Facts Database Research Articles

Assessing efficiency of public health and medical care provision in OECD countries after a decade of reform

The objective of this study was to examine the change in efficiency of health care systems of 34 OECD countries between 2000 and 2012, a period marked by significant health reform in most OECD countries. This paper uses a novel Dynamic Network Data Envelopment Analysis (DNDEA) model to analyze the efficiency of the public health system and the medical care system of these OECD countries independently along with assessing the efficiency of their overall health system. This helps understand the relative priorities for improving the overall health system. The data for this study was obtained from the OECD Health Facts database. The study findings suggest that countries which improved their public health system were more likely to show overall improvement in efficiency.

Associations Between Cardiovascular Events and Nonergot Dopamine Agonists in Parkinson's Disease.

Knowledge of possible cardiovascular risks from Parkinson's disease (PD) medications is critical to informing safe and effective treatment decisions. The objective of our study was to determine whether PD patients treated with nonergot dopamine agonists (DAs) are at increased risk of adverse cardiovascular or cerebrovascular outcomes, relative to PD patients receiving other treatments. Matched case-control studies were conducted within a cohort of 14,122 inpatients receiving treatment for PD who were identified in the Cerner Health Facts database. Primary outcomes were associations between nonergot DA use and diagnosis of adverse cardiovascular events (acute myocardial infarction, heart failure [HF], hypotension, and valvulopathy). Secondary outcomes included associations between nonergot DA use and diagnosis of adverse cerebrovascular events (cerebrovascular accident and ischemic stroke) and odds of significant exposure-outcome relationships by patient factors. HF was the only adverse event that demonstrated a significant association with nonergot DA use. Individuals treated with pramipexole were more likely to be diagnosed with HF, relative to no use (adjusted odds ratio [AOR]: 1.28; 95% confidence interval [CI]: 1.07-1.53). The association between pramipexole and HF was greater among individuals treated with pramipexole monotherapy (relative to levodopa monotherapy) (AOR, 1.50; 95% CI: 1.09-2.06). Compared to nonusers, men and older individuals treated with pramipexole were more likely to be diagnosed with HF. Results from our study suggest an association between pramipexole use and HF. Findings warrant replication; however, individuals with PD and independent risk factors for, or a history of, HF may benefit from limited use of this drug.

Temporal trends of esophageal disorders by age in the Cerner Health Facts database

PurposeEsophageal adenocarcinoma incidence has increased approximately 600% over the last 4 decades in the United States. Little research has been conducted on the temporal trends of gastroesophageal reflux disease (GERD) and Barrett's esophagus (BE), yet it is important to establish whether these conditions have also increased with time or differ by age. MethodsThe Cerner Health Facts® database contains information on 35 million patients between 2001 and 2010. GERD, BE, and esophageal cancer (EC) cases were defined using International Classification of Diseases, ninth edition codes. We calculated age-adjusted rates and 95% confidence intervals for GERD, BE, and EC. ResultsIn this population, the overall, all-age rate per 100,000 encounters for GERD was 711.9, BE was 21.6, and EC was 6.1. During 2001–2010, GERD rates increased by approximately 50% and EC rates more than doubled, but BE rates declined by approximately 40%. Trends were similar by age, and all rates were higher in Caucasians and males. ConclusionsThese data indirectly support the idea that increased incidence of EC may be partially due to GERD and raise the provocative hypothesis that BE rates may be decreasing possibly as a forerunner of continued stabilization of esophageal adenocarcinoma rates and a possible subsequent decline.

Spectr-W3 online database on atomic properties of multicharged ions

The Spectr-W3 information-reference system was developed in 2001-2013 and realized as an online Web resource based on the factual atomic database Spectr-W3. The information accumulated in the Spectr-W3 atomic database contains about 450,000 records and includes the experimental and theoretical data on ionization potentials, energy levels, wavelengths, radiation transition probabilities, and oscillator strengths, and the parameters of analytical approximations of electron-collisional cross-sections and rates for atoms and ions. Those data were extracted from publications in physical journals, proceedings of the related conferences, special-purpose publications on atomic data, provided directly by authors. The information is supplied with references to the original sources and comments, elucidating the details of experimental measurements or calculations. To date, the Spectr-W3 atomic database is still the largest factual database in the world, containing the information on spectral properties of multicharged ions. In 2014, the new stage in the development of the Spectr-W3 atomic database started. The purpose of this stage is the creation of a new information section of the Spectr-W3 database. This section would contain the information on the x-ray emission spectrograms registered from various plasma sources.

