or past smokers (1,168, 61.4%), as seen in Table 1. Most of the patients were admitted from home (1,660, 84.9%). Of all patients, 266 (13.6%) died during hospitalization, 687 (35.1%) were discharged home, 747 (38.2%) were discharged to long-term care settings (e.g., home health, hospice, long-term care facility or skilled nursing facility), and 256 (13.1%) were discharged to other facilities (e.g., court, custody). Comorbidities in the cohort included 793 (35.4%) COPD patients, 562 (28.7%) heart failure patients, 148 (7.6%) stroke patients, 632 (32.3%) type 2 diabetes patients, and 1,105 (57.3%) patients who were overweight or obese. Median (IQR) Charlson comorbidity index was 4 (2–7). Out of 1,285 patients who had respiratory sample cultures, 354 (27.6%) had Pseudomonas, 87 (6.8%) had multi-drug resistant (MDR) Pseudomonas as defined in previous literature (11), and 342 (26.6%) had MRSA isolates.
Table 1:
General characteristics by medical intensive care unit admission, median length of hospital stay (time in days from admission to discharge), in-hospital all-cause mortality, and 1-year readmission in community-acquired bacterial pneumonia
| Characteristics | Total | MICU Admission, N (%) | Median LOS in days (IQR) | Mortality, N (%) | 1-year readmission, N (%) |
|---|---|---|---|---|---|
| All | 1,956 | 513 (26.2) | 11 (6 – 23) | 266 (13.6) | 117 (6.0) |
| COPD | |||||
| yes | 693 (35.4) | 185 (26.7) | 9 (5 – 17) | 86 (12.4) | 58 (8.4) |
| no | 1,263 (64.6) | 328 (26.0) | 13 (6 – 25) | 180 (14.3) | 59 (4.7) |
| Heart failure | |||||
| yes | 562 (28.7) | 193 (34.3) | 12 (7 – 23) | 107 (19.0) | 30 (5.3) |
| no | 1,394 (71.3) | 320 (23.0) | 11 (5 – 22) | 159 (11.4) | 87 (6.2) |
| Stroke | |||||
| yes | 148 (7.6) | 43 (29.1) | 18 (9 – 29) | 36 (24.3) | 9 (6.1) |
| no | 1,808 (92.4) | 470 (26.0) | 11 (5 – 22) | 230 (12.7) | 108 (6.0) |
| Diabetes Mellitus type 2 | |||||
| yes | 632 (32.3) | 192 (30.4) | 12 (6 – 24) | 98 (15.5) | 43 (6.8) |
| no | 1,324 (67.7) | 321 (24.2) | 11 (5 – 22) | 168 (12.7) | 74 (5.6) |
| Charlson comorbidity index | |||||
| 0 – 3 | 1,015 (51.9) | 238 (23.5) | 11 (5 – 24) | 90 (8.9) | 48 (4.7) |
| ≥ 4 | 941 (48.1) | 275 (29.2) | 11 (6 – 22) | 176 (18.7) | 69 (7.3) |
| Admission source | |||||
| physician’s office | 296 (15.1) | 53 (17.9) | 9 (5 – 20) | 37 (12.5) | 29 (9.8) |
| home | 1,660 (84.9) | 460 (27.7) | 12 (6 – 23) | 229 (13.8) | 88 (5.3) |
| Smoking | |||||
| ever | 1,168 (61.4) | 289 (24.7) | 11 (5 – 23) | 147 (12.6) | 76 (6.5) |
| never | 679 (35.7) | 184 (27.1) | 11 (5 – 22) | 99 (14.6) | 41 (6.0) |
| unknown | 54 (2.8) | 24 (44.4) | 12 (8 – 23) | 6 (11.1) | 0 (0) |
| Body mass index | |||||
| Underweight | 172 (8.9) | 47 (27.3) | 8 (5 – 18) | 29 (16.9) | 20 (11.6) |
| Normal weight | 651 (33.8) | 158 (24.3) | 9 (5 – 19) | 78 (12.0) | 40 (6.1) |
| Overweight | 517 (26.8) | 126 (24.4) | 12 (6 – 25) | 75 (14.5) | 22 (4.3) |
| Obese | 588 (30.5) | 180 (30.6) | 14 (7 – 27) | 81 (13.8) | 33 (5.6) |
| Age (years) | |||||
| ≥ 65 | 660 (33.7) | 184 (27.9) | 10 (5 – 19) | 119 (18.0) | 37 (5.6) |
| < 65 | 1,296 (66.3) | 329 (25.4) | 12 (6 – 25) | 147 (11.3) | 80 (6.2) |
| Race | |||||
| Black | 730 (37.4) | 205 (28.1) | 12 (6 – 23) | 114 (15.6) | 41 (5.6) |
| White | 1,149 (58.8) | 288 (25.1) | 11 (5 – 22) | 140 (12.2) | 74 (6.4) |
| others | 75 (3.8) | 19 (25.3) | 13 (6 – 28) | 12 (16.0) | 2 (2.7) |
| Sex | |||||
| male | 1,178 (60.2) | 288 (24.5) | 12 (6 – 24) | 162 (13.8) | 71 (6.0) |
| Female | 778 (39.8) | 225 (28.9) | 10 (5 – 21) | 104 (13.4) | 46 (5.9) |
| Health Insurance | |||||
| Medicaid | 293 (15.0) | 80 (27.3) | 14 (7 – 32) | 30 (10.2) | 19 (6.5) |
| Medicare | 825 (42.2) | 228 (27.6) | 10 (5 – 20) | 134 (16.2) | 50 (6.1) |
| financial assistance | 52 (2.7) | 13 (25.0) | 11 (5 – 20) | 7 (13.5) | 2 (3.9) |
| private | 604 (30.9) | 148 (24.5) | 12 (6 – 25) | 79 (13.1) | 35 (5.8) |
| others | 182 (9.3) | 44 (24.2) | 11 (5 – 20) | 16 (8.8) | 11 (6.0) |
| Pseudomonas in respiratory culture* | |||||
| yes | 354 (27.6) | 76 (21.5) | 15 (7 – 32) | 52 (14.7) | 35 (9.9) |
| no | 931 (72.5) | 306 (32.9) | 15 (8 – 27) | 161 (17.3) | 53 (5.7) |
| MDR Pseudomonas in respiratory culture* | |||||
| yes | 87 (6.8) | 21 (24.1) | 16 (8 – 36) | 13 (14.9) | 10 (11.5) |
| no | 1,198 (93.2) | 361 (30.1) | 15 (5 – 28) | 200 (16.7) | 78 (6.5) |
| MRSA in respiratory culture* | |||||
| yes | 342 (26.6) | 112 (32.8) | 18 (9 – 32) | 51 (14.9) | 29 (8.5) |
| no | 943 (73.4) | 270 (28.6) | 10 (7 – 27) | 162 (17.2) | 59 (6.3) |
Cohort N = 1,956 LOS: length of hospital stay (in days); COPD: chronic obstructive pulmonary disease; MICU: medical intensive care unit; MDR: multi-drug resistance; MRSA: methicillin-resistant Staphylococcus aureus. BMI categories: Underweight (< 18.5), Normal weight (18.5 – 24.9), Overweight (25.0 – 29.9), Obese (≥ 30.0). Subgroup counts for MICU admission, in-hospital all-cause mortality, and 1-year readmission may not total 1,956 because of missing data. %s may not add to 100% due to approximation.
