On the impact of disproportional samples in credit scoring models: An application to a Brazilian bank data

Statistical methods have been widely employed to assess the capabilities of credit scoring classification models in order to reduce the risk of wrong decisions when granting credit facilities to clients. The predictive quality of a classification model can be evaluated based on measures such as sensitivity, specificity, predictive values, accuracy, correlation coefficients and information theoretical measures, such as relative entropy and mutual information. In this paper we analyze the performance of a naive logistic regression model (Hosmer & Lemeshow, 1989) and a logistic regression with state-dependent sample selection model (Cramer, 2004) applied to simulated data. Also, as a case study, the methodology is illustrated on a data set extracted from a Brazilian bank portfolio. Our simulation results so far revealed that there is no statistically significant difference in terms of predictive capacity between the naive logistic regression models and the logistic regression with state-dependent sample selection models. However, there is strong difference between the distributions of the estimated default probabilities from these two statistical modeling techniques, with the naive logistic regression models always underestimating such probabilities, particularly in the presence of balanced samples. (C) 2012 Elsevier Ltd. All rights reserved.

An epidemiological scoring system for tooth wear and dental erosive wear

Background: The prevalence and severity of tooth wear and dental erosion is rising in children and there is no consensus about an index to be employed. Aim: To assess the reliability of an epidemiological scoring system dental wear index (DWI) to measure tooth wear and dental erosive wear. Design: An epidemiological cross-sectional survey was conducted to evaluate and compare tooth wear and dental erosion using the dental wear index and erosion wear index (EWI). The study was conducted with randomised samples of 2,371 children aged between 4 years and 12 years selected from the State of São Paulo, Brazil. Records were used for calculating tooth wear and dental erosion; the incisal edge and canine cusp were excluded. Results: As the schoolchildren's ages increased the severity of primary tooth wear increased in canines (P = 0.0001, OR = 0.34) and molars (P = 0.0001, OR = 2.47) and erosion wear increased in incisal/occlusal (P = 0.0001, OR = 5.18) and molars (P = 0.0001, OR = 2.47). There was an increased prevalence of wear in the permanent teeth of older schoolchildren, particularly on the incisal/occlusal surfaces (P = 0.0001, OR = 7.03). Conclusion: The prevalence of tooth wear and dental erosion increased as age increased in children. The epidemiological scoring system Dental Wear Index is able to measure both tooth wear and dental erosive wear. This index should be used to monitor the progression of non-carious lesions and to evaluate the levels of disease in the population.

Duodenal stenting for malignant gastric outlet obstruction: Prospective study

AIM: To evaluate the results of duodenal stenting for palliation of gastroduodenal malignant obstruction by using a gastric outlet obstruction score (GOOS). METHODS: A prospective, non-randomized study was performed at a tertiary center between August 2005 and April 2010. Patients were eligible if they had malignant gastric outlet obstruction (GOO) and were not candidates for surgical treatment. Medical history and patient demographics were collected at baseline. Scheduled interviews were made on the day of the procedure and 15, 30, 90 and 180 d later or unscheduled as necessary. RESULTS: Fifteen patients (6 male, 9 female; median age 61 years) with GOO who had undergone duodenal stenting were evaluated. Ten patients had metastasis at baseline (66.6%) and 14 were unable to accept oral intake (93.33%), including 7 patients who were using a feeding tube. Laboratory data showed biliary obstruction in eight cases (53.33%); all were submitted to biliary drainage. Two patients developed obstructive symptoms due to tumor ingrowth after 30 d and another due to tumor overgrowth after 180 d. Two cases of stent migration occurred. A good response to treatment was observed, with a mean time of approximately 1 d (19 h) until toleration of a liquid diet and slightly more than 2 d for both soft solids (51 h) and a solid food/normal diet (55 h). The mean time to first failure to maintain liquid intake (GODS >= 1) was 93 d. During follow-up, the mean time to first failure to maintain the previously achieved GODS of 2-3 (solid/semi-solid food), considered technical failure, was 71 d. On the basis of oral intake a GODS is defined: 0 for no oral intake; 1 for liquids only; 2 for soft solids only; 3 for low-residue or full diet. CONCLUSION: Enteral stenting to alleviate gastroduodenal malignant obstruction improves quality of life in patients with limited life expectancy, which can be evaluated by using a GOO scoring system. (C) 2012 Baishideng. All rights reserved.

