Regression-based techniques for statistical decision making in single-case designs

The present study evaluates the performance of four methods for estimating regression coefficients used to make statistical decisions regarding intervention effectiveness in single-case designs. Ordinary least squares estimation is compared to two correction techniques dealing with general trend and one eliminating autocorrelation whenever it is present. Type I error rates and statistical power are studied for experimental conditions defined by the presence or absence of treatment effect (change in level or in slope), general trend, and serial dependence. The results show that empirical Type I error rates do not approximate the nominal ones in presence of autocorrelation or general trend when ordinary and generalized least squares are applied. The techniques controlling trend show lower false alarm rates, but prove to be insufficiently sensitive to existing treatment effects. Consequently, the use of the statistical significance of the regression coefficients for detecting treatment effects is not recommended for short data series.

Early Recurrence and Cerebral Bleeding in Patients With Acute Ischemic Stroke and Atrial Fibrillation: effect of Anticoagulation and Its Timing: the RAF Study.

BACKGROUND AND PURPOSE: The best time for administering anticoagulation therapy in acute cardioembolic stroke remains unclear. This prospective cohort study of patients with acute stroke and atrial fibrillation, evaluated (1) the risk of recurrent ischemic event and severe bleeding; (2) the risk factors for recurrence and bleeding; and (3) the risks of recurrence and bleeding associated with anticoagulant therapy and its starting time after the acute stroke. METHODS: The primary outcome of this multicenter study was the composite of stroke, transient ischemic attack, symptomatic systemic embolism, symptomatic cerebral bleeding and major extracranial bleeding within 90 days from acute stroke. RESULTS: Of the 1029 patients enrolled, 123 had 128 events (12.6%): 77 (7.6%) ischemic stroke or transient ischemic attack or systemic embolism, 37 (3.6%) symptomatic cerebral bleeding, and 14 (1.4%) major extracranial bleeding. At 90 days, 50% of the patients were either deceased or disabled (modified Rankin score ≥3), and 10.9% were deceased. High CHA2DS2-VASc score, high National Institutes of Health Stroke Scale, large ischemic lesion and type of anticoagulant were predictive factors for primary study outcome. At adjusted Cox regression analysis, initiating anticoagulants 4 to 14 days from stroke onset was associated with a significant reduction in primary study outcome, compared with initiating treatment before 4 or after 14 days: hazard ratio 0.53 (95% confidence interval 0.30-0.93). About 7% of the patients treated with oral anticoagulants alone had an outcome event compared with 16.8% and 12.3% of the patients treated with low molecular weight heparins alone or followed by oral anticoagulants, respectively (P=0.003). CONCLUSIONS: Acute stroke in atrial fibrillation patients is associated with high rates of ischemic recurrence and major bleeding at 90 days. This study has observed that high CHA2DS2-VASc score, high National Institutes of Health Stroke Scale, large ischemic lesions, and type of anticoagulant administered each independently led to a greater risk of recurrence and bleedings. Also, data showed that the best time for initiating anticoagulation treatment for secondary stroke prevention is 4 to 14 days from stroke onset. Moreover, patients treated with oral anticoagulants alone had better outcomes compared with patients treated with low molecular weight heparins alone or before oral anticoagulants.

Impact of Study Design on Reported Incidences of Acute Mountain Sickness: A Systematic Review.

AIMS: Published incidences of acute mountain sickness (AMS) vary widely. Reasons for this variation, and predictive factors of AMS, are not well understood. We aimed to identify predictive factors that are associated with the occurrence of AMS, and to test the hypothesis that study design is an independent predictive factor of AMS incidence. We did a systematic search (Medline, bibliographies) for relevant articles in English or French, up to April 28, 2013. Studies of any design reporting on AMS incidence in humans without prophylaxis were selected. Data on incidence and potential predictive factors were extracted by two reviewers and crosschecked by four reviewers. Associations between predictive factors and AMS incidence were sought through bivariate and multivariate analyses for different study designs separately. Association between AMS incidence and study design was assessed using multiple linear regression. RESULTS: We extracted data from 53,603 subjects from 34 randomized controlled trials, 44 cohort studies, and 33 cross-sectional studies. In randomized trials, the median of AMS incidences without prophylaxis was 60% (range, 16%-100%); mode of ascent and population were significantly associated with AMS incidence. In cohort studies, the median of AMS incidences was 51% (0%-100%); geographical location was significantly associated with AMS incidence. In cross-sectional studies, the median of AMS incidences was 32% (0%-68%); mode of ascent and maximum altitude were significantly associated with AMS incidence. In a multivariate analysis, study design (p=0.012), mode of ascent (p=0.003), maximum altitude (p<0.001), population (p=0.002), and geographical location (p<0.001) were significantly associated with AMS incidence. Age, sex, speed of ascent, duration of exposure, or history of AMS were inconsistently reported and therefore not further analyzed. CONCLUSIONS: Reported incidences and identifiable predictive factors of AMS depend on study design.

