RECERCAT - Estadística

North African populations carry the signature of admixture with Neandertals

North African populations carry the signature of admixture with Neandertals Sánchez Quinto, F.; Rodríguez Botigué, L..; Civit Vives, Sergi; Arenas Solà, Concepción; Ávila-Arcos, M.C.; Bustamante, C.D.; Comas, D.; Lalueza Fox, Carles, 1965- One of the main findings derived from the analysis of the Neandertal genome was the evidence for admixture between Neandertals and non-African modern humans. An alternative scenario is that the ancestral population of non-Africans was closer to Neandertals than to Africans because of ancient population substructure. Thus, the study of North African populations is crucial for testing both hypotheses. We analyzed a total of 780,000 SNPs in 125 individuals representing seven different North African locations and searched for their ancestral/derived state in comparison to different human populations and Neandertals. We found that North African populations have a significant excess of derived alleles shared with Neandertals, when compared to sub-Saharan Africans. This excess is similar to that found in non-African humans, a fact that can be interpreted as a sign of Neandertal admixture. Furthermore, the Neandertal's genetic signal is higher in populations with a local, pre-Neolithic North African ancestry. Therefore, the detected ancient admixture is not due to recent Near Eastern or European migrations. Sub-Saharan populations are the only ones not affected by the admixture event with Neandertals.

Fragmentation of contaminant and endogenous DNA in ancient samples determined by shotgun sequencing; prospects for human palaeogenomics

Fragmentation of contaminant and endogenous DNA in ancient samples determined by shotgun sequencing; prospects for human palaeogenomics García Garcerà, M.; Gigli, E.; Sanchez Quinto, F.; Ramirez, O.; Calafell, F.; Civit Vives, Sergi; Lalueza Fox, C. Despite the successful retrieval of genomes from past remains, the prospects for human palaeogenomics remain unclear because of the difficulty of distinguishing contaminant from endogenous DNA sequences. Previous sequence data generated on high-throughput sequencing platforms indicate that fragmentation of ancient DNA sequences is a characteristic trait primarily arising due to depurination processes that create abasic sites leading to DNA breaks.

Search for Specific Biomarkers of IFNb Bioactivity in Patients with Multiple Sclerosis

Search for Specific Biomarkers of IFNb Bioactivity in Patients with Multiple Sclerosis Malhotra, S.; Bustamante, Marta F.; Perez Miralles, F.; Riol, J.; Ruiz de Villa, M.C.; Vegas, E.; Nonell, L.; Deisenhammer, F.; Fissolo, N.; Nurtdinov, Ramil N.; Montalban, X.; Comabella, M. Myxovirus A (MxA), a protein encoded by the MX1 gene with antiviral activity, has proven to be a sensitive measure of IFNβ bioactivity in multiple sclerosis (MS). However, the use of MxA as a biomarker of IFNβ bioactivity has been criticized for the lack of evidence of its role on disease pathogenesis and the clinical response to IFNβ. Here, we aimed to identify specific biomarkers of IFNβ bioactivity in order to compare their gene expression induction by type I IFNs with the MxA, and to investigate their potential role in MS pathogenesis. Gene expression microarrays were performed in PBMC from MS patients who developed neutralizing antibodies (NAB) to IFNβ at 12 and/or 24 months of treatment and patients who remained NAB negative. Nine genes followed patterns in gene expression over time similar to the MX1, which was considered the gold standard gene, and were selected for further experiments: IFI6, IFI27, IFI44L, IFIT1, HERC5, LY6E, RSAD2, SIGLEC1, and USP18. In vitro experiments in PBMC from healthy controls revealed specific induction of selected biomarkers by IFNβ but not IFNγ, and several markers, in particular USP18 and HERC5, were shown to be significantly induced at lower IFNβ concentrations and more selective than the MX1 as biomarkers of IFNβ bioactivity. In addition, USP18 expression was deficient in MS patients compared with healthy controls (p = 0.0004). We propose specific biomarkers that may be considered in addition to the MxA to evaluate IFNβ bioactivity, and to further explore their implication in MS pathogenesis.

