Abstract
Objectives
To examine the effect of pre-existing atrial fibrillation (AF) and associated therapy on the risk of arterial thromboembolism (ATE) and death following pneumonia.
Design, setting and participants
Population-based cohort study (1997–2012) of 88 315 patients with first-time hospitalisation with pneumonia in Northern Denmark.
Results
Of the included patients (median age 73.4 years), 8880 (10.1%) had pre-existing AF. The risk of ATE within 30 days of admission was 5.2% in patients with AF and 3.6% in patients without AF. After adjustment for higher age and comorbidity, the adjusted HR (aHR) with AF was 1.06 (95% CI 0.96 to 1.18). Among patients with AF, reduced risk of ATE was observed in vitamin-K antagonist users compared with non-users (aHR 0.74 (95% CI 0.61 to 0.91)). Thirty-day mortality was 20.1% in patients with AF and 13.9% in patients without AF. Corresponding 1-year mortalities were 43.7% and 30.3%. The aHRs for 30-day and 1-year mortality with AF were 1.00 (95% CI 0.94 to 1.05) and 1.01 (95% CI 0.98 to 1.05). In patients with AF, reduced mortality risk was observed in users of vitamin-K antagonists (aHR 0.70 (95% CI 0.63 to 0.77)) and β-blockers (aHR 0.77 (95% CI 0.70 to 0.85). Increased mortality was found in digoxin users (aHR 1.16 (95% CI 1.06 to 1.28)).
Conclusions
Pre-existing AF is frequent in patients hospitalised with pneumonia and a marker of increased risk of ATE and death, explained by higher patient age and comorbidity. Prognosis is closely related to preadmission medical treatment for AF.
Keywords: EPIDEMIOLOGY
Strengths and limitations of this study.
Large cohort including 88 315 consecutive patients with first pneumonia hospitalisation.
Complete follow-up data and reliable end points.
Conduct within an equal access healthcare system.
Use of detailed, high-quality registries allowed for extensive confounder adjustment.
As in all observational studies, the presence of residual or unmeasured confounding cannot be excluded.
Introduction
Atrial fibrillation (AF) is the most common cardiac arrhythmia. This condition affects 1.0–1.8% of the adult population in western countries and >10% of individuals ≥80 years of age.1 2 Since the prevalence of AF increases with age, the numbers of patients with AF are expected to increase dramatically in the future.1 2 While AF has been recognised as a major risk factor for all-cause death and stroke in the general population,3–5 very little is known regarding the effect of AF on the prognosis of other acute disease, including severe infections.
Pneumonia remains a leading cause of death globally.6 Survival has not improved and increasing incidence rates have been reported in recent years, probably because of the ageing population and increasing prevalence of comorbid conditions.7 8 New onset of AF during pneumonia has been linked to poor prognosis,9 but the prognostic effect of pre-existing AF on mortality and complications in patients with pneumonia has not been investigated. The pathophysiological alterations that occur during pneumonia may provoke complications (eg, haemodynamic instability and thromboembolic events) in patients with AF. A worsening of pre-existing cardiac diseases often occurs during pneumonia,10 and the risk of stroke temporarily increases threefold in patients with pneumonia.11 While AF might lead to increased mortality and complications following pneumonia, medications frequently prescribed to patients with AF such as vitamin-K antagonists and β-blockers might alter these associations.
Since both AF and pneumonia incidence rates are increasing, a negative prognostic impact of AF on the clinical course of pneumonia has important clinical and public health implications. We conducted a large population-based cohort study to examine the effects of pre-existing AF and concomitant drug therapy on the risk of arterial thromboembolism and death in patients with pneumonia.
Methods
Setting
This cohort study was conducted using prospectively collected data obtained from medical registries within the North and Central Denmark Regions. This geographic area includes approximately 1.8 million inhabitants. Every Danish citizen is assigned a unique personal identification number that allows unambiguous cross-linking of registry data at the individual level. Tax-funded, unrestricted healthcare is provided for all Danish citizens through a national health insurance programme.
Identification of the study cohort
We included all first-time cases of hospitalised pneumonia (1 January 1997 to 31 December 2012) in patients aged ≥15 years in the study cohort. We used the Danish National Patient Register (DNPR) to identify the cohort members. The DNPR includes all hospital admissions and hospital outpatient clinic contacts from 1977 and 1995, respectively. The registry maintains data on admission and discharge dates, surgical procedure codes and diagnostic codes.12
Identification of patients with AF
Using DNPR data, we identified all cohort members who had received a diagnosis of AF within 5 years before the date of hospital admission for pneumonia. We included AF diagnoses given during previous hospitalisations and previous hospital outpatient clinic visits. Since very few cardiologists practise outside the Danish public hospital system, most patients with known AF are likely to be registered in the DNPR.
Comorbidity data
Adjustment for comorbidities in the mortality analyses was performed using data retrieved from the DNPR. These data included the 19 conditions in the Charlson Index, which has been validated for prediction of mortality following hospital admission.13 14 We also noted previous diagnoses of valvular heart disease, alcoholism and obesity. We included the specific thromboembolism risk factors included in the CHA2DS2-VASc score (ie, congestive heart failure, hypertension, age, diabetes and previous stroke/transient cerebral ischaemic attack, vascular disease and sex) in the thromboembolism risk analyses. This scoring system has been validated for prediction of stroke risk in patients with AF.15 We also summarised data on previous episodes of gastrointestinal bleeding and head injuries because these conditions could contraindicate anticoagulant treatment.