The predictive value of preoperative B-type natriuretic peptide in children undergoing cardiac surgery

Introduction: B-type natriuretic peptide (BNP) is a biomarker of myocardial stress. In children, the value of preoperative BNP on postoperative outcome is unclear. The aim of this study was to determine the predictive value of preoperative NT-proBNP on postoperative outcome in children after congenital heart surgery. Results: Ninety-seven patients were included in the study with a median age of 3.3 years [0.7-5.2]. Preoperative median NT-proBNP was 412 pg/ml [164-1309]. NT-proBNP was above the P95 reference value for age in 56 patients (58%). Preoperative NT-proBNP was significantly higher in patients who had mechanical ventilation duration of more than 2 days (1156 pg/ml [281-1951] vs. 267 pg/ml [136-790], p=0.003) and who stayed more than 6 days in the pediatric intensive care unit (727 pg/ml [203-1951] vs. 256 pg/ml [136-790], p=0.007). However, preoperative NT-proBNP was not significantly higher in patients with an increased inotropic score, a prolonged cardiopulmonary bypass time or an increased surgical risk category. Conclusions: An elevated preoperative NT-proBNP reflects hemodynamic status and cardiac dysfunction, and therefore is a valuable adjunct in predicting a complicated postoperative course. ___________________________________ Introduction: Le peptide natriurétique type B (BNP) est un marqueur reflétant le stress myocardique. Dans la population pédiatrique, la signification des valeurs préopératoire de BNP, en particulier sur l'évolution postopératoire, n'est pas clairement établie. Le but de l'étude est de déterminer la valeur prédictive de la partie NT sérique du BNP (NT-proBNP) sur l'évolution post opératoire d'enfants porteur d'une cardiopathie congénitale et ayant eu une chirurgie cardiaque. Résultats: Nonante-sept enfants ont été inclus dans l'étude, avec un âge médian de 3.3 ans [0.7-5.2]. La valeur médiane du NT-proBNP préopératoire était de 412 pg/ml [164-1309]. Le NT-proBNP préopératoire était supérieur au P95 des valeurs de référence pour l'âge chez 56 patients (58%). Le NT-proBNP préopératoire était significativement plus élevé chez les patients ayant eu plus de deux jours de ventilation mécanique dans la période postopératoire (1156 pg/ml [281-1951] vs. 267 pg/ml [136-790], p=0.003) et ayant été hospitalisés plus de 6 jours dans l'unité de soins intensifs pédiatrique (727 pg/ml [203-1951] vs. 256 pg/ml [136-790], p=0.007). Par contre, le NT-proBNP préopératoire n'était pas significativement plus élevé chez les patients ayant eu un score d'inotrope élevé pendant leur hospitalisation aux soins intensifs, un temps de circulation extracorporelle prolongé ou ayant subi une chirurgie avec un risque chirurgical élevé. Conclusions: Un NT-proBNP sérique élevé en préopératoire reflète l'importance du stress myocardique induit par l'hémodynamique et la dysfonction myocardique, il est un marqueur qui permet d'améliorer l'identification des patients à risque d'avoir une évolution post opératoire compliquée.

Prenatal Risk Factors and Outcomes in Gastroschisis: A Meta-Analysis.

