Risk for Myocardial Infarction and Stroke After Community-Acquired BacteremiaCLINICAL PERSPECTIVE
A 20-Year Population-Based Cohort Study
Background—Infections may trigger acute cardiovascular events, but the risk after community-acquired bacteremia is unknown. We assessed the risk for acute myocardial infarction and ischemic stroke within 1 year of community-acquired bacteremia.
Methods and Results—This population-based cohort study was conducted in Northern Denmark. We included 4389 hospitalized medical patients with positive blood cultures obtained on the day of admission. Patients hospitalized with bacteremia were matched with up to 10 general population controls and up to 5 acutely admitted nonbacteremic controls, matched on age, sex, and calendar time. All incident events of myocardial infarction and stroke during the following 365 days were ascertained from population-based healthcare databases. Multivariable regression analyses were used to assess relative risks with 95% confidence intervals (CIs) for myocardial infarction and stroke among bacteremia patients and their controls. The risk for myocardial infarction or stroke was greatly increased within 30 days of community-acquired bacteremia: 3.6% versus 0.2% among population controls (adjusted relative risk, 20.86; 95% CI, 15.38–28.29) and 1.7% among hospitalized controls (adjusted relative risk, 2.18; 95% CI, 1.80–2.65). The risks for myocardial infarction or stroke remained modestly increased from 31 to 180 days after bacteremia in comparison with population controls (adjusted hazard ratio, 1.64; 95% CI, 1.18–2.27), but not versus hospitalized controls (adjusted hazard ratio, 0.95; 95% CI, 0.69–1.32). No differences in cardiovascular risk were seen after >6 months. Increased 30-day risks were consistently found for a variety of etiologic agents and infectious foci.
Conclusions—Community-acquired bacteremia is associated with increased short-term risk of myocardial infarction and stroke.
Each year, >1 000 000 Americans experience acute myocardial infarction (AMI) or acute ischemic stroke (AIS).1 Any role of acute infection in triggering acute cardiovascular events is of major clinical and public health interest.2–5 Infections may increase the risk of AMI and AIS6–14 by inducing demand ischemia, decreasing myocardial contractility, or causing endothelial dysfunction, coagulation disturbance, or direct platelet activation.2–4,15–18 The magnitude and duration of the increased cardiovascular risk is debatable. Cohort studies have reported short-term risks of AMI and stroke varying from 0.2% to >10% after patients have been hospitalized with pneumonia, sepsis, endocarditis, or meningitis.7–11,19–21
Editorial see p 1375
Clinical Perspective on p 1396
Case-only studies suggest a 10- to 50-fold increased risk for AMI or stroke shortly after patients have been hospitalized with infection,6,9,12 and a 2- to 5-fold increased risk shortly after infection diagnosed by general practitioners.13,14 Only 1 cohort study of 206 patients with pneumonia included a comparison group,9 and we are aware of only 3 studies that included microbiological test results.6,9,10 The lack of laboratory confirmation of infection may have falsely inflated the effect estimates if cardiovascular events were initially misdiagnosed as infections. Community-acquired bacteremia (CAB) is a well-defined clinical entity that embraces a wide range of mechanisms whereby infection may trigger cardiovascular events. We conducted a 20-year population-based cohort study in Denmark to assess the short- and longer-term risks of AMI and AIS among medical patients with CAB in comparison with the background population and with other acutely admitted patients.
The study was conducted in Northern Denmark from 1992 to 2011. This area had a stable urban/rural catchment population of ≈500 000 inhabitants who received universal tax-financed primary and secondary care, free at the point of delivery. Throughout the study period, Aalborg University Hospital was the only referral hospital, and all regional hospitals relied on its Department of Clinical Microbiology for blood culture analyses.