Ultra-Deep Pyrosequencing Detects Conserved Genomic Sites and Quantifies Linkage of Drug-Resistant Amino Acid Changes in the Hepatitis B Virus Genome

Ultra-Deep Pyrosequencing Detects Conserved Genomic Sites and Quantifies Linkage of Drug-Resistant Amino Acid Changes in the Hepatitis B Virus Genome Rodriguez Frías, F.; Tabernero, D.; Quer, J.; Esteban, J.I.; Ortega, I.; Domingo, E.; Cubero, M.; Camos, S.; Ferrer Costa, C.; Sànchez, Àlex (Sànchez Pla); Jardí, R.; Schaper, M.; Homs, M.; García Cehic, M.; Guardia, J.; Esteban, R.; Buti, M. Selection of amino acid substitutions associated with resistance to nucleos(t)ide-analog (NA) therapy in the hepatitis B virus (HBV) reverse transcriptase (RT) and their combination in a single viral genome complicates treatment of chronic HBV infection and may affect the overlapping surface coding region. In this study, the variability of an overlapping polymerase-surface region, critical for NA resistance, is investigated before treatment and under antiviral therapy, with assessment of NA-resistant amino acid changes simultaneously occurring in the same genome (linkage analysis) and their influence on the surface coding region.

Similar and Divergent Effects of ppGpp and DksA Deficiencies on Transcription in Escherichia coli.

Similar and Divergent Effects of ppGpp and DksA Deficiencies on Transcription in Escherichia coli. Aberg, A.; Fernández Vázquez, Jorge; Cabrer-Panes, J.D.; Sànchez, Àlex (Sànchez Pla); Balsalobre Parra, Carlos The concerted action of ppGpp and DksA in transcription has been widely documented. In disparity with this model, phenotypic studies showed that ppGpp and DksA might also have independent and opposing roles in gene expression in Escherichia coli. In this study we used a transcriptomic approach to compare the global transcriptional patterns of gene expression in strains deﬁcient in ppGpp (ppGpp0) and/or DksA ( dksA). Approximately 6 and 7% of all genes were signiﬁcantly affected by more than twofold in ppGpp- and DksAdeﬁcient strains, respectively, increasing to 13% of all genes in the ppGpp0 dksA strain. Although the data indicate that most of the affected genes were copositively or conegatively regulated by ppGpp and DksA, some genes that were independently and/or differentially regulated by the two factors were found. The large functional group of chemotaxis and ﬂagellum synthesis genes were notably differentially affected, with all genes being upregulated in the DksA-deﬁcient strain but 60% of them being downregulated in the ppGpp-deﬁcient strain. Revealingly, mutations in the antipausing Gre factors suppress the upregulation observed in the DksA-deﬁcient strain, emphasizing the importance of the secondary channel of the RNA polymerase for regulation and ﬁne-tuning of gene expression in E. coli.

White eye phenotypes and their genetic analysis

White eye phenotypes and their genetic analysis Mestres i Naval, Francesc; Arenas Solà, Concepción An interesting case for undergraduate students of general Genetics is to consider that different genes can produce the same or similar phenotypes. We present here an experiment to discover that the same phenotype could be produced by different genes, and then, to carry out the genetic analysis of these genes. For this laboratory study we have used the following Drosophila melanogaster strains: white (white eyes) and scarlet
brown (white eyes).

El análisis canónico y su aplicación en Geologia

El análisis canónico y su aplicación en Geologia Cuadras, C. M. (Carlos María); Campá Viñeta, Juan A.; Montoriol-Pous, Joaquín, 1924-2011 Se exponen los fundamentos del análisis canóiiico, combinación del análisis de la varianza y del factorial, así como sus posibles aplicaciones a los problemas geológicos.

