Background: Streptococci frequently cause infective endocarditis (IE), yet the prevalence of IE in patients with bloodstream infections (BSIs) caused by different streptococcal species is unknown. We aimed to investigate the prevalence of IE at species level in patients with streptococcal BSIs. Methods: We investigated all patients with streptococcal BSIs, from 2008 to 2017, in the Capital Region of Denmark. Data were crosslinked with Danish nationwide registries for identification of concomitant hospitalization with IE. In a multivariable logistic regression analysis, we investigated the risk of IE according to streptococcal species adjusted for age, sex, ≥3 positive blood culture bottles, native valve disease, prosthetic valve, previous IE, and cardiac device. Results: Among 6506 cases with streptococcal BSIs (mean age 68.1 years [SD 16.2], 52.8% men) the IE prevalence was 7.1% (95% CI, 6.5–7.8). The lowest IE prevalence was found with Streptococcus pneumoniae ( S pneumoniae ) 1.2% (0.8–1.6) and Spyogenes 1.9% (0.9–3.3). An intermediary IE prevalence was found with Sanginosus 4.8% (3.0–7.3), Ssalivarius 5.8% (2.9–10.1), and Sagalactiae 9.1% (6.6–12.1). The highest IE prevalence was found with Smitis/oralis 19.4% (15.6–23.5), Sgallolyticus (formerly Sbovis ) 30.2% (24.3–36.7), Ssanguinis 34.6% (26.6–43.3), Sgordonii 44.2% (34.0–54.8), and Smutans 47.9% (33.3–62.8). In multivariable analysis using S pneumoniae as reference, all species except S pyogenes were associated with significantly higher IE risk, with the highest risk found with S gallolyticus odds ratio (OR) 31.0 (18.8–51.1), S mitis/oralis OR 31.6 (19.8–50.5), S sanguinis OR 59.1 (32.6–107), S gordonii OR 80.8 (43.9–149), and S mutans OR 81.3 (37.6–176). Conclusions: The prevalence of IE in streptococcal BSIs is species dependent with S mutans, S gordonii, S sanguinis, S gallolyticus , and S mitis/oralis having the highest IE prevalence and the highest associated IE risk after adjusting for IE risk factors.

The role of Pseudomonas aeruginosa in the long-term prognosis of chronic obstructive pulmonary disease (COPD) is unknown. The purpose of this study was to determine whether P. aeruginosa is associated with increased risk of exacerbations or death in patients with COPD. This is a multiregional epidemiological study based on complete data on COPD outpatients between 1 January 2010 and 31 October 2017 and corresponding microbiology and national register data. Time-dependent Cox proportional hazards models and propensity matching was used to estimate hospitalization-demanding exacerbations and death after 2 years, separately and in combination. A total of 22 053 COPD outpatients were followed for a median of 1082 days (interquartile-range: 427-1862). P. aeruginosa was present in 905 (4.1%) patients. During 730 days of follow-up, P. aeruginosa strongly and independently predicted an increased risk of hospitalization for exacerbation or all-cause death (HR 2.8, 95%CI 2.2-3.6; p <0.0001) and all-cause death (HR 2.7, 95%CI 2.3-3.4; p <0.0001) in analyses adjusted for known and suspected confounders. The signal remained unchanged in unadjusted analyses as well as propensity-matched subgroup analyses. Among patients 'ever colonized' with P. aeruginosa, the incidence of hospital-demanding exacerbations doubled after the time of the first colonization. COPD patients in whom P. aeruginosa can be cultured from the airways had a markedly increased risk of exacerbations and death. It is still not clear whether this risk can be reduced by offering patients targeted antipseudomonal antibiotics. A randomized trial is currently recruiting patients to clarify this (ClinicalTrials.gov: NCT03262142).

Objective: To investigate the distribution of antibiotic prescriptions in primary health care among children aged 0–6 years and its association with socioeconomic factors. Design: A cross-sectional study describing antibiotic prescriptions and socioeconomic factors, using different population-based registers from Statistics Denmark. Setting: Antibiotic prescriptions in 2012 from primary health care in the Capital Region of Denmark. Subjects: The population of children aged 0–6 years (n = 139,398) in the Capital Region of Denmark. Main outcome measures: High use of antibiotics identified by number of antibiotic prescriptions (≥ 3 prescriptions per year) and defined daily doses (DDD). A multinomial logistic regression analysis estimating the association between high antibiotic use and parents’ education, employment status, income, child’s sex, and ethnic background. Results: Ten percent of children accounted for 25% of the total use DDD. There was a clear tendency that the risk for high antibiotic use increased as parental educational level decreased. The risk for high use was the highest among children of mothers and fathers with basic schooling ≤10 years (OR 1.60, 95% CI 1.29–1.98, and OR 1.60, 95% CI 1.34–1.91, respectively). Low income and unemployment were not associated with high antibiotic use. Conclusion: Socioeconomic factors can only partially explain differences in antibiotic use. Further research is needed to clarify the unequal distribution of antibiotic prescribing and the association between high antibiotic use and low educational level. This would provide valuable information in the planning of strategies to promote rational use of antibiotics among children.