A new long-term lifetime distribution induced by a latent complementary risk framework

In this paper, we proposed a new three-parameter long-term lifetime distribution induced by a latent complementary risk framework with decreasing, increasing and unimodal hazard function, the long-term complementary exponential geometric distribution. The new distribution arises from latent competing risk scenarios, where the lifetime associated scenario, with a particular risk, is not observable, rather we observe only the maximum lifetime value among all risks, and the presence of long-term survival. The properties of the proposed distribution are discussed, including its probability density function and explicit algebraic formulas for its reliability, hazard and quantile functions and order statistics. The parameter estimation is based on the usual maximum-likelihood approach. A simulation study assesses the performance of the estimation procedure. We compare the new distribution with its particular cases, as well as with the long-term Weibull distribution on three real data sets, observing its potential and competitiveness in comparison with some usual long-term lifetime distributions.

Is podoplanin expression associated with the proliferative activity of ameloblastomas?

Oral Diseases (2012) 18, 673679 Objectives: The aim of this study was to investigate the relationship between podoplanin expression and proliferative activity of ameloblastomas and remnants of the odontogenic epithelium from dental follicles (DF) of unerupted teeth. Subjects and methods: Thirty-three paraffin-embedded ameloblastomas and thirty-two DF obtained of unerupted teeth were analyzed by immunohistochemistry using anti-human podoplanin and anti-Ki-67 antibodies. Podoplanin expression in odontogenic epithelial cells was evaluated using a scoring method, and the Ki-67 labeling index was determined by the percentage of positive odontogenic cells. Results: All ameloblastomas displayed podoplanin expression in ameloblast-like cells of the epithelial islands. Membranous expression of podoplanin in ameloblastomas was stronger than in the remnants of odontogenic epithelium (P = 0.001). Statistically significant difference was observed between the cytoplasmic and membranous expression of podoplanin in the remnants of odontogenic epithelium (P = 0.001). The index of epithelial odontogenic proliferative activity, verified by Ki-67 expression, was higher in ameloblastomas vs remnants of odontogenic epithelium (P < 0.001). No statistically significant correlation was identified between podoplanin and the cellular odontogenic proliferative activity in ameloblastomas and DF (P > 0.05). Conclusions: These results provide evidence that there is no connection between podoplanin immunostaining and odontogenic cellular proliferative activity and suggest a role for membranous podoplanin expression in the local invasion of ameloblastomas.

Clinical Predictors of Different Grades of Nonalcoholic Fatty Liver Disease

Nonalcoholic fatty liver disease (NAFLD) is one of the comorbidities related to obesity. Liver biopsy has been used as the "gold standard" for the diagnosis, grading, and prognosis of obese patients. The objective of the present study was to evaluate clinical predictors of more advanced stages of NAFLD. In this retrospective study we assessed several physical and laboratorial factors, including some cytokines, in morbidly obese patients submitted to Roux-en-Y gastric bypass that could be related to the diagnosis and staging of NAFLD. Fragments of the livers were obtained from wedge biopsies during operation. The medical records of 259 patients were studied. The patients were divided into four groups: normal hepatic biopsy, steatosis, mild nonalcoholic steatohepatitis (NASH), and moderate and severe NASH. There were no differences in cytokine levels among groups. The triglyceride levels were the only variable that could stratify the grades of NAFLD and also differentiate from normal livers in the female patients. Also in this group, the aminotransferases and GGT levels and fasting glucose were predictors of the more advanced stages of NASH, while BMI and weight were predictors of the more advanced stages of NASH in male patients. There are no available markers in clinical practice to detect the initial stages of NAFLD. It is very important to perform a liver biopsy in all patients submitted to bariatric surgery and in obese patients with no indication to be operated in the presence of elevated blood levels of aminotransferases, GGT, and fasting glucose.