An expert system for acoustic diagnosis of power circuit breakers and on-load tap changers

Most of the faults on power Circuit Breakers (CBs) and On-Load Tap Changers (OLTCs) on high voltage transformers are of mechanical origin. Mechanical malfunction, mechanical wear and other types of abnormal behaviors can be detected as changes in the acoustic signatures. In this paper, an Expert System (ES) with the addition of a signal processor module and factual database module to the conventional ES has been proposed for the diagnosis of CBs and OLTCs. The feature extraction of the acoustic signatures has been done by decomposing them into voiced and silent portions in time domain and through FFT spectrum analysis in frequency domain. The OLTC's motor current has been decomposed into inrush, steady-state and extended currents to locate the cause of different anomalies in the OLTCs and acoustic-current pair has been used for the synchronization testing. Each extracted feature of the testing signal has been compared to the similar feature of a reference signal to not only identify the health of the testing device but also to locate the cause of each anomaly. Several test signals have been compared against the reference signature and the proposed expert system has proved to be a suitable, effective and reliable system for the acoustic diagnosis of OLTCs and CBs.

Trends in inpatient antiparkinson drug use in the USA, 2001-2012.

PurposeAlthough therapeutic options and clinical guidelines for Parkinson’s disease (PD) have changed significantly in the past 15 years, prescribing trends in the USA remain unknown. The purpose of this population-based cohort study was to examine patterns of inpatient antiparkinson drug use between January 2001 and December 2012 in relation to clinical guideline publication, drug introduction/withdrawal, and emerging safety concerns.MethodsA total of 16,785 inpatients receiving pharmacological treatment for PD were identified in the Cerner Health Facts database. Our primary outcome was standardized (age, sex, race, and census region) annual prevalence of antiparkinson drug use. We also examined antiparkinson medication trends and polypharmacy by age and sex.ResultsThe most frequently prescribed antiparkinson drugs between 2001 and 2012 were levodopa (85 %) and dopamine agonists (28 %). Dopamine agonist use began declining in 2007, from 34 to 27 % in 2012. The decline followed publication of the American Academy of Neurology’s practice parameter refuting levodopa toxicity, pergolide withdrawal, and pramipexole label revisions. Despite safety concerns for cognitive impairment and falls, individuals ≥80 years of age demonstrated stable rates of dopamine agonist use from 2001 to 2012. Polypharmacy was most common in younger patients.ConclusionsDopamine agonist use declined from 2007 to 2012, suggesting that increased awareness of safety issues and practice guidelines influenced prescribing. These events appear to have minimally influenced treatment provided to older PD patients. Antiparkinson prescribing trends indicate that safety and best practice information may be communicated effectively.Electronic supplementary materialThe online version of this article (doi:10.1007/s00228-015-1881-4) contains supplementary material, which is available to authorized users.

Association of persistent and transient worsening renal function with mortality risk, readmissions risk, length of stay, and costs in patients hospitalized with acute heart failure