Objectives We evaluated the effect of an antibiotic stewardship programme (ASP) on the use of antibiotics and resistance levels of Escherichia coli using a method that allowed direct comparison between an intervention hospital and a control hospital. Methods The study was conducted as a retrospective controlled interrupted time series (ITS) at two university teaching hospitals, intervention and control, with 736 and 552 beds, respectively. The study period was between January 2008 and September 2014. We used ITS analysis to determine significant changes in antibiotic use and resistance levels of E. coli. Results were directly compared with data from the control hospital utilizing a subtracted time series (STS). Results Direct comparison with the control hospital showed that the ASP was associated with a significant change in the level of use of cephalosporins [−151 DDDs/1000 bed-days (95% CI −177, −126)] and fluoroquinolones [−44.5 DDDs/1000 bed-days (95% CI −58.9, −30.1)]. Resistance of E. coli showed a significant change in slope for cefuroxime [−0.13 percentage points/month (95% CI −0.21, −0.057)] and ciprofloxacin [−0.15 percentage points/month (95% CI −0.26, −0.038)]. Conclusions The ASP significantly reduced the use of cephalosporins and fluoroquinolones, with concomitant decreasing levels of E. coli resistance to cefuroxime and ciprofloxacin. The same development was not observed at the control hospital.

Introduction: Penicillin allergy is suspected in 10% of hospital inpatients but can be disproved in 90% of cases. Direct oral provocation without preceding tests among low-risk patients has proven to be safe in studies of both children and adults and is gaining use across the world. The aims of this study were to investigate the rate of severe allergic reactions to direct oral drug provocation, without preceding tests, in penicillin allergy patients stratified to be at low risk, as well as to examine if these patients have barriers to penicillin allergy de-labeling and future use of penicillins. Methods: Adult patients referred to a university hospital allergy clinic with a suspected penicillin allergy were prospectively risk evaluated. Patients stratified to be at low risk were offered a direct oral provocation with a single-dose amoxicillin followed by 4 days of continued treatment. The same risk stratification criteria were applied to a larger retrospective cohort. Results: In the prospective study population, 202 patients had a direct oral drug provocation and 20 (10%) were positive. There were no cases of anaphylaxis or severe delayed hypersensitivity. Fifteen reactions were benign rashes with onset >1 day after initial dosing, and 13 of these were maculopapular rashes. The same low-risk criteria were applied retrospectively to patients in a drug provocation database, and 1,759 patients fulfilled the criteria; of these, 10% had positive provocations, and there were no cases of anaphylaxis or severe delayed hypersensitivity. De-labeled patients in the prospective study reported not to fear future penicillin intake, after prolonged provocation. Conclusion: The risk stratification criteria for identifying low-risk patients for the oral drug provocation test without prior skin testing were safe in terms of avoiding anaphylaxis or severe delayed hypersensitivity. Benign delayed skin reactions still occurred, and access to allergy advice and follow-up is necessary.

Objectives: To compare the therapeutic effect of pivmecillinam and other common oral antibiotics for community-acquired urinary tract infections (UTIs) caused by Extended Spectrum Beta-Lactamase (ESBL)- or non-ESBL-producing Escherichia coli.
Methods: Retrospective cohort study from 2010 to mid-2016 with data from the regional Laboratory Database and three national databases on antibiotic prescriptions, hospital admission, and mortality, respectively. Primary care patients (≥18 years) empirically treated for UTI caused by non-ESBL- or ESBL-producing E. coli (non-ESBL and ESBL E. coli) were included. Seven antibiotics, commonly used empirically for UTI, were investigated. Treatment failure measured as the redemption of a new antibiotic prescription or admission to hospital due to UTI. Cox proportional hazard ratios and adjusted risk differences along with 95% confidence intervals were calculated for 14 and 30 days, respectively. Results: Thirty-six thousand two hundred and ninety-three (95.7%) and 1624 (4.3%) cases were included in the non-ESBL and ESBL groups, respectively. Male sex, high age, ESBL production, and resistance to empirical therapy were found to independently increase the risk of treatment failure. Compared to pivmecillinam, ciprofloxacin had significantly lower treatment failure for non-ESBL E. coli, but significantly higher treatment failure in ESBL E. coli. There was no significant difference between nitrofurantoin and pivmecillinam. Conclusion: All antibiotics seem to have a higher risk of treatment failure for UTI caused by ESBL-producing E. coli as compared to non-ESBL-producing E. coli. At present, nitrofurantoin and pivmecillinam seem to be the most relevant orally available therapies for E. coli UTI. Local resistance data should guide which of the two that should be the contemporary first-line option.