A total of 1,285 patients had respiratory cultures and were considered to estimate the percentages of patients who had Pseudomonas, MDR Pseudomonas, and MRSA isolates. Estimates in bold font had p-value < 0.05
Length of Hospital Stay
The median (IQR) LOS was11 days (6 – 23). As shown in Table 1, patients aged ≥ 65 years had shorter median LOS compared to those younger (median LOS in days: 10 vs 12; p-value = 0.001); this was further seen in an adjusted model (Table 2) where the LOS of patients aged ≥ 65 years was 15% lower when compared to those younger (TR: 0.85; 95% CI: 0.76,0.96; p-value = 0.029). We believed this finding could be explained by examining discharge disposition. In a post-hoc analysis that compared patients that were discharged home (n = 687) to those discharged to long-term care settings (n = 746), the proportion of patients aged ≥ 65 years was higher among those discharged to long-term care setting (46.0% vs 22.3%; p-value < 0.0001). Also, female patients had shorter LOS when compared to males (TR: 0.81; 95% CI: 0.72,0.90; p-value: 0.0002). Black patients had longer LOS than white (TR: 1.17; 95 % CI: 1.04,1.31; p-value: 0.011).
Table 2:
Association of length of hospital stay with comorbidities known for community-acquired bacterial pneumonia hospitalization and socio-demographic characteristics
| Risk factors | TR | p-value |
|---|---|---|
| Comorbidity/behavioral characteristics | ||
| COPD | ||
| yes | 0.73 (0.65,0.82) | < 0.0001 |
| no | ref | ref |
| Heart failure | ||
| yes | 1.27 (1.12,1.43) | 0.0002 |
| no | ref | ref |
| Stroke | ||
| yes | 1.90 (1.54,2.35) | < 0.0001 |
| no | ref | ref |
| Diabetes Mellitus type 2 | ||
| yes | 1.20 (1.07,1.36) | 0.003 |
| no | ref | ref |
| Smoking | ||
| ever | 0.97 (0.86,1.10) | 0.453 |
| never | ref | ref |
| BMI | ||
| underweight (BMI < 18.5) | 0.98 (0.80,1.20) | 0.823 |
| overweight (BMI 25.0 – 29.9) | 1.26 (1.09,1.45) | 0.002 |
| obese or worse (BMI ≥ 30.0) | 1.50 (1.31,1.72) | < 0.0001 |
| normal (BMI 18.5 – 24.9) | ref | ref |
| Demographic characteristics | ||
| Age | ||
| ≥ 65 | 0.85 (0.76,0.96) | 0.008 |
| < 65 | ref | ref |
| Sex | ||
| female | 0.81 (0.72,0.90) | 0.0002 |
| male | ref | ref |
| Race | ||
| black | 1.17 (1.04,1.31) | 0.011 |
| others | 1.16 (0.85,1.59) | 0.352 |
| white | ref | ref |
| Admission source | ||
| home | 1.03 (0.88,1.20) | 0.698 |
| physician office | ref | ref |
Death was censored, and all patients were censored at 30 days. Patients admitted for more than 100 days were excluded. Accelerated failure time model with log-normal distribution was used for estimation. Model adjusted for medical intensive care unit admission, Charlson comorbidity index, age, sex, race, smoking status. TR: time ratio estimate comparing ratio of time from admission to discharge. ref: reference group. COPD: chronic obstructive pulmonary disease. COPD diagnosis was based on time of bacterial pneumonia admission. Chronic lung diseases included COPD, asthma, and bronchiectasis.