Adaptación española del cuestionario E.P.I. de Eysenck

Adaptación española del cuestionario E.P.I. de Eysenck Sánchez Turet, Miguel, 1942-2008; Cuadras, C. M. (Carlos María) A partir de la evolución de la estructura del gasto alimentario en los hogares españoles a lo largo de los últimos cuarenta, el artículo analiza la relación entre los cambios sociales, económicos, tecnológicos, demogrdficos, ideológicos, sanitarios, de estética corporal, etc. y los cambios habidos en 10s comportamientos alimentarios. Estos cambios permiten caracterizar una ccmodemidad alimentarias cuyos contenidos referirían a una articulación dinámica y compleja entre una revalorizacibn de las tradiciones culinarias y una udlización cada vez más frecuente de los productos elaborados por las industrias alimentarias.

El tratamiento numérico de la realidad. Reflexiones sobre la importancia actual de la estadística en la Sociedad de la Información.

El tratamiento numérico de la realidad. Reflexiones sobre la importancia actual de la estadística en la Sociedad de la Información. Monleón Getino, Toni La estadística se ha convertido en un método práctico para describir los valores de datos económicos, políticos, sociales, psicológicos, médicos, biológicos y físicos, como herramienta para relacionar y analizar dichos datos. El trabajo del estadístico no consiste sólo en obtener, reunir o tabular los datos, sino sobre todo el proceso de interpretación de esa información solo o en colaboración con los expertos en cada ámbito. A pesar de que es imposible entender la Sociedad de la Información sin la estadística, es preocupante la situación actual de las ciencias matemáticas y más concretamente de la estadística, ya que a pesar de su imprescindibilidad existe una falta de vocación entre los jóvenes por su estudio y uso, ya sea por su dificultad matemática, desconocimiento u otros motivos. Este artículo pretende reflexionar sobre la importancia de la misma en la sociedad actual que permite tratar numéricamente la realidad.

Applications of the bootstrap to survival analysis

Applications of the bootstrap to survival analysis Utzet i Civit, Frederic; Sànchez, Àlex (Sànchez Pla) En este trabajo se revisan algunas de las aplicaciones clásicas del bootstrap al análisis de la supervivencia. Se consideran en primer lugar el estimador bootstrap de la varianza y el estimador de la mediana corregido para el sesgo del estimador de Kaplan-Meier de la función de supervivencia. A continuación se consideran algunos aspectos mas recientes, tales como métodos para construir bandas de confianza para el estimador de la funcidn de supervivencia y contrastes aproximados para la comparación de funciones de supervivencia. En ambas situaciones el bootstrap resulta de gran utilidad para la aproximación de 10s valores críticos necesarios.

Recent statistical methods based on distances

Recent statistical methods based on distances Arenas Solà, Concepción; Cuadras, C. M. (Carlos María) El concepte de distància s'ha utilizat en diferents camps i és bàsic en alguns mètodes estadístics recents, vàlids per a variables no numèriques així com per a una barreja de diferents tipus de variables. A aquest article expliquem i il·lustrem mètodes recents basats en distàncies: regressió basada en distàncies; discriminant basat en distàncies, related metric scaling i continuous scaling. Aquests mètodes formen part de la recerca actual del grup d'anàlisi multivariant, dirigit per C. M. Cuadras, del Departament d' Estadística de la Universitat de Barcelona. Aquest grup, que treballa a les Facultats de Biologia i Matemàtiques, està format per set investigadors: C. Arenas, C. M. Cuadras, D. Cuadras, A. Esteve, J. Fortiana, A. Grané i F. Oliva.

R commands secondary production

R commands secondary production Gaudes Saez, Ainhoa; Ocaña i Rebull, Jordi; Muñoz Gràcia, Isabel R commands to calculate the secondary production estimates using the size-frequency method after Hynes and Coleman (1968), Benke (1979) and Huryn (1996).

Evaluación taxonómica de la variabilidad intraespecífica de Galium brockmannii briq. (Rubiaceae).

Evaluación taxonómica de la variabilidad intraespecífica de Galium brockmannii briq. (Rubiaceae). Sáez, Llorenç; Galbany Casals, Mercè; Cáceres Ainsa, Miquel de 5 poblaciones diferentes, utilizados como referencia externa en este análisis, se muestran claramente agrupados y separados de las poblaciones de G. brockmannii. En resumen, todos los dados disponibles sugieren que G. brockmannii subsp. aterratzense debe ser considerado un sinónimo de G. brockmannii.