Data on preadmission prescriptions
The Aarhus University Prescription Database receives information on filled prescriptions from all pharmacies in the study area. We acquired data on filled prescriptions for the most commonly prescribed drugs in AF treatment (ie, vitamin K antagonists, aspirin, β-blockers, non-dihydropyridine calcium-channel blockers, amiodarone and digoxin) from this database.16 17 We also included data on prescriptions for statins, which have been associated with favourable outcome in pneumonia.18
Markers of frailty and health awareness
We used data obtained from the National Health Service Register to further control for potential differences in patient frailty and health awareness. This registry of information on contacts with general practitioners (GPs) supplied data on preventive consultations, social-medicine-related consultations, conversational therapy at the GP within 1 year preceding pneumonia admission, influenza vaccination and application for reimbursement due to chronic or terminal illness.19
Outcomes
Outcomes were any hospitalised episodes of arterial thromboembolism within 30 days (index admission included), and death within 30 days and 1 year following the hospital admission date for pneumonia. Information on arterial thromboembolism was obtained from the DNPR. We defined arterial thromboembolism as an in-hospital diagnosis of non-haemorrhage stroke, or thrombosis or embolism in arteries of the extremities, the mesenteric arteries or in unspecified arteries. We assessed the vital statistics of each cohort member using the Civil Registration System (CRS). This database includes information on all individuals who have lived in Denmark at any time since 1968. The CRS is updated daily and records the date of birth and exact date of death or emigration.20
Statistical analyses
Follow-up began at hospital admission with pneumonia and continued until 1 January 2013, or until death or migration, whichever occurred first. We computed the cumulative incidence (risk) of arterial thromboembolism within 30 days of admission in patients with and without pre-existing AF and accounted for the competing risk of death in the analysis.21 Further comparisons were performed using Cox regression analyses to estimate crude HRs, and HRs adjusted for the thromboembolism risk factors in the CHA2DS2-VASc score. We also performed an analysis stratified by the presence or absence of previously diagnosed arterial thromboembolism to examine whether an association between pre-existing AF and development of arterial thromboembolism was related to repeated events. Among patients with and without AF, the effect of preadmission treatment with vitamin K antagonists or aspirin on risk of arterial thromboembolism was evaluated by comparing users to non-users. We repeated the analyses for patients without contraindications for anticoagulant therapy. All diagnoses coded during a given admission were assigned to the same discharge date in the DNPR. Consequently, we were unable to assess the actual temporal relationship between pneumonia onset and arterial thromboembolism in patients treated for the two conditions during the same admission. Arterial thromboembolism might have preceded pneumonia in some patients. Thus, we reassessed the 30-day risk of arterial thromboembolism, and only considered diagnoses assigned after discharge from the index pneumonia admission.
Cumulative mortality risks at 30 days and 1 year after admission with pneumonia in patients with and without AF were assessed using the Kaplan-Meier method and compared using Cox regression. The HRs were adjusted for sex and age. Further adjustments were performed for the conditions included in the Charlson Index, and for valvular heart disease, alcoholism and obesity, and GP contacts regarding preventive consultations, social-medicine-related consultations, conversational therapy, vaccination for influenza and reimbursement due to chronic or terminal illness. To examine whether the association between AF and pneumonia mortality was related to coexisting cardiovascular diseases or pneumonia severity, the analyses were repeated using stratification by previous myocardial infarction, congestive heart failure, treatment with mechanical ventilation, and admission to the intensive care unit during the index admission.
We also stratified the data by AF status and compared mortality at 30 days and 1 year in users and non-users of vitamin K antagonists, aspirin, β-blockers, calcium-channel blockers, digoxin, amiodarone and statins using the Kaplan-Meier method and Cox regression analyses. Crude HRs and HRs adjusted for the potential confounders described above were estimated from the Cox regression analyses.
All Cox analyses were preceded by a graphical verification of the proportional hazards assumption. Statistical analyses were performed using Stata V.11.2 software (StataCorp, College Station, Texas, USA).
Ethics
The study protocol was approved by the Danish Data Protection Agency (record numbers 2009-41-3866 and 2013-41-1924). Informed consent from the participants was not required.
Results
Of the 88 315 patients hospitalised for pneumonia between 1997 and 2012, 8880 (10.1%) had a previous diagnosis of AF (table 1). Follow-up data were incomplete for 149 individuals. Patients with AF were older than patients without AF (median age 80.0 vs 72.3 years), yet the proportion of males was higher among patients with AF than among patients without AF (58.0% vs 52.5%). Compared with patients without AF, patients with AF had a substantially higher burden of comorbidity and were more likely to have coexisting cardiac diseases (eg, previous myocardial infarction (18.6% vs 8.3%) and congestive heart failure (34.8% vs 7.2%)). Patients with AF were as likely to be admitted to the intensive care unit (ICU; 7.5% vs 7.0%) and to be treated with mechanical ventilation (4.9% vs 5.3%) as patients without AF.
Table 1.
Characteristics of 88 315 patients admitted to hospital with pneumonia according to pre-existing atrial fibrillation
| No AF, n (%) | AF, n (%) | Total, n (%) | |
|---|---|---|---|
| All patients | 79 435 (89.9) | 8880 (10.1) | 88 315 (100.0) |
| Age, median (IQR), years | 72.3 (59.2–81.5) | 80.0 (73.3–85.4) | 73.4 (60.6–82.1) |
| Female | 37 751 (47.5) | 3738 (42.1) | 41 489 (47.0) |
| Male | 41 684 (52.5) | 5142 (58.0) | 46 826 (53.0) |
| Admission period | |||