We used high-quality population-based databases with prospectively collected data: the Danish Civil Registration System, The North Denmark Bacteremia Research Database,22 the Aarhus University Prescription Database,23 the regional hospital discharge registry,24 and the clinical laboratory information system research database.25 The Civil Registration System contains general personal data for all citizens, updated daily. Unique Civil Registration System numbers, assigned to every Danish resident and used for all healthcare contacts, facilitated the linkage between databases. The study was approved by the Danish Data Protection Agency (2011-41-5864).
Bacteremia and Control Cohorts
Eligibility criteria for study inclusion were age ≥15 years, no hospitalization within the previous 30 days, no record of previous bacteremia (since 1981), and study area residence for ≥1 year.
We assembled 3 study cohorts. First, we used the Bacteremia Database to identify all adult patients who had a first-time positive blood culture taken on the day of admission to a medical ward during 1992 to 2010. We defined CAB as the presence of viable bacteria or fungi in the bloodstream, determined by blood cultures performed on the day of admission, among clinically ill patients who had not been admitted to the hospital within the previous 30 days. The Bacteremia Database has registered all bacteremia cases in the study area since 1981, with prospective data collection since 1992, and is described in detail elsewhere.22 It provided information on the date of blood culture sampling, the number of positive culture bottles, the focus of infection, and the etiologic agent(s). To assess the level of inflammation, we retrieved information on index-date white blood cell counts and C-reactive protein levels for admissions since 1998 from the laboratory database. We categorized white blood cell counts as decreased (<3.5×109/L), normal (3.5–10×109/L), or increased (>10×109/L), and the C-reactive protein levels as normal (<10 mg/L), increased (10–100 mg/L), or highly increased (>100 mg/L).
Second, to assess the effect of hospitalization with CAB on the risk for AMI/AIS, we used the Civil Registration System to assemble a matched population control cohort of individuals at risk for their first CAB as of the admission date of their matched bacteremia patient (the index date). Up to 10 population controls were randomly selected for each bacteremic patient, and matched for year of birth, sex, and calendar time.
Third, because acute medical hospitalization itself may increase the risk for subsequent AMI and AIS, we assembled an additional matched hospitalized control cohort of up to 5 randomly selected acutely admitted medical patients for each CAB patient. Hospitalized controls were patients who were acutely admitted for reasons other than a positive blood culture or a primary diagnosis of cardiovascular disease or rehabilitation (International Classification of Diseases [ICD], ICD-8: 390–458, ICD-10: DI00-99, DZ50). They were matched to CAB patients on sex, year of birth, and calendar year of hospital admission.
Data on Cardiovascular Outcomes and Comorbidity
The hospital discharge registry provided data on all events of hospitalization with AMI and AIS and on preexisting comorbid conditions. It has recorded complete diagnosis codes from all inpatient hospitalizations and hospital outpatient clinic contacts in Denmark since 1977 and 1995, respectively. Hospital discharge registry uses 2 versions of the World Health Organization’s International Classification of Diseases (ICD-8 until the end of 1993 and ICD-10 thereafter).
One primary discharge diagnosis and up to 20 secondary diagnoses are assigned by physicians at the treating hospital department. The primary diagnosis refers to the condition that prompted patient admission and the main condition responsible for the completed diagnosis and treatment course. The secondary diagnoses refer to conditions that affect the diagnosis and treatment course.
To assess if AMI/AIS did precede CAB in some patients admitted with CAB and a primary diagnosis of AMI/AIS, we examined available electronic hospital files for 21 of the 60 bacteremic patients who had a primary discharge diagnosis of AMI or AIS. In all 21 patients, AMI/AIS developed >24 hours after admission for suspected infection (n=7) or occurred on the day of admission in patients with ongoing infection (evidenced by microbiological tests performed by the patient’s general practitioner, recent antibiotic prescriptions, and medical history taking [n=14]). Therefore, we used primary and secondary discharge diagnosis codes to identify episodes of AMI and AIS (Table I in the online-only Data Supplement).