Quantification of human adenoviruses in European recreational waters

Quantification of human adenoviruses in European recreational waters Bofill Mas, Sílvia; Calgua, Byron; Clemente Casares, Pilar; La Rosa, Giuseppina; Iaconelli, Marcello; Muscillo, Michele; Rutjes, Saskia; Roda Husman, Ana Maria de; Grünert, Andreas; Graver, Ingeburg; Verani, Marco; Carducci, Annalaura; Calvo, Miquel; Wyn-Jones, A. Peter; Gironès Llop, Rosina The presence of human adenoviruses in recreational water might cause disease in the population upon exposure. Human adenoviruses detected by PCR could also serve as indicators of the virological water quality. In order to assess the applicability of human adenoviruses to the evaluation of the faecal contamination in European bathing waters, a real-time quantitative PCR assay was developed for the quantification of human adenoviruses in 132 samples collected from 24 different recreational marine and freshwater sites in nine European countries.Selected samples presenting positive nested-PCR results for human adenoviruses were analyzed using quantitative PCR and 80 samples from a total of 132 produced quantitative results with mean values of 3.2x102 10 per 100 ml of water, human adenovirus 41 being the most prevalent serotype. Human adenoviruses were quantified in samples from all 15 surveillance laboratories. Statistical analysis showed no homogeneous linear relation between humanadenoviruses and E. coli, intestinal enterococci or somatic coliphages concentrations in the tested samples when considering all the data together. Significant correlations between human adenoviruses and at least one of the other indicators were observed only when data from individual Laboratories were considered. The quantification of human adenoviruses may provide complementary information in relation to the use of bacterial standards in the control of water quality in bathing water.

Differential equations driven by fractional Brownian motion

Differential equations driven by fractional Brownian motion Nualart, David, 1951-; Rascanu, Aurel A global existence and uniqueness result of the solution for multidimensional, time dependent, stochastic differential equations driven by a fractional Brownian motion with Hurst parameter H& is proved. It is shown, also, that the solution has finite moments. The result is based on a deterministic existence and uniqueness theorem whose proof uses a contraction principle and a priori estimates.

Uso de la estadística en la oncología y la hematología

Uso de la estadística en la oncología y la hematología Monleón Getino, Toni En este artículo abordamos el uso y la importancia de las herramientas estadísticas que se utilizan principalmente en los estudios médicos del ámbito de la oncología y la hematología, pero aplicables a muchos otros campos tanto médicos como experimentales o industriales. El objetivo del presente trabajo es presentar de una manera clara y precisa la metodología estadística necesaria para analizar los datos obtenidos en los estudios rigurosa y concisamente en cuanto a las hipótesis de trabajo planteadas por los investigadores. La medida de la respuesta al tratamiento elegidas en al tipo de estudio elegido determinarán los métodos estadísticos que se utilizarán durante el análisis de los datos del estudio y también el tamaño de muestra. Mediante la correcta aplicación del análisis estadístico y de una adecuada planificación se puede determinar si la relación encontrada entre la exposición a un tratamiento y un resultado es casual o por el contrario, está sujeto a una relación no aleatoria que podría establecer una relación de causalidad. Hemos estudiado los principales tipos de diseño de los estudios médicos más utilizados, tales como ensayos clínicos y estudios observacionales (cohortes, casos y controles, estudios de prevalencia y estudios ecológicos). También se presenta una sección sobre el cálculo del tamaño muestral de los estudios y cómo calcularlo, ¿Qué prueba estadística debe utilizarse?, los aspectos sobre fuerza del efecto ¿odds ratio¿ (OR) y riesgo relativo (RR), el análisis de supervivencia. Se presentan ejemplos en la mayoría de secciones del artículo y bibliografía más relevante.; In this article we address the use and importance of statistical tools that are mainly used in medical studies in the field of oncology and hematology, but applicable to many other medical fields as both experimental and industrial. The objective of this paper is to present a clear and accurate statistical methodology to analyze the data obtained inrigorous studies and concisely in terms of working hypotheses planned by researchers. The measure of response to the treatment of choice in the study determines the statistical methods used for the analysis of data and sample size. Through proper application of statistical analysis and proper planning you can determine whether the relationship foundbetween exposure to treatment and outcome is casual or on the contrary, is subject to a random connection that could not establish a causal relationship. We have studied the major types of study designs on medical use, such as clinical trials and observational studies 3 (cohorts, case control studies, prevalence and ecological studies). It also presents a section on calculating the sample size of studies and how to calculate it, what statistical test should be used, the strength of the effect on odds ratio (OR) and relative risk (RR), survival analysis.