Considering the topmost comorbidities known to be associated with CAP hospitalization, COPD comorbidity in patients was found to be associated with shorter LOS (median LOS in days: 9 vs 13; p-value < 0.0001) when compared to patient with no COPD, as seen in Table 1. Also, after excluding in-hospital deaths, patients with COPD had shorter median LOS (median LOS in days: 8 vs 13; p-value < 0.0001) when compared to those with no COPD. In adjusted models that censored death during hospitalization and LOS greater than 30 days (Table 2), the LOS of patients with COPD was 27% lower than those with no COPD (TR: 0.73; 95% CI: 0.65,0.82; p-value < 0.0001). However, other comorbidities including stroke, heart failure, and type 2 diabetes were each associated with higher LOS (Tables 1 and 2). In the adjusted model in Table 2, LOS for patients with pre-existing stroke diagnosis was 90% higher when compared to those with no stroke (TR: 1.90; 95% CI: 1.54,2.35; p-value < 0.0001). LOS was 27% higher in patients with heart failure (TR: 1.27; 95% CI: 1.12,1.43; p-value = 0.0002) when compared to those with no heart failure, and 20% higher in patients with type 2 diabetes (TR: 1.20; 95% CI: 1.07,1.36; p-value = 0.003) when compared to patients with no type 2 diabetes. There was a dose-response relationship between BMI and LOS, with increasing BMI categories being associated with increasing LOS (normal weight: 9 days; overweight: 12 days; obese: 14 days; p-value < 0.0001). This trend was also observed in the adjusted model for the association between BMI categories and LOS (overweight vs normal weight: TR: 1.26; 95%CI: 1.09,1.45; p-value: 0.002; obese vs normal weight: TR: 1.50; 95%CI: 1.31,1.72; p-value < 0.0001). In the unadjusted model in Table 1, patients with MRSA isolates in their respiratory samples had higher LOS when compared to those with no MRSA isolates (median LOS in days: 18 vs 10; p-value = 0.0002).
Severe pneumonia
A total of 513 patients (26.2%) were admitted to the medical intensive care unit (MICU). In the adjusted models in Table 3, comorbidity with heart failure when compared with no heart failure (RR: 1.38; 95%CI: 1.17,1.62; p-value < 0.0001) and obesity when compared with normal BMI (RR: 1.26; 95%CI: 1.04,1.52; p-value = 0.018) were associated with higher risk of MICU admission. Also, those who were admitted from home, when compared with admission from physician office, had higher risk of MICU admission (RR: 1.45; 95%CI: 1.11,1.88; p-value = 0.006). In the unadjusted model in Table 1, the proportion of MICU admission was lower in patients who had Pseudomonas isolates when compared to those who did not have Pseudomonas isolates (21.5% vs 32.9%; p-value < 0.0001).
Table 3:
Association of medical intensive care unit admission with comorbidities known for community-acquired bacterial pneumonia hospitalization and socio-demographic characteristics
| Risk factors | RR | p-value |
|---|---|---|
| Comorbidity/behavioral characteristics | ||
| COPD | ||
| yes | 0.97 (0.82,1.15) | 0.732 |
| no | ref | ref |
| Heart failure | ||
| yes | 1.38 (1.17,1.62) | < 0.0001 |
| no | ref | ref |
| Stroke | ||
| yes | 0.97 (0.73,1.27) | 0.829 |
| no | ref | ref |
| Diabetes Mellitus type 2 | ||
| yes | 1.14 (0.96,1.34) | 0.114 |
| no | ref | ref |
| Smoking | ||
| ever | 0.95 (0.81,1.12) | 0.554 |
| never | ref | ref |
| BMI | ||
| underweight (BMI < 18.5) | 1.00 (0.82,1.46) | 0.528 |
| overweight (BMI 25.0 – 29.9) | 1.02 (0.83,1.27) | 0.831 |
| obese or worse (BMI ≥ 30.0) | 1.26 (1.04,1.52) | 0.018 |
| normal (BMI 18.5 – 24.9) | ref | ref |
| Demographic characteristics | ||
| Age | ||
| ≥ 65 | 1.04 (0.89,1.22) | 0.625 |
| < 65 | ref | ref |
| Sex | ||
| female | 1.14 (0.98,1.34) | 0.093 |
| male | ref | ref |
| Race | ||
| black | 1.09 (0.93,1.29) | 0.283 |
| others | 0.99 (0.63,1.54) | 0.954 |
| white | ref | ref |
| Admission source | ||
| home | 1.45 (1.11,1.88) | 0.006 |
| physician office | ref | ref |
Model adjusted for medical intensive care unit admission, Charlson comorbidity index, age, sex, race, smoking status. RR: adjusted risk ratio estimate from modified Poisson regression model. ref: reference group. COPD: chronic obstructive pulmonary disease. COPD diagnosis was based on time of bacterial pneumonia admission.
In-hospital All-cause Mortality
A total of 266 (13.6%) patients died during hospitalization. In the unadjusted model in Table 1, stroke was strongly associated with the risk of in-hospital mortality (24.3% vs 12.7%; p-value < 0.0001). The association was maintained in the adjusted model in Table 4 where pre-existing stroke comorbidity increased the risk of in-hospital mortality by 45% (TR: 1.45; 95 % CI: 1.03,2.04; p-value: 0.032), when compared to those without stroke. Heart failure (19.0% vs 11.4%; p-value < 0.0001) and higher Charlson comorbidity index (score ≥ 4 vs < 4: 18.7% vs 8.9%; p-value < 0.0001) were associated with higher in-hospital all-cause mortality in the unadjusted models in Table 1. Older patients (≥ 65 years) had higher in-hospital all-cause mortality compared to younger ones in both unadjusted (18.0% vs 11.3%; p-value < 0.0001) and adjusted (RR: 1.34; 95% CI: 1.06,1.68; p-value = 0.014) models. In an unadjusted model, there was no difference in the in-hospital all-cause mortality between COPD patients and those without COPD (12.4% vs 14.3%; p-value = 0.256). However, in the adjusted model in Table 4, patients with COPD comorbidity had lower risk of in-hospital death when compared to those with no COPD (RR: 0.73; 95% CI: 0.57,0.93; p-value = 0.011). Patients who had Medicaid insurance, which is a correlate of socioeconomic status, had lower in-hospital all-cause mortality when compared to those with private insurance (10.2% vs 13.1%; p-value = 0.0220), in the unadjusted model in Table 1.
