SEARCH · PubMed Health

Results for “Multiple Classification Analysis”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 775 records · Page 43Linked to original sources

PubMed1983

The analysis of dependence in cross-classifications having ordered categories, using log-linear models for frequencies and log-linear models for odds.

To analyse the dependence of a qualitative (dichotomous or polytomous) response variable upon one or more qualitative explanatory variables, log-linear models for frequencies are compared with log-linear models for odds, when the categories of the response variable are ordered and the categories of each explanatory variable may be either ordered or unordered. The log-linear models for odds express the odds (or log odds) pertaining to adjacent response categories in terms of appropriate multiplicative (or additive) factors. These models include the 'null log-odds model', the 'uniform log-odds model', the 'parallel log-odds model', and other log-linear models for the odds. With these models, the dependence of the response variable (with ordered categories) can be analyzed in a manner analogous to the usual multiple regression analysis and related analysis of variance and analysis of covariance. Application of log-linear models for the odds sheds light on earlier applications of log-linear models for the frequencies in contingency tables with ordered categories.

Biometry↗

PubMed2003

Evidence for multiple hybrid groups in Trypanosoma cruzi.

A role for parasite genetic variability in the spectrum of Chagas disease is emerging but not yet evident, in part due to an incomplete understanding of the population structure of Trypanosoma cruzi. To investigate further the observed genotypic variation at the sequence and chromosomal levels in strains of standard and field-isolated T. cruzi we have undertaken a comparative analysis of 10 regions of the genome from two isolates representing T. cruzi I (Dm28c and Silvio X10) and two from T. cruzi II (CL Brener and Esmeraldo). Amplified regions contained intergenic (non-coding) sequences from tandemly repeated genes. Multiple nucleotide polymorphisms correlated with the T. cruzi I/T. cruzi II classification. Two intergenic regions had useful polymorphisms for the design of classification probes to test on genomic DNA from other known isolates. Two adjacent nucleotide polymorphisms in HSP 60 correlated with the T. cruzi I and T. cruzi II distinction. 1F8 nucleotide polymorphisms revealed multiple subdivisions of T. cruzi II: subgroups IIa and IIc displayed the T. cruzi I pattern; subgroups IId and IIe possessed both the I and II patterns. Furthermore, isolates from subgroups IId and IIe contained the 1F8 polymorphic markers on different chromosome bands supporting a genetic exchange event that resulted in chromosomes V and IX of T. cruzi strain CL Brener. Based on these analyses, T. cruzi I and subgroup IIb appear to be pure lines, while subgroups IIa/IIc and IId/IIe are hybrid lines. These data demonstrate for the first time that IIa/IIc are hybrid, consistent with the hypothesis that genetic recombination has occurred more than once within the T. cruzi lines.

Animals↗

PubMed1989

Prognostic factors in multiple myeloma: definition of risk groups in 410 previously untreated patients: a Grupo Argentino de Tratamiento de la Leucemia Aguda study.

Four hundred ten previously untreated multiple myeloma patients entered onto two consecutive Grupo Argentino de Tratamiento de la Leucemia Aguda (GATLA) protocols were analyzed to identify significant prognostic factors influencing survival. The univariate analysis selected the following variables: performance status, renal function, percentage of bone marrow plasma cells at diagnosis, hemoglobin, and age. A multivariate analysis showed that performance status, renal function, percentage of bone marrow plasma cells, hemoglobin, and age were the best predictive variables for survival. A score was assigned to each patient according to these variables, which led to their classification in three groups: good, intermediate, and poor risk, with a probability of survival of 26% and 10% at 96 months, and 5% at 56 months, and median survival of 60, 37, and 14 months, respectively (P = .0000). In our patient population, this model proved to be superior to the Durie-Salmon staging system in defining prognostic risk groups, and separating patients with significantly different risks within each Durie-Salmon stage.

