Análisis bayesiano del maltrato infantil en adolescentes peruanos

Introducción: El maltrato infantil es un problema psicológico frecuente en niños y adolescentes de todo el mundo. La exposición a diversos tipos de maltrato puede conducir a otros problemas psicológicos y físicos, siendo necesario identificar qué tipo de maltrato repercute con mayor probabilidad en hombres y mujeres. Objetivo: Analizar las diferencias del maltrato infantil según sexo en adolescentes peruanos. Método: Estudio transversal, enfoque cuantitativo y comparativo realizado en los meses de julio y agosto de 2022. Se analizó una muestra por conveniencia de 1 376 adolescentes, los cuales respondieron un autoinforme sobre maltrato infantil. Para comparar ambos grupos, se utilizó la prueba U de Mann-Whitney de muestras independientes bayesianas. Resultados: Entre los diversos tipos de maltrato infantil, solo se encontró evidencia a favor de la hipótesis alternativa en el maltrato emocional (BF10=190,38; δ=0,31; IC 95 %: 0,20-0,48), con una evidencia muy fuerte. Esta diferencia fue superior en mujeres (M=8,28) en comparación con los hombres (M=7,09). En la negligencia emocional y física, se reportó evidencia fuerte para la hipótesis nula. El análisis robusto del factor Bayes confirmó los mismos resultados. Conclusiones: El maltrato emocional se presenta mayormente en mujeres adolescentes, siendo el género que requiere mayor apoyo psicológico en este tipo de maltrato infantil. En los otros tipos de maltrato, no se encontró diferencias respecto al sexo. En consecuencia, los profesionales de la salud deben considerar el rol del género en el maltrato emocional, pero no descuidar los otros tipos de maltrato infantil, enfatizando en las particularidades específicas de cada caso.

Introduction: Child maltreatment is a common psychological problem of children and adolescents around the world. Presence of various form of abuse can lead to other psychological and physical problems, so, it is necessary to identify which form of abuse is more likely to affect men and women. Objective: To analyze the differences observed in child maltreatment according to sex in Peruvian adolescents. Method: A cross-sectional, and quantitative and comparative approach study was carried out in the months of July and August 2022. A convenience sample of 1.376 adolescents who responded to a self-report on child abuse was analyzed. To compare both groups (men and women), the Mann-Whitney U-test for Bayesian independent samples was used. Results: Among the various forms of child maltreatment, only was found in favor and with high evidence as alternative hypothesis, the emotional maltreatment (BF10=190.38; δ=0.31; 95 % CI: 0.20-0.48). This difference was higher in females (M=8.28) compared to males (M=7.09). In emotional and physical neglect, strong evidence was reported for the null hypothesis. Robust Bayes factor analysis confirmed the same results. Conclusions: Emotional maltreatment occurs mostly in adolescent females, being the gender that requires more psychological support. In the other forms of maltreatment, no differences were found in correspondence with gender. Consequently, health professionals should consider the role of gender in emotional maltreatment, but should not neglect the other forms of child maltreatment, emphasizing the specific particularities of each case.

Introdução: O abuso infantil é um problema psicológico frequente em crianças e adolescentes em todo o mundo. A exposição a vários tipos de abuso pode levar a outros problemas psicológicos e físicos, tornando-se necessário identificar qual tipo de abuso tem maior probabilidade de afetar homens e mulheres. Objetivo: Analisar as diferenças no abuso infantil de acordo com o sexo em adolescentes peruanos. Método: Estudo transversal, abordagem quantitativa e comparativa realizado nos meses de julho e agosto de 2022. Foi analisada uma amostra de conveniência de 1.376 adolescentes, que responderam a um autorrelato sobre abuso infantil. Para comparar os dois grupos, foi utilizado o teste U de Mann-Whitney para amostras independentes bayesianas. Resultados: Entre os vários tipos de abuso infantil, foram encontradas evidências a favor da hipótese alternativa apenas no abuso emocional (BF10=190,38; δ=0,31; IC 95%: 0,20-0,48), com evidências muito fortes. Essa diferença foi maior nas mulheres (M=8,28) em relação aos homens (M=7,09). Na negligência emocional e física, fortes evidências foram relatadas para a hipótese nula. A análise fatorial robusta de Bayes confirmou os mesmos resultados. Conclusões: O abuso emocional ocorre maioritariamente em mulheres adolescentes, sendo o género que mais requer apoio psicológico neste tipo de abuso infantil. Nos outros tipos de abuso, não foram encontradas diferenças quanto ao sexo. Consequentemente, os profissionais de saúde devem considerar o papel do gênero no abuso emocional, mas não negligenciar outros tipos de abuso infantil, enfatizando as especificidades de cada caso.

Added value of PET Bayesian penalized likelihood reconstruction algorithm in the diagnosis of solitary pulmonary nodules/masses / 中华核医学与分子影像杂志