| 1997–2001 | 20 555 (25.9) | 1730 (19.5) | 22 285 (25.2) |
| 2002–2006 | 26 552 (33.4) | 2839 (32.0) | 29 391 (33.3) |
| 2007–2012 | 32 328 (40.7) | 4311 (48.6) | 36 639 (41.5) |
| Comorbidities, Charlson Index | |||
| Myocardial infarction | 6605 (8.3) | 1654 (18.6) | 8259 (9.4) |
| Congestive heart failure | 5724 (7.2) | 3090 (34.8) | 8814 (9.98) |
| Peripheral vascular disease | 6496 (8.2) | 1365 (15.4) | 7861 (8.9) |
| Cerebrovascular disease | 11 234 (14.1) | 2434 (27.4) | 13 668 (15.5) |
| Dementia | 1764 (2.2) | 346 (3.9) | 2110 (2.4) |
| Chronic pulmonary disease | 15 101 (19.0) | 2305 (26.0) | 17 406 (19.7) |
| Connective tissue disease | 4165 (5.2) | 624 (7.0) | 4789 (5.4) |
| Ulcer disease | 6542 (8.2) | 1013 (11.4) | 7555 (8.6) |
| Mild liver disease | 1540 (1.9) | 132 (1.5) | 1672 (1.9) |
| Diabetes, types I and II | 6159 (7.8) | 1315 (14.8) | 7474 (8.5) |
| Hemiplegia | 545 (0.7) | 47 (0.5) | 592 (0.7) |
| Moderate to severe renal disease | 2885 (3.6) | 790 (8.9) | 3675 (4.2) |
| Diabetes with end-organ failure | 3224 (4.1) | 791 (8.6) | 3985 (4.5) |
| Any tumour | 12 536 (15.8) | 1743 (19.6) | 14 279 (16.2) |
| Leukaemia | 855 (1.1) | 87 (1.0) | 942 (1.1) |
| Lymphoma | 1561 (2.0) | 178 (2.0) | 1.739 (2.0) |
| Moderate to severe liver disease | 463 (0.6) | 47 (0.5) | 510 (0.6) |
| Metastatic solid tumour | 2.649 (3.3) | 267 (3.0) | 2916 (3.3) |
| AIDS | 105 (0.1) | 2 (0.0) | 107 (0.1) |
| Comorbidities, other types | |||
| Valvular heart disease | 2798 (3.5) | 1309 (14.7) | 4107 (4.7) |
| Hypertension | 13 905 (17.5) | 3623 (40.8) | 17 528 (19.8) |
| Transient cerebral ischaemic attack | 3101 (3.9) | 732 (8.2) | 3833 (4.3) |
| Obesity | 3144 (4.0) | 577 (6.5) | 3721 (4.2) |
| Alcoholism | 3865 (4.9) | 301 (3.4) | 4166 (4.7) |
| Frailty and health awareness markers | |||
| Influenza vaccination given | 13 456 (16.9) | 2506 (28.2) | 15 962 (18.1) |
| Preventive GP consultation | 13 148 (16.6) | 2065 (23.3) | 15 213 (17.2) |
| Social medicine GP consultation | 9908 (12.5) | 1722 (19.4) | 11 630 (13.2) |
| GP conversational therapy | 4610 (5.8) | 452 (5.1) | 5062 (5.7) |
| Reimbursement for chronic illness | 2097 (2.6) | 294 (3.3) | 2391 (2.7) |
| Reimbursement for terminal illness | 281 (0.4) | 34 (0.4) | 315 (0.4) |
| CHA2DS2-VASc score* | |||
| 0 | 11 420 (14.4) | 183 (2.1) | 11 603 (13.1) |
| 1 | 17 576 (22.1) | 580 (6.5) | 18 156 (20.6) |
| 2+ | 50 439(63.5) | 8117 (91.4) | 58 556 (66.3) |
| Admission variables | |||
| Intensive care unit admission | 5588 (7.0) | 669 (7.5) | 6257 (7.1) |
| Mechanical ventilation | 4189 (5.3) | 438 (4.9) | 4627 (5.2) |
| Non-invasive ventilation | 1666 (2.1) | 197 (2.2) | 1863 (2.1) |
| Preadmission cardiovascular drug use | |||
| Statins | 10 145 (12.8) | 2097 (22.6) | 12 242 (13.9) |
| Aspirin | 14 944 (18.8) | 3251 (36.6) | 18 195 (20.6) |
| Vitamin K-antagonists | 2207 (2.8) | 3396 (38.2) | 5603 (6.3) |
| β-Blockers | 11 967 (15.1) | 3878 (43.7) | 15 845 (17.9) |
| Non-dihydropyridine CCBs† | 2415 (3.0) | 859 (9.7) | 3274 (3.7) |
| Digoxin | 3307 (4.2) | 3622 (40.8) | 6929 (7.9) |
| Amiodarone | 319 (0.4) | 685 (7.7) | 1004 (1.1) |
*Disease categories in the CHA2DS2-VASc score include congestive heart failure, hypertension, age >75 years (double weight), diabetes, previous stroke (double weight), vascular disease, age >65 years, and sex category (female=1, male=0).
†Non-dihydropyridine calcium-channel blockers.
AF, atrial fibrillation; CCBs, calcium-channel blockers; GP, general practitioner.
Risk of arterial thromboembolism
Within 30 days from admission, the cumulative incidence of arterial thromboembolism was 3.6% in patients without AF and 5.2% in patients with AF (table 2). The HR for arterial thromboembolism was 1.61 (95% CI 1.46 to 1.78). After adjustment for prevalence of the risk factors in the CHA2DS2-VASc-score, the HR decreased substantially to 1.06 (95% CI 0.96 to 1.18). In patients without previous stroke, the HR for arterial thromboembolism adjusted for the CHA2DS2-VASc risk factors was 0.97 (95% CI 0.83 to 1.14), and the adjusted HR for patients who had a previous stroke was 1.17 (95% CI 1.02 to 1.35). The incidence for episodes of arterial thromboembolism recorded after discharge from the index pneumonia admission was 0.8% in patients with AF and 0.5% in patients without AF (aHR=1.13 (95% CI 0.87 to 1.47).
Table 2.
Risk of arterial thromboembolism* within 30 days following admission for pneumonia, by atrial fibrillation status
| No AF | AF | |
|---|---|---|
| Overall | ||
| Cumulative incidence, % | 3.6 (3.1 to 3.4) | 5.2 (4.7 to 5.7) |
| HRs | ||
| Crude | 1.00 (ref) | 1.61 (1.46 to 1.78) |
| Adjusted for CHA2DS2-VASc† | 1.00 (ref) | 1.06 (0.96 to 1.18) |
| Vitamin K antagonists | ||
| Cumulative incidence, % | ||
| Non-users | 3.2 (3.1 to 3.4) | 5.7 (5.1 to 6.4) |
| Users | 4.2 (3.4 to 5.0) | 4.3 (3.7 to 5.0) |
| HRs | ||
| Non-user | 1.00 (ref) | 1.00 (ref) |
| Users, Crude | 1.29 (1.05 to 1.59) | 0.74 (0.61 to 0.90) |
| Adjusted for CHA2DS2-VASc† | 0.97 (0.79 to 1.20) | 0.74 (0.61 to 0.91) |
| Aspirin | ||
| Cumulative incidence, % | ||
| Non-users | 2.9 (2.8 to 3.0) | 5.3 (4.8 to 5.9) |
| Users | 4.8 (4.5 to 5.2) | 5.0 (4.3 to 5.8) |
| HRs | ||
| Non-users | 1.00 (ref) | 1.00 (ref) |
| Users, crude | 1.69 (1.55 to 1.84) | 0.94 (0.78 to 1.14) |
| Adjusted for CHA2DS2-VASc† | 0.96 (0.88 to 1.06) | 0.83 (0.68 to 1.01) |
Cumulative incidence values were computed with non-stroke associated death as a competing risk.
95% CIs are presented in parentheses.
*Defined as a diagnostic code for ischaemic stroke, unspecified stroke, thrombosis, or embolism in arteries of the extremities, mesenteric arteries, or unspecified arteries assigned during index admission or subsequent admissions.
†Adjusted for prevalence of the risk factors included in the CHA2DS2-VASc-score (ie, congestive heart failure, hypertension, age, diabetes, previous stroke, vascular disease and sex).
AF, atrial fibrillation; HR, hazard ratio.
Effect of preadmission drug use on risk of arterial thromboembolism
In patients with AF, users of vitamin K antagonists had a markedly lower risk of arterial thromboembolism compared with non-users (aHR=0.74 (95% CI 0.61 to 0.91; table 2). Users of aspirin had an adjusted HR of 0.83 (95% CI 0.68 to 1.01). The exclusion of patients with potential contraindications for anticoagulant therapy did not change the results (see online supplementary table S2).