Adult↗

PubMed1990

Variation and classification of human age patterns of mortality: analysis using competing hazards models.

The structure of variation in human mortality patterns is explored using a five-parameter competing hazards model and standard multivariate taxonomic procedures. The data consist of 281 national life tables representing a wide range of environmental and cultural regions of the world. A general or average age pattern of mortality was generated for the entire sample using multiple regression procedures. A K mean cluster analysis conducted on the residuals of the regression analysis identified seven distinct models of mortality that differ in characteristic ways from the general pattern. Four of the seven clusters have age patterns of mortality similar to the north, east, south, and west regions of the Coale and Demeny model life tables. The remaining three clusters represent regions of the world and age patterns of mortality that are not represented in the Coale and Demeny model life tables.

Age Factors↗

PubMed2002

Quantifying psychiatric comorbidity--lessions from chronic disease epidemiology.

BACKGROUND: Comorbidity research in psychiatric epidemiology mostly uses measures of association like odds or risk ratios to express how strongly disorders are linked. In contrast, chronic disease epidemiologists increasingly use measures of clustering, like multimorbidity (cluster) coefficients, to study comorbidity. This article compares measures of association and clustering. METHODS: Narrative review, algebraical examples, a secondary analysis of an existing dataset and a pooled analysis of published data. RESULTS: Odds and risk ratios, but the former more than the latter, confound clustering with coincidental comorbidity. Multimorbidity coefficients provide a pure estimate of clustering which is the proportion of the association between disorders that is of etiological interest. Odds and risk ratios can express comorbidity between no more than two disorders, whilst clustering coefficients, although computationally laboursome, can capture multimorbidity of any number of disorders. Cluster coefficients depend less on the prevalence of illness in study groups than measures of association. CONCLUSION: Odds and risk ratios are well suited for comorbidity research which focuses on which sets of disorders or syndromes tend to occur in combination and the implications of this for, for instance, nosological classification, a traditional interest of psychiatric epidemiology. However, the cluster coefficient is to be preferred if the interest is more aetiological, addressing for example why certain individuals are prone to multiple health problems.

Alcoholism↗

PubMed2005

A predictive model of rectal tumor response to preoperative radiotherapy using classification and regression tree methods.

PURPOSE: The ability to predict rectal tumor response to preoperative radiotherapy before treatment would significantly impact patient selection. In this study, classification and regression tree (CART) methods were used to model tumor response to preoperative conformal high-dose rate brachytherapy by assessing the predictive value of vascular endothelial growth factor (VEGF), Bcl-2, p21, p53, and APAF-1. EXPERIMENTAL DESIGN: Immunohistochemistry was used to detect VEGF, Bcl-2, p21, p53, and APAF-1 from 62 pretreatment rectal tumor biopsies. Scores were assigned as percentages of positive tumor cell staining and were used in CART analysis to identify the proteins that best predicted response to radiotherapy. Ten-fold cross-validation was used to prevent overfitting and multiple cross-validation experiments were run to estimate the prediction error. RESULTS: Postoperative pathologic evaluation of the irradiated tumor bed revealed 43 responsive tumors [20 with complete response (T(0)) and 23 with partial response] and 19 nonresponsive tumors. The optimal tree resulting from CART analysis had five terminal nodes with a misclassification rate of 18%. Of the five proteins selected for their predictive value, VEGF and Bcl-2 contributed most to the classification of responsive and nonresponsive tumors. All 10 tumors with no VEGF were completely responsive (T(0)) to radiotherapy; 85% of those with VEGF and negative for Bcl-2 were responsive to therapy. CONCLUSIONS: VEGF and Bcl-2 status in pretreatment rectal tumor biopsies may be predictive of response to preoperative high-dose rate brachytherapy.

Apoptotic Protease-Activating Factor 1↗

PubMed2006

The validity of an empirical delineation of heterogeneity in chronic unexplained fatigue.