Objective:To investigate the effects of silicon photomutipliers (SiPM) detector and Bayesian penalized likelihood (BPL) reconstruction algorithm on semiquantitative parameters of 18F-FDG PET/CT and diagnostic efficiency for solitary pulmonary nodules/masses compared with traditional photomultiplier tube (PMT) and ordered subsets expectation maximization (OSEM). Methods:From March 2020 to January 2022, 118 patients (76 males, 42 females, age (63.0±10.1) years) newly diagnosed with solitary pulmonary nodules/masses in First Hospital of Shanxi Medical University were prospectively enrolled and underwent 18F-FDG PET/CT imaging with two different PET/CT scanners successively. The images were divided into PMT+ OSEM, SiPM+ OSEM and SiPM+ BPL groups according to PET detector and reconstruction algorithms. The SUV max, SUV mean, metabolic tumor volume (MTV) and total lesion glycolysis (TLG) of pulmonary nodules/masses were measured, then signal-to-noise ratio (SNR) and signal-to-background ratio (SBR) were calculated. One-way analysis of variance and Kruskal-Wallis rank sum test were performed to compare differences of above parameters among groups. ROC curve analysis was used to analyze the optimal threshold of SUV max for the differential diagnosis of pulmonary nodules/masses and AUCs were obtained. Results:There were 83 malignant nodules and 35 benign nodules. The image quality of SiPM+ BPL group (4.23±0.64) was better than that of SiPM+ OSEM group (3.57±0.50) or PMT+ OSEM group (3.58±0.51; F=54.85, P<0.001). There were significant differences in SUV max (7.57(3.86, 15.61) vs 4.95(2.22, 10.48)), SUV mean (4.43(2.28, 9.12) vs 2.84(1.21, 5.71)), MTV (3.54(1.57, 7.67) vs 5.09(2.83, 11.79)), SNR (28.12(12.55, 54.38) vs 20.16(8.29, 41.45)) and SBR (4.03(1.83, 7.75) vs 2.32(0.96, 5.03)) between SiPM+ BPL and SiPM+ OSEM groups ( H values: 16.63-37.05, all P<0.001). The optimal threshold values of SUV max in SiPM+ BPL, SiPM+ OSEM and PMT+ OSEM were 3.31, 2.21, 2.05 with AUCs of 0.686, 0.689, 0.615 for nodules < 2 cm, and were 10.29, 6.49, 4.33 with AUCs of 0.775, 0.782, 0.774 for nodules/masses ≥2 cm. Conclusions:Image quality and parameters of pulmonary nodules/masses are mainly affected by the reconstruction algorithms. BPL can improve SUV max, SUV mean, SBR and SNR, but reduce MTV without significant effect on liver parameters. SiPM+ BPL has a higher diagnostic threshold of SUV max than SiPM+ OSEM and PMT+ OSEM.

Dedução, Indução e a Arte do Raciocínio Clínico na Educação Médica: Revisão Sistemática e Proposta Bayesiana / Deduction, Induction and the Art of Clinical Reasoning in Medical Education: Systematic Review and Bayesian Proposal

Resumo Fundamento O raciocínio clínico está no centro da prática médica e emaranhado em uma confusão conceitual.A teoria da dualidade da probabilidade permite analisar seus aspectos objetivos e subjetivos. Objetivos Fazer revisão sistemática da literatura sobre o raciocínio clínico para tomada de decisão na educação médica e uma proposta chamada "Pensamento Conforme a Regra de Bayes" (PCRB). Métodos A revisão sistemática foi realizada na base PubMed até a data de 27/02/2022, seguindo metodologia rigorosa, por pesquisador experiente em revisão sistemática. A proposta PCRB, apresentada na discussão, foi elaborada no trabalho de conclusão de graduação em Filosofia na Universidade Federal de Minas Gerais. Usou-se a insuficiência cardíaca como exemplo. Resultados De 3340 artigos encontrados, incluíram-se 154 artigos: 24 tratando da condição de incerteza; 87 tratando de conceitos vagos (discussão de casos, heurística, lista de vieses cognitivos, escolha com sabedoria) incluídos sob o termo 'arte'; e 43 discutindo a ideia geral de raciocínio indutivo ou dedutivo. PCRB oferece regras de coerência e reprodutibilidade, inferência sob incerteza e regra de aprendizado, e pode, por meio da perspectiva subjetiva sobre a probabilidade, incorporar aqueles termos vagos classificados como 'arte', bem como argumentos e evidências. Conclusões A revisão sistemática mostra que o raciocínio é fundado na incerteza, predominantemente probabilístico; além de mostrar algumas possibilidades de erro do pensamento hipotético-dedutivo. O PCRB é um pensamento probabilístico em duas etapas que pode ser ensinado. A regra de Bayes é uma ferramenta linguística, uma regra geral de raciocínio, de diagnóstico, de comunicação científica e de revisão do conhecimento médico conforme novas evidências.

Abstract Background Clinical reasoning is at the core of medical practice and entangled in a conceptual confusion. The duality theory in probability allows to evaluate its objective and subjective aspects. Objectives To conduct a systematic review of the literature about clinical reasoning in decision making in medical education and to propose a "reasoning based on the Bayesian rule" (RBBR). Methods A systematic review on PubMed was conducted (until February 27, 2022), following a strict methodology, by a researcher experienced in systematic review. The RBBR, presented in the discussion section, was constructed in his undergraduate dissertation in Philosophy at Minas Gerais Federal University. Heart failure was used as example. Results Of 3,340 articles retrieved, 154 were included: 24 discussing the uncertainty condition, 87 on vague concepts (case discussion, heuristics, list of cognitive biases, choosing wisely) subsumed under the term "art", and 43 discussing the general idea of inductive or deductive reasoning. RBBR provides coherence and reproducibility rules, inference under uncertainty, and learning rule, and can incorporate those vague terms classified as "art", arguments and evidence, from a subjective perspective about probability. Conclusions This systematic review shows that reasoning is grounded in uncertainty, predominantly probabilistic, and reviews possible errors of the hypothetico-deductive reasoning. RBBR is a two-step probabilistic reasoning that can be taught. The Bayes theorem is a linguistic tool, a general rule of reasoning, diagnosis, scientific communication and review of medical knowledge according to new evidence.