Table 4:
Association of in-hospital all-cause mortality with comorbidities known for community-acquired bacterial pneumonia hospitalization and socio-demographic characteristics
| Risk factors | RR | p-value |
|---|---|---|
| Comorbidity/behavioral characteristics | ||
| COPD | ||
| yes | 0.73 (0.57,0.93) | 0.011 |
| no | ref | ref |
| Heart failure | ||
| yes | 1.12 (0.87,1.43) | 0.376 |
| no | ref | ref |
| Stroke | ||
| yes | 1.45 (1.03,2.04) | 0.032 |
| no | ref | ref |
| Diabetes Mellitus type 2 | ||
| yes | 0.80 (0.63,1.02) | 0.073 |
| no | ref | ref |
| Smoking | ||
| ever | 0.81 (0.64,1.02) | 0.079 |
| never | ref | ref |
| BMI | ||
| underweight (BMI < 18.5) | 1.42 (0.98,2.05) | 0.061 |
| overweight (BMI 25.0 – 29.9) | 1.23 (0.92,1.66) | 0.167 |
| obese or worse (BMI ≥ 30.0) | 1.07 (0.80,1.44) | 0.633 |
| normal (BMI 18.5 – 24.9) | ref | ref |
| Demographic characteristics | ||
| Age | ||
| ≥ 65 | 1.34 (1.06,1.68) | 0.014 |
| < 65 | ref | ref |
| Sex | ||
| female | 0.90 (0.71,1.13) | 0.357 |
| male | ref | ref |
| Race | ||
| black | 1.17 (0.92,1.48) | 0.190 |
| others | 1.96 (1.20,3.21) | 0.008 |
| white | ref | ref |
| Admission source | ||
| home | 1.02(0.73,1.42) | 0.921 |
| physician office | ref | ref |
Model adjusted for medical intensive care unit admission, Charlson comorbidity index, age, sex, race, smoking status. RR: adjusted risk ratio estimate from modified Poisson regression model. ref: reference group. COPD: chronic obstructive pulmonary disease. COPD diagnosis was based on time of bacterial pneumonia admission.
One-year risk of readmission with bacterial CAP diagnosis
Of all the 1,956 patients initially hospitalized with bacterial CAP diagnosis, all-cause readmission within one year occurred in 836 (42.7%) patients, and readmission with bacterial CAP diagnosis within 1 year occurred in 117 (6.0%) patients. COPD, when compared with those with no COPD, was associated with higher risk of readmission for bacterial CAP within one year of discharge after the initial hospitalization for bacterial CAP (RR 1.55;95%CI 1.05–2.27; p = 0.027), as seen in Table 5. Similarly, being underweight when compared with normal weight was associated with higher risk of readmission (RR 1.74;95%CI 1.04–2.90; p = 0.034). Among the socio-demographic characteristics, those who were admitted from physician office were more likely to have bacterial CAP readmission when compared to those who were admitted from home in an adjusted model (RR: 1.64; 95% CI: 1.07,2.49; p-value = 0.022). In Table 1, patients who had Pseudomonas (9.9% vs 5.7%; p-value = 0.008) in their respiratory cultures had higher risk of 1-year readmission with bacterial CAP diagnosis.
Table 5:
Readmission with bacterial CAP diagnosis within 1 year of previous admission and its association with comorbidities known for community-acquired bacterial pneumonia hospitalization and socio-demographic characteristics
| Risk factors | RR | p-value |
|---|---|---|
| Comorbidity/behavioral characteristics | ||
| COPD | ||
| yes | 1.55 (1.05,2.77) | 0.027 |
| no | ref | ref |
| Heart failure | ||
| yes | 0.77 (0.51,1.18) | 0.228 |
| no | ref | ref |
| Stroke | ||
| yes | 0.99 (0.51,1.93) | 0.970 |
| no | ref | Ref |
| Diabetes Mellitus type 2 | ||
| yes | 1.12 (0.76,1.66) | 0.553 |
| no | ref | ref |
| Smoking | ||
| ever | 1.06 (0.72,1.55) | 0.786 |
| never | ref | ref |
| BMI | ||
| underweight (BMI < 18.5) | 1.74 (1.04,2.90) | 0.034 |
| overweight (BMI 25.0 – 29.9) | 0.74 (0.44,1.22) | 0.235 |
| obese or worse (BMI ≥ 30.0) | 0.97 (0.62,1.52) | 0.886 |
| normal (BMI 18.5 – 24.9) | ref | ref |
| Socio-demographic characteristics | ||
| Age | ||
| ≥ 65 | 0.80 (0.55,1.18) | 0.267 |
| < 65 | ref | ref |
| Sex | ||
| female | 0.99 (0.68,1.44) | 0.947 |
| male | ref | ref |
| Race | ||
| black | 0.81 (0.55,1.19) | 0.280 |
| others | 0.47 (0.12,1.87) | 0.286 |
| white | ref | ref |
| Admission source | ||
| home | 0.61 (0.40,0.93) | 0.022 |
| physician office | ref | ref |
Model adjusted for medical intensive care unit admission, Charlson comorbidity index, age, sex, race, smoking status. RR: adjusted risk ratio estimate from modified Poisson regression model. ref: reference group. COPD: chronic obstructive pulmonary disease. COPD diagnosis was based on time of bacterial pneumonia admission.