BACKGROUND AND OBJECTIVE: Gastroschisis is a congenital anomaly with increasing incidence, easy prenatal diagnosis and extremely variable postnatal outcomes. Our objective was to systematically review the evidence regarding the association between prenatal ultrasound signs (intraabdominal bowel dilatation [IABD], extraabdominal bowel dilatation, gastric dilatation [GD], bowel wall thickness, polyhydramnios, and small for gestational age) and perinatal outcomes in gastroschisis (bowel atresia, intra uterine death, neonatal death, time to full enteral feeding, length of total parenteral nutrition and length of in hospital stay). METHODS: Medline, Embase, and Cochrane databases were searched electronically. Studies exploring the association between antenatal ultrasound signs and outcomes in gastroschisis were considered suitable for inclusion. Two reviewers independently extracted relevant data regarding study characteristics and pregnancy outcome. All meta-analyses were computed using individual data random-effect logistic regression, with single study as the cluster unit. RESULTS: Twenty-six studies, including 2023 fetuses, were included. We found significant positive associations between IABD and bowel atresia (odds ratio [OR]: 5.48, 95% confidence interval [CI] 3.1-9.8), polyhydramnios and bowel atresia (OR: 3.76, 95% CI 1.7-8.3), and GD and neonatal death (OR: 5.58, 95% CI 1.3-24.1). No other ultrasound sign was significantly related to any other outcome. CONCLUSIONS: IABD, polyhydramnios, and GD can be used to an extent to identify a subgroup of neonates with a prenatal diagnosis of gastroschisis at higher risk to develop postnatal complications. Data are still inconclusive on the predictive ability of several signs combined, and large prospective studies are needed to improve the quality of prenatal counseling and the neonatal care for this condition.

Biomarkers of Host Response Predict Primary End-Point Radiological Pneumonia in Tanzanian Children with Clinical Pneumonia: A Prospective Cohort Study.

BACKGROUND: Diagnosing pediatric pneumonia is challenging in low-resource settings. The World Health Organization (WHO) has defined primary end-point radiological pneumonia for use in epidemiological and vaccine studies. However, radiography requires expertise and is often inaccessible. We hypothesized that plasma biomarkers of inflammation and endothelial activation may be useful surrogates for end-point pneumonia, and may provide insight into its biological significance. METHODS: We studied children with WHO-defined clinical pneumonia (n = 155) within a prospective cohort of 1,005 consecutive febrile children presenting to Tanzanian outpatient clinics. Based on x-ray findings, participants were categorized as primary end-point pneumonia (n = 30), other infiltrates (n = 31), or normal chest x-ray (n = 94). Plasma levels of 7 host response biomarkers at presentation were measured by ELISA. Associations between biomarker levels and radiological findings were assessed by Kruskal-Wallis test and multivariable logistic regression. Biomarker ability to predict radiological findings was evaluated using receiver operating characteristic curve analysis and Classification and Regression Tree analysis. RESULTS: Compared to children with normal x-ray, children with end-point pneumonia had significantly higher C-reactive protein, procalcitonin and Chitinase 3-like-1, while those with other infiltrates had elevated procalcitonin and von Willebrand Factor and decreased soluble Tie-2 and endoglin. Clinical variables were not predictive of radiological findings. Classification and Regression Tree analysis generated multi-marker models with improved performance over single markers for discriminating between groups. A model based on C-reactive protein and Chitinase 3-like-1 discriminated between end-point pneumonia and non-end-point pneumonia with 93.3% sensitivity (95% confidence interval 76.5-98.8), 80.8% specificity (72.6-87.1), positive likelihood ratio 4.9 (3.4-7.1), negative likelihood ratio 0.083 (0.022-0.32), and misclassification rate 0.20 (standard error 0.038). CONCLUSIONS: In Tanzanian children with WHO-defined clinical pneumonia, combinations of host biomarkers distinguished between end-point pneumonia, other infiltrates, and normal chest x-ray, whereas clinical variables did not. These findings generate pathophysiological hypotheses and may have potential research and clinical utility.

Single centre experience of the application of self navigated 3D whole heart cardiovascular magnetic resonance for the assessment of cardiac anatomy in congenital heart disease.