Estrategia de predicción pacientes con casos de coronavirus; reporte de experiencia Pereira, Risaralda, marzo-abril 2020 / Prediction strategy for patients with coronavirus cases; experience report Pereira, Risaralda, March- April 2020

Resumen Introducción: la infección por COVID 19 corresponde actualmente al evento infeccioso con mayor impacto en salud púbica a nivel mundial, en Colombia, al 30 de abril de 2020 se registraron 6465 casos acumulados, 360 defunciones y 2186 casos recuperados, dado el aumento en los casos reportados mediante los sistemas de vigilancia epidemiológica se precisa de herramientas que faciliten el diagnóstico oportuno y la predicción en el comportamiento de los casos a nivel nacional. Objetivos: proponer un modelo estadístico que permita predecir la probabilidad de cursar con diagnóstico de COVID-19 en la población atendida por sospecha de infección por el mismo en una institución de tercer nivel del municipio de Pereira- Risaralda entre marzo y abril de 2020. Materiales y métodos: se presenta un estudio descriptivo de corte trasversal en el cual se analizaron 82 casos, se realizó un modelo predictivo basado en compuertas lógicas AND y OR, y análisis por estadística descriptiva e inferencial. Resultados: de los 82 registros analizados se encontró una relación hombre: mujer de 1:2; el 6% de los pacientes tuvo alta probabilidad para diagnóstico de COVID 19, el 20% tuvo probabilidad intermedia y el 72% registró baja probabilidad para COVID19, la concordancia del modelo con los resultados de las pruebas fue inferior a 0,5. Conclusiones: el modelo estadístico planteado fue insuficiente para lograr la predicción de la totalidad de los casos de COVID-19 basados en el perfil de riego de la población, se precisan nuevas investigaciones con tamaños de muestra superiores, diseños y análisis distintos. MÉD.UIS.2022;35(1): 57-69.

Abstract Introduction: COVID 19 infection currently corresponds to the infectious event with the greatest impact on public health worldwide, in Colombia, as of April 30, 2020, 6465 accumulated cases, 360 deaths and 2186 recovered cases were registered, given the increase in cases reported through epidemiological surveillance systems, tools are needed to facilitate timely diagnosis and prediction in the behavior of cases at the national level. Objectives: to propose a statistical model that allows predicting the probability of a diagnosis of COVID-19 in the population treated for suspected coronavirus infection in a third-level institution in the population of Pereira-Risaralda between March and April 2020. Materials and methods: a descriptive cross-sectional study is presented, in which 82 cases were analyzed, a predictive model based on AND and OR logic gates, analyzes by descriptive and inferential statistics were performed. Results: of the 82 records analyzed, a male: female ratio of 1: 2 was found; 6% of the patients had a high probability for the diagnosis of COVID 19, 20% had an intermediate probability and 72% had a low probability for COVID19, the agreement of the model with the test results was less than 0.5. Conclusions: the proposed statistical model was insufficient to achieve the prediction of all the cases of COVID-19 based on the irrigation profile of the population. New investigations are required with larger sample sizes associated with longitudinal designs and combined statistical analyzes that allow to refine the proposed model. MÉD.UIS.2022;35(1): 57-69.

El Enfoque Bayesiano en las Ciencias de la Salud / The Bayesian Approach in the Health Sciences

En estadística existen dos enfoques básicos, la estadística frecuentista que es la corriente principal y la estadística bayesiana. La mayoría de los principales métodos estadísticos son frecuentistas siendo el enfoque bayesiano más desconocido entre los investigadores. En el presente artículo se exponen los fundamentos lógicos del enfoque bayesiano y su uso mediante un ejemplo de aplicación. En este contexto, más que presentar un debate entre la lógica clásica y la bayesiana, se pretende mostrar de manera introductoria las enormes posibilidades que el enfoque bayesiano puede aportar a la investigación en las Ciencias de la Salud.

In the stadistic field there are two basic approaches, the Frequentist Statistics which is the primary one, and the Bayesian Statistics. The most used statistical methods are the Frequentist methods, being the Bayesian approach the most popular among researchers. In this article, the logical basis of the Bayendian approach and its use are exposed through an application example. In this context, rather than presenting a debate between the classic and the Bayensian logic, it is intended to demonstrate in an introductory method the considerable possibilities how Bayesian approach can contribute to Health and Sciences research.

Clinical efficacy and safety of ex-vivo liver resection and liver auto-transplantation: a bayesian single-arm Meta-analysis / 中华肝胆外科杂志

Objective:To evaluate the clinical efficacy and safety of ex vivo liver resection and autotransplantation (ELRA) by using a Bayesian single-arm Meta-analysis.Methods:Databases of PubMed, Embase, Cochrane Library, Web of Science, CNKI, and Wanfang were searched from January 1, 1990 to December 30, 2021 on ELRA studies. The Bayesian one-arm Meta-analysis was performed by using the statistical software of R (V4.1.2) and the Markov chain-Monte Carlo method was used to simulate the posterior distribution. The mortality rate within 30 days after operation, 1-year survival rate, major postoperative complications, R 0 resection rate and other related indexes were analyzed. Results:A total of 20 studies with 436 patients were included. Bayesian single-arm Meta-analysis showed that the 1-year survival rate after ELRA was 83.24% [95% highest posterior density ( HPD): 72.40%-92.05%]. The 1-year survival rates after surgery were 88.66% (95% HPD: 81.52%-94.50%) for patients with hepatic alveolar echinococcosis and 61.29% (95% HPD: 38.53%-93.68%) for patients with hepatic malignancies, respectively. The mortality rate within 30 d after surgery, the incidence of significant postoperative complications, and the R 0 resection rate were 6.96% (95% HPD: 4.47%-10.15%), 27.91% (95% HPD: 19.00%-38.30%), and 99.84% (95% HPD: 37.61%-100.00%), respectively. Renal failure was the most frequent cause of death after ELRA. Conclusion:ELRA is indicated for hepatic malignancies and hepatic alveolar echinococcosis when intrahepatic resection cannot be accomplished in vivo. The greatest benefit is observed in patients with hepatic alveolar echinococcosis, while only some patients with hepatic malignancies can benefit. The indications for ELRA for hepatic malignancies need to be further studied to define the subgroup of patients who can benefit from this operation.