In patients with AF without and with previous stroke, users of vitamin K antagonists had similar adjusted HRs for arterial thromboembolism (HR=0.74 (95% CI 0.54 to 1.01) and HR=0.74 (95% CI 0.57 to 0.96), respectively), compared with non-users of vitamin K antagonists. Users of aspirin had adjusted HRs for arterial thromboembolism of 0.89 (95% CI 0.66 to 1.20) and 0.78 (95% CI 0.61 to 1.01), compared with non-users of aspirin.
Mortality
The 30-day mortality rate was 20.1% in patients with AF and 13.9% in patients without AF (table 3). The effect of AF differed little between men and women. The presence or absence of previous myocardial infarction had limited effect on the estimates. Mortality rates were substantially higher in patients with congestive heart failure compared with patients without heart failure, but coexisting AF and heart failure in patients only resulted in a minor increase in mortality rate compared with patients with heart failure only (23.7% vs 21.2%). In patients treated with mechanical ventilation, 30-day mortality was 34.6% in patients with AF and 27.8% in patients without AF.
Table 3.
Mortality rates and HRs at 30 and 365 days following admission, by atrial fibrillation status
| No AF, % | AF, % | Crude HR* | Adjusted HR, step 1*† | Adjusted HR, step 2*‡ | |
|---|---|---|---|---|---|
| 30-day mortality | |||||
| Overall | 13.9 (13.7 to 14.1) | 20.1 (19.3 to 21.0) | 1.49 (1.42 to 1.57) | 1.08 (1.03 to 1.14) | 1.00 (0.94 to 1.05) |
| Sex | |||||
| Male | 14.5 (14.2 to to 14.9) | 20.4 (19.3 to 21.5) | 1.45 (1.36 to 1.55) | 1.09 (1.02 to 1.16) | 1.01 (0.95 to 1.09) |
| Female | 13.3 (12.9 to to 13.6) | 19.7 (18.5 to 21.0) | 1.53 (1.42 to 1.66) | 1.07 (0.99 to 1.16) | 0.97 (0.90 to 1.06) |
| Previous myocardial infarction | |||||
| Not present | 13.7 (13.4 to 13.9) | 20.1 (19.1 to 21.0) | 1.52 (1.44 to 1.60) | 1.08 (1.02 to 1.14) | 0.98 (0.93 to 1.04) |
| Present | 16.8 (15.9 to 17.7) | 20.4 (18.5 to 22.4) | 1.24 (1.09 to 1.40) | 1.12 (0.99 to 1.27) | 1.06 (0.93 to 1.20) |
| Congestive heart failure | |||||
| Not present | 13.4 (13.1 to 13.6) | 18.2 (17.2 to 19.2) | 1.40 (1.31 to 1.49) | 1.01 (0.95 to 1.08) | 0.98 (0.92 to 1.05) |
| Present | 21.2 (20.1 to 22.2) | 23.7 (22.2 to 25.2) | 1.13 (1.03 to 1.24) | 1.06 (0.97 to 1.16) | 1.04 (0.95 to 1.14) |
| ICU admission | |||||
| No | 13.1 (12.9 to 13.4) | 19.1 (18.3 to 20.0) | 1.50 (1.43 to 1.59) | 1.08 (1.02 to 1.14) | 0.99 (0.93 to 1.04) |
| Yes | 24.8 (23.7 to 25.9) | 32.1 (28.7 to 35.8) | 1.38 (1.20 to 1.59) | 1.05 (0.91 to 1.22) | 1.05 (0.90 to 1.22) |
| Mechanical ventilation | |||||
| No | 13.2 (12.9 to 13.4) | 19.4 (18.5 to 20.2) | 1.52 (1.44 to 1.60) | 1.08 (1.03 to 1.14) | 0.99 (0.94 to 1.05) |
| Yes | 27.8 (26.5 to 29.2) | 34.6 (30.4 to 39.3) | 1.33 (1.12 to 1.58) | 1.05 (0.89 to 1.25) | 1.06 (0.89 to 1.27) |
| 365-day mortality | |||||
| Overall | 30.3 (30.0 to 30.6) | 43.7 (42.7 to 44.7) | 1.57 (1.52 to 1.63) | 1.14 (1.10 to 1.18) | 1.01 (0.98 to 1.05) |
| Sex | |||||
| Male | 31.8 (31.3 to 32.2) | 43.9 (41.8 to 45.0) | 1.50 (1.43 to 1.57) | 1.12 (1.07 to 1.17) | 1.01 (0.96 to 1.06) |
| Female | 28.7 (28.2 to 29.1) | 43.4 (42.5 to 45.3) | 1.66 (1.58 to 1.75) | 1.17 (1.11 to 1.24) | 1.01 (0.96 to 1.07) |
| Previous myocardial infarction | |||||
| Not present | 29.5 (29.2 to 29.9) | 43.6 (42.4 to 44.7) | 1.61 (1.55 to 1.68) | 1.15 (1.10 to 1.19) | 1.02 (0.98 to 1.06) |
| Present | 38.7 (37.5 to 39.9) | 44.3 (41.9 to 46.7) | 1.20 (1.10 to 1.30) | 1.09 (1.01 to 1.19) | 0.99 (0.91 to 1.08) |
| Congestive heart failure | |||||
| Not present | 29.0 (28.7 to 29.4) | 40.6 (39.3 to 41.9) | 1.50 (1.44 to 1.57) | 1.08 (1.04 to 1.13) | 1.02 (0.98 to 1.07) |
| Present | 46.7 (45.4 to 47.9) | 49.5 (47.7 to 51.3) | 1.09 (1.02 to 1.16) | 1.03 (0.97 to 1.10) | 1.01 (0.95 to 1.08) |
| ICU admission | |||||
| No | 29.5 (29.2 to 29.7) | 42.3 (41.9 to 44.1) | 1.58 (1.53 to 1.65) | 1.14 (1.10 to 1.18) | 1.00 (0.97 to 1.04) |
| Yes | 40.5 (39.2 to 41.8) | 52.3 (48.6 to 56.2) | 1.43 (1.27 to 1.60) | 1.09 (0.97 to 1.22) | 1.07 (0.95 to 1.20) |
| Mechanical ventilation | |||||
| No | 29.6 (29.27 to 29.9) | 43.1 (42.0 to 44.2) | 1.59 (1.54 to 1.65) | 1.14 (1.10 to 1.18) | 1.01 (0.97 to 1.04) |
| Yes | 42.8 (41.3 to 44.3) | 54.9 (50.2 to 59.7) | 1.41 (1.24 to 1.62) | 1.10 (0.96 to 1.26) | 1.11 (0.96 to 1.28) |
*Reference group: patients without AF.