OBJECTIVES: To validate a latent class structure derived empirically from a clinical data set obtained from persons with chronic medically unexplained fatigue. METHODS: The strategies utilized in this validation study included: recalculating latent class analysis (LCA) results varying random seeds and the number of initial random starting sets; recalculating LCA results by substituting alternate variables to demonstrate a robust solution; determining the statistical significance of between-class differences on disability, fatigue and demographic measures omitted from the data set used for LCA; cross-classifying class membership using established Centers for Disease Control and Prevention (CDC) research criteria for chronic fatigue syndrome (CFS) to compare the relative proportions of subjects designated CFS, chronic fatigue (not CFS) or healthy controls captured by the latent classes. RESULTS: Recalculation of results and substitution of variables for low-loading variables demonstrated a robust LCA result. Highly significant between-class differences were confirmed between Class 2 (well) and those interpreted as ill/fatigued. Analysis of between-class differences for the fatigue groups revealed significant differences for all disability and fatigue variables, but with equivalent levels of reported activity and reduction in motivation. Cross-classification against established CDC criteria demonstrated that 89% of subjects constituting Class 2 (well) were indeed nonfatigued controls. A general tendency for grouping CFS cases in the multiple symptomatic classes was noted. CONCLUSION: This study established reasonably good validity for an empirically-derived latent class solution reflecting considerable heterogeneity among subjects with medically unexplained chronic fatigue. This work strengthens the growing understanding of CFS as a heterogeneous entity comprised of several conditions with different underlying pathophysiological mechanisms.

Adult↗

PubMed2000

Evaluation of a multiple-antigen enzyme-linked immunosorbent assay for detection of Mycobacterium tuberculosis infection in captive elephants.

Mycobacterium tuberculosis has become an important agent of disease in the captive elephant population of the United States, although current detection methods appear to be inadequate for effective disease management. This investigation sought to validate a multiple-antigen enzyme-linked immunosorbent assay (ELISA) for screening of M. tuberculosis infection in captive elephants and to document the elephant's serologic response over time using a cross-sectional observational study design. Serum samples were collected from 51 Asian elephants (Elephas maximus) and 26 African elephants (Loxodonta africana) from 16 zoos and circuses throughout the United States. Infection status of each animal was determined by mycobacterial culture of trunk washes. Reactivity of each serum sample against six antigens was determined, and the linear combination of antigens that accurately predicted the infection status of the greatest number of animals was determined by discriminant analysis. The resulting classification functions were used to calculate the percentage of animals that were correctly classified (i.e., specificity and sensitivity). Of the 77 elephants sampled, 47 fit the criteria for inclusion in discriminant analysis. Of these, seven Asian elephants were considered infected; 25 Asian elephants and 15 African elephants were considered noninfected. The remaining elephants had been exposed to one or more infected animals. The specificity and sensitivity of the multiple-antigen ELISA were both 100% (91.9-100% and 54.4-100%, respectively) with 95% confidence intervals. Mycobacterium bovis culture filtrate showed the highest individual antigen specificity (95%; 83.0-100%) and sensitivity (100%; 54.4-100%). Serum samples from 34 elephants were analyzed over time by the response to the culture filtrate antigen; four of these elephants were culture positive and had been used to calculate the discriminant function. Limitations such as sample size, compromised ability to ascertain each animal's true infection status, and absence of known-infected African elephants suggest that much additional research needs to be conducted regarding the use of this ELISA. However, the results indicate that this multiple-antigen ELISA would be a valuable screening test for detecting M. tuberculosis infection in elephant herds.

Animal Diseases↗

PubMed2005

Phenotypic and molecular characterisation of a de novo 5q deletion that includes the APC gene.