A Novel Method of Evaluating Ureteropelvic Junction Obstruction: Dynamic Near Infrared Fluorescence Imaging Compared to Standard Modalities to Assess Urinary Obstruction in a Swine Model

Purpose: Dynamic near infrared fluorescence imaging of the urinary tract provides a promising way to diagnose ureteropelvic junction obstruction. Initial studies demonstrated the ability to visualize urine flow and peristalsis in great detail. We analyzed the efficacy of near infrared imaging in evaluating ureteropelvic junction obstruction, renal involvement and the anatomical detail provided compared to conventional imaging modalities. Materials and Methods: Ten swine underwent partial or complete unilateral ureteral obstruction. Groups were survived for the short or the long term. Imaging was performed with mercaptoacetyltriglycine diuretic renogram, magnetic resonance urogram, excretory urogram, ultrasound and near infrared imaging. Scoring systems for ureteropelvic junction obstruction were developed for magnetic resonance urogram and near infrared imaging. Physicians and medical students graded ureteropelvic junction obstruction based on magnetic resonance urogram and near infrared imaging results. Results: Markers of vascular and urinary dynamics were quantitatively consistent among control renal units. The same markers were abnormal in obstructed renal units with significantly different times of renal phase peak, start of pelvic phase and start of renal uptake. Such parameters were consistent with those obtained with mercaptoacetyltriglycine diuretic renography. Near infrared imaging provided live imaging of urinary flow, which was helpful in identifying the area of obstruction for surgical planning. Physicians and medical students categorized the degree of obstruction appropriately for fluorescence imaging and magnetic resonance urogram. Conclusions: Near infrared imaging offers a feasible way to obtain live, dynamic images of urine flow and ureteral peristalsis. Qualitative and quantitative parameters were comparable to those of conventional imaging. Findings support fluorescence imaging as an accurate, easy to use method of diagnosing ureteropelvic junction obstruction.

Expression of calcium binding protein S100 A7 (psoriasin) in laryngeal carcinoma

Many studies have reported increased expression of S100 A7 (psoriasin) in neoplastic lesions. Among them are studies on breast carcinoma, bladder squamous cell carcinoma, skin tumors and oral cavity squamous cell carcinoma. The expression of S100 A7 has not been described for laryngeal cancer. Objective: This study aims to identify the expression of the calcium-binding protein S100 A7 and its correlation with squamous cell carcinomas of the larynx. Material and Methods: Specimens from 63 patients were submitted to immunohistochemistry testing with antibody S100 A7. Results were classified and compared. Results: The group with highly differentiated tumors had the highest treatment failure scores. Moderately differentiated tumors had higher treatment failure scores than poorly differentiated tumors. Higher scores were predominantly seen on stages I and II in moderately differentiated tumors, whereas score distribution was more homogeneous in advanced stage disease (III and IV). Regarding failure in treatment, the group scoring zero (3/4 complications: 75%) differed significantly from the remaining groups (13/59: 22%). Conclusions: S100 A7 marker was expressed in 93.7% of laryngeal cancer cases, with higher positive correlation rates in more differentiated tumors and significantly lower rates of treatment failure. Scores had no impact on survival rates.

A critical and descriptive approach to interictal behavior with the Neurobehavior Inventory (NBI)

Purpose: The purpose of this study was to test the psychometric properties of the Neurobehavior Inventory (NBI) in a group of temporal lobe epilepsy (TLE) patients from a tertiary care center, correlating its scores with the presence of psychiatric symptoms. Methods: Clinical and sociodemographic data from ninety-six TLE outpatients were collected, and a neuropsychiatric evaluation was performed with the following instruments: Mini-Mental State Examination (MMSE), structured psychiatric interview (MINI-PLUS), Neurobehavior Inventory (NBI), and Hamilton Depression Rating Scale (HAM-D). Results: Some traits evaluated by the NBI showed adequate internal consistency (mean inter-item correlation between 0.2 and 0.4) and were frequent, such as religiosity (74%) and repetitiveness (60.4%). Principal component analysis showed three factors, named here as emotions (Factor 1), hyposexuality (Factor 2), and unusual ideas (Factor 3). Depressive symptoms on HAM-D showed a strong association with emotions and hyposexuality factors. When patients with left TLE and right TLE were compared, the former exhibited more sadness (p=0.017), and the latter, a greater tendency toward sense of personal destiny (p=0.028). Conclusion: Depression influences NBI scoring, mainly emotionality and hyposexuality traits. Neurobehavior Inventory subscales can be better interpreted with an appropriate evaluation of comorbid mood and anxiety disorders. Compromise in left temporal mesial structures is associated with increased tendency toward sad affect, whereas right temporal pathology is associated with increased beliefs in personal destiny. (C) 2012 Elsevier Inc. All rights reserved.