BACKGROUND: For free-breathing cardiovascular magnetic resonance (CMR), the self-navigation technique recently emerged, which is expected to deliver high-quality data with a high success rate. The purpose of this study was to test the hypothesis that self-navigated 3D-CMR enables the reliable assessment of cardiovascular anatomy in patients with congenital heart disease (CHD) and to define factors that affect image quality. METHODS: CHD patients ≥2 years-old and referred for CMR for initial assessment or for a follow-up study were included to undergo a free-breathing self-navigated 3D CMR at 1.5T. Performance criteria were: correct description of cardiac segmental anatomy, overall image quality, coronary artery visibility, and reproducibility of great vessels diameter measurements. Factors associated with insufficient image quality were identified using multivariate logistic regression. RESULTS: Self-navigated CMR was performed in 105 patients (55% male, 23 ± 12y). Correct segmental description was achieved in 93% and 96% for observer 1 and 2, respectively. Diagnostic quality was obtained in 90% of examinations, and it increased to 94% if contrast-enhanced. Left anterior descending, circumflex, and right coronary arteries were visualized in 93%, 87% and 98%, respectively. Younger age, higher heart rate, lower ejection fraction, and lack of contrast medium were independently associated with reduced image quality. However, a similar rate of diagnostic image quality was obtained in children and adults. CONCLUSION: In patients with CHD, self-navigated free-breathing CMR provides high-resolution 3D visualization of the heart and great vessels with excellent robustness.

Prediction of multiple infections after severe burn trauma: a prospective cohort study.

OBJECTIVE: To develop predictive models for early triage of burn patients based on hypersusceptibility to repeated infections. BACKGROUND: Infection remains a major cause of mortality and morbidity after severe trauma, demanding new strategies to combat infections. Models for infection prediction are lacking. METHODS: Secondary analysis of 459 burn patients (≥16 years old) with 20% or more total body surface area burns recruited from 6 US burn centers. We compared blood transcriptomes with a 180-hour cutoff on the injury-to-transcriptome interval of 47 patients (≤1 infection episode) to those of 66 hypersusceptible patients [multiple (≥2) infection episodes (MIE)]. We used LASSO regression to select biomarkers and multivariate logistic regression to built models, accuracy of which were assessed by area under receiver operating characteristic curve (AUROC) and cross-validation. RESULTS: Three predictive models were developed using covariates of (1) clinical characteristics; (2) expression profiles of 14 genomic probes; (3) combining (1) and (2). The genomic and clinical models were highly predictive of MIE status [AUROCGenomic = 0.946 (95% CI: 0.906-0.986); AUROCClinical = 0.864 (CI: 0.794-0.933); AUROCGenomic/AUROCClinical P = 0.044]. Combined model has an increased AUROCCombined of 0.967 (CI: 0.940-0.993) compared with the individual models (AUROCCombined/AUROCClinical P = 0.0069). Hypersusceptible patients show early alterations in immune-related signaling pathways, epigenetic modulation, and chromatin remodeling. CONCLUSIONS: Early triage of burn patients more susceptible to infections can be made using clinical characteristics and/or genomic signatures. Genomic signature suggests new insights into the pathophysiology of hypersusceptibility to infection may lead to novel potential therapeutic or prophylactic targets.

Hire me: Computational inference of hirability in employment interviews based on nonverbal behavior

Understanding the basis on which recruiters form hirability impressions for a job applicant is a key issue in organizational psychology and can be addressed as a social computing problem. We approach the problem from a face-to-face, nonverbal perspective where behavioral feature extraction and inference are automated. This paper presents a computational framework for the automatic prediction of hirability. To this end, we collected an audio-visual dataset of real job interviews where candidates were applying for a marketing job. We automatically extracted audio and visual behavioral cues related to both the applicant and the interviewer. We then evaluated several regression methods for the prediction of hirability scores and showed the feasibility of conducting such a task, with ridge regression explaining 36.2% of the variance. Feature groups were analyzed, and two main groups of behavioral cues were predictive of hirability: applicant audio features and interviewer visual cues, showing the predictive validity of cues related not only to the applicant, but also to the interviewer. As a last step, we analyzed the predictive validity of psychometric questionnaires often used in the personnel selection process, and found that these questionnaires were unable to predict hirability, suggesting that hirability impressions were formed based on the interaction during the interview rather than on questionnaire data.

CBS: an open platform that integrates predictive methods and epigenetics information to characterize conserved regulatory features in multiple Drosophila genomes.