†Step 1: adjusted for age and sex.
‡Step 2: adjusted for age, sex, the 19 individual disease categories in the Charlson Index, hypertension, heart valve disease, transient cerebral ischaemic attack, obesity, alcoholism, and contacts with the general practitioner regarding preventive consultations, social-medicine-related consultations, conversational therapy, vaccination for influenza, and reimbursement due to chronic or terminal illness.
AF, atrial fibrillation; HR, hazard ratio; ICU, intensive care unit.
95% CIs are presented in parentheses.
The overall HR for death at 30 days following admission was 1.49 (95% CI 1.42 to 1.57) (table 3). After controlling for the effect of age and sex, the estimate decreased to 1.08 (95% CI 1.03 to 1.14). Further adjustment for comorbid conditions and lifestyle factors resulted in an HR of 1.00 (95% CI 0.94 to 1.05). Likewise, analyses comparing patients with and without AF and stratified by sex, previous myocardial infarction, congestive heart failure, ICU admission and mechanical ventilation yielded HRs close to a value of one in the fully adjusted model.
One-year mortality was 43.7% in patients with AF and 30.3% in patients without AF. The corresponding fully adjusted HR was 1.01 (95% CI 0.98 to 1.05). Similar estimates were obtained when the analyses were stratified according to sex, previous myocardial infarction, heart failure and ICU admission (table 3).
Effect of preadmission drug use on mortality
At 30 days of follow-up in patients with AF, the adjusted HR for death comparing users to non-users of vitamin K antagonists was 0.70 (95% CI 0.63 to 0.77) (table 4). Similarly, reduced mortality was observed in patients with AF who used β-blockers (aHR=0.77 (95% CI 0.70 to 0.85)) and statins (aHR=0.70 (95% CI 0.61 to 0.80)). No difference in mortality was observed in users compared with non-users of aspirin (aHR=0.98 (95% CI 0.89 to 1.08)). The values for the estimates changed very little when the follow-up period was extended to 1 year (table 4).
Table 4.
Effects of preadmission drug use on mortality at 30 and 365 days, by atrial fibrillation status
| Non-users, % | Users, % | Crude HR* | Adjusted HR, step 1*† | Adjusted HR, step 2*‡ | |
|---|---|---|---|---|---|
| 30-day mortality | |||||
| Vitamin K antagonists | |||||
| No AF | 13.9 (13.7 to 14.2) | 13.8 (12.4 to 15.3) | 0.99 (0.88 to 1.11) | 0.87 (0.77 to 0.97) | 0.83 (0.74 to 0.93) |
| AF | 23.2 (22.1 to 24.3) | 15.2 (14.0 to 16.4) | 0.62 (0.56 to 0.69) | 0.69 (0.62 to 0.76) | 0.70 (0.63 to 0.77) |
| Aspirin | |||||
| No AF | 13.2 (13.0 to 13.5) | 17.0 (16.4 to 17.6) | 1.31 (1.25 to 1.37) | 0.95 (0.90 to 0.99) | 0.89 (0.85 to 0.93) |
| AF | 19.6 (18.6 to 20.7) | 21.0 (19.6 to 22.4) | 1.07 (0.97 to 1.18) | 0.99 (0.89 to 1.08) | 0.98 (0.89 to 1.08) |
| β-Blockers | |||||
| No AF | 13.7 (13.5 to 14.0) | 15.0 (13.3 to 15.6) | 1.09 (1.04 to 1.15) | 0.93 (0.88 to 0.97) | 0.90 (0.86 to 0.96) |
| AF | 22.4 (21.2 to 23.6) | 17.2 (16.0 to 18.4) | 0.74 (0.68 to 0.82) | 0.78 (0.71 to 0.86) | 0.77 (0.70 to 0.85) |
| Calcium-channel blockers | |||||
| No AF | 13.8 (13.6 to 14.1) | 16.7 (15.27 to 18.2) | 1.23 (1.11 to 1.36) | 1.00 (0.91 to 1.11) | 1.05 (0.95 to 1.16) |
| AF | 20.0 (19.2 to 20.9) | 20.1 (18.4 to 23.9) | 1.06 (0.90 to 1.23) | 1.10 (0.95 to 1.29) | 1.17 (1.00 to 1.36) |
| Statins | |||||
| No AF | 14.3 (14.0 to 14.6) | 11.3 (10.7 to 12.0) | 0.78 (0.73 to 0.83) | 0.75 (0.70 to 0.80) | 0.69 (0.64 to 0.73) |
| AF | 21.67 (20.7 to 22.7) | 15.0 (13.5 to 16.6) | 0.67 (0.59 to 0.75) | 0.75 (0.66 to 0.84) | 0.70 (0.61 to 0.80) |
| Amiodarone | |||||
| No AF | 13.9 (13.7 to 14.1) | 19.8 (15.8 to 24.6) | 1.46 (1.14 to 1.87) | 1.25 (0.98 to 1.60) | 1.23 (0.96 to 1.58) |
| AF | 20.1 (19.2 to 21.0) | 20.3 (17.5 to 23.5) | 1.02 (0.86 to 1.21) | 1.21 (1.02 to 1.44) | 1.18 (1.00 to 1.42) |
| Digoxin | |||||
| No AF | 13.5 (13.2 to 13.7) | 24.0 (22.5 to 25.5) | 1.90 (1.76 to 2.04) | 1.26 (1.17 to 1.36) | 1.23 (1.15 to 1.33) |
| AF | 18.4 (17.4 to 19.5) | 22.5 (21.2 to 23.9) | 1.25 (1.14 to 1.37) | 1.19 (1.08 to 1.31) | 1.16 (1.06 to 1.28) |
| 365-day mortality | |||||
| Vitamin K antagonists | |||||
| No AF | 30.2 (29.9 to 30.6) | 34.3 (32.3 to 36.3) | 1.15 (1.07 to 1.24) | 0.99 (0.92 to 1.07) | 0.92 (0.85 to 0.99) |
| AF | 48.9 (47.6 to 50.3) | 35.2 (33.4 to 36.9) | 0.64 (0.60 to 0.69) | 0.70 (0.65 to 0.75) | 0.72 (0.67 to 0.77) |
| Aspirin | |||||
| No AF | 28.7 (28.3 to 29.0) | 37.5 (36.7 to 38.3) | 1.38 (1.34 to 1.42) | 1.00 (0.97 to 1.03) | 0.91 (0.88 to 0.94) |
| AF | 43.1 (41.8 to 44.5) | 44.7 (42.3 to 46.4) | 1.06 (0.99 to 1.13) | 0.98 (0.91 to 1.04) | 0.98 (0.91 to 1.05) |
| β-Blockers | |||||
| No AF | 29.7 (29.4 to 30.1) | 33.5 (32.6 to 34.3) | 1.15 (1.11 to 1.19) | 0.96 (0.93 to 1.00) | 0.91 (0.88 to 0.94) |
| AF | 46.6 (45.2 to 48.0) | 39.8 (38.3 to 41.4) | 0.80 (0.75 to 0.85) | 0.83 (0.78 to 0.89) | 0.83(0.78 to 0.89) |
| Calcium-channel blockers | |||||
| No AF | 30.1 (29.8 to 30.5) | 35.9 (33.4 to 37.8) | 1.24 (1.16 to 1.33) | 1.01 (0.94 to 1.08) | 1.04 (0.97 to 1.11) |
| AF | 43.9 (42.8 to 45.0) | 41.7 (38.5 to 45.1) | 0.94 (0.85 to 1.05) | 0.99 (0.88 to 1.10) | 1.03 (0.93 to 1.15) |
| Statins | |||||