We report on a 12-year-old female patient with mild dysmorphic signs, including bilateral epicanthal folds, low-set dysplastic ears, a short nose with anteverted nostrils, conically shaped fingers, generalised increase of subcutaneous fat, multiple fine venous teleangiectasia on her back, mild pectus carinatum, and a general muscular hypotonia. Cytogenetic analysis and fluorescence in situ hybridisation (FISH) studies using region-specific BAC and YAC clones indicated a de novo interstitial deletion of the long arm of chromosome 5, resulting in monosomy 5q21.1-q23.1. Molecular analysis of polymorphic markers helped to narrow down the breakpoints and demonstrated that the derivative chromosome 5 is of paternal origin. By using the same panel of polymorphic markers, a reinvestigation of a similar, already published, 5q deletion case [Raedle et al. (2001) Am J Gastroenterol 96:3016-3020] was performed, allowing a more detailed genotype-phenotype correlation. Phenotypic classification was also carried out. Several known genes, including APC and MCC, were found to map to the common deleted genomic segment. Genetic counselling based on the molecular analysis data was performed for the index family.

Abnormalities, Multiple↗

PubMedDate not supplied

Carbon monoxide-related injury estimation using ICD-coded data: methodologic implications for public health surveillance.

Estimates of unintentional deaths from carbon monoxide (CO) poisoning can be obtained from national mortality data. We explored ways of accurately estimating CO-related deaths from International Classification of Diseases, 9th Revision (ICD-9) coded U.S. mortality data. We evaluated and identified CO-related ICD-9 codes and created five classes of codes for case ascertainment that represented a continuum of the degree of certainty that the ICD-coded death was truly CO-related. We conducted single (underlying) cause-of-death and multiple-cause-of-death analysis using 20 years of data (1979-1998), and calculated sensitivity and positive predictive value using different criteria for case ascertainment. Single-cause analysis provided accurate estimates only when we used CO-exclusive E-codes, however this method failed to identify approximately one third of the CO-related deaths over the study period. Single-cause analysis overestimated the number of CO-related deaths when we used E-codes that were not exclusive to CO exposure. Identification of true CO-related deaths required multiple cause-of-death analysis and use of the CO nature-of-injury code, N986, to confirm suspected cases. Sensitivity of N986 was 99.5%, and positive predictive value of the individual E-codes in single cause-of-death analysis ranged from 1.5% to 92%. Estimating CO-related deaths from ICD-coded data requires a thorough understanding of the ICD codes, coding rules, and of the limitations imposed by case selection criteria and single cause-of-death analysis.

Air Pollutants↗

PubMed2000

Serum syndecan-1: a new independent prognostic marker in multiple myeloma.

Serum samples drawn at diagnosis from 174 myeloma patients were analyzed for the presence of the heparan [corrected] sulfate proteoglycan, syndecan-1. Syndecan-1 was elevated in 79% of patients (median, 643 units/mL) compared with 40 healthy controls (median, 128 units/mL), P <.0001. Serum syndecan-1 correlated with the following: serum creatinine, secretion of urine M-component over the course of 24 hours, soluble interleukin-6 (IL-6) receptor, C-terminal telopeptide of type I collagen, beta(2)-microglobulin, percentage of plasma cells in the bone marrow, disease stage, and serum M-component concentration. In order to evaluate syndecan-1 as a prognostic marker in multiple myeloma, it was entered into a multivariate Cox regression model. Data from 138 patients were available for this analysis. As a continuous variable, syndecan-1 was an independent prognostic parameter in addition to serum beta(2)-microglobulin and World Health Organization performance status. When syndecan-1 was dichotomized by the best cutoff (66th percentile, 1170 units/mL), the survival difference between the groups was highly significant: "high" syndecan-1 group had a median survival of 20 months, and the "low" syndecan-1 group had a median of 44 months (P <.0001). We conclude that syndecan-1 is a new independent prognostic parameter in multiple myeloma, and its role in prognostic classification systems should be further investigated. (Blood. 2000;95:388-392)

Biomarkers↗

PubMed2002

Prediction of protein solvent accessibility using support vector machines.