Background: Information about the composition of regulatory regions is of great value for designing experiments to functionally characterize gene expression. The multiplicity of available applications to predict transcription factor binding sites in a particular locus contrasts with the substantial computational expertise that is demanded to manipulate them, which may constitute a potential barrier for the experimental community. Results: CBS (Conserved regulatory Binding Sites, http://compfly.bio.ub.es/CBS) is a public platform of evolutionarily conserved binding sites and enhancers predicted in multiple Drosophila genomes that is furnished with published chromatin signatures associated to transcriptionally active regions and other experimental sources of information. The rapid access to this novel body of knowledge through a user-friendly web interface enables non-expert users to identify the binding sequences available for any particular gene, transcription factor, or genome region. Conclusions: The CBS platform is a powerful resource that provides tools for data mining individual sequences and groups of co-expressed genes with epigenomics information to conduct regulatory screenings in Drosophila.

Effects of unit - based pricing on the waste collection demand : a meta - regression analysis

We perform a meta - analysis of 21 studies that estimate the elasticity of the price of waste collection demand upon waste quantities, a prior literature review having revealed that the price elasticity differs markedly. Based on a meta - regression with a total of 65 observations, we find no indication that municipal data give higher estimates for price elasticities than those associated with household data. Furthermore, there is no evidence that treating prices as exogenous underestimates the price elasticity. We find that much of the variation can be explained by sample size, the use of a weight - based as opposed to a volume - based pricing system, and the pricing of compostable waste. We also show that price elasticities determined in the USA and point estimations of elasticities are more elastic, but these effects are not robust to the changing of model specifications. Finally, our tests show that there is no evidence of publication bias while there is some evidence of the existence of genuine empirical effect.

Characterizing the vulnerability of frequent emergency department users by applying a conceptual framework: a controlled, cross-sectional study.

BACKGROUND: Frequent emergency department (ED) users meet several of the criteria of vulnerability, but this needs to be further examined taking into consideration all vulnerability's different dimensions. This study aimed to characterize frequent ED users and to define risk factors of frequent ED use within a universal health care coverage system, applying a conceptual framework of vulnerability. METHODS: A controlled, cross-sectional study comparing frequent ED users to a control group of non-frequent users was conducted at the Lausanne University Hospital, Switzerland. Frequent users were defined as patients with five or more visits to the ED in the previous 12 months. The two groups were compared using validated scales for each one of the five dimensions of an innovative conceptual framework: socio-demographic characteristics; somatic, mental, and risk-behavior indicators; and use of health care services. Independent t-tests, Wilcoxon rank-sum tests, Pearson's Chi-squared test and Fisher's exact test were used for the comparison. To examine the -related to vulnerability- risk factors for being a frequent ED user, univariate and multivariate logistic regression models were used. RESULTS: We compared 226 frequent users and 173 controls. Frequent users had more vulnerabilities in all five dimensions of the conceptual framework. They were younger, and more often immigrants from low/middle-income countries or unemployed, had more somatic and psychiatric comorbidities, were more often tobacco users, and had more primary care physician (PCP) visits. The most significant frequent ED use risk factors were a history of more than three hospital admissions in the previous 12 months (adj OR:23.2, 95%CI = 9.1-59.2), the absence of a PCP (adj OR:8.4, 95%CI = 2.1-32.7), living less than 5 km from an ED (adj OR:4.4, 95%CI = 2.1-9.0), and household income lower than USD 2,800/month (adj OR:4.3, 95%CI = 2.0-9.2). CONCLUSIONS: Frequent ED users within a universal health coverage system form a highly vulnerable population, when taking into account all five dimensions of a conceptual framework of vulnerability. The predictive factors identified could be useful in the early detection of future frequent users, in order to address their specific needs and decrease vulnerability, a key priority for health care policy makers. Application of the conceptual framework in future research is warranted.

Screening of mental health and substance users in frequent users of a general Swiss emergency department.