Selection of Top 10 Quality Indicators for Nutrition Therapy

Background: The identification of useful quality indicators for nutrition therapy (QINTs) is of great interest and a challenge. This study attempted to identify the 10 QINTs that best suit the practice of quality control in nutrition therapy (NT) by evaluating the opinion of experts in NT with the use of psychometric techniques and statistical tools. Methods: Thirty-six QINTs available for clinical application in Brazil were assessed in 2 distinct phases. In phase 1, 26 nutrition experts ranked QINTs by scoring 4 attributes (utility, simplicity, objectivity, low cost) to assess each QINT on a 5-point Likert scale. The top 10 QINTs were identified from the 10 best scores obtained, and the reliability of expert opinion for each indicator was assessed by Cronbach's alpha. In phase 2, experts provided feedback regarding the selected top 10 QINTs by answering 2 closed questions. Results: The top 10 QINTs, in descending order, are the frequency of nutrition screening of hospitalized patients, diarrhea, involuntary withdrawal of enteral feeding tubes, feeding tube obstruction, fasting longer than 24 hours, glycemic dysfunction, estimated energy expenditure and protein needs, central venous catheter infection, compliance of NT indication, and frequency of application of subjective global assessment. Opinions were consistent among the interviewed experts. During feedback, 96% of experts were satisfied with the top 10 QINTs, and 100% had considered them in accordance with their previous opinion. Conclusion: The top 10 QINTs were identified according to their usefulness in clinical practice by obtaining adequate agreement and representativeness of opinion of nutrition experts. (Nutr Clin Pract. 2012;27:261-267)

The Truncated Inflated Beta Distribution

The study of proportions is a common topic in many fields of study. The standard beta distribution or the inflated beta distribution may be a reasonable choice to fit a proportion in most situations. However, they do not fit well variables that do not assume values in the open interval (0, c), 0 < c < 1. For these variables, the authors introduce the truncated inflated beta distribution (TBEINF). This proposed distribution is a mixture of the beta distribution bounded in the open interval (c, 1) and the trinomial distribution. The authors present the moments of the distribution, its scoring vector, and Fisher information matrix, and discuss estimation of its parameters. The properties of the suggested estimators are studied using Monte Carlo simulation. In addition, the authors present an application of the TBEINF distribution for unemployment insurance data.

Orthorexia nervosa behavior in a sample of Brazilian dietitians assessed by the Portuguese version of ORTO-15

BACKGROUND: Orthorexia nervosa (ON) is described as an obsessive pathological behavior characterized by a strong preoccupation with healthy eating and the avoidance of foods or ingredients considered unhealthy by the subject. Although it is still not officially recognized as an eating disorder, previous studies have discussed its frequency in some groups and a fifteen-question test (ORTO-15) was developed elsewhere to assess ON behavior. OBJECTIVE: The present study aimed to evaluate ON behavior in a sample of Brazilian dietitians after testing the psychometric properties of the Portuguese version of ORTO-15. METHODS: A total of 392 dietitians answered an online version of the test. The answers were analyzed regarding ON tendency, according with the scoring grid proposed by its authors. Exploratory factor analysis was performed and internal consistency was assessed. RESULTS: It was found that three questions of the test presented loadings lower than 0.5. The 12 remaining question formed 3 factors with internal consistency of -0.51, 0.63 and 0.47. The answers of the participants to these questions revealed a tendency to orthorexic behavior, mainly regarding aspects such as: making food choices conditioned by worry about health status, evaluating food rather from nutritional quality than from its taste, believing that consuming healthy food may improve appearance, discrediting the influence of mood on eating behavior and banning food choices considered by them as eating transgressions. CONCLUSION: There is no evidence of the validity and reliability of the ORTO-15 with the initial psychometric evaluation performed. Further analyses are needed. Nevertheless, it was possible to observe a high frequency of orthorexic behavior among the studied Brazilian dietitians. However, additional studies are needed to completely understand dietitians behavior toward ON. (Eat. Weight Disord. 17: e29-c35, 2012). (C) 2012, Editrice Kurtis