For all 3 cohorts, we obtained data on preexisting diseases recorded before the index date, including the 19 disease categories in the Charlson Comorbidity Index,26 and other conditions described as risk factors for AMI/AIS in the literature.1 Moreover, we used the Prescription Database to ascertain the use of medications before the index date that may affect AMI/AIS risk. The database has Anatomic Therapeutic Chemical classification codes on all reimbursed prescriptions since 1991 (Table I in the online-only Data Supplement).23
We followed all study subjects from the index date until first hospitalization for AMI/AIS, death, emigration out of Denmark, or January 1, 2012, whichever occurred first. Because we expected the risk of AMI and AIS to be highest shortly after the onset of infection, we split follow-up into 3 time periods: 0 to 30, 31 to 180, and 181 to 365 days after the index date. For each time period, we computed absolute risks for our predefined outcome AMI/AIS, and for AMI and AIS separately. We also computed relative risks (RRs) of AMI/AIS, AMI, and AIS with 95% confidence intervals (CIs) for CAB patients versus their 2 matched control cohorts. Because the hospital discharge registry does not include the exact date of AMI/AIS, we used conditional Poisson regression with robust variance estimation to compute RRs for any hospital admission with AMI/AIS within 0 to 30 days (index admission included). We used Cox proportional hazards models and stsplit in Stata to compare hazard rates of hospital admission with AMI/AIS during 31 to 180 and 181 to 365 days after the index date among CAB patients still alive and at risk of first incident AMI/AIS on day 31 and 181, respectively, and their at-risk matched controls. Because death was a competing risk for AMI/AIS, we modeled the cause-specific hazards of AMI/AIS. To account for the matched design, Cox models were stratified on matched sets.
Because endocarditis is a common cause of septic emboli to the brain, and sometimes the coronary vessels, we performed a supplementary analysis in which we excluded all CAB patients with endocarditis. Because AMI/AIS might have preceded CAB in some patients despite our sample validation results, we conducted supplementary sensitivity analyses in which we excluded matched groups if the CAB patient had a primary discharge code of AMI/AIS. In regression models we controlled for a priori potential confounders: age, sex, calendar time, marital status, previous AMI, previous cerebrovascular disease, diabetes mellitus, chronic pulmonary disease, other cardiovascular diseases, other comorbid conditions, and the use of medications for cardiovascular disease (see Table I in the online-only Data Supplement for further detail).
We further examined AMI/AIS RRs for predefined CAB subgroups (age group, sex, study period, etiologic agent, focus of infection, and levels of inflammatory markers) and in strata of previous cardiovascular disease. Analyses within strata of previous cardiovascular disease required that we ignored the matching, so we used modified Poisson regression. For 2005 to 2010, when intensive care unit (ICU) data were available, we did a supplementary subgroup analysis restricted to CAB patients who had an ICU stay, as a proxy for severe sepsis. Stratified and subgroup analyses were controlled for age, sex, calendar time, and any comorbidity (except when stratified on comorbidity). Next, we included interaction terms in regression models and used the Wald test to assess whether the effect of CAB on AMI/AIS differed by subgroup.
Finally, because of possible residual confounding when comparing hospitalized CAB patients with healthy population controls, we performed a sensitivity analysis to estimate how much a potential strong unmeasured confounder might have influenced the observed association (see Online-only Data Supplement).
For Cox models, the proportional hazards assumption was checked with log-minus-log plots. Stata 11.2 for Windows (Stata Corp, College Station, TX) was used for all data analyses.
Study Subject Characteristics
The study included 4389 CAB patients, 43 831 matched population controls, and 21 893 matched hospitalized controls. Median study participant age was 73 years (interquartile range, 61–82 years). CAB patients had a substantially higher burden of preexisting disease and filled drug prescriptions than background population controls, and a burden similar to other hospitalized patients (Table 1). Among CAB patients, the 30-day mortality was 15.7%, nearly twice that of hospitalized controls (7.9%), and after 1 year it was 29.7% (versus 5.8% for population controls and 23.9% for hospitalized controls; Table II in the online-only Data Supplement).