| No AF | 30.1 (30.3 to 31.0) | 27.7 (26.8 to 28.6) | 0.88 (0.84 to 0.91) | 0.82 (0.79 to 0.86) | 0.72 (0.69 to 0.75) |
| AF | 46.3 (45.2 to 47.6) | 34.7 (32.69 to 36.9) | 0.68 (0.63 to 0.74) | 0.76 (0.70 to 0.83) | 0.72 (0.66 to 0.79) |
| Amiodarone | |||||
| No AF | 30.2 (29.9 to 30.6) | 43.8 (38.5 to 49.5) | 1.56 (1.32 to 1.85) | 1.38 (1.12 to 1.57) | 1.26 (1.06 to 1.49) |
| AF | 43.7 (42.6 to 44.8) | 43.8 (40.1 to 47.6) | 1.01 (0.90 to 1.14) | 1.19 (1.06 to 1.34) | 1.15 (1.02 to 1.30) |
| Digoxin | |||||
| No AF | 29.5 (29.2 to 29.9) | 47.9 (46.2 to 49.6) | 1.86 (1.77 to 1.96) | 1.25 (1.18 to 1.31) | 1.21 (1.15 to 1.27) |
| AF | 40.7 (39.6 to 42.1) | 48.0 (45.3 to 49.6) | 1.24 (1.16 to 1.32) | 1.17 (1.10 to 1.25) | 1.15 (1.08 to 1.23) |
95% CIs are presented in parentheses.
*Reference group: patients in the stratum without the given drug exposure.
†Step 1: adjusted for age and sex.
‡Step 2: adjusted for age, sex, the 19 individual disease categories in the Charlson Index, hypertension, heart valve disease, transient cerebral ischaemic attack, obesity, alcoholism, and contacts with a general practitioner regarding preventive consultations, social-medicine-related consultations, conversational therapy, vaccination for influenza, and reimbursement due to chronic or terminal illness.
AF, atrial fibrillation.
Increased 30-day mortality was observed in patients with AF who were users of amiodarone (aHR=1.18 (95% CI 1.00 to 1.42)) and users of digoxin (aHR=1.16 (95% CI 1.06 to 1.28)) (table 4). Uses of amiodarone and digoxin were also associated with increased 1-year mortality. Also, the use of calcium-channel blockers was associated with an increased 30-day mortality (aHR=1.17 (95% CI 1.00 to 1.36)), but no difference was observed for 1-year mortality (aHR=1.03 (95% CI 0.93 to 1.15) (table 4).
Discussion
In this population-based study of 88 315 patients hospitalised with pneumonia, we found that compared with patients without AF, patients with pre-existing AF had a 60% higher risk of arterial thromboembolism within 30 days of admission and 50% higher mortality after 30 days and 1 year. However, after controlling for other prognostic factors associated with AF, the excess risk of arterial thromboembolism and mortality was found to be related to major differences in age and coexisting diseases, rather than to AF directly.
There was a robust association between preadmission use of vitamin K antagonists and reduced risk of arterial thromboembolism following pneumonia in patients with AF. We also observed markedly reduced risks of death in patients with AF who were treated with statins, β-blockers or vitamin K antagonists. However, compared with non-users, there was an increased risk of death in users of amiodarone and digoxin.
Strengths and limitations
The risk of selection bias was minimal in this study because we used population-based data with virtually complete follow-up. A positive predictive value of 90% has been estimated for the pneumonia diagnoses recorded in the DNPR; 87% of the cases represent community-acquired infections.22 Atrial fibrillation is also accurately coded in the DNPR. Positive predictive value estimates range from 93% to 99%.23–25 Since AF and atrial flutter share the same code in the DNPR, we were unable to distinguish between these arrhythmias. However, of patients coded with atrial fibrillation/flutter, only 5–6% have atrial flutter.23–25 The coding system also did not allow for differentiation between paroxysmal, persistent and permanent AF. Finally, the DNPR's positive predictive value for ischaemic stroke is 88–100%. Of the patients coded with unspecified stroke, 57–70% have ischaemic stroke.26 27
Since we used filled prescriptions as a measure of actual drug intake, non-adherence or treatment discontinuation before admission could have biased the effect of any given drug towards the null value. However, we concluded that this source of bias was a low concern for the results of this study. We did not have data on in-hospital medications; thus, we were unable to assess whether pre-admission treatments were continued, discontinued or altered (eg, change of vitamin K antagonists to low-molecular heparins) during admission. Any bias introduced by the discontinuation of a drug during admission would be directed towards the null. Use of the prophylactic drugs (eg, statins and β-blockers) could be markers for unmeasured, but greater, health awareness among the patient population. However, the welfare structure of the Danish society reduces the risk of confounding by differences in socioeconomic status; in Denmark, statin users have unhealthier lifestyles than statin non-users.28 29
We adjusted our estimates for a broad range of potentially confounding factors, and the major diseases included in the Charlson Index have a mean positive predictive value of 98%.30 However, confounding by unmeasured variables and residual confounding might have affected the results of this observational study. Since we were unable to identify conditions managed only by a patient's GP, the prevalence values for hypertension, obesity and alcoholism are most likely underestimated in our study.