BackgroundData comparing effects of transient worsening renal function (WRFt) and persistent WRF (WRFp) on outcomes in patients hospitalized with acute heart failure (AHF) are lacking. We determined the characteristics of hospitalized AHF patients who experienced no worsening renal function (non-WRF), WRFt, or WRFp, and the relationship between cohorts and AHF-related outcomes.Methods and resultsA patient’s first AHF hospitalization (index) was identified in the Cerner Health Facts® database (January 2008−March 2011). Patients had WRF if serum creatinine (SCr) was ≥0.3 mg/dL and increased ≥25% from baseline, and they were designated as WRFp if present at discharge or WRFt if not present at discharge. A total of 55,436 patients were selected (non-WRF =77%, WRFp =10%, WRFt =13%). WRFp had greater comorbidity burden than WRFt. At index hospitalization, WRFp patients had the highest mortality, whereas WRFt patients had the longest length of stay (LOS) and highest costs. These trends were observed at 30, 180, and 365 days postdischarge and confirmed by multivariable analyses. WRF patients had more AHF-related readmissions than non-WRF patients. In sensitivity analyses of the patient subset with live index hospitalization discharges, postdischarge LOS and costs were highest in WRFt patients, whereas mortality associated with a HF hospitalization was significantly higher for WRF patients vs non-WRF patients, with no difference between WRFp and WRFt.ConclusionIn patients hospitalized for AHF, WRFp was associated with the highest mortality, whereas WRFt was associated with the highest LOS and costs. WRF patients had higher readmissions than non-WRF patients. Transient increases in SCr appear to be associated with detrimental outcomes, especially longer LOS and higher costs.

To Begin at the Beginning: Determining Priorities for Pediatric Palliative Care Fast Facts with a Modified Delphi Study (TH346-A)

Discuss the Fast Fact topics that pediatric palliative care physicians and pediatric residency program directors agree are most important for the education of pediatric interns. Original Research Background: Palliative care Fast Facts—a collection of concise educational summaries on palliative care topics— enhance knowledge of internal medicine residents. There is no existing database of pediatric palliative care Fast Facts and no agreement on the core topics for which these summaries are needed. Research Objectives: To determine the core topics for pediatric palliative care Fast Facts for the pediatric intern. Methods: of experts on an issue where none previously existed. We surveyed expert pediatric palliative care physicians and expert pediatric residency program directors to reach consensus on the highest yield topics for pediatric Fast Facts for the pediatric intern. In the first round survey, experts rated (1-5 Likert scale) a list of 93 pediatric palliative care topics identified via literature review to indicate the top Fast Facts that should be written for the pediatric intern and suggested additional topics. Topics were categorized

Pharmacoepidemiological study of drug-drug interactions in onco-hematological pediatric patients.

Onco-hematological patients are particularly susceptible to drug-drug interactions (DDIs) because they often undergo multiple combined treatments. Some studies have analyzed the frequency of DDIs in adult patients with cancer; however, the prevalence of DDIs in children, and especially among pediatric cancer patients, remains unknown. To determine the prevalence of DDIs in treatment sheets comparing two commonly used drug interaction databases, to describe the most common clinically relevant DDIs (CR-DDIs) and to investigate the risk factors associated with them. An onco-hematological pediatric unit from a tertiary hospital in Spain. A prospective, observational and descriptive study was carried out from November 2012 to February 2013. Twice a week, every patient's treatment sheet was collected. Each medication list was screened through two databases: Thomson Micromedex™ and Drug Interaction Facts™. All identified DDIs were graded by their level of severity. Summary statistics were used to describe patient and disease characteristics, most often prescribed drugs, and frequency, types and classification of CR-DDIs. Multivariate analysis was used to identify risk factors associated with CRDDIs. Prevalence of CR-DDIs was measured as percentage. A total of 506 potential DDIs were detected in 150 treatment sheets. The prevalence of CR-DDIs by Micromedex database and Drug Interaction Facts database were 44.7 and 51.3% respectively. Amikacin, azole antifungals, antiemetics and cyclosporine were the most frequent drugs involved in CR-DDIs. In multivariate analysis, the main risk factor associated with increased odds for CR-DDIs was a higher number of drugs. The frequency of potential DDIs was related to a higher number of drugs, being immunosuppressant and azole antifungal agents the most commonly involved drugs. The lack of agreement between different databases enhances the complexity to detect drug interactions in clinical practice.