Risk for AMI/AIS
Within the first 30 days of follow-up, 3.6% of CAB patients had an AMI/AIS (versus 0.2% for population controls and 1.7% for hospitalized controls), 1.7% had an AMI (versus 0.1% for population controls and 0.8% for hospitalized controls), and 2.1% had an AIS (versus 0.1% for population controls and 0.9% for hospitalized control) (Table 2). Approximately 80% of these early events among CAB patients occurred during the index admission (57/73 AMI events and 72/91 AIS events).
The adjusted 0- to 30-day RR of AMI/AIS in CAB patients versus population controls was 20.86 (95% CI, 15.38–28.29) and versus population controls 2.18 (95% CI, 1.80–2.65). Similar short-term risk increases were observed when AMI and AIS were analyzed separately (see Table 2). An increased hazard rate ratio of AMI/AIS remained for CAB patients in comparison with population controls during 31 to 180 days (AMI/AIS, adjusted hazard ratio [HR], 1.64; 95% CI, 1.18–2.27), with an adjusted HR of 1.90 (95% CI, 1.26–2.89) for AIS and 1.42 (95% CI, 0.86–2.34) for AMI. In comparison with hospitalized controls, HRs were close to one within 31 to 180 days. One hundred eighty-one to 365 days after admission, the adjusted AMI/AIS HRs were close to one both in comparison with population and hospital controls (Table 2).
Risk for AMI/AIS in Subgroups
In general, the finding of an increased risk of AMI and AIS was consistent across subgroups. As expected, CAB patients and controls without previous cardiovascular disease had lower absolute risks of AMI and AIS within 0 to 30 days (Tables 3 and 4). However, the 30-day AMI and AIS RR increase tended to be higher in CAB patients without previous cardiovascular disease, related to a low baseline risk among their controls. In comparison, the absolute risk increase of AMI/AIS tended to be similar in those with and without previous cardiovascular disease. Table III in the online-only Data Supplement details the 0- to 30-day AMI/AIS risk by subgroups and strata of prevalent cardiovascular disease, and Tables IV to VI in the online-only Data Supplement show data on longer-term AMI and AIS risks.
Patients with gram-positive infections were younger than patients with gram-negative CAB (median age, 71.0 versus 75.9 years), and they had a lower prevalence of previous AMI (6.5% versus 9.1%) and stroke (9.8% versus 15.7%; Table II in the online-only Data Supplement). Gram-positive CAB yielded a 1.2% risk for AMI and a 2.5% risk for AIS. For gram-negative CAB, both the AMI and the AIS risk were 1.9%. Of note, Streptococcus pneumoniae was a major contributor to the gram-positive CAB group, as was Escherichia coli for the gram-negative group. Therefore, infection with these pathogens mirrored the findings from the respective Gram stain groups. Patients with Staphylococcus aureus infection had an elevated 30-day risk of AMI and, in particular, AIS (Tables 3 and 4), especially if they had endocarditis (4.5% with AMI and 9.1% with AIS), yet also without endocarditis (1.8% with AMI and 3.2% with AIS). The corresponding adjusted HRs for AIS were high with S aureus infection, and remained increased during 31 to 180 days of follow-up in comparison with hospitalized controls (HR, 7.12; 95% CI, 1.63–31.03) and population controls (HR, 7.39; 95% CI, 2.21–24.69) (Table VI in the online-only Data Supplement).
Patients who had elevated white blood cell counts, increased C-reactive protein levels, or all blood culture bottles positive had high 30-day risk increases for AMI and AIS (Tables 3 and 4; also see online-only Data Supplement).
Supplementary and Sensitivity Analyses
CAB patients who needed ICU treatment had a 6.0% 30-day risk of AMI/AIS (adjusted RR versus hospitalized controls, 4.43; 95% CI, 1.82–10.74; and versus population controls, 84.53; 95% CI, 10.41–686.60) and those without an ICU stay had a 3.0% risk (adjusted RR versus hospitalized controls, 2.24; 95% CI, 1.59–3.15; and versus population controls, 19.39; 95% CI, 11.81–31.81).