Interpretation
To the best of our knowledge, this study was the first to examine the effect of pre-existing AF and its treatment on the prognosis for pneumonia patients. This study added to previous studies reporting a high risk of thromboembolic complications and mortality in pneumonia patients.7 11 Our results indicated that there was a similar adjusted risk of arterial thromboembolism and an equal adjusted mortality in hospitalised pneumonia patients with and without AF, after confounders were accounted for. We also did not find that there were different rates in ICU admission or treatment with mechanical ventilation for patients with AF compared with patients without AF. Our finding of similar prognosis in pneumonia patients with and without pre-existing AF conflicts with previous findings of increased mortality in pneumonia patients with new-onset AF.9 We can only speculate on the mechanisms behind this difference, yet as opposed to patients with new-onset AF, patients with pre-existing AF are more likely to receive well-balanced treatment for AF at the time of pneumonia; possibly, current management approaches can counterbalance the potentially deleterious effects of AF. Indeed, we have shown that preadmission medical treatment for AF appears to have great prognostic impact in pneumonia patients with this condition. Alternatively, new-onset AF during pneumonia may not be associated with worse prognosis per se, but rather is a marker of clinically more severe pneumonia.
Our study adds to the growing number of observational studies that support a protective role for the preadmission use of statins in hospitalised pneumonia patients. Our findings for the prognostic effect measures of statin therapy are comparable to previously published findings.18 31–34
We also found that there was a substantial beneficial effect of β-blocker therapy in pneumonia patients with AF. β-blockers may improve the prognosis for pneumonia patients with AF by protecting against uncontrolled tachycardia and reducing myocardial oxygen consumption. In patients with paroxysmal AF, β-blockers could protect against AF relapse provoked by increased sympathetic tonus. β-blockers can also dampen the metabolic changes towards catabolism that occur during critical illness.35 Pre-admission β-blocker use was also related to improved prognosis in patients without AF indicating that the beneficial effects of β-blockers are not exclusive to patients with AF. This result conflicts, however, with the results of a previous study of 1007 patients hospitalised with pneumonia (including 188 β-blocker users) that reported an increased 30-day mortality in β-blocker users, compared with non-users (aOR=1.97 (95% CI 1.03 to 3.78)).34 Explanations for this discrepancy are unclear but may include more extensive control for confounding factors in the present study.
The observed excess mortality related to amiodarone use should be interpreted within the context that only 7.7% of patients with AF are users of this drug. In this study cohort, prescription of amiodarone may have been reserved for special circumstances, such as for patients with impaired left ventricular function who poorly tolerate drugs with negative inotrope properties. Even though we adjusted for diagnosed heart failure, the mortality difference between amiodarone users and non-users could be related to differences in cardiac performance because the data set did not include quantitative measures of heart failure severity. A well-described side effect of amiodarone treatment is the development of pulmonary fibrosis,36 which may have been a factor that contributed to the excess mortality observed among amiodarone users.
The finding of increased mortality in patients with AF using digoxin was alarming. Since digoxin was the single most used AF drug in our cohort, it seems unlikely that digoxin users represent a selected sample of patients with particularly severe AF.
The effect of cardiac-acting calcium-channel blockers on the prognosis of acute disease has not been investigated. We found that there was a slightly increased 30-day mortality in patients with AF using calcium-channel blockers, but there were no mortality differences 1 year after admission.
We found that compared with non-users, patients with AF treated with vitamin K antagonists had a markedly reduced mortality rate and risk of arterial thromboembolism. The role of preadmission treatment with vitamin K antagonists has not been investigated previously in patients with pneumonia. However, severe and uncomplicated infections induce increased activity of the coagulation system and the degree of coagulation abnormality at hospital admission is correlated with the outcome of community-acquired pneumonia.37 38 Thus, we speculate that the beneficial effect of vitamin K antagonist treatment may be related to protection from hypercoagulation induced by systemic inflammation. This may also explain the finding of favourable prognosis with preadmission use of vitamin K antagonists in patients without AF.
We did not find that mortality of aspirin users was reduced in patients with pneumonia with AF. Furthermore, aspirin use by patients with AF was associated with only a modest risk reduction from arterial thromboembolisms. Our findings agree with the findings of clinical studies that indicate that aspirin use in patients with AF has little benefit for prevention of stroke.39 Of note, we found preadmission use of aspirin to be associated with reduced mortality in patients without AF, which is in line with a previous study.34
Finally, the results of our study indicated that there was an increased risk of thromboembolism in patients with AF not taking anticoagulants, even after controlling for the established risk factors for stroke (ie, the risk factors included in the CHA2DS2-VASc-score). Our study results indicate that severe infection should be considered to be a novel risk factor for stroke in patients with AF.
Conclusion
In conclusion, pre-existing AF is a frequent condition in patients admitted to the hospital with pneumonia, and marks increased risk of death and arterial thromboembolism. This effect is attributable to the more advanced ages and higher burdens of coexisting disease that are present in patients with AF. Our results also showed that the prognosis for patients with AF with pneumonia was substantially influenced by preadmission drug treatment, which suggests that treatment protocols could be improved.
Supplementary Material
Acknowledgments
The authors are grateful to Rikke Beck Nielsen, MSc for performing data extraction and management.
Footnotes
Contributors: JG performed the statistical analyses, interpreted the data, and wrote the first draft of the manuscript. CFC and RWT helped in interpreting the data. LHR and RWT provided the funding. RWT supervised the study. All authors were involved in conceptualising and designing the study. All authors participated in critically reviewing and writing the manuscript. All authors have read and approved the final draft of the manuscript.
Funding: The study was supported by research grants from the Danish Council for Strategic Research (grant number 09-066965), the Clinical Epidemiological Research Foundation, the Karen Elise Jensen Foundation, and the Heinrich Kopp Foundation.
Competing interests: None.
Ethics approval: Danish Data Protection Agency.
Provenance and peer review: Not commissioned; externally peer reviewed.
Data sharing statement: No additional data are available.