BACKGROUND: The objectives of this study were to determine the proportions of psychiatric and substance use disorders suffered by emergency departments' (EDs') frequent users compared to the mainstream ED population, to evaluate how effectively these disorders were diagnosed in both groups of patients by ED physicians, and to determine if these disorders were predictive of a frequent use of ED services. METHODS: This study is a cross-sectional study with concurrent and retrospective data collection. Between November 2009 and June 2010, patients' mental health and substance use disorders were identified prospectively in face-to-face research interviews using a screening questionnaire (i.e. researcher screening). These data were compared to the data obtained from a retrospective medical chart review performed in August 2011, searching for mental health and substance use disorders diagnosed by ED physicians and recorded in the patients' ED medical files (i.e. ED physician diagnosis). The sample consisted of 399 eligible adult patients (≥18 years old) admitted to the urban, general ED of a University Hospital. Among them, 389 patients completed the researcher screening. Two hundred and twenty frequent users defined by >4 ED visits in the previous twelve months were included and compared to 169 patients with ≤4 ED visits in the same period (control group). RESULTS: Researcher screening showed that ED frequent users were more likely than members of the control group to have an anxiety, depressive disorder, post-traumatic stress disorder (PTSD), or suffer from alcohol, illicit drug abuse/addiction. Reviewing the ED physician diagnosis, we found that the proportions of mental health and substance use disorders diagnosed by ED physicians were low both among ED frequent users and in the control group. Using multiple logistic regression analyses to predict frequent ED use, we found that ED patients who screened positive for psychiatric disorders only and those who screened positive for both psychiatric and substance use disorders were more likely to be ED frequent users compared to ED patients with no disorder. CONCLUSIONS: This study found high proportions of screened mental health and/or substance use disorders in ED frequent users, but it showed low rates of detection of such disorders in day-to-day ED activities which can be a cause for concern. Active screening for these disorders in this population, followed by an intervention and/or a referral for treatment by a case-management team may constitute a relevant intervention for integration into a general ED setting.

Vestibular Schwannoma: The Evolution of Hearing and Tumor Size in Natural Course and after Treatment by LINAC Stereotactic Radiosurgery.

OBJECTIVE: To review the natural course of tumor size and hearing during conservative management of 151 patients with unilateral vestibular schwannoma (VS), and to evaluate the same parameters for the part of the group (n = 84) who were treated by LINAC stereotactic radiosurgery (SRS). METHODS: In prospectively collected data, patients underwent MRI and complete audiovestibular tests at inclusion, during the conservative management period and after SRS. Hearing was graded according to the Gardner-Robertson (GR) scale and tumor size according to Koos. Statistics were performed using Kaplan-Meier survival analysis and multivariate analyses including linear and logistic regression. Specific insight was given to patients with serviceable hearing. RESULTS: During the conservative management period (mean follow-up time: 24 months, range: 6-96), the annual risk of GR class degradation was 6% for GRI and 15% for GR II patients. Hearing loss as an initial symptom was highly predictive of further hearing loss (p = 0.003). Tumor growth reached 25%. For SRS patients, functional hearing preservation was 51% at 1 year and 36% at 3 years. Tumor control was 94 and 91%, respectively. CONCLUSION: In VS patients, hearing loss at the time of diagnosis is a predictor of poorer hearing outcome. LINAC SRS is efficient for tumor control. Patients who preserved their pretreatment hearing presented less hearing loss per year after SRS than before treatment, suggesting a protective effect of SRS when cochlear function can be preserved.

The relation between internet use and overweight among adolescents: a longitudinal study in Switzerland.

OBJECTIVE: This longitudinal study aimed to investigate the characteristics and predictive risk factors of overweight among adolescents. The hypothesis was that baseline overweight predicted most overweight over time compared to other factors, especially excessive internet use. SUBJECTS: A sample of 621 youths were followed from age 14 (T0 Spring 2012) to age 16 (T1 Spring 2014) in Switzerland. Participants were divided into two groups according to their weight at the final assessment: overweight and non-overweight. At T0, participants reported demographic, health, substance use and internet use data. A logistic regression was performed to assess the explanatory variables of overweight at T1. Data are presented as adjusted odds ratios (aORs) with 95% confidence interval. RESULTS: The 2-year evolution showed a net BMI increase of 4.8%. Overweight adolescents were significantly more likely to be male, to live in an urban area, to be on a diet and to report using the internet more than 2 h per day on weekends at T0. However, with the addition of baseline overweight, only the excessive use of internet on weekends remained as an explanatory variable. An adolescent who was already overweight at T0 had a more than 20-fold risk (aOR 21.04) of being overweight 2 years later. Moreover, among adolescents becoming overweight between T0 and T1, internet use did not show any significant effect. CONCLUSION: The risk of being overweight is mostly influenced by weight status at baseline compared to excessive internet use. Thus, our results do not confirm the negative effect of internet on healthier activities. Internet use could at most reinforce an already existing risk of being overweight.