Exclusion of 160 patients with endocarditis and their matched controls gave 30-day absolute risk estimates for AMI/AIS of 3.5% for CAB patients versus 1.7% for hospitalized controls (adjusted RR, 2.08; 95% CI, 1.71–2.55) and 0.2% for population controls (adjusted RR, 20.08; 95% CI, 14.71–27.43).
Exclusion of 61 matched groups in which the CAB patients had a primary discharge diagnosis of AMI/AIS lowered the adjusted 0- to 30-day RR of AMI/AIS to 1.36 (95% CI, 1.08–1.72) in comparison with hospitalized controls, and to 12.98 (95% CI, 9.33–18.06) in comparison with population controls.
We estimated that if a strong unmeasured confounder had a prevalence of 5% among population controls and 50% in CAB, and independently increased the 30-day AMI/AIS risk by a factor of 20, the true risk for AMI/AIS following CAB would still be increased 5.73- fold (5.09-fold for AMI and 6.34-fold for AIS).
This study provides evidence that acute admission with CAB is associated with a transient >2-fold increased risk of AMI and AIS in comparison with other acutely hospitalized patients and a 20-fold increased risk in comparison with the background population. Importantly, a >60% increased risk for AMI/AIS persists for 1 to 6 months after CAB hospitalization in comparison with population controls. Patients with S aureus bacteremia are at particularly increased risk for both early and late AIS. However, all types of CAB increase the short-term risk for cardiovascular events.
To our knowledge, this is the largest cohort study to evaluate the risk of AMI and AIS after microbiologically verified infection. Our 30-day absolute risk estimates for AMI (1.7%) and AIS (2.1%) are consistent with findings from previous smaller cohort studies. Levine et al7 pooled data from 3 clinical trials in patients with severe sepsis and septic shock and found that 0.5% to 1.5% had an AMI and 1.0% to 2.7% had an ischemic stroke within 28 days. For patients with respiratory tract infection, we observed absolute risks consistent with previous pneumonia studies.5,8–11 In the largest cohort study conducted to date, Perry et al11 evaluated the risk of cardiovascular events among 50 119 patients admitted for pneumonia. Within 30 days following admission, 1.2% of patients experienced a first-time AMI, and 0.2% experienced a first-time stroke. For endocarditis and meningitis, AMI and AIS risk estimates in our study were particularly high but were less than the ≥10% reported in previous studies that included hemorrhagic stroke outcomes.20,21 Most previous cohort studies on the association between infection and cardiovascular events lacked a comparison group or did not have adequate long-term follow-up. One small cohort study compared 208 patients hospitalized for pneumonia with 395 hospitalized controls and found an 8-fold risk increase for acute coronary syndrome within 15 days,9 higher than our 2.2-fold increased RR within 30 days.
Case-only study designs have been used to examine short-term risk of acute coronary syndrome and stroke after hospitalization for infection.6,9,12 Patients were included in these studies if they had both a transient exposure (infection) and an acute outcome of interest (AMI or stroke) during an observation period. The outcome risk was compared for different time periods with each patient serving as his/her own control. In 32 patients, the risk of acute coronary syndrome increased 50-fold within a 15-day period after hospitalization for pneumonia.9 Likewise, in 42 patients the risk of AMI was 35-fold higher within 2 days after the recognition of S aureus bacteremia,6 and in 669 patients the risk of stroke was 8-fold higher within 2 weeks after infection.12 These estimates are comparable to our 20-fold increased RR versus population controls.