Abstract 234: Prevalence and Prognosis of Hyperkalemia in Dialysis and Non-Dialysis Patients with Acute Myocardial Infarction

Background: Hyperkalemia is a potentially dangerous complication during hospitalization for acute myocardial infarction (AMI), but its prevalence and prognostic importance in contemporary practice have not been described. Methods: We studied consecutive patients with AMI from 67 hospitals in the Cerner Health Facts database to assess the association between maximum in-hospital potassium levels (max K) and in-hospital mortality. Patients were stratified by dialysis status, and grouped into categories of max K (<5 mEq/L, 5-<5.5 mEq/L, 5.5-<6.0 mEq/L, 6.0-<6.5 mEq/L, and ≥ 6.5 mEq/L). Multivariable hierarchical logistic regression was used to adjust for multiple patient and site characteristics. Results: Of 38,689 patients, 886 (2.3%) were on dialysis. Hyperkalemia (max K ≥ 5.0 mEq/L) occurred in 23.6% of the overall cohort; 22.6% of non-dialysis patients; and 66.8% of dialysis patients. Moderate-severe hyperkalemia (max K ≥ 5.5 mEq/L) occurred in 9.8% of patients. There was an increase in mortality with higher max K levels; this relationship was linear in non-dialysis patients, while a plateau was observed in dialysis patients (Figure). Regardless of dialysis status, in-hospital mortality exceeded 15% once max K ≥5.5 mEq/L. The relationship between max K and mortality persisted after multivariable adjustment (adjusted odds ratios [OR] for max K of 5-<5.5, 5.5-<6.0, 6.0-<6.5, ≥ 6.5 vs. <5 were 2.03 (95% CI 1.79-2.31), 2.79 (2.35-3.31), 4.62 (3.74-5.71), 4.97 (3.96-6.25) respectively for dialysis patients; and 1.62 (1.41-1.87), 2.02 (1.68-2.43), 3.18 (2.51-4.03), 3.37 (2.60-4.36) respectively for non-dialysis patients). Conclusions: Hyperkalemia occurs in >20% of non-dialysis and >65% of dialysis patients with AMI. There is a graded mortality increase with higher max K levels, particularly in non-dialysis patients, with marked increases in mortality at even mild levels of hyperkalemia. Additional research into the determinants, pathophysiology, and clinical scenarios in which hyperkalemia contributes to mortality is warranted.

Operational rules for the implementation of INN prescribing

PurposeThe aim is to describe the process, principles and results of the International Nonproprietary Name (INN) prescribing project in Belgium. The purpose of this project was to operationalize electronic INN prescribing for outpatient care in daily medical practice and to develop a factual database which can be used in electronic INN prescribing applications. MethodsThe operationalization process consisted of three phases: (1) expert consultation, (2) review by regulatory authorities and (3) test phase with stakeholders and end-users. ResultsThe INN prescribing project resulted into (1) operational rules for electronic INN prescribing and (2) a reference database to be implemented in commercial medical software. The operational rules for electronic INN prescribing define valid INN groups as sets of equivalent medicinal products, described by three elements: the therapeutic moiety (the active part of the therapeutic ingredient) or combination of therapeutic moieties, the strength (with standardized denominators), and the method of administration (with simplified but standardized options). The operational rules also define two categories of exemptions for INN prescribing: INN groups where the first choice of treatment should be continued throughout the therapy period (NO SWITCH) and medicinal product groups not suitable for INN prescribing (NO INN). The reference database is the result of the virtual classification of the Belgian therapeutic arsenal into INN groups, according to the operational rules. ConclusionsDefining the operational rules for INN prescribing for and with different stakeholders was a difficult yet feasible assignment. The INN prescribing project resulted into explicit operational rules and a reference database. The Belgian experience may provide important information for other countries planning to operationalize or refine electronic INN prescribing. It can also be used for a thorough evaluation of the impact of the new concept of INN prescribing on daily practice and on medical education.