Prevalencia bayesiana de Staphylococcus aureus en vacas lecheras en el Valle del Cauca, Colombia / Bayesian prevalence of Staphylococcus aureus in Valle del Cauca, Colombia dairy cows / Prevalência bayesiana de Staphylococcus aureus em vacas leiteiras no Vale do Cauca, Colômbia

Resumen El Staphylococcus aureus es uno de los principales microorganismos causantes de mastitis subclínica en las vacas lecheras. Determinar la prevalencia del S. aureus se hace necesario en vacas de hatos lecheros grandes y pequeños ubicados en la región norte, centro y sur del departamento del Valle del Cauca, Colombia. Se utilizó la metodología del National Mastitis Council recolectando 1070 muestras de leche de 273 vacas en 17 hatos mediante un muestreo polietápico, durante el periódo 2015 - 2016. Se realizó cultivo con todas las muestras recolectadas independientemente del resultado de la prueba de California mastitis test. Se utilizó el análisis bayesiano de los datos para obtener estimaciones más precisas. La prevalencia bayesiana del patógeno S. aureus en vacas lecheras en el Valle del Cauca fue de 31,94% y se tiene un 95% de probabilidad de que el intervalo (30,03 - 37,62%) contenga el valor de dicha prevalencia. La región Norte, Centro y Sur evidenciaron prevalencias de 33%, 34% y 24% respectivamente. En cuanto al tamaño del hato se evidenció que las vacas de los hatos pequeños presentan mayor prevalencia con 47% que las vacas de los hatos grandes 17% a la presencia del patógeno. La alta prevalencia encontrada en las vacas genera un alto impacto en la salud pública debido a una posible contaminación zoonótica por este patógeno.

Abstract Staphylococcus aureus is one of the main microorganisms responsible of subclinical mastitis in dairy cows. To determine the bayesian prevalence of S. aureus of cows in large and small dairy farms located in the northern, central and southern region of the Valle del Cauca province in Colombia. We Followed the National Mastitis Council methodology, 1070 milk samples from 273 cows in 17 dairy farms by means of a multi-stage sampling, were collected during the period 2015-2016. Cultivation was carried out with all the samples. Bayesian inference was used. The prevalence of the pathogen S. aureus in dairy cows in Valle del Cauca is 31,94% and there is a 95% probability that the interval (30,03 - 37,62%). The North, Central and South regions showed prevalence values of 33%, 34% and 24% respectively. Small dairy farms have prevalence, 47%, large dairy farms 17%. The high prevalence found in the cows studied generates a high impact on public health due to possible zoonotic contamination by this pathogen.

Resumo O Staphylococcus aureus é um dos principais microrganismos causadores de mastite subclínica em vacas leiteiras. Determinar a prevalência de S. aureus em vacas de leite em pequenas e grandes rebanhos leiteiros localizados região no norte, centro e sul do estado do Valle del Cauca, Colômbia. A metodologia do Conselho Nacional de Mastite 1070 foi utilizado para coletar amostras de leite de 273 vacas em 17 rebanhos através de amostragem de vários estágios durante o período 2015 - 2016. O cultivo foi realizado com todas as amostras coletadas, independentemente do resultado do teste de mastite Califórnia. O análise Bayesiana dos dados foi utilizada para obter estimativas mais precisas. Bayesian prevalência de agentes patogénicos S. aureus em vacas leiteiras no Valle foi 31,94% e tem uma probabilidade de 95% com um intervalo 30,03% - 37,62%. As regiões Norte, Central e Sul apresentaram prevalências de 33%, 34% e 24%, respectivamente. Em relação ao tamanho do rebanho, ficou evidente que as vacas dos pequenos rebanhos apresentam maior prevalência do patogeno com 47% comparado com 17% nas vacas dos grandes rebanhos. A alta prevalência encontrada nas vacas estudadas gera alto impacto na saúde pública, agravado pelo fato de metade do leite consumido na Colômbia ser cru.

Uso del factor Bayes durante el análisis estadístico: un ejemplo / Use of the Bayes factor during statistical analysis: an example

El factor de Bayes resulta una prueba recomendable para la comprobación de las hipótesis esta-dísticas atendiendo al estado de los p valores, empleando la escala de clasificación de Jeffreys preferiblemente

The Bayes factor is a recommended test for the verification of statistical hypotheses taking into account the state of the p values, preferably using the Jeffreys classification scale.

Análise dos padrões espaciais e caracterização dos suicídios no Brasil entre 1990 e 2015 / Analysis of spatial patterns and characterization of suicides in Brazil from 1990 to 2015 / Análisis de los patrones espaciales y caracterización de los suicidios en Brasil entre 1990 y 2015

Resumo: O objetivo deste estudo foi analisar o padrão espacial das taxas de mortalidade por suicídio no Brasil entre 1990 e 2015. Realizou-se análise espacial das taxas de mortalidade por suicídio dos municípios, em triênios, por meio de inferência bayesiana e análise de clusters, segundo risco de óbito ajustado por sexo e faixa etária. A taxa de mortalidade por suicídio aumentou de 3,5 óbitos em 1990 para 5,3 óbitos/100 mil habitantes em 2015 quando, a cada 64 minutos, uma morte foi registrada. Houve predominância de óbitos masculinos em todos os períodos, e as variáveis idade e raça/cor apresentaram alterações. Alta taxa de mortalidade entre jovens e indígenas foi observada no último triênio. Observamos perda de áreas sem notificação, bem como redução de áreas com taxas baixas de mortalidade em oposição ao aumento de áreas com taxas médias em todas as regiões. As taxas altas, que se concentravam no Sul, se dispersaram para outras regiões. Taxas muito altas surgiram no sul de Mato Grosso do Sul. Clusters de maior verossimilhança em todos os períodos são observados no Sul. As taxas de mortalidade por suicídio aumentaram, com alteração no padrão espacial no período estudado. Apesar da concentração inicial das altas taxas de mortalidade no Sul, houve dispersão dos óbitos para as demais regiões com diferentes magnitudes. Por um lado, houve redução dos óbitos entre idosos no Sul; por outro lado, houve aumento de óbitos no Nordeste, de adultos e pessoas negras no Centro-oeste e entre jovens e indígenas no noroeste do Amazonas.

