Bodil Svennblad

Reliability of Bayesian Posterior Probabilities and Bootstrap Frequencies in Phylogenetics

Many empirical studies have revealed considerable differences between nonparametric bootstrapping and Bayesian posterior probabilities in terms of the support values for branches, despite claimed predictions about their approximate equivalence. We investigated this problem by simulating data, which were then analyzed by maximum likelihood bootstrapping and Bayesian phylogenetic analysis using identical models and reoptimization of parameter values. We show that Bayesian posterior probabilities are significantly higher than corresponding nonparametric bootstrap frequencies for true clades, but also that erroneous conclusions will be made more often. These errors are strongly accentuated when the models used for analyses are underparameterized. When data are analyzed under the correct model, nonparametric bootstrapping is conservative. Bayesian posterior probabilities are also conservative in this respect, but less so.

Association Between Adoption of Evidence-Based Treatment and Survival for Patients With ST-Elevation Myocardial Infarction

CONTEXT Only limited information is available on the speed of implementation of new evidence-based and guideline-recommended treatments and its association with survival in real life health care of patients with ST-elevation myocardial infarction (STEMI). OBJECTIVE To describe the adoption of new treatments and the related chances of short- and long-term survival in consecutive patients with STEMI in a single country over a 12-year period. DESIGN, SETTING, AND PARTICIPANTS The Register of Information and Knowledge about Swedish Heart Intensive Care Admission (RIKS-HIA) records baseline characteristics, treatments, and outcome of consecutive patients with acute coronary syndrome admitted to almost all hospitals in Sweden. This study includes 61,238 patients with a first-time diagnosis of STEMI between 1996 and 2007. MAIN OUTCOME MEASURES Estimated and crude proportions of patients treated with different medications and invasive procedures and mortality over time. RESULTS Of evidence-based treatments, reperfusion increased from 66% (95%, confidence interval [CI], 52%-79%) to 79% (95% CI, 69%-89%; P < .001), primary percutaneous coronary intervention from 12% (95% CI, 11%-14%) to 61% (95% CI, 45%-77%; P < .001), and revascularization from 10% (96% CI, 6%-14%) to 84% (95% CI, 73%-95%; P < .001). The use of aspirin, clopidogrel, β-blockers, statins, and angiotensin-converting enzyme (ACE) inhibitors all increased: clopidogrel from 0% to 82% (95% CI, 69%-95%; P < .001), statins from 23% (95% CI, 12%-33%) to 83% (95% CI, 75%-91%; P < .001), and ACE inhibitor or angiotensin II receptor blockers from 39% (95% CI, 26%-52%) to 69% (95% CI, 58%-70%; P < .001). The estimated in-hospital, 30-day and 1-year mortality decreased from 12.5% (95% CI, 4.3%-20.6%) to 7.2% (95% CI, 1.7%-12.6%; P < .001); from 15.0% (95% CI, 6.2%-23.7%) to 8.6% (95% CI, 2.7%-14.5%; P < .001); and from 21.0% (95% CI, 11.0%-30.9%) to 13.3% (95% CI, 6.0%-20.4%; P < .001), respectively. After adjustment, there was still a consistent trend with lower standardized mortality over the years. The 12-year survival analyses showed that the decrease of mortality was sustained over time. CONCLUSION In a Swedish registry of patients with STEMI, between 1996 and 2007, there was an increase in the prevalence of evidence-based treatments. During this same time, there was a decrease in 30-day and 1-year mortality that was sustained during long-term follow-up.

Eptifibatide Is Noninferior to Abciximab in Primary Percutaneous Coronary Intervention: Results From the SCAAR (Swedish Coronary Angiography and Angioplasty Registry)