Characteristics of patients with atrial fibrillation and history of stroke or transient ischemic attack - some results from the FACTS Database

Background and purpose. Atrial fibrillation (AF) is the most common arrhythmia in adult population and represents an independent risk factor for stroke with an annual risk for thromboembolic events for chronic fibrillation estimated at three to six percentage. The risk of stroke is five time more frequent in AF patients and the outcome of the ischemic stroke in AF patients is worse than other forms of ischemic stroke. There are few data regarding the specific management and outcome of the 200,000 estimated AF patients in Romania. The aim of the present study (part of the FACTS program) was to determine the characteristics of patients with AF/AFl and their management. In this paper we have analyzed the characteristics of the subgroup of patients with history of stroke or transient ischemic attack (TIA). Methods and results. From March 2011 to April 2012 a total of 3700 patients from 27 sites, ambulatories or hospitals, were enrolled including regions all over the country. Of this cohort 374 patients had documented history of stroke or TIA. Hypertension, sedentary lifestyle, associated diseases (cardiovascular and non-cardiovascular) statistically correlate with history of stroke/TIA. CHADS2 and CHA2DS2Vasc scores had higher values for patients with stroke/TIA. Conclusion. This database (part of the Romanian AF Registry) offers an insight into patient’s characteristics with AF/AFl and stoke/TIA history and provides comprehensive data that will allow a range of evaluation at national level and contribute to better understand the management of AF/AFl in Romania. Next evaluation aims to analyze anticoagulant therapy in patients with AF and especially of those with stroke/TIA history.

Learning Dependency-Based Compositional Semantics

Suppose we want to build a system that answers a natural language question by representing its semantics as a logical forxm and computing the answer given a structured database of facts. The core part of such a system is the semantic parser that maps questions to logical forms. Semantic parsers are typically trained from examples of questions annotated with their target logical forms, but this type of annotation is expensive. Our goal is to instead learn a semantic parser from question–answer pairs, where the logical form is modeled as a latent variable. We develop a new semantic formalism, dependency-based compositional semantics (DCS) and define a log-linear distribution over DCS logical forms. The model parameters are estimated using a simple procedure that alternates between beam search and numerical optimization. On two standard semantic parsing benchmarks, we show that our system obtains comparable accuracies to even state-of-the-art systems that do require annotated logical forms.

Risk factors for readmission after lower extremity procedures for peripheral artery disease

As pressure to contain health care costs increases, there has been greater scrutiny of readmissions in the vascular surgery population. The objective of this study was to evaluate postoperative readmissions after open and endovascular lower extremity (LE) procedures for peripheral artery disease (PAD). A retrospective cohort study was performed of elective adult inpatients with PAD and LE procedures which were selected from the Health Facts database (Cerner Corporation, Kansas City, Mo) between October 2008 and December 2010 using International Classification of Disease, 9th Clinical Modification diagnosis codes (claudication, rest pain, and ulceration/gangrene) and procedure codes for LE revascularization (endovascular and open). Multivariable logistic regression and χ(2) were used to compare patients who received endovascular and open procedures. The Charlson Comorbidity Index, comorbid diagnoses, and laboratory results were used to adjust for confounding. The main outcome measure evaluated was readmission ≤ 30 days of discharge. Of 463,362 index admissions, 16,574 patients were identified with a diagnosis of PAD. Combining PAD with elective LE procedures during the index admission, 777 underwent open and 681 underwent endovascular procedures. Unadjusted readmission rates for open and endovascular procedures for claudication, rest pain, and ulceration/gangrene were 10.2% vs 11.3% (P = .69), 14.0% vs 18.2% (P = .43), and 21.1% vs 19.5% (P = .69), respectively. Readmission increased by the severity of the diagnosis for open and endovascular procedures (P = .0006). Men comprised 58% of the cohort; readmission rates were not statistically different by sex (P = .19). Race was not associated with procedure performed (P = .16), but nonwhite race was associated with more severe PAD (P < .0001). After adjusting for race, sex, comorbidities, length of stay, and laboratory values outside of normal reference ranges, the association between an endovascular procedure and readmission was not statistically significant (odds ratio, 1.28; 95% confidence interval, 0.94-1.75). Less invasive endovascular procedures were not associated with decreased readmission rates compared with open surgery. The overall readmission rate for claudicant patients was 10.7%, which was unexpectedly high. Predictors of readmission included male sex, longer hospital stays, hospital infection, elevated aspartate aminotransferase, and high numbers of medications ordered and dispensed. Further examination exploring reasons for readmission are required to decrease readmission rates in the vascular surgery population.