Abstract: This study aimed to analyze the spatial pattern in mortality rates from suicide in Brazil from 1990 to 2015. A spatial analysis was performed for the mortality rates from suicide in municipalities (counties) in three-year periods using Bayesian inference and clusters according to risk of death adjusted by sex and age bracket. The mortality rate from suicide increased from 3.5 deaths in 1990 to 5.3 deaths/100,000 inhabitants in 2015, when there was one death from suicide every 64 minutes. There was a predominance of suicide deaths in males in all the periods, while the variables age and race/color showed some changes. High suicide mortality rates among youth and indigenous people were seen in the most recent three-year period. We observed a loss of areas without suicide reporting, a reduction in areas with low suicide mortality rates, as opposed to an increase in areas with medium rates in all the regions of Brazil. The high rates, previously concentrated in the South of Brazil, had expanded to other regions of the country. Very high rates emerged in southern Mato Grosso do Sul state. Clusters with higher likelihood in all the periods were observed in the South. Suicide mortality rates increased, with changes in the spatial during the 25-year period. Despite the initial concentration of high suicide mortality rates in the South, the deaths spread to other regions of the country with different magnitudes. There was a reduction in suicide deaths among the elderly in the South and an increase in the Northeast, and an increase in deaths in adults and black Brazilians in the Central and among youth and indigenous persons in northwestern Amazonas state.

Resumen: El objetivo de este estudio fue analizar el patrón espacial de la tasa de mortalidad por suicidio en Brasil entre 1990 y 2015. Se realizó un análisis espacial de las tasas de mortalidad por suicidio en municipios durante trienios, mediante inferencia bayesiana y análisis de clústeres, según el riesgo de fallecimiento, ajustado por sexo y franja de edad. La tasa de mortalidad por suicidio aumentó de 3,5 óbitos en 1990 a 5,3 óbitos/100 mil habitantes en 2015, cuando cada 64 minutos se registró una muerte. Hubo predominancia de fallecimientos masculinos en todos los períodos, mientras que las variables edad y raza/color presentaron alteraciones. Una alta tasa de mortalidad entre jóvenes e indígenas se observó en el último trienio. Observamos pérdida de áreas sin notificación, reducción de áreas con tasas bajas de mortalidad, frente al aumento de áreas con tasas medias en todas las regiones. Las tasas altas, que se concentraban en el Sur, se dispersaron hacia otras regiones. Aparecieron tasas muy altas en el sur de Mato Grosso do Sul. Los clústeres con mayor verosimilitud durante todos los períodos se observan en el Sur. Las tasas de mortalidad por suicidio aumentaron, con alteración en el patrón espacial durante el período estudiado. A pesar de la concentración inicial de las altas tasas de mortalidad en el Sur, hubo una dispersión de los fallecimientos hacia las demás regiones con diferentes magnitudes. Hubo una reducción de los óbitos entre ancianos en el Sur y un aumento en el Nordeste, así como un aumento de óbitos de adultos y personas negras en el Centro-oeste, y entre jóvenes e indígenas en el noroeste del Amazonas.

Enfermedades crónicas no transmisibles prioritarias en la Región Caribe Colombiana en el periodo 2008-2015: un análisis de la mortalidad / Chronic noncommunicable diseases priority in the Colombian Caribbean Region in the period 2008-2015: An analysis of mortality

RESUMEN Objetivo: Estimar el riesgo de muerte por enfermedades crónicas no transmisibles prioritarias en la región Caribe colombiana en el periodo 2008-2015. Materiales y métodos: Mediante un estudio ecológico se analizaron los datos de mortalidad disponibles en el Departamento Administrativo Nacional de Estadística (DANE). Con el ajuste de las tasas de mortalidad mediante razones estandarizadas de mortalidad (REM) por edad y sexo y mediante un modelo bayesiano, se estimó el riesgo suavizado de morir por causa de las enfermedades priorizadas. Se valoró su evolución temporal con tasas ajustadas anuales de mortalidad. Resultados: De 2008 a 2015 ocurrieron 148.331 muertes, de las cuales 76 201 (51,4 %) ocurrieron en hombres. El 58.1 % (86 185 muertes) corresponden a trastornos en el sistema circulatorio, seguido de los tumores malignos con un 24.4 % (36 188 muertes). Las enfermedades isquémicas del corazón ocupan el primer lugar en las causas de muerte, con un aumento significativo (p<0,001) del riesgo a partir de 2011, al igual que la tendencia en neoplasias de próstata (p<0,001) y mama en mujeres (p=0,022). Conclusión: La mortalidad debida a las enfermedades crónicas no transmisibles estudiadas aumentó en la región en el período 2008-2015. Con una mayor tendencia en el riesgo en los hombres para la mayoría de las enfermedades, lo que genera información relevante para que los tomadores de decisiones en salud ajusten los programas y servicios de prevención de la enfermedad, promoción, atención y rehabilitación en salud acorde a la realidad del ente territorial.

ABSTRACT Objective: To estimate the risk of mortality due to Chroic noncommunicable diseases priority in the Colombian Caribbean region using Bayesian methods for the 2008-2015 period. Instruments and methods: Through an ecological study, the mortality data available in the National Administrative Department of Statistics (DANE) database was analyzed. Adjusting the mortality rates using standardized mortality ratios (REM) by age and gender and using a Bayesian model, the smoothed risk of dying due to prioritized diseases was estimated. Its temporal evolution was assessed with annual adjusted mortality rates. Results: For the 2008 - 2015 period, 148,331 people died, 76 201 (51.4 %) of which were men. 58.1 % (86 185 deaths) correspond to disorders in the circulatory system, followed by malignant tumors with 24.4 % (36 188 deaths). Ischemic heart diseases occupy the first place in the causes of death with a significant increase (p<0,001) of risk as of 2011, likewise for prostate neoplasms (p<0,001) and breast neoplasms in women (p=0,022). Conclusion: Mortality due to the observed chronic noncommunicable diseases increased in the region during the 2008-2015 period. There is a greater tendency of risk for men for the majority of the diseases which generates relevant information so that the decisión makers in healthcare adjust the programs and services of disease prevention, promotion, attention and rehabilitation according to the reality of the territorial entity.