References
- 1.Go AS, Hylek EM, Phillips KA et al. Prevalence of diagnosed atrial fibrillation in adults. JAMA 2001;285:2370–5. [DOI] [PubMed] [Google Scholar]
- 2.Krijthe BP, Kunst A, Benjamin EJ et al. Projections on the number of individuals with atrial fibrillation in the European Union, from 2000 to 2060. Eur Heart J 2013;34:2746–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Benjamin EJ, Wolf PA, D'Agostino RB et al. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946–52. [DOI] [PubMed] [Google Scholar]
- 4.Friberg L, Hammar N, Pettersson H et al. Increased mortality in paroxysmal atrial fibrillation: report from the Stockholm Cohort-Study of Atrial Fibrillation (SCAF). Eur Heart J 2007;28:2346–53. [DOI] [PubMed] [Google Scholar]
- 5.Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991;22:983–8. [DOI] [PubMed] [Google Scholar]
- 6.World Health Organization. The global burden of disease: 2004 update. Geneva, 2008. [Google Scholar]
- 7.Søgaard M, Nielsen RB, Schønheyder HC et al. Nationwide trends in pneumonia hospitalization rates and mortality, Denmark 1997–2011. Respir Med 2014;108:1214–22. [DOI] [PubMed] [Google Scholar]
- 8.Fry AM, Shay DK, Holman RC et al. Trends in hospitalizations for pneumonia among persons aged 65 years or older in the United States, 1988–2002. JAMA 2005;294:2712–19. [DOI] [PubMed] [Google Scholar]
- 9.Mandal P, Chalmers JD, Choudhury G et al. Vascular complications are associated with poor outcome in community-acquired pneumonia. QJM 2011;104:489–5. [DOI] [PubMed] [Google Scholar]
- 10.Corrales-Medina VF, Musher DM, Wells GA et al. Cardiac complications in patients with community-acquired pneumonia: incidence, timing, risk factors, and association with short-term mortality. Circulation 2012;125:773–81. [DOI] [PubMed] [Google Scholar]
- 11.Smeeth L, Thomas SL, Hall AJ et al. Risk of Myocardial Infarction and Stroke after Acute Infection or Vaccination. N Engl J Med 2004;351:2611–18. [DOI] [PubMed] [Google Scholar]
- 12.Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health 2011;39(7 Suppl):30–3. [DOI] [PubMed] [Google Scholar]
- 13.Charlson ME, Pompei P, Ales KL et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83. [DOI] [PubMed] [Google Scholar]
- 14.De Groot V, Beckerman H, Lankhorst GJ et al. How to measure comorbidity: a critical review of available methods. J Clin Epidemiol 2003;56:221–9. [DOI] [PubMed] [Google Scholar]
- 15.Olesen JB, Lip GYH, Hansen ML et al. Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study. BMJ 2011;342:d124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ehrenstein V, Antonsen S, Pedersen L. Existing data sources for clinical epidemiology: Aarhus University Prescription Database. Clin Epidemiol 2010;2:273–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hansen ML, Gadsbøll N, Gislason GH et al. Atrial fibrillation pharmacotherapy after hospital discharge between 1995 and 2004: a shift towards beta-blockers. Europace 2008;10:395–402. [DOI] [PubMed] [Google Scholar]
- 18.Nielsen AG, Nielsen RB, Riis AH et al. The impact of statin use on pneumonia risk and outcome: a combined population-based case-control and cohort study. Crit Care 2012;16:R122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Andersen JS, Olivarius NDF, Krasnik A. The Danish National Health Service Register. Scand J Public Health 2011;39:34–7. [DOI] [PubMed] [Google Scholar]
- 20.Pedersen CB, Gøtzsche H, Møller JO et al. The Danish Civil Registration System. A cohort of eight million persons. Dan Med Bull 2006;53:441–9. [PubMed] [Google Scholar]
- 21.Andersen PK, Geskus RB, de Witte T et al. Competing risks in epidemiology: possibilities and pitfalls. Int J Epidemiol 2012;41:861–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Thomsen RW, Riis A, Nørgaard M et al. Rising incidence and persistently high mortality of hospitalized pneumonia: a 10-year population-based study in Denmark. J Intern Med 2006;259:410–17. [DOI] [PubMed] [Google Scholar]
- 23.Rix TA, Riahi S, Overvad K et al. Validity of the diagnoses atrial fibrillation and atrial flutter in a Danish patient registry. Scand Cardiovasc J 2012;46:149–53. [DOI] [PubMed] [Google Scholar]
- 24.Frost L, Vestergaard P. Alcohol and risk of atrial fibrillation or flutter: a cohort study. Arch Intern Med 2004;164:1993–8. [DOI] [PubMed] [Google Scholar]
- 25.Frost L, Andersen LV, Vestergaard P et al. Trend in mortality after stroke with atrial fibrillation. Am J Med 2007;120:47–53. [DOI] [PubMed] [Google Scholar]
- 26.Johnsen SP, Overvad K, Sørensen HT et al. Predictive value of stroke and transient ischemic attack discharge diagnoses in The Danish National Registry of Patients. J Clin Epidemiol 2002;55:602–7. [DOI] [PubMed] [Google Scholar]
- 27.Krarup L-H, Boysen G, Janjua H et al. Validity of stroke diagnoses in a National Register of Patients. Neuroepidemiology 2007;28:150–4. [DOI] [PubMed] [Google Scholar]
- 28.Thomsen RW, Johnsen SP, Olesen AV et al. Socioeconomic gradient in use of statins among Danish patients: population-based cross-sectional study. Br J Clin Pharmacol 2005;60:534–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Thomsen RW, Nielsen RB, Nørgaard M et al. Lifestyle profile among statin users. Epidemiology 2013;24:619–20. [DOI] [PubMed] [Google Scholar]
- 30.Thygesen SK, Christiansen CF, Christensen S et al. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol 2011;11:83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Thomsen RW, Riis A, Kornum JB et al. Preadmission use of statins and outcomes after hospitalization with pneumonia: population-based cohort study of 29,900 patients. Arch Intern Med 2008;168:2081–7. [DOI] [PubMed] [Google Scholar]
- 32.Douglas I, Evans S, Smeeth L. Effect of statin treatment on short term mortality after pneumonia episode: cohort study. BMJ 2011;342:d1642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Mortensen EM, Nakashima B, Cornell J et al. Population-based study of statins, angiotensin II receptor blockers, and angiotensin-converting enzyme inhibitors on pneumonia-related outcomes. Clin Infect Dis 2012;55:1466–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chalmers JD, Singanayagam A, Murray MP et al. Prior statin use is associated with improved outcomes in community-acquired pneumonia. Am J Med 2008;121:1002–1007.e1. [DOI] [PubMed] [Google Scholar]
- 35.Norbury W, Jeschke M, Herndon D. Metabolism modulators in sepsis: propranolol. Crit Care Med 2007;35:s616–620. [DOI] [PubMed] [Google Scholar]
- 36.Ernawati DK, Stafford L, Hughes JD. Amiodarone-induced pulmonary toxicity. Br J Clin Pharmacol 2008;66:82–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Milbrandt EB, Reade MC, Lee M et al. Prevalence and significance of coagulation abnormalities in community-acquired pneumonia. Mol Med 2009;15:438–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Amaral A, Opal SM, Vincent J-L. Coagulation in sepsis. Intensive Care Med 2004;30:1032–40. [DOI] [PubMed] [Google Scholar]
- 39.Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007;146:857–67. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.