A Support Vector Machine learning system has been trained to predict protein solvent accessibility from the primary structure. Different kernel functions and sliding window sizes have been explored to find how they affect the prediction performance. Using a cut-off threshold of 15% that splits the dataset evenly (an equal number of exposed and buried residues), this method was able to achieve a prediction accuracy of 70.1% for single sequence input and 73.9% for multiple alignment sequence input, respectively. The prediction of three and more states of solvent accessibility was also studied and compared with other methods. The prediction accuracies are better than, or comparable to, those obtained by other methods such as neural networks, Bayesian classification, multiple linear regression, and information theory. In addition, our results further suggest that this system may be combined with other prediction methods to achieve more reliable results, and that the Support Vector Machine method is a very useful tool for biological sequence analysis.

Bayes Theorem↗

PubMed1991

Adjustment for non-differential misclassification error in the generalized linear model.

It is well known that estimates of association between an outcome variable and a set of categorical covariates, some of which are measured with misclassification, tend to be biased upon application of the usual methods of estimation that ignore the classification error. We propose a method to adjust for misclassification in covariates when one applies the generalized linear model. In the case where one can observe some true covariates only through surrogates, we combine a latent class analysis with the approach to incorporate multiple surrogates into the model. We include discussion on the efficacy of repeated measurements which one can view as a special case of multiple surrogates with identical distribution. We provide two examples to demonstrate the applicability of the method and the efficacy of multiple replicates for a covariate subject to misclassification in a regression framework.

Adolescent↗

PubMed1998

Association of insulin resistance to electrocardiographic changes in non obese Asian Indian subjects with hypertension.

To investigate the relationship between insulin resistance and electrocardiographic changes in hypertension in the absence of confounding influences, plasma glucose and insulin responses to oral glucose were studied in 26 normotensive and 38 hypertensive subjects. Resting ECG was taken and classified into normal or abnormal using the Minnesota code. Among the 38 subjects, 16 had ECG abnormalities. All the hypertensive subjects had normal glucose tolerance. Serum insulin response of hypertensive subjects with ECG changes was 43% higher than those of hypertensive subjects without ECG changes and of normotensive subjects. The ratio AUC glucose/AUC insulin, a measure of insulin sensitivity was significantly reduced in subjects with abnormal ECG. Serum LDL cholesterol was significantly elevated and was the highest in hypertensive subjects with abnormal ECG. The ratio, Total Cholesterol/HDL Cholesterol was elevated to 5.81+/-0.47. I(125)-insulin binding to erythrocytes from 6 normotensive subjects, and 16 hypertensive subjects (8 with and 8 without ECG abnormalities) indicated 50% reduction in insulin receptor number in both the groups of hypertensive subjects compared to normotensive subjects. Multiple logistic regression analysis using mean blood pressure, serum total cholesterol, LDL cholesterol/HDL cholesterol, sex, insulin level at 60 min in OGTT, treatment, serum triglyceride, presence of family history of diabetes, CHD, hypertension and tobacco as independent variables causing ECG changes, revealed correct classification in 84% of cases. Among the variables, insulin level in OGTT contributed the most to ECG abnormalities. The data suggest that in the non obese, non diabetic Asian Indian hypertensive subjects, the presence of electrocardiographic abnormalities might be partly related to hyperinsulinemia or insulin resistance in them.

Area Under Curve↗

PubMed2006

[Retrospective application of adjusted clinical groups (ACGs) at a primary care centre].