Few data are available on the long-term risks of AMI/AIS following infection. A case-only study by Smeeth et al14 found that out-of-hospital respiratory tract or urinary tract infection was associated with a 1.2-fold increased risk of AMI or stroke within 29 to 91 days after infection. In comparison, we found 1.4- to 1.9-fold increased risks for AMI and stroke, respectively, during 31 to 180 days when CAB patients were compared with population controls. Variations in study populations and the definition and severity of infection, AMI, and stroke make it difficult to directly compare risk estimates from various studies. Nonetheless, infection is associated with 30-day AMI and stroke risk in the present study and in previous studies.
Several mechanisms explain why bacterial infections can trigger cardiovascular events. In theory, any severe infection that causes hypotension may disturb the balance between myocardial or cerebral metabolic supply and demand and trigger an ischemic event.15 Toxins from gram-negative and gram-positive bacteria can impair myocardial function16 and damage the endothelium.17 Furthermore, bacteria can provoke inflammation-induced activation of the coagulatory system and directly activate platelets,18 which may lead to thrombosis. Moreover, septic embolism may be important. The markedly increased risk only within the first 30 days after bacteremia supports a pathogenic link between the acute inflammation associated with bacterial infection and vascular events. In a previous study, Grau et al27 found that an increased neutrophil count heralds a short period at increased risk for recurrent ischemic events. Moreover, previous studies have shown that more severe in-hospital infections are associated with higher risk increases for AMI and AIS than less severe infections treated in primary care.2,4,13,14 We found that high levels of inflammatory markers such as C-reactive protein levels and white blood cell counts, and high bacteremia severity as reflected by multiple culture bottles being positive or the need for ICU stay predicted high risk increases for AMI/AIS. However, because of the limited precision of estimates, we cannot conclude that more severe inflammation or more severe disease confers a greater risk for AMI/AIS than lower levels of inflammation and severity.
Finally, our findings suggest that patients with S aureus infection may have a particularly high propensity for thrombosis.28
The strengths of the present study included large sample size, population-based design, and complete follow-up of all study subjects. We had access to high-quality microbiological data and captured all diagnosed cases of hospitalized CAB over a 20-year period in the study population.22 Our data on AMI/AIS had high validity.29–31 Because all data were prospectively collected in independent databases, recall and investigator bias was negligible.
Our study had some limitations. Heart failure and stroke are known risk factors for infection, with risk of reverse causation bias. To increase the likelihood that infection preceded AMI/AIS, we limited our study to patients who had a positive blood culture on the day of admission. Furthermore, our medical record review did not reveal reverse causation bias in any patient, and CAB remained associated with AMI/AIS in various sensitivity analyses.
CAB patients were already receiving medical attention and therefore potentially more likely to be diagnosed with AMI/AIS. However, AMI and AIS are serious acute conditions that usually lead to hospital contact. Out-of-hospital deaths from AMI/AIS may have decreased risk estimates for population controls. On the other hand, severe infection may lead to death before AMI/AIS is diagnosed in hospitalized patients. A third possible limitation is that troponin spill during severe infection may have falsely inflated AMI risk estimates.32 However, the AMI risk was similar before and after the introduction of troponin assays in clinical practice. Because death was a competing risk in this study, the usual 1-to-1 correspondence between risk and rate was lost.33 Therefore, hazard ratios of AMI and AIS during 31 to 180 days and 181 to 365 days should be interpreted with some caution. Although we were able to adjust for a wide range of confounders, residual and unmeasured confounding may have occurred. For example, we lacked information on smoking status and relied on proxy variables for smoking (eg, chronic pulmonary disease). However, to nullify our findings, an unmeasured confounder would have to be extremely strong. The fact that there was no adjusted risk increase for AMI/AIS in CAB patients versus both control cohorts after >180 days argues against substantial unmeasured confounding factors.
In conclusion, patients admitted with CAB had a transient increased risk of AMI and AIS. The risk of AMI/AIS was greatest during the first 30 days after the infection, although a modestly elevated risk, in particular, for AIS was observed for 6 months postinfection. There is a need for a better understanding of the mechanisms including metabolic supply/demand mismatch and embolic events that may increase the risk for cardiovascular events following severe infection. Improving our understanding of these mechanisms through future experimental and observational studies may lead to more targeted prevention and treatment strategies.