DISCUSSION
In a clinical cohort of hospitalized patients with bacterial CAP diagnosis, we identified that MICU admission was associated with admission from home, heart failure, and obesity; higher LOS was associated with black race, male, heart failure, pre-existing stroke, type 2 diabetes, obesity, but not with older age and COPD; In-hospital all-cause mortality was associated with older age and pre-existing stroke, but not with COPD. Bacterial CAP readmission within a year was associated with COPD and underweight BMI. Comorbidity and socio-demographic characteristics have varying but important impacts on bacterial CAP prognosis during and after hospitalization. Preventive and early treatment public health measures targeting these factors could help reduce the disease burden of CAP including its poor prognosis in at-risk population.
The current study shared similar observations with previous literature on how heart failure, stroke, type 2 diabetes, and obesity affect CAP prognosis during and after hospitalization. Infections like CAP could trigger acute exacerbation of heart failure, thereby worsening pneumonia and heart failure prognosis.(13–16) Among stroke patients, complications like dysphagia can increase the risk and severity of aspiration events including pneumonia. Like in the current study, pneumonia was shown in previous study to be associated with poor outcomes including longer LOS and mortality in stroke patients by other studies.(17–19) Additionally, in a previous study of patients hospitalized with bacterial CAP, we observed that those who had stroke comorbidity had higher risk of having MDR Pseudomonas isolates in their respiratory samples when compared to those who did not have stroke.(11) Managing the burden of antibiotic resistance in stroke patients with bacterial pneumonia could further worsen prognosis during or after hospitalization. In a study of patients hospitalized to the ICU with severe CAP, type 2 diabetes patients spent more days in the ICU than those with no type 2 diabetes, though there was no difference in overall hospital LOS.(20) Also, in non-critical-care diabetic CAP inpatients, hyperglycemic blood glucose level was found to be associated with adverse clinical course which was a composite of ICU admission and all-cause death.(21) But in another study of patient hospitalized with CAP from a Spanish National Hospital Discharge Database, there was no difference in LOS between patients with type 2 diabetes comorbidity and those who did not have type 2 diabetes.(22)
The link between obesity and major cardiovascular risk factors - like heart failure – could explain why obesity was associated with poor in-hospital CAP prognosis as observed in the current study (higher risk of MICU admission and higher LOS).(23) Interestingly, we also found that being underweight was associated with higher risk of 1-year bacterial CAP readmission. In a previous study, we found that patients who had COPD and those who were underweight were more likely to be admitted with Pseudomonas CAP. This suggests a possible correlation between underweight and COPD.(11) Also, previous studies found an association between underweight and severe COPD, especially class 3 or 4 COPD severity according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) classification.(24–26). It is possible that the increased risk of 1-year bacterial CAP readmission observed in the current study in underweight patients was driven by a high proportion of severe COPD. Future larger epidemiology studies could further examine this unique association between BMI and CAP readmission, and the potential interaction between BMI, CAP hospitalization, and COPD comorbidity.
In our cohort, COPD was associated with lower LOS and lower in-hospital all-cause mortality. Prior literature has shown LOS to be longer in COPD patients.(27) In other studies, there were no associations between COPD and LOS.(28–34) Also, Molinos et al. showed that COPD increased 30-day mortality in patients with CAP.(35) Similarly, Adamuz et al. reported that COPD was a risk factor for 1-year mortality in patients with CAP after discharge.(36) However, there are other studies that did not find an association between COPD and 30-day mortality.(5, 31, 37–40). The unexpected shorter LOS of COPD patients when compared to those with no COPD, as observed in the current study, could be because of clinicians’ bias favoring early discharge of patients with COPD so that they could start or continue close follow-up clinic visits with pulmonologists, or access medical devices and therapies in the home setting. Intrinsic factors like altered tolerance of respiratory symptoms could also play a role. Additionally, COPD treatment like azithromycin is known to have anti-inflammatory actions which may help reduce exacerbations in COPD.(41–44). Though we observed that patients who had COPD had shorter LOS in the current study, we also observed that they had an increased risk of bacterial CAP readmission within 1 year of the first CAP admission. An unadjusted model from a study in a UK population showed there was an association between COPD and 30-day readmission.(45) COPD was also found to be associated with pneumonia-unrelated 30-day readmission rate in a South-Korean population.(46). Though the findings in these studies were consistent with the results in the current study, the reasons for readmission in those studies were not necessarily related to CAP. Overall, preventive public health measures like vaccination against pneumonia pathogens, healthy-living advocacy, and lifestyle modifications – e.g., exercise, healthy diet and weight - to reduce the risk of comorbidities like heart failure, stroke, type 2 diabetes, COPD, and obesity would be beneficial in reducing the disease burden and poor prognosis of CAP. Also, access to healthcare and adequate healthcare management of comorbidities will reduce CAP burden.
Among the socio-demographic characteristics considered in the current study, we found that patients aged ≥ 65 years, females, and white had lower LOS. This may be attributed to higher proportions of older patients, females, and whites among patients with COPD comorbidity. We showed earlier that those who had COPD comorbidity had shorter LOS. In a post-hoc analysis, we found that more older patients were discharged to long-term care settings when compared to those discharged home. There have been reports about how female sex hormones - e.g., estradiol – may enhance immunity against certain infections while male sex hormones – e.g. testosterone – may suppress the immune system and make males more prone to certain infections than females.(47, 48) This may play a role in the observed longer LOS for males during CAP hospitalization in the current study. Additionally, a study by Kaplan et al. found that hospitalization for CAP and requirement of ICU admission were higher for men.(49) The difference in LOS between males and females that we observed in the current study population is unique, and there is a need for future large epidemiologic studies to further examine sex differences in CAP disease prognosis. There is substantial evidence on the cardiovascular disease burden among blacks.(50, 51) It is possible that higher burden of cardiovascular diseases among black patients might have contributed to the longer LOS of black patients when compared to whites observed in the current study. Furthermore, we observed that patients admitted from home had higher risk of MICU admission when compared to those admitted from physician’s office, and those admitted from physician office are more likely to be readmitted with bacterial CAP diagnosis. This was a distinct and notable finding because admission sources, as used in the current study, could provide some perspectives to healthcare access and healthcare-seeking behavior. Timely access to healthcare among those admitted from physician office or limited healthcare access among those admitted from home may explain the observed differences in MICU admission and CAP readmission. Healthcare-seeking behaviors including possession of health insurance and regular physician visits for routine checkup and chronic disease management may also account for the differences. Lastly, we found that patients who had Medicaid insurance had longer LOS but lower mortality when compared to those who had other healthcare insurance (e.g., private healthcare insurance) in unadjusted models. In the adult population considered in the current study, Medicaid insurance could be considered as a correlate of low socio-economic status. It would be important to understand how its healthcare coverage for indigent population impact CAP disease prognosis in future studies that adjusted for potential confounders.