OBJECTIVES The aim of this study was to test the noninferiority of eptifibatide relative to abciximab in patients with ST-segment elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (PCI). BACKGROUND Glycoprotein IIb/IIIa inhibitors are recommended by international guidelines in patients with acute coronary syndromes undergoing PCI. Abciximab is recommended with a higher level of evidence than eptifibatide in patients with STEMI. No large, prospective, randomized trial comparing abciximab and eptifibatide has been published. METHODS All (n = 11,479) STEMI patients in Sweden who underwent primary PCI and received either eptifibatide or abciximab from 2004 to 2007 were derived from the SCAAR (Swedish Coronary Angiography and Angioplasty Registry). The primary end point was death or myocardial infarction (MI) during 1-year follow-up, with adjustment for baseline differences with a multivariate logistic regression analysis including propensity score. The pre-specified noninferiority margin was set to 1.29. RESULTS The combined end point occurred in 353 of 2,355 patients (15.0%) treated with eptifibatide and in 1,432 of 9,124 patients (15.7%) treated with abciximab. The unadjusted odds ratio (OR) for eptifibatide versus abciximab was 0.95 (95% confidence interval [CI]: 0.84 to 1.08). Multivariate adjustment (n = 11,317) confirmed noninferiority, with an OR of 0.94 (95% CI: 0.82 to 1.09). The adjusted secondary end points of death and MI separately also showed noninferiority, with ORs of 0.99 (95% CI: 0.82 to 1.19) and 0.88 (95% CI: 0.73 to 1.05), respectively. CONCLUSIONS This large registry study suggests that eptifibatide is noninferior to abciximab in patients with STEMI undergoing primary PCI with respect to death or MI during 1 year, thereby supporting the use of either drug in clinical practice.

Radical reduction of cephalosporin use at a tertiary hospital after educational antibiotic intervention during an outbreak of extended-spectrum β-lactamase-producing Klebsiella pneumoniae

OBJECTIVES During an outbreak of extended-spectrum β-lactamase (ESBL)-producing Klebsiella pneumoniae at our hospital, we performed an educational antibiotic intervention aimed at reducing prescriptions of second- and third-generation cephalosporins and preventing increased use of fluoroquinolones and carbapenems. In this report, we describe the implementation strategy used and evaluate the intervention effect according to Cochrane recommendations. METHODS New recommendations for empirical intravenous antibiotic treatment were communicated to prescribers throughout the hospital by infectious diseases physicians working with Strama (the Swedish strategic programme against antibiotic resistance). No restrictive measures were used. The intervention effect was analysed with interrupted time series (ITS) regression analysis of local and national monthly antibiotic sales data. RESULTS A radical immediate and sustained reduction was demonstrated for the cephalosporins targeted in the intervention, whereas consumption of piperacillin/tazobactam and penicillin G increased substantially. Fluoroquinolone and carbapenem use was essentially unchanged. The ESBL outbreak subsided and no increased resistance to piperacillin/tazobactam was detected in K. pneumoniae, Escherichia coli or Pseudomonas aeruginosa blood isolates during the 2.5 year follow-up. CONCLUSIONS Our study clearly demonstrates that an educational intervention can have an immediate and profound effect on antibiotic prescription patterns at a large tertiary hospital. ITS regression analysis of local and national antibiotic sales data was valuable to readily assess the immediate and sustained effects of the intervention.

Blood pressure levels and risk of cardiovascular events and mortality in type-2 diabetes: cohort study of 34 009 primary care patients.

Objective: The optimal blood pressure (BP) in persons with type-2 diabetes is debated. We investigated shapes of the associations of SBP and DBP levels with risk of cardiovascular events and mortality in a large primary care-based sample of diabetic patients. Methods: We investigated all 34 009 consecutive cardiovascular disease-free type-2 diabetes patients aged 35 years or older (mean age 64 years) at 84 primary care centers in central Sweden between 1999 and 2008. We followed this cohort until the end of 2009 in national registries for the incidence of major cardiovascular events (a composite endpoint of myocardial infarction, stroke, heart failure, or cardiovascular mortality) or total mortality. Results: During up to 11 years of follow-up, 6344 patients (18.7%) had a first cardiovascular event, and 6235 died (18.3%). The associations of annually updated SBP and DBP with risk of major cardiovascular events were U-shaped. The lowest risk of cardiovascular events was observed at a SBP of 135–139 mmHg and a DBP of 74–76 mmHg, and the lowest mortality risk at a SBP of 142–150 mmHg and a DBP of 78–79 mmHg, in both antihypertensive drug-untreated and drug-treated persons. Conclusion: In a large primary care-based sample of patients with type-2 diabetes, associations of SBP and DBP with risk of major cardiovascular events and mortality were U-shaped. This may have implications for risk stratification of persons with diabetes.

Associations of HbA1c and educational level with risk of cardiovascular events in 32 871 drug-treated patients with Type 2 diabetes: a cohort study in primary care

To explore the association of HbA1c and educational level with risk of cardiovascular events and mortality in patients with Type 2 diabetes.