We thank Lena Mortensen, Department of Clinical Microbiology, Aalborg University Hospital, for meticulous assistance with The North Denmark Bacteremia Research Database. We thank Rikke Mortensen, MSc, and Jacob Bonde Jacobsen, MSc, Department of Clinical Epidemiology, Aarhus University Hospital, for help with data preparation and statistical guidance.
Sources of Funding
This study was supported by The Karen Elise Jensen, Heinrich Kopp, Svend Andersen, and Helga and Peter Korning Foundations, and the North Denmark Health Sciences Research Foundation. The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.113.006699/-/DC1.
- Received June 11, 2013.
- Accepted January 9, 2014.
- © 2014 American Heart Association, Inc.
- Roger VL,
- Go AS,
- Lloyd-Jones DM,
- Benjamin EJ,
- Berry JD,
- Borden WB,
- Bravata DM,
- Dai S,
- Ford ES,
- Fox CS,
- Fullerton HJ,
- Gillespie C,
- Hailpern SM,
- Heit J a.,
- Howard VJ,
- Kissela BM,
- Kittner SJ,
- Lackland DT,
- Lichtman JH,
- Lisabeth LD,
- Makuc DM,
- Marcus GM,
- Marelli A,
- Matchar DB,
- Moy CS,
- Mozaffarian D,
- Mussolino ME,
- Nichol G,
- Paynter NP,
- Soliman EZ,
- Sorlie PD,
- Sotoodehnia N,
- Turan TN,
- Virani SS,
- Wong ND,
- Woo D,
- Turner MB
- Corrales-Medina VF,
- Musher DM,
- Wells GA,
- Chirinos JA,
- Chen L,
- Fine MJ
- Musher DM,
- Rueda AM,
- Kaka AS,
- Mapara SM
- Elkind MS,
- Carty CL,
- O’Meara ES,
- Lumley T,
- Lefkowitz D,
- Kronmal RA,
- Longstreth WT Jr..
- Owens P,
- O’Brien E
- Grandel U,
- Bennemann U,
- Buerke M,
- Hattar K,
- Seeger W,
- Grimminger F,
- Sibelius U
- Lynge E,
- Sandegaard JL,
- Rebolj M
- Grau AJ,
- Boddy AW,
- Dukovic DA,
- Buggle F,
- Lichy C,
- Brandt T,
- Hacke W
- Andersen PK,
- Geskus RB,
- de Witte T,
- Putter H
Recent research suggests that patients who acquire an infection may have a transient increased risk of cardiovascular events. Acute myocardial infarction (AMI) and acute ischemic stroke (AIS) are common, and any role of acute infections in triggering them is of major clinical interest. We investigated short- and longer-term risks of AMI and AIS among >4000 patients with microbiologically verified community-acquired bacteremia. Within the first 30 days, 3.6% of bacteremia patients experienced an AMI or AIS. Their adjusted risk was >2-fold higher than in other acutely hospitalized patients, and 20-fold increased in comparison with the background population. A modestly increased risk in particular for AIS persisted up to 6 months after bacteremia. High risks of AMI and AIS were seen in patients with endocarditis, Staphylococcus aureus bacteremia, high levels of inflammatory markers, or intensive care unit stay. A high absolute risk for AMI and AIS was also found among those with previous cardiovascular disease, which underlines current vaccine recommendations. Further studies are needed to clarify the mechanisms, including metabolic supply/demand mismatch and embolic events that may increase the risk for cardiovascular events following severe infection. Such knowledge may improve prevention, early diagnosis, and targeted management of AMI and AIS in patients with ongoing bacteremia. As the burden of hospitalization with bacteremia continues to grow, clinicians should be aware of the possible link between bacteremia and thromboembolic events.