Our analysis has several strengths. We studied a timeframe that immediately preceded the COVID-19 pandemic to reflect the current state of practice. We used an analytical method that compared LOS of COPD and non-COPD patients, and censored patients that died in the hospital. This methodology distinguishes the present study from previous studies that compared the means of the LOS of patients who had specific comorbidity – e.g., COPD – and those who did not. Also, there were substantial number of patients with COPD in our cohort. To have a more homogenous population, we excluded cystic fibrosis patients from our cohort. We included unique socio-demographic characteristics - e.g. admission source - in our analyses to see how they affect prognosis during and after CAP hospitalization. One of the limitations of our study is the reliance on EMR for diagnosis and other information. Though one of the first in the south, the current study represents a single region which limits the generalizability of its findings. This necessitates the need for future larger cohort studies which are nationally representative and incorporate diverse, regional, cultural, and societal factors to confirm the study findings. Lastly, due to data limit, we did not identify patients who had been on medication to manage specific comorbidity - e.g., COPD - before hospitalization and the type of medication and adherence to treatment. This could provide useful information in future studies on the impact of the treatment of comorbidities like COPD on disease prognosis in patients hospitalized with bacterial pneumonia. Despite these limitations, the EMR data used in this study provided rich demographic and pragmatic clinical information which can be used to improve healthcare through research.
Overall, among patients hospitalized with community-acquired bacterial pneumonia, those who had heart failure, stroke, type 2 diabetes mellitus, and obesity had higher risk of poor bacterial CAP prognosis including MICU admission, LOS, and in-hospital all-cause mortality. Those who had COPD comorbidity, when compared to those without COPD, had lower LOS and in-hospital all-cause mortality, but higher risk of readmission with bacterial CAP diagnosis. Like COPD, those who were underweight had higher risk of readmission when compared to normal weight. Lastly, socio-demographic characteristics had varying impacts on these measures of disease prognosis. This study highlights the importance of controlling comorbidities and socio-demographic characteristics that may worsen CAP prognosis. It also necessitates the need for a larger cohort study to confirm the study findings to help with clinical and public health interventions.
Supplementary Material
Figure 1:
Cohort’s flowchart
Acknowledgement
The University of Alabama at Birmingham i2b2 (Informatics for Integrating Biology and the Bedside) group provided the data used for this project. The i2b2 instance at the UAB is supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UL1TR003096. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Funding
The authors received no financial support for the research.
Footnotes
Conflict of interest
No conflicts exist for AJI, MLB, HWW, RLG, SS, and RAL.
Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request.
REFERENCES
- 1.Ramirez JA, Wiemken TL, Peyrani P, Arnold FW, Kelley R, Mattingly WA, et al. Adults Hospitalized With Pneumonia in the United States: Incidence, Epidemiology, and Mortality. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2017;65(11):1806–12. [DOI] [PubMed] [Google Scholar]
- 2.Torres A, Blasi F, Dartois N, Akova M. Which individuals are at increased risk of pneumococcal disease and why? Impact of COPD, asthma, smoking, diabetes, and/or chronic heart disease on community-acquired pneumonia and invasive pneumococcal disease. Thorax. 2015;70(10):984–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jain S, Self WH, Wunderink RG, Fakhran S, Balk R, Bramley AM, et al. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Adults. The New England journal of medicine. 2015;373(5):415–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dela Cruz CS, Wunderink RG, Christiani DC, Cormier SA, Crothers K, Doerschuk CM, et al. Future Research Directions in Pneumonia. NHLBI Working Group Report. American journal of respiratory and critical care medicine. 2018;198(2):256–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jiang HL, Chen HX, Liu W, Fan T, Liu GJ, Mao B. Is COPD associated with increased mortality and morbidity in hospitalized pneumonia? A systematic review and meta-analysis. Respirology (Carlton, Vic). 2015;20(7):1046–54. [DOI] [PubMed] [Google Scholar]
- 6.Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weissfeld LA, et al. Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis. Jama. 1996;275(2):134–41. [PubMed] [Google Scholar]
- 7.Restrepo MI, Sibila O, Anzueto A. Pneumonia in Patients with Chronic Obstructive Pulmonary Disease. Tuberc Respir Dis (Seoul). 2018;81(3):187–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ramirez JA, Wiemken TL, Peyrani P, Arnold FW, Kelley R, Mattingly WA, et al. Adults Hospitalized With Pneumonia in the United States: Incidence, Epidemiology, and Mortality. Clinical Infectious Diseases. 2017;65(11):1806–12. [DOI] [PubMed] [Google Scholar]
- 9.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. Journal of chronic diseases. 1987;40(5):373–83. [DOI] [PubMed] [Google Scholar]