Desarrollo de algoritmos clínicos para el diagnóstico del dengue en Colombia / Development of clinical algorithms for the diagnosis of dengue in Colombia

Resumen Introducción. Dado el aumento de la incidencia y la mortalidad por dengue, su diagnóstico es relevante para los países endémicos. Las clasificaciones clínicas y las pruebas de laboratorio existentes tienen un desempeño variable en la práctica clínica, pues su sensibilidad fluctúa entre 45 y 98 %, y su especificidad, entre 4 y 98 %, lo cual se debe, en parte, a la diversidad de contextos en los que se utilizan. Objetivo. Desarrollar algoritmos clínicos para el diagnóstico del dengue en el contexto colombiano. Materiales y métodos. Se hizo un estudio transversal a partir de fuentes secundarias. Se construyeron algoritmos clínicos de diagnóstico del dengue con base en métodos bayesianos que combinaron síntomas, signos y parámetros del hemograma, y se comparó su exactitud diagnóstica con la de las pruebas de referencia. Se hizo una validación externa del algoritmo de mayor exactitud y sensibilidad, comparándolo con la clasificación clínica de la Organización Mundial de la Salud de 1997 y la del 2009, con la guía colombiana del 2010 y con la escala diagnóstica propuesta por el Ministerio de Salud y Protección Social de Colombia en el 2013. Resultados. Se generaron cuatro algoritmos, dos de signos y síntomas y dos que incluyeron el conteo de leucocitos (≤4.500/mm3) o de plaquetas (≤60.000/mm3). El algoritmo de mayor exactitud incluyó los parámetros del hemograma, con una sensibilidad de 76,5 % (IC95% 71,9-80,5) y una especificidad de 46,0 % (IC95% 37,6-54,7). En la validación externa, la sensibilidad fue de 11,1 % (IC95% 4,9-20,7) y la especificidad fue de 91,9 % (IC95% 87,5-93,9). La escala del Ministerio tuvo una sensibilidad de 76,4 % (IC95% 64,9-85,6) y una especificidad de 38,0 % (IC95% 32,8-43,4). Conclusión. La inclusión de los parámetros del hemograma mejoró la sensibilidad de los algoritmos de diagnóstico basados en los signos y los síntomas. Sin embargo, el diagnóstico clínico del dengue sigue siendo un reto para la investigación en salud.

Abstract Introduction: Due to the increase in dengue incidence and mortality, its diagnosis is relevant for endemic countries. Clinical classifications and laboratory tests have a variable performance in clinical practice with a sensitivity level between 45% and 98%, and a specificity level between 4% and 98% partly due to the variety of contexts where they are applied. Objective: To develop clinical algorithms for the diagnosis of dengue in the Colombian context. Materials and methods: A cross-sectional study was conducted based on secondary sources. We constructed clinical diagnostic algorithms of dengue based on Bayesian methods combining symptoms, signs, and blood count parameters, and then we compared them in terms of diagnostic accuracy with gold standard tests. In addition, an external validation of the algorithm with greater accuracy and sensibility was performed comparing it with the WHO-1997 and the WHO-2009 clinical classifications, the Colombian guide for 2010, and the diagnostic scale recommended by the Ministerio de Salud y Protección Social of Colombia for 2013. Results: Four algorithms were generated, two for signs and symptoms, and two that included leukocytes (≤4,500/mm3) and/or platelets (≤160,000/mm3) counts. The most accurate algorithm included blood count parameters with a sensitivity of 76.5% (95%CI: 71.9-80.5) and a specificity of 46.0% (95%CI: 37.6-54.7). In the external validation we found a sensitivity of 11.1% (95%CI: 4.9-20.7) and a specificity of 91.9% (95%CI: 87.5- 93.9). The scale of the Ministerio de Salud had a sensitivity of 76.4% (95%CI: 64.9-85.6) and a specificity of 38.0% (95%CI: 32.8-43.4). Conclusion: The inclusion of blood count parameters improved the sensitivity of diagnostics algorithms based on signs and symptoms. Clinical diagnosis of dengue remains a challenge for health research.

Bases científicas del razonamiento clínico: inferencia Bayesiana / Bayesian inference in clinical reasoning

A conceptual analysis of diagnostic reasoning in clinical practice is carried out. Using Bayesian inference as an alternative to frequentist inference usually used in science, clinical reasoning uses the scientific method step by step. The concepts of scientific method, probability, statistics and Bayesian inference are reviewed, highlighting their fundamental differences with the frequentist approach. This paper shows how the diagnostic approach proceeds in a Bayesian sense, ending with a basic example of application.

What Is Appropriate Upper Endoscopic Interval Among Dyspeptic Patients With Previously Normal Endoscopy? A Multicenter Study With Bayesian Change Point Analysis

BACKGROUND/AIMS: Appropriate interval for performing follow-up endoscopy among dyspeptic patients without abnormal findings on previous endoscopy is unclear. We analyzed the multicenter-collected data from the Korean Society of Neurogastroenterology and Motility. METHODS: We collected clinical data of the patients who visited the gastroenterology department and underwent 2 or more sessions of upper endoscopy during 2012–2017 at 6 university hospitals in Korea. Patients with endoscopic interval between 90 days and 760 days were included. For those with multiple endoscopic sessions, only the first 2 were analyzed. Positive outcome was defined as adenoma or cancer in the upper gastrointestinal tract. To identify the point of change and estimate the properties of the stochastic process before and after the change, we used Bayesian regression with Metropolis-Hastings algorithm. RESULTS: There were 1595 patients. Mean age was 58.8 years (standard deviation, 12.8). Median interval of endoscopy was 437 days (standard deviation, 153). On follow-up endoscopy, there were 12 patients (0.75%) who had neoplasia (4 with gastric cancer and 8 with gastric adnoma). As with the prior hypothesis, we presumed the change point (CP) of increase in frequency of organic lesion as 360 days. After random-walk Metropolis-Hastings sampling with Markov-Chain Monte Carlo iterations of 5000, the CP was 560 days (95% credible interval, 139–724). Estimated average of frequency of dysplastic lesions increased by a factor of 4.4 after the estimated CP. CONCLUSION: To rule out dysplastic lesions among dyspeptic patients who had previously normal endoscopy, a 2-year interval could be offered as follow-up interval for repeat upper endoscopy.