Fatalities and Injuries Due to Motorcycle-Deer Collisions in Michigan

Motorcycle-deer collisions have been described in several small case series. Risk factors for motorcycle-deer collisions and opportunities for intervention have yet to be described. The aim of this case-control study is to determine risk factors associated with fatal and injury-related motorcycle-deer collisions. The study group was created from Michigan State Police reports and were abstracted from the Michigan Traffic Crash Facts Database for motorcycle-deer collisions occurring from 2004 to 2011 (n = 2692). A total of 50 fatal (1.9%) and 2642 nonfatal (98.1%) collisions were reported during the study period. Persons involved in fatal collisions were older than those in nonfatal collisions (54.3 vs 44.3, P = .028). Injury severity was evenly distributed (20.3% incapacitating, 31.7% nonincapacitating, 22.2% possible injury, and 25.8% not injured). Transfer to hospital was highest among persons with incapacitating injury (90.7%), fatalities (68%), and nonincapacitating injury (63.7%). Among persons transported to the hospital, helmet use was significantly less for fatal collisions (P = .0009). Motorcycle damage was significantly greater among fatal crashes (P < .0001). Swerving to avoid a crash resulted in higher proportions of nonincapacitating injury but zero fatalities (P < .0001). The majority of collisions (80%) occurred between June and October. November saw a higher proportion of fatal crashes (12% vs 4%, P = .004). The highest risk time of day for all collisions was between 5:00 am and 7:00 am (18% vs 13.6%) and 7:00 pm and 11:00 pm (42% vs 36.5%). Both groups were primarily male, had similar alcohol and drug use, and experienced similar weather, light, and road conditions. Fatal motorcycle-deer collisions were more likely to involve older persons, less helmet usage, and do more damage to the motorcycle. Swerving to avoid collision may prevent a fatal crash. Opportunities for intervention should focus on limiting deer access to roadways and increasing safe driving practices.

Automatic Learning of Ontologies for the Semantic Web: experiment lexical learning

This paper proposes the design of a System for Automatic Learning of Ontologies and Lexical Information (SALOX) for the Dynamic Semantic Ontological Framework for the Semantic Web (DSOFSW). DSOFSW interprets query in natural language (Spanish) to the web, and is composed by five parts; a linguistic ontology for the grammar of Spanish, a lexicon for the lexical information, a database of facts about the system experiences, a task ontology for the linguistic analysis process, and an interpretative ontology of the context. SALOX integrates several methods, approaches and techniques for information extraction, discovery and actualization (pragmatic (user profile, context knowledge), lexical and semantic linguistic information, etc.) in order to update the knowledge used for DSOFSW. SALOX has a component to map the sources of learning with the learning methods, and another to update the linguistic ontology and the lexicon of the DSOFSW. Specifically, in this paper we present the design of the learning unit of lexical information. Keywords: Natural language processing, ontological semantic, machine learning, ontological leaning, semantic web.

Trends in the Incidence of Acute Kidney Injury in Patients Hospitalized With Acute Myocardial Infarction

Acute kidney injury (AKI) is common in patients with acute myocardial infarction (AMI) and is associated with permanent renal impairment and death. Although guidelines increasingly emphasize AKI prevention, whether increased awareness has translated into reduced AKI rates is unclear. Among 33,249 consecutive hospitalizations in 31,532 unselected patients with AMI across 56 US centers from Cerner Corporation's Health Facts database, we examined the temporal trends in AKI incidence from 2000 to 2008. Acute kidney injury was defined as an absolute increase in creatinine level of at least 0.3 mg/dL or a relative increase of at least 50% during hospitalization. From 2000 to 2008, the mean age of patients increased (from 66.5 to 68.6 years), as did the known AKI risk factors, including chronic kidney disease, cardiogenic shock, diabetes mellitus, heart failure, coronary angiography, and percutaneous coronary intervention. Despite this, AKI incidence declined from 26.6% in 2000 to 19.7% in 2008 (P < .001). After multivariate adjustment, the trend of decreasing AKI rates persisted (4.4% decline per year; P < .001). In addition, in-hospital mortality also declined over time among patients developing AKI, from 19.9% in 2000 to 13.8% in 2008 (P = .003). In a large national study, AKI incidence in patients hospitalized with AMI declined significantly from 2000 to 2008 despite the aging population and rising prevalence of AKI risk factors. These findings may reflect increased clinician awareness, better risk stratification, or greater use of AKI prevention efforts during this time period.