- 10.Austin SR, Wong Y-N, Uzzo RG, Beck JR, Egleston BL. Why Summary Comorbidity Measures Such As the Charlson Comorbidity Index and Elixhauser Score Work. Med Care. 2015;53(9):e65–e72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Idigo AJ, Wells JM, Brown ML, Wiener HW, Griffin RL, Cutter G, et al. Clinical risk factors for admission with Pseudomonas and multidrug-resistant Pseudomonas community-acquired pneumonia. Antimicrobial resistance and infection control. 2022;11(1):95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wei LJ. The accelerated failure time model: a useful alternative to the Cox regression model in survival analysis. Statistics in medicine. 1992;11(14–15):1871–9. [DOI] [PubMed] [Google Scholar]
- 13.Corrales-Medina VF, Suh KN, Rose G, Chirinos JA, Doucette S, Cameron DW, et al. Cardiac complications in patients with community-acquired pneumonia: a systematic review and meta-analysis of observational studies. PLoS medicine. 2011;8(6):e1001048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bezati S, Velliou M, Ventoulis I, Simitsis P, Parissis J, Polyzogopoulou E. Infection as an under-recognized precipitant of acute heart failure: prognostic and therapeutic implications. Heart failure reviews. 2023;28(4):893–904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Feldman C, Normark S, Henriques-Normark B, Anderson R. Pathogenesis and prevention of risk of cardiovascular events in patients with pneumococcal community-acquired pneumonia. Journal of internal medicine. 2019;285(6):635–52. [DOI] [PubMed] [Google Scholar]
- 16.Aliberti S, Ramirez JA. Cardiac diseases complicating community-acquired pneumonia. Current opinion in infectious diseases. 2014;27(3):295–301. [DOI] [PubMed] [Google Scholar]
- 17.Finlayson O, Kapral M, Hall R, Asllani E, Selchen D, Saposnik G. Risk factors, inpatient care, and outcomes of pneumonia after ischemic stroke. Neurology. 2011;77(14):1338–45. [DOI] [PubMed] [Google Scholar]
- 18.Westendorp WF, Nederkoorn PJ, Vermeij JD, Dijkgraaf MG, van de Beek D. Post-stroke infection: a systematic review and meta-analysis. BMC neurology. 2011;11:110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ingeman A, Andersen G, Hundborg HH, Svendsen ML, Johnsen SP. In-hospital medical complications, length of stay, and mortality among stroke unit patients. Stroke. 2011;42(11):3214–8. [DOI] [PubMed] [Google Scholar]
- 20.Huang D, He D, Gong L, Wang W, Yang L, Zhang Z, et al. Clinical characteristics and risk factors associated with mortality in patients with severe community-acquired pneumonia and type 2 diabetes mellitus. Critical care (London, England). 2021;25(1):419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schuetz P, Friedli N, Grolimund E, Kutz A, Haubitz S, Christ-Crain M, et al. Effect of hyperglycaemia on inflammatory and stress responses and clinical outcome of pneumonia in non-critical-care inpatients: results from an observational cohort study. Diabetologia. 2014;57(2):275–84. [DOI] [PubMed] [Google Scholar]
- 22.López-de-Andrés A, de Miguel-Díez J, Jiménez-Trujillo I, Hernández-Barrera V, de Miguel-Yanes JM, Méndez-Bailón M, et al. Hospitalisation with community-acquired pneumonia among patients with type 2 diabetes: an observational population-based study in Spain from 2004 to 2013. BMJ open. 2017;7(1):e013097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Viasus D, Pérez-Vergara V, Carratalà J. Effect of Undernutrition and Obesity on Clinical Outcomes in Adults with Community-Acquired Pneumonia. Nutrients. 2022;14(15). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Steuten LM, Creutzberg EC, Vrijhoef HJ, Wouters EF. COPD as a multicomponent disease: inventory of dyspnoea, underweight, obesity and fat free mass depletion in primary care. Primary care respiratory journal : journal of the General Practice Airways Group. 2006;15(2):84–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Eriksson B, Backman H, Bossios A, Bjerg A, Hedman L, Lindberg A, et al. Only severe COPD is associated with being underweight: results from a population survey. ERJ open research. 2016;2(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kwan HY, Maddocks M, Nolan CM, Jones SE, Patel S, Barker RE, et al. The prognostic significance of weight loss in chronic obstructive pulmonary disease-related cachexia: a prospective cohort study. Journal of cachexia, sarcopenia and muscle. 2019;10(6):1330–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dusemund F, Chronis J, Baty F, Albrich WC, Brutsche MH. The outcome of community-acquired pneumonia in patients with chronic lung disease: a case-control study. Swiss medical weekly. 2014;144:w14013. [DOI] [PubMed] [Google Scholar]
- 28.Restrepo MI, Mortensen EM, Pugh JA, Anzueto A. COPD is associated with increased mortality in patients with community-acquired pneumonia. The European respiratory journal. 2006;28(2):346–51. [DOI] [PubMed] [Google Scholar]
- 29.Snijders D, van der Eerden M, de Graaff C, Boersma W. The influence of COPD on mortality and severity scoring in community-acquired pneumonia. Respiration; international review of thoracic diseases. 2010;79(1):46–53. [DOI] [PubMed] [Google Scholar]
- 30.Crisafulli E, Menéndez R, Huerta A, Martinez R, Montull B, Clini E, et al. Systemic inflammatory pattern of patients with community-acquired pneumonia with and without COPD. Chest. 2013;143(4):1009–17. [DOI] [PubMed] [Google Scholar]
- 31.Liapikou A, Polverino E, Ewig S, Cillóniz C, Marcos MA, Mensa J, et al. Severity and outcomes of hospitalised community-acquired pneumonia in COPD patients. The European respiratory journal. 2012;39(4):855–61. [DOI] [PubMed] [Google Scholar]