Network meta-analysis: application and practice using R software / 한국역학회지

The objective of this study is to describe the general approaches to network meta-analysis that are available for quantitative data synthesis using R software. We conducted a network meta-analysis using two approaches: Bayesian and frequentist methods. The corresponding R packages were “gemtc” for the Bayesian approach and “netmeta” for the frequentist approach. In estimating a network meta-analysis model using a Bayesian framework, the “rjags” package is a common tool. “rjags” implements Markov chain Monte Carlo simulation with a graphical output. The estimated overall effect sizes, test for heterogeneity, moderator effects, and publication bias were reported using R software. The authors focus on two flexible models, Bayesian and frequentist, to determine overall effect sizes in network meta-analysis. This study focused on the practical methods of network meta-analysis rather than theoretical concepts, making the material easy to understand for Korean researchers who did not major in statistics. The authors hope that this study will help many Korean researchers to perform network meta-analyses and conduct related research more easily with R software.

Network meta-analysis: application and practice using R software / 한국역학회지

The objective of this study is to describe the general approaches to network meta-analysis that are available for quantitative data synthesis using R software. We conducted a network meta-analysis using two approaches: Bayesian and frequentist methods. The corresponding R packages were “gemtc” for the Bayesian approach and “netmeta” for the frequentist approach. In estimating a network meta-analysis model using a Bayesian framework, the “rjags” package is a common tool. “rjags” implements Markov chain Monte Carlo simulation with a graphical output. The estimated overall effect sizes, test for heterogeneity, moderator effects, and publication bias were reported using R software. The authors focus on two flexible models, Bayesian and frequentist, to determine overall effect sizes in network meta-analysis. This study focused on the practical methods of network meta-analysis rather than theoretical concepts, making the material easy to understand for Korean researchers who did not major in statistics. The authors hope that this study will help many Korean researchers to perform network meta-analyses and conduct related research more easily with R software.

Predicting needlestick and sharps injuries and determining preventive strategies using a Bayesian network approach in Tehran, Iran / 한국역학회지

OBJECTIVES: Recent studies have shown that the rate of needlestick and sharps injuries (NSIs) is unacceptably high in Iranian hospitals. The aim of the present study was to use a systematic approach to predict and reduce these injuries.METHODS: This cross-sectional study was conducted in 5 hospitals in Tehran, Iran. Eleven variables thought to affect NSIs were categorized based on the Human Factors Analysis and Classification System (HFACS) framework and modeled using a Bayesian network. A self-administered validated questionnaire was used to collect the required data. In total, 343 cases were used to train the model and 50 cases were used to test the model. Model performance was assessed using various indices. Finally, using predictive reasoning, several intervention strategies for reducing NSIs were recommended.RESULTS: The Bayesian network HFACS model was able to predict 86% of new cases correctly. The analyses showed that safety motivation and fatigue were the most important contributors to NSIs. Supervisors' attitude toward safety and working hours per week were the most important factors in the unsafe supervision category. Management commitment and staffing were the most important organizational-level factors affecting NSIs. Finally, promising intervention strategies for reducing NSIs were identified and discussed.CONCLUSIONS: To reduce NSIs, both management commitment and sufficient staffing are necessary. Supervisors should encourage nurses to engage in safe behavior. Excessive working hours result in fatigue and increase the risk of NSIs.

Causal Inference Network of Genes Related with Bone Metastasis of Breast Cancer and Osteoblasts Using Causal Bayesian Networks

BACKGROUND: The causal networks among genes that are commonly expressed in osteoblasts and during bone metastasis (BM) of breast cancer (BC) are not well understood. Here, we developed a machine learning method to obtain a plausible causal network of genes that are commonly expressed during BM and in osteoblasts in BC. METHODS: We selected BC genes that are commonly expressed during BM and in osteoblasts from the Gene Expression Omnibus database. Bayesian Network Inference with Java Objects (Banjo) was used to obtain the Bayesian network. Genes registered as BC related genes were included as candidate genes in the implementation of Banjo. Next, we obtained the Bayesian structure and assessed the prediction rate for BM, conditional independence among nodes, and causality among nodes. Furthermore, we reported the maximum relative risks (RRs) of combined gene expression of the genes in the model. RESULTS: We mechanistically identified 33 significantly related and plausibly involved genes in the development of BC BM. Further model evaluations showed that 16 genes were enough for a model to be statistically significant in terms of maximum likelihood of the causal Bayesian networks (CBNs) and for correct prediction of BM of BC. Maximum RRs of combined gene expression patterns showed that the expression levels of UBIAD1, HEBP1, BTNL8, TSPO, PSAT1, and ZFP36L2 significantly affected development of BM from BC. CONCLUSIONS: The CBN structure can be used as a reasonable inference network for accurately predicting BM in BC.