A measure of stability as a criterion for the verification and analysis of simulation models

A measure of stability as a criterion for the verification and analysis of simulation models Monleón Getino, Toni; Ruiz de Villa Jubany, Ma. Carmen; Ocaña i Rebull, Jordi The aim of this study is to define a new statistic, PVL, based on the relative distance between the likelihood associated with the simulation replications and the likelihood of the conceptual model. Our results coming from several simulation experiments of a clinical trial show that the PVL statistic range can be a good measure of stability to establish when a computational model verifies the underlying conceptual model. PVL improves also the analysis of simulation replications because only one statistic is associated with all the simulation replications. As well it presents several verification scenarios, obtained by altering the simulation model, that show the usefulness of PVL. Further simulation experiments suggest that a 0 to 20 % range may define adequate limits for the verification problem, if considered from the viewpoint of an equivalence test.

Supplementary material to Ocaña, J., Sánchez, M.P., Sánchez, A. and Carrasco, J.L. ¿On equivalence and bioequivalence testing¿, SORT, 32(2), 2008

Supplementary material to Ocaña, J., Sánchez, M.P., Sánchez, A. and Carrasco, J.L. ¿On equivalence and bioequivalence testing¿, SORT, 32(2), 2008 Ocaña i Rebull, Jordi; Sánchez Olavarría, Maria Pilar; Sànchez, Àlex (Sànchez Pla); Carrasco Jordan, Josep Lluís

Impact of incorrect assumptions on the covariance structure of random effects and/or residuals in nonlinear mixed models for repeated measures data

Impact of incorrect assumptions on the covariance structure of random effects and/or residuals in nonlinear mixed models for repeated measures data El Halimi, Rachid; Ocaña i Rebull, Jordi In this paper we analyse, using Monte Carlo simulation, the possible consequences of incorrect assumptions on the true structure of the random effects covariance matrix and the true correlation pattern of residuals, over the performance of an estimation method for nonlinear mixed models. The procedure under study is the well known linearization method due to Lindstrom and Bates (1990), implemented in the nlme library of S-Plus and R. Its performance is studied in terms of bias, mean square error (MSE), and true coverage of the associated asymptotic confidence intervals. Ignoring other criteria like the convenience of avoiding over parameterised models, it seems worst to erroneously assume some structure than do not assume any structure when this would be adequate.

On the consequences of misspecifing assumptions concerning residuals distribution in a repeated measures and nonlinear mixed modelling context

On the consequences of misspecifing assumptions concerning residuals distribution in a repeated measures and nonlinear mixed modelling context El Halimi, Rachid; Ocaña i Rebull, Jordi In this paper we describe the results of a simulation study performed to elucidate the robustness of the Lindstrom and Bates (1990) approximation method under non-normality of the residuals, under different situations. Concerning the fixed effects, the observed coverage probabilities and the true bias and mean square error values, show that some aspects of this inferential approach are not completely reliable. When the true distribution of the residuals is asymmetrical, the true coverage is markedly lower than the nominal one. The best results are obtained for the skew normal distribution, and not for the normal distribution. On the other hand, the results are partially reversed concerning the random effects. Soybean genotypes data are used to illustrate the methods and to motivate the simulation scenarios