- 32.Rello J, Rodriguez A, Torres A, Roig J, Sole-Violan J, Garnacho-Montero J, et al. Implications of COPD in patients admitted to the intensive care unit by community-acquired pneumonia. The European respiratory journal. 2006;27(6):1210–6. [DOI] [PubMed] [Google Scholar]
- 33.Gómez-Junyent J, Garcia-Vidal C, Viasus D, Millat-Martínez P, Simonetti A, Santos MS, et al. Clinical features, etiology and outcomes of community-acquired pneumonia in patients with chronic obstructive pulmonary disease. PloS one. 2014;9(8):e105854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ruiz De Oña JM, Gómez Fernández M, Celdrán J, Puente-Maestu L. [Pneumonia in the patient with chronic obstructive pulmonary disease. Levels of severity and risk classification]. Archivos de bronconeumologia. 2003;39(3):101–5. [DOI] [PubMed] [Google Scholar]
- 35.Molinos L, Clemente MG, Miranda B, Alvarez C, del Busto B, Cocina BR, et al. Community-acquired pneumonia in patients with and without chronic obstructive pulmonary disease. The Journal of infection. 2009;58(6):417–24. [DOI] [PubMed] [Google Scholar]
- 36.Adamuz J, Viasus D, Jiménez-Martínez E, Isla P, Garcia-Vidal C, Dorca J, et al. Incidence, timing and risk factors associated with 1-year mortality after hospitalization for community-acquired pneumonia. The Journal of infection. 2014;68(6):534–41. [DOI] [PubMed] [Google Scholar]
- 37.Naucler P, Darenberg J, Morfeldt E, Örtqvist Å, Henriques Normark B. Contribution of host, bacterial factors and antibiotic treatment to mortality in adult patients with bacteraemic pneumococcal pneumonia. Thorax. 2013;68(6):571–9. [DOI] [PubMed] [Google Scholar]
- 38.Kothe H, Bauer T, Marre R, Suttorp N, Welte T, Dalhoff K. Outcome of community-acquired pneumonia: influence of age, residence status and antimicrobial treatment. The European respiratory journal. 2008;32(1):139–46. [DOI] [PubMed] [Google Scholar]
- 39.Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weissfeld LA, et al. Prognosis and Outcomes of Patients With Community-Acquired Pneumonia: A Meta-analysis. Jama. 1996;275(2):134–41. [PubMed] [Google Scholar]
- 40.Loke YK, Kwok CS, Wong JM, Sankaran P, Myint PK. Chronic obstructive pulmonary disease and mortality from pneumonia: meta-analysis. International journal of clinical practice. 2013;67(5):477–87. [DOI] [PubMed] [Google Scholar]
- 41.Suzuki T, Yamaya M, Sekizawa K, Hosoda M, Yamada N, Ishizuka S, et al. Erythromycin Inhibits Rhinovirus Infection in Cultured Human Tracheal Epithelial Cells. American journal of respiratory and critical care medicine. 2002;165(8):1113–8. [DOI] [PubMed] [Google Scholar]
- 42.Seemungal TAR, Wilkinson TMA, Hurst JR, Perera WR, Sapsford RJ, Wedzicha JA. Long-term Erythromycin Therapy Is Associated with Decreased Chronic Obstructive Pulmonary Disease Exacerbations. American journal of respiratory and critical care medicine. 2008;178(11):1139–47. [DOI] [PubMed] [Google Scholar]
- 43.Uzun S, Djamin RS, Kluytmans JA, Mulder PG, van’t Veer NE, Ermens AA, et al. Azithromycin maintenance treatment in patients with frequent exacerbations of chronic obstructive pulmonary disease (COLUMBUS): a randomised, double-blind, placebo-controlled trial. The Lancet Respiratory medicine. 2014;2(5):361–8. [DOI] [PubMed] [Google Scholar]
- 44.Albert RK, Connett J, Bailey WC, Casaburi R, Cooper JA Jr., Criner GJ, et al. Azithromycin for prevention of exacerbations of COPD. The New England journal of medicine. 2011;365(8):689–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Chakrabarti B, Lane S, Jenks T, Higgins J, Kanwar E, Allen M, et al. Predictors of 30-day readmission following hospitalisation with community-acquired pneumonia. BMJ open respiratory research. 2021;8(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Jang JG, Ahn JH. Reasons and Risk Factors for Readmission Following Hospitalization for Community-acquired Pneumonia in South Korea. Tuberc Respir Dis (Seoul). 2020;83(2):147–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Vázquez-Martínez ER, García-Gómez E, Camacho-Arroyo I, González-Pedrajo B. Sexual dimorphism in bacterial infections. Biol Sex Differ. 2018;9(1):27-. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Caceres F, Welch VL, Kett DH, Scerpella EG, Peyrani P, Ford KD, et al. Absence of gender-based differences in outcome of patients with hospital-acquired pneumonia. Journal of women’s health (2002). 2013;22(12):1069–75. [DOI] [PubMed] [Google Scholar]
- 49.Kaplan V, Angus DC, Griffin MF, Clermont G, Scott Watson R, Linde-Zwirble WT. Hospitalized Community-acquired Pneumonia in the Elderly. American journal of respiratory and critical care medicine. 2002;165(6):766–72. [DOI] [PubMed] [Google Scholar]
- 50.Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and Socioeconomic Differences in Cardiovascular Disease Risk FactorsFindings for Women From the Third National Health and Nutrition Examination Survey, 1988–1994. Jama. 1998;280(4):356–62. [DOI] [PubMed] [Google Scholar]
- 51.Kleindorfer DO, Khoury J, Moomaw CJ, Alwell K, Woo D, Flaherty ML, et al. Stroke incidence is decreasing in whites but not in blacks: a population-based estimate of temporal trends in stroke incidence from the Greater Cincinnati/Northern Kentucky Stroke Study. Stroke. 2010;41(7):1326–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used during the current study are available from the corresponding author on reasonable request.