The role of pneumonia scores in the emergency room in patients infected by 2009 H1N1 infection

Despite the severity of pneumonia in patients with pandemic influenza A infection (H1N1), no validated risk scores associated with H1N1 pneumonia were tested. In this prospective observational study, we analyzed data of consecutive patients in our emergency room, hospitalized because of pneumonia between July and August 2009 in a public hospital in Brazil. The following pneumonia scoring systems were applied: the SMART-COP rule; the Pneumonia Severity Index; and the CURB-65 rule. Of 105 patients with pneumonia, 53 had H1N1 infection. Among them, only 9.5% that had a low risk according to SMART-COP were admitted to ICU, compared with 36.8% of those with the Pneumonia Severity Index score of 1-2 and 49% of those with CURB-65 score of 0-1. The SMART-COP had an accuracy of 83% to predict ICU admission. The SMART-COP rule presented the best performance to indicate ICU admission in patients with H1N1 pneumonia. European Journal of Emergency Medicine 19: 200-202 (C) 2012 Wolters Kluwer Health vertical bar Lippincott Williams & Wilkins.

Interphase Chromosome Flow-FISH

A 2-day method using flow cytometry and FISH for interphase cells was developed to detect monosomy 7 cells in myelodysplastic syndrome patients. The method, Interphase Chromosome Flow-FISH (IC Flow-FISH), involves fixation of leukocytes from blood, membrane permeabilization, hybridization of cellular DNA with peptide nucleic acid probes with cells intact, and analysis by flow cytometry. Hundreds to thousands of monosomy 7 cells were consistently detected from 10-20 mL of blood in patients with monosomy 7. Proportions of monosomy 7 cells detected in IC Flow-FISH were compared with results from conventional cytogenetics; identification of monosomy 7 populations was verified with FACS; and patient and donor cells were mixed to test for sensitivity. IC Flow-FISH allows for detecting monosomy 7 without requiring bone marrow procurement or the necessity of metaphase spreads, and wider applications to other chromosomal abnormalities are in development. (Blood. 2012; 120(15): e54-e59)

Decreasing the number of false positives in sequence classification

Abstract Background A large number of probabilistic models used in sequence analysis assign non-zero probability values to most input sequences. To decide when a given probability is sufficient the most common way is bayesian binary classification, where the probability of the model characterizing the sequence family of interest is compared to that of an alternative probability model. We can use as alternative model a null model. This is the scoring technique used by sequence analysis tools such as HMMER, SAM and INFERNAL. The most prevalent null models are position-independent residue distributions that include: the uniform distribution, genomic distribution, family-specific distribution and the target sequence distribution. This paper presents a study to evaluate the impact of the choice of a null model in the final result of classifications. In particular, we are interested in minimizing the number of false predictions in a classification. This is a crucial issue to reduce costs of biological validation. Results For all the tests, the target null model presented the lowest number of false positives, when using random sequences as a test. The study was performed in DNA sequences using GC content as the measure of content bias, but the results should be valid also for protein sequences. To broaden the application of the results, the study was performed using randomly generated sequences. Previous studies were performed on aminoacid sequences, using only one probabilistic model (HMM) and on a specific benchmark, and lack more general conclusions about the performance of null models. Finally, a benchmark test with P. falciparum confirmed these results. Conclusions Of the evaluated models the best suited for classification are the uniform model and the target model. However, the use of the uniform model presents a GC bias that can cause more false positives for candidate sequences with extreme compositional bias, a characteristic not described in previous studies. In these cases the target model is more dependable for biological validation due to its higher specificity.