Spectral analysis of the luteinizing hormone in the blood samples

Spectral analysis of the luteinizing hormone in the blood samples Liutsko, Liudmila Generally, medicine books are concentrated almost exclusively in explaining methodology that analyzes fixed measures, measures done in a certain moment, nevertheless the evolution of the measurement and correct interpretation of the missed values are very important and sometimes can give the key information of the results obtained. Thus, the analysis of the temporary series and spectral analysis or analysis of the time series in the dominion of frequencies can be regarded as an appropriate tool for this kind of studies.In this work the frequency of the pulsating secretion of luteinizing hormone LH (thatregulates the fertile life of women) were analyzed in order to determine the existence of the significant frequencies obtained by analysis of Fourier. Detection of the frequencies, with which the pulsating secretion of the LH takes place, is a quite difficult question due topresence of the random errors in measures and samplings, i.e. that pulsating secretions of small amplitude are not detected and disregarded. In physiology it is accepted that cyclical patterns in the secretion of the LH exist and in the results of this research confirm this pattern and determine its frequency presented in the corresponded periodograms to each of studied cycle. The obtained results can be used as key pattern for future sampling frequencies in order to ¿catch¿ the significant picks of the luteinizing hormone and reflect on time forproductivity treatment of women.; Diploma d'Estudis Avançats - Programa de doctorat en Estadística. 2008. Tutors: Martín Ríos Alcolea

What trees tell us: dendrochronological and statistical analysis of the data

What trees tell us: dendrochronological and statistical analysis of the data Liutsko, Liudmila Trees are a great bank of data, named sometimes for this reason as the "silentwitnesses" of the past. Due to annual formation of rings, which is normally influenced directly by of climate parameters (generally changes in temperature and moisture or precipitation) and other environmental factors; these changes, occurred in the past, are"written" in the tree "archives" and can be "decoded" in order to interpret what hadhappened before, mainly applied for the past climate reconstruction.Using dendrochronological methods for obtaining samples of Pinus nigra fromthe Catalonian PrePirineous region, the cores of 15 trees with total time spine of about 100 - 250 years were analyzed for the tree ring width (TRW) patterns and had quite high correlation between them (0.71 ¿ 0.84), corresponding to a common behaviour for the environmental changes in their annual growth.After different trials with raw TRW data for standardization in order to take outthe negative exponential growth curve dependency, the best method of doubledetrending (power transformation and smoothing line of 32 years) were selected for obtaining the indexes for further analysis.Analyzing the cross-correlations between obtained tree ring width indexes andclimate data, significant correlations (p&0.05) were observed in some lags, as forexample, annual precipitation in lag -1 (previous year) had negative correlation with TRW growth in the Pallars region. Significant correlation coefficients are between 0.27- 0.51 (with positive or negative signs) for many cases; as for recent (but very short period) climate data of Seu d¿Urgell meteorological station, some significant correlation coefficients were observed, of the order of 0.9.These results confirm the hypothesis of using dendrochronological data as aclimate signal for further analysis, such as reconstruction of climate in the past orprediction in the future for the same locality.; Diploma d'Estudis Avançats - Programa de doctorat en Estadística. 2008. Tutor: Dr. Antoni Monleón

Anàlisi factorial confirmatòria per a variables categòriques: Aplicació al qüestionari de discapacitat WHODAS-II

