Death certificates may lack accuracy and misclassify the cause of death. The validity of proxy-reported cause of death is not well established. The authors examined death records on 336 participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study, a national cohort study of 30,239 community-dwelling US adults (2003-2010). Trained experts used study data, medical records, death certificates, and proxy reports to adjudicate causes of death. The authors computed agreement on cause of death from the death certificate, proxy, and adjudication, as well as sensitivity and specificity for certain diseases. Adjudicated cause of death had a higher rate of agreement with proxy reports (73%; Cohen's kappa (κ) statistic = 0.69) than with death certificates (61%; κ = 0.54). The agreement between proxy reports and adjudicators was better than agreement with death certificates for all disease-specific causes of death. Using the adjudicator assessments as the "gold standard," for disease-specific causes of death, proxy reports had similar or higher specificity and higher sensitivity (sensitivity = 50%-89%) than death certificates (sensitivity = 31%-81%). Proxy reports may be more concordant with adjudicated causes of death than with the causes of death listed on death certificates. In many settings, proxy reports may represent a better strategy for determining cause of death than reliance on death certificates.

We investigate the sensitivity of the Fréchet Inception Distance (FID) score to inconsistent and often incorrect implementations across different image processing libraries. FID score is widely used to evaluate generative models, but each FID implementation uses a different low-level image processing process. Image resizing functions in commonlyused deep learning libraries often introduce aliasing artifacts. We observe that numerous subtle choices need to be made for FID calculation and a lack of consistencies in these choices can lead to vastly different FID scores. In particular, we show that the following choices are significant: ( 1) selecting what image resizing library to use, ( 2) choosing what interpolation kernel to use, ( 3) what encoding to use when representing images. We additionally outline numerous common pitfalls that should be avoided and provide recommendations for computing the FID score accurately. We provide an easy-to-use optimized implementation of our proposed recommendations in the accompanying code.

In a real-world setting, despite achieving better anticoagulation control, elderly patients had a higher risk of major hemorrhagic events. As the population ages and the candidacy for oral anticoagulation increases, strategies that mitigate the elevated risk of hemorrhage need to be identified.

BackgroundInternational guidelines recommend that the decision threshold for troponin should be the 99th percentile of a normal population, or, if the laboratory assay is not sufficiently precise at this low level, the level at which the assay achieves a 10% or better coefficient of variation (CV). Our objectives were to examine US hospital laboratory troponin reports to determine whether either the 99th percentile or the 10% CV level were clearly indicated, and whether nonconcordance with these guidelines was a potential barrier to detecting clinically important microscopic or ‘microsize’ myocardial infarctions (MIs). To confirm past reports of the clinical importance of microsize MIs, we also contrasted in-hospital, 28-day and 1-year mortality among those with microsize and nonmicrosize MI.MethodsIn the REasons for Geographic And Racial Differences in Stroke national prospective cohort study (n=30,239), 1029 participants were hospitalized for acute coronary syndrome (ACS) between 2003–2009. For each case, we recorded all thresholds of abnormal troponin on the laboratory report and whether the 99th percentile or 10% CV value were clearly identified. All cases were expert adjudicated for presence of MI. Peak troponin values were used to classify MIs as microsize MI (< five times the lowest listed upper limit of normal) and nonmicrosize MI.ResultsParticipants were hospitalized at 649 acute care US hospitals, only 2% of whose lab reports clearly identified the 99th percentile or the 10% CV level; 52% of reports indicated an indeterminate range, a practice that is no longer recommended. There were 183 microsize MIs and 353 nonmicrosize MIs. In-hospital mortality tended to be lower in the microsize than in the nonmicrosize MI group (1.1 vs. 3.6%, p = 0.09), but 28-day and 1-year mortality were similar (2.5% vs. 2.7% [p = 0.93] and 5.2% vs. 4.3% [p = 0.64], respectively).ConclusionsCurrent practices in many US hospitals created barriers to the clinical recognition of microsize MI, which was common and clinically important in our study. Improved hospital troponin reporting is warranted.

Individuals with cardiovascular disease (CVD) living in Health Professional Shortage Areas (HPSA) may receive less preventive care than others. The Reasons for Geographic And Racial Differences in Stroke Study (REGARDS) surveyed 30,221 African American (AA) and White individuals older than 45 years of age between 2003–2007. We compared medication use for CVD prevention by HPSA and insurance status, adjusting for sociodemographic factors, health behaviors, and health status. Individuals residing in partial HPSA counties were excluded. Mean age was 64±9 years, 42% were AA, 55% were women, and 93% had health insurance; 2,545 resided in 340 complete HPSA counties and 17,427 in 1,145 non-HPSA counties. Aspirin, beta-blocker, and ACE-inhibitor use were similar by HPSA and insurance status. Compared with insured individuals living in non-HPSA counties, statin use was lower among uninsured participants living in non-HPSA and HPSA counties. Less medication use for CVD prevention was not associated with HPSA status, but less statin use was associated with lack of insurance.