Predicting needlestick and sharps injuries and determining preventive strategies using a Bayesian network approach in Tehran, Iran / 한국역학회지

OBJECTIVES: Recent studies have shown that the rate of needlestick and sharps injuries (NSIs) is unacceptably high in Iranian hospitals. The aim of the present study was to use a systematic approach to predict and reduce these injuries. METHODS: This cross-sectional study was conducted in 5 hospitals in Tehran, Iran. Eleven variables thought to affect NSIs were categorized based on the Human Factors Analysis and Classification System (HFACS) framework and modeled using a Bayesian network. A self-administered validated questionnaire was used to collect the required data. In total, 343 cases were used to train the model and 50 cases were used to test the model. Model performance was assessed using various indices. Finally, using predictive reasoning, several intervention strategies for reducing NSIs were recommended. RESULTS: The Bayesian network HFACS model was able to predict 86% of new cases correctly. The analyses showed that safety motivation and fatigue were the most important contributors to NSIs. Supervisors' attitude toward safety and working hours per week were the most important factors in the unsafe supervision category. Management commitment and staffing were the most important organizational-level factors affecting NSIs. Finally, promising intervention strategies for reducing NSIs were identified and discussed. CONCLUSIONS: To reduce NSIs, both management commitment and sufficient staffing are necessary. Supervisors should encourage nurses to engage in safe behavior. Excessive working hours result in fatigue and increase the risk of NSIs.

Modelo de predicción probabilística de deterioro en jamón de cerdo cocido / Probabilistic prediction model of deterioration of pork cooked ham

Antecedentes: el deterioro del jamón cocido es un proceso complejo que puede manifestarse por alteraciones de tipo microbiológico, fisicoquímico y sensorial, que hasta hoy han sido determinadas por metodologías tradicionales que incluyen largos periodos de seguimiento y un número importante de muestras. Objetivos: en el presente estudio se propuso un modelo para la predicción probabilística del deterioro de jamón cocido, por medio de análisis de componentes principales (ACP) y regresión logística con aproximación bayesiana. Métodos: para realizar el análisis, se evaluaron muestras provenientes de 300 lotes independientes de jamón de cerdo cocido, los cuales se conservaron en cámaras de almacenamiento a temperaturas de 12±1°C. Se determinaron experimentalmente tres tipos de variables: fisicoquímicas (pH, sinéresis, nitritos residuales, porcentaje de ácido láctico), de textura (adhesividad, dureza, trabajo de punción y firmeza instrumental) y microbiológicas (mesófilos aerobios y bacterias acidolácticas) en dos tiempos de medición: 3 y 40 días. Para realizar la creación del modelo se seleccionaron aleatoriamente el 60% de los datos y para la validación el 40% restante. Los datos obtenidos fueron procesados con el paquete estadístico R Core Team 2012. Resultados: los nitritos residuales y la sinéresis fueron las variables más representativas, ya que su distribución, correlación y carga, fueron las más significativas, con mayor poder discriminante del fenómeno de deterioro en los dos momentos de medición. El modelo desarrollado permitió correlacionar los parámetros estudiados y predecir la probabilidad de deterioro del jamón, además de clasificarlo de acuerdo a su estado de calidad. La primera componente principal (CP1) (variables bioquímicas) explicó el 73,3 % de la variación total de los datos, siendo los nitritos residuales y la sinéresis los factores más relacionados con el deterioro. Conclusión: el modelo logístico con aproximación bayesiana permitió obtener la probabilidad de deterioro del jamón cocido, almacenado a 12°C usando parámetros fisicoquímicos. La aplicación del ACP permitió correlacionar y clasificar los factores de deterioro del producto.

Background: the spoilage of cooked ham is a complex process that may be manifested by alterations of microbiological, physicochemical and sensory type, which has been determined by traditional methodologies including long periods of monitoring with a significant number of samples. Objective: in the present study a model for probabilistic prediction of spoilage of cooked ham was proposed by principal component analysis (PCA) and logistic regression with Bayesian approach. Methods: to perform the analysis, independent samples from 300 batches of cooked pork ham were evaluated, which were stored in cold at temperatures 12 ± 1°C. Three types of variables were experimentally determined: physicochemical (pH, syneresis, residual nitrite, lactic acid percentage), texture (adhesiveness, toughness, puncture and instrumental firmness) and microbiological (aerobic mesophilic bacteria and lactic acid bacteria) in two times: 3 and 40 days. To make this model were randomly selected 60% of the data and to validate it the remaining 40%. The obtained data were processed with statistical package R package statistical Core Team 2012. Results: residual nitrite and syneresis were the most representative variables, since its distribution, correlation and loadings were the most significant, more discriminant of the phenomenon of spoilage in the two times of measurement. The developed model allowed correlating the parameters and predicting the spoilage probability of cooked ham, and classifying them according to their quality status. The first principal component (PC1) (biochemical variables) explained 73.3 % of the total variation, the residual nitrite and syneresis were the factors that related the spoilage. Conclusions: the logistic regression model with Bayesian approach allowed obtaining the probability of deterioration of cooked ham, stored at 12°C using physicochemical parameters. Applying the PCA permited to correlate and classify the factors of spoilage.

Applying Bayesian Model to Predict Socio-demographic and Occlusal Determinants of Early Childhood Caries (ECC)

Objective: To use the Bayesian statistical Model approach to predict the most important socio-demographic and occlusal factors pertinent to high prevalence of ECC. Material and Methods: A questionnaire and an oral examination was conducted on children who attended a pediatric dental clinic in Nairobi during the period of study. The parents provided information on socio-demographic and oral habits of the children. The oral examination for presence of dental caries was recorded for each child. Descriptive statistics were obtained for dental caries, oral hygiene, using plaque score, and malocclusion. The results of the questionnaire and presence of dental caries were analyzed and the results subjected to Bayesian statistical analysis to determine any predictive factors for ECC. Results: 55% of the children had plaque accumulating on more than one third but less than two thirds of tooth surfaces. The highest plaque scores were reported among children whose fathers (48.2%) and mothers (42.0%) had completed secondary, and whose fathers were in non-formal employment 73.2%. The overall prevalence of dental caries in the study group was 95.5% with a mean dmft of 8.53 (+ 5.52 SD), with the male children having higher dmft 8.65 (SD+5.54) than the female children 8.37 (SD+ 5.50). The prevalence of malocclusion among children in the study was 55%. The majority had mesial step, 51.5% (n=140) and flush terminal plane 28.3% (n=77). Conclusion: The Bayesian Model, with a correct assumption, can be used to determine the important factors involved in high prevalence of ECC.