Anàlisi factorial confirmatòria per a variables categòriques: Aplicació al qüestionari de discapacitat WHODAS-II Vilagut Saiz, Gemma Objective: To describe the methodology of Confirmatory Factor Analyis for categorical items and to apply this methodology to evaluate the factor structure and invariance of the WHO-Disability Assessment Schedule (WHODAS-II) questionnaire, developed by the World HealthOrganization.Methods: Data used for the analysis come from the European Study of Mental Disorders(ESEMeD), a cross-sectional interview to a representative sample of the general population of 6 european countries (n=8796). Respondents were administered a modified version of theWHODAS-II, that measures functional disability in the previous 30 days in 6 differentdimensions: Understanding and Communicating; Self-Care, Getting Around, Getting Along withOthers, Life Activities and Participation. The questionnaire includes two types of items: 22severity items (5 points likert) and 8 frequency items (continuous). An Exploratory factoranalysis (EFA) with promax rotation was conducted on a random 50% of the sample. Theremaining half of the sample was used to perform a Confirmatory Factor Analysis (CFA) inorder to compare three different models: (a) the model suggested by the results obtained in theEFA; (b) the theoretical model suggested by the WHO with 6 dimensions; (c) a reduced modelequivalent to model b where 4 of the frequency items are excluded. Moreover, a second orderfactor was also evaluated. Finally, a CFA with covariates was estimated in order to evaluatemeasurement invariance of the items between Mediterranean and non-mediterranean countries.Results: The solution that provided better results in the EFA was that containing 7 factors. Twoof the frequency items presented high factor loadings in the same factor, and one of thempresented factor loadings smaller than 0.3 with all the factors. With regard to the CFA, thereduced model (model c) presented the best goodness of fit results (CFI=0.992,TLI=0.996,RMSEA=0.024). The second order factor structure presented adequate goodness of fit (CFI=0.987,TLI=0.991, RMSEA=0.036). Measurement non-invariance was detected for one of the items of thequestionnaire (FD20 ¿ Embarrassment due to health problems).Conclusions: AFC confirmed the initial hypothesis about the factorial structure of the WHODAS-II in 6factors. The second order factor supports the existence of a global dimension of disability. The use of 4of the frequency items is not recommended in the scoring of the corresponding dimensions.; Diploma d'Estudis Avançats - Programa de doctorat en Estadística, Anàlisi de dades i bioestadística. 2008. Tutors: Josep Fortiana i Jordi Alonso

Estudi de l'estat de Salut autopercebut: Modelització de l'índex d'utilitat EQ-5D mitjançant un model tobit

Estudi de l'estat de Salut autopercebut: Modelització de l'índex d'utilitat EQ-5D mitjançant un model tobit Vilagut Saiz, Gemma Objective: Health status measures usually have an asymmetric distribution and present a highpercentage of respondents with the best possible score (ceiling effect), specially when they areassessed in the overall population. Different methods to model this type of variables have beenproposed that take into account the ceiling effect: the tobit models, the Censored Least AbsoluteDeviations (CLAD) models or the two-part models, among others. The objective of this workwas to describe the tobit model, and compare it with the Ordinary Least Squares (OLS) model,that ignores the ceiling effect.Methods: Two different data sets have been used in order to compare both models: a) real datacomming from the European Study of Mental Disorders (ESEMeD), in order to model theEQ5D index, one of the measures of utilities most commonly used for the evaluation of healthstatus; and b) data obtained from simulation. Cross-validation was used to compare thepredicted values of the tobit model and the OLS models. The following estimators werecompared: the percentage of absolute error (R1), the percentage of squared error (R2), the MeanSquared Error (MSE) and the Mean Absolute Prediction Error (MAPE). Different datasets werecreated for different values of the error variance and different percentages of individuals withceiling effect. The estimations of the coefficients, the percentage of explained variance and theplots of residuals versus predicted values obtained under each model were compared.Results: With regard to the results of the ESEMeD study, the predicted values obtained with theOLS model and those obtained with the tobit models were very similar. The regressioncoefficients of the linear model were consistently smaller than those from the tobit model. In thesimulation study, we observed that when the error variance was small (s=1), the tobit modelpresented unbiased estimations of the coefficients and accurate predicted values, specially whenthe percentage of individuals wiht the highest possible score was small. However, when theerrror variance was greater (s=10 or s=20), the percentage of explained variance for the tobitmodel and the predicted values were more similar to those obtained with an OLS model.Conclusions: The proportion of variability accounted for the models and the percentage ofindividuals with the highest possible score have an important effect in the performance of thetobit model in comparison with the linear model.; Diploma d'Estudis Avançats - Programa de doctorat en Estadística, Anàlisi de dades i bioestadística. 2008. Tutors: Josep Fortiana i Jordi Alonso