Association of Spironolactone Use With All-Cause Mortality in Heart Failure A Propensity Scored Cohort Study

Background—In 3 randomized controlled trials in heart failure (HF), mineralocorticoid receptor antagonists reduced mortality. The net benefit from randomized controlled trials may not be generalizable, and eplerenone was, but spironolactone was not, studied in mild HF. We tested the hypothesis that spironolactone is associated with reduced mortality also in a broad unselected contemporary population with HF and reduced ejection fraction, in particular New York Heart Association (NYHA) I–II. Methods and Results—We prospectively studied 18 852 patients (age 71±12 years; 28% women) with NYHA I–IV and ejection fraction <40% who were registered in the Swedish Heart Failure Registry between 2000 and 2012 and who were (n=6551) or were not (n=12 301) treated with spironolactone. We derived propensity scores for spironolactone treatment based on 41 covariates. We assessed survival by Cox regression with adjustment for propensity scores and with matching based on propensity score. We performed sensitivity and residual confounding analyses and analyzed the NYHA I–II and III–IV subgroups separately. One-year survival was 83% versus 84% in treated versus untreated patients (log rank P<0.001). After adjustment for propensity scores, the hazard ratio for spironolactone was 1.05 (95% confidence interval, 1.00–1.11; P=0.054). Spironolactone interacted with NYHA (P<0.001). In the NYHA I–II subgroup, after adjustment for propensity scores, the hazard ratio for spironolactone was 1.11 (95% confidence interval, 1.02–1.21; P=0.019). Conclusions—In an unselected contemporary population of HF with reduced ejection fraction, spironolactone was not associated with reduced mortality. The net benefits of spironolactone may be lower outside the clinical trial setting and in milder HF.

Prehospital Trauma Life Support (PHTLS) training of ambulance caregivers and impact on survival of trauma victims

BACKGROUND The Prehospital Trauma Life Support (PHTLS) course has been widely implemented and approximately half a million prehospital caregivers in over 50 countries have taken this course. Still, the effect on injury outcome remains to be established. The objective of this study was to investigate the association between PHTLS training of ambulance crew members and the mortality in trauma patients. METHODS A population-based observational study of 2830 injured patients, who either died or were hospitalized for more than 24 h, was performed during gradual implementation of PHTLS in Uppsala County in Sweden between 1998 and 2004. Prehospital patient records were linked to hospital-discharge records, cause-of-death records, and information on PHTLS training and the educational level of ambulance crews. The main outcome measure was death, on scene or in hospital. RESULTS Adjusting for multiple potential confounders, PHTLS training appeared to be associated with a reduction in mortality, but the precision of this estimate was poor (odds ratio, 0.71; 95% confidence interval, 0.42-1.19). The mortality risk was 4.7% (36/763) without PHTLS training and 4.5% (94/2067) with PHTLS training. The predicted absolute risk reduction is estimated to correspond to 0.5 lives saved annually per 100,000 population with PHTLS fully implemented. CONCLUSIONS PHTLS training of ambulance crew members may be associated with reduced mortality in trauma patients, but the precision in this estimate was low due to the overall low mortality. While there may be a relative risk reduction, the predicted absolute risk reduction in this population was low.

Fundamental differences between the methods of maximum likelihood and maximum posterior probability in phylogenetics

Using a four-taxon example under a simple model of evolution, we show that the methods of maximum likelihood and maximum posterior probability (which is a Bayesian method of inference) may not arrive at the same optimal tree topology. Some patterns that are separately uninformative under the maximum likelihood method are separately informative under the Bayesian method. We also show that this difference has impact on the bootstrap frequencies and the posterior probabilities of topologies, which therefore are not necessarily approximately equal. Efron et al. (Proc. Natl. Acad. Sci. USA 93:13429-13434, 1996) stated that bootstrap frequencies can, under certain circumstances, be interpreted as posterior probabilities. This is true only if one includes a non-informative prior distribution of the possible data patterns, and most often the prior distributions are instead specified in terms of topology and branch lengths. [Bayesian inference; maximum likelihood method; Phylogeny; support.].

Outcomes of antibiotic prophylaxis in acute cholecystectomy in a population-based gallstone surgery registry.

The aim of this study was to assess the effect of antibiotic prophylaxis (AP) on postoperative infections in acute cholecystectomy.