OBJECTIVE: To evaluate retrospectively the application of adjusted clinical groups (ACGs) at a primary care centre. DESIGN: Descriptive, retrospective examination. SETTING: Urban. PARTICIPANTS: All patients seen by the team in 2000 were included in the study. Those who moved or died during the study period were excluded. MAIN MEASUREMENTS: Universal variables (age and sex), dependant variables (visits, episodes, and costs), and case-mix or comorbidity variables (ACGs 2.0) were measured. The model of cost per patient was established by distinguishing the costs of the PCC from the variables. The ICPC was converted to the ICD-9-CM and a multiple linear regression analysis was performed to predict the models. RESULTS: The total number of patients studied was 15 983, with an average of 5.0+/-3.2 episodes and 8.0+/-7.7 visits during the year. The power of explanation of the variability of the classification between the number of episodes was 71.9%; of the visits, 50.0% (with refinement, 56.3%); and cost, 30.2% (with refinement, 55.0%) (P=.000). CONCLUSIONS: ACGs were shown to be an acceptable system for classifying patients according to the consumption of resources used in primary care. In addition, the methodology used was adequate for integrating clinical and economic information at the PCC.

Adult↗

PubMed1989

The epidemiology of pulmonary function and COPD mortality in the multiple risk factor intervention trial.

The potential determinants of the changes in chronic obstructive pulmonary disease (COPD) mortality were evaluated using both the Multiple Risk Factor Intervention Trial (MRFIT) screenees, the longitudinal analysis of the participants, and the differences in special intervention (SI) and usual care (UC) groups. COPD was the underlying cause for only one third of all death certificates listing COPD. Small changes in classification will have a major impact on reported COPD death rates. Cigarette smoking is clearly the primary determinant of COPD mortality. Decreased pulmonary function is an independent risk factor for coronary heart disease (CHD) mortality. Smoking cessation results in a slower rate of decline in pulmonary function over time, especially among heavy smokers. Careful evaluation of smoking cessation, including repeat chemical measures, suggest that the percentage of long-term quitters, especially among heavy smokers has been overestimated. The low percentage of quitters substantially reduced the power to detect an intervention effect. The increased cigarette smoking among recent older cohorts, and failure to substantially reduce smoking, especially among heavy smokers, may be an important factor accounting for the failure to note a decline in COPD mortality among older persons.

Adult↗

PubMed2001

Attentional and neuromotor deficits in ADHD.

In order to classify attention-deficit-hyperactivity disorder (ADHD) in 11-year-old children, the role of specific attentional and motor deficits was examined. Participants comprised 22 children with ADHD (19 male, 3 female; median age 11 years, range 8.8 to 13.5 years) and 20 control children (17 male, 3 female; median age 10.6 years, range 8.2 to 12.6 years). Neuromotor assessment indicated that while both groups needed more time to complete finger compared to hand movements, this increase was more pronounced in children with ADHD. Reaction-time testing with continuous-force recording identified both motor and attentional deficits in children with ADHD. Longer intervals between force onset and force peak, and higher rate of responses with multiple force peaks (particularly in the bilateral condition) revealed specific deficits in the speed and quality of their motor output. Increase in errors and variability of force onsets indicated attentional deficits. Prediction analysis indicated that force-onset variability contributed significantly to group classification which was 85.7% correct. Neither neuromotor assessment nor specific motor deficits contributed significantly to classification, indicating that pure motor-speed measures play a minor role in characterizing ADHD in this age range.

Adolescent↗

PubMed1984

Biological markers for endogenous depression in series and parallel.

Several authors have shown that endogenous depressed patients have biological abnormalities which exist to a lesser extent in nonendogenous depressives and in normals, and have suggested that these biological abnormalities may be used as markers to classify depressed patients as endogenous or nonendogenous. We have investigated several of these biological abnormalities, and have explored the diagnostic classifications based on individual markers and on two or more markers in combination. The evaluation of combinations of diagnostic tests requires different statistical techniques than does the evaluation of single tests, and we discuss two such techniques: stepwise multiple regression and contingency table analysis. We report here our results with the dexamethasone suppression test (DST), EEG studies of sleep, and the amphetamine-stimulated release of growth hormone separately and in combination. The amphetamine-stimulated release of growth hormone was not diagnostically useful.

Amphetamine↗