Serum Potassium Levels and Mortality in Acute Myocardial Infarction

Clinical practice guidelines recommend maintaining serum potassium levels between 4.0 and 5.0 mEq/L in patients with acute myocardial infarction (AMI). These guidelines are based on small studies that associated low potassium levels with ventricular arrhythmias in the pre-β-blocker and prereperfusion era. Current studies examining the relationship between potassium levels and mortality in AMI patients are lacking. To determine the relationship between serum potassium levels and in-hospital mortality in AMI patients in the era of β-blocker and reperfusion therapy. Retrospective cohort study using the Cerner Health Facts database, which included 38,689 patients with biomarker-confirmed AMI, admitted to 67 US hospitals between January 1, 2000, and December 31, 2008. All patients had in-hospital serum potassium measurements and were categorized by mean postadmission serum potassium level (<3.0, 3.0-<3.5, 3.5-<4.0, 4.0-<4.5, 4.5-<5.0, 5.0-<5.5, and ≥5.5 mEq/L). Hierarchical logistic regression was used to determine the association between potassium levels and outcomes after adjusting for patient- and hospital-level factors. All-cause in-hospital mortality and the composite of ventricular fibrillation or cardiac arrest. There was a U-shaped relationship between mean postadmission serum potassium level and in-hospital mortality that persisted after multivariable adjustment. Compared with the reference group of 3.5 to less than 4.0 mEq/L (mortality rate, 4.8%; 95% CI, 4.4%-5.2%), mortality was comparable for mean postadmission potassium of 4.0 to less than 4.5 mEq/L (5.0%; 95% CI, 4.7%-5.3%), multivariable-adjusted odds ratio (OR), 1.19 (95% CI, 1.04-1.36). Mortality was twice as great for potassium of 4.5 to less than 5.0 mEq/L (10.0%; 95% CI, 9.1%-10.9%; multivariable-adjusted OR, 1.99; 95% CI, 1.68-2.36), and even greater for higher potassium strata. Similarly, mortality rates were higher for potassium levels of less than 3.5 mEq/L. In contrast, rates of ventricular fibrillation or cardiac arrest were higher only among patients with potassium levels of less than 3.0 mEq/L and at levels of 5.0 mEq/L or greater. Among inpatients with AMI, the lowest mortality was observed in those with postadmission serum potassium levels between 3.5 and <4.5 mEq/L compared with those who had higher or lower potassium levels.

U-Shaped Association Between Potassium Level and Acute-MI Mortality

On the basis of small studies with surrogate outcomes, current practice guidelines for managing ST-elevation myocardial infarction (STEMI) recommend maintaining a potassium level greater than 4.0 mEq/L. To explore the association of potassium levels with mortality, investigators used the Cerner Health Facts database to conduct a retrospective cohort study involving 38,689 patients with biomarker-confirmed acute MI and one or more …

Complexity of verification of nondeterministic probabilistic multiagent systems

Probabilistic systems of interacting nondeterministic intelligent agents are considered. The states of agents in such systems are probabilistic databases (of facts), and their actions are controlled by probabilistic logical programs. Besides, communication channels between agents are also probabilistic. It is shown how such systems can be transformed in poynomial time to equivalent finite Markov decision processes. This makes it possible to translate the known results on the verification of the dynamic properties of the finite Markov processes to the probabilistic multiagent systems of the considered type.