SEARCH · PubMed Health

Results for “likelihood”

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 397 records · Page 22Linked to original sources

PubMed2005

Stress-related racial discrimination and hypertension likelihood in a population-based sample of African Americans: the Metro Atlanta Heart Disease Study.

OBJECTIVE: Exposure to racial discrimination has been identified as an adverse biopsychosocial stressor that may be related to the prevalence of hypertension in African Americans. The overall objective of this research was to estimate aspects of the effects of self-reported exposure to stress-related racial discrimination and hypertension likelihood in a sample of African Americans. DATA: Physiologic and self-reported cross-sectional data were collected on a stratified sample of hypertensive (n=174) and normotensive (n=182) African Americans, aged > or =21 years, residing in metropolitan Atlanta, Georgia from 1999 through 2001. METHOD: Separate, multivariate logistic regression models were fit, including relevant covariates, to ascertain the effects of exposure to incidents of stress-related racial discrimination and level of generally derived stress with hypertension likelihood. RESULTS: Exposure to incidents of racial discrimination was not significantly associated with an increased likelihood of hypertension. More than 74% of respective normotensive and hypertensive respondents reportedly experienced stress-provoking encounters of racial discrimination. Magnitude of stress generally derived from exposure was, however, a highly significant predictor. Respondents reporting "moderate" and a "high to very high" level of derived stress were more than twice as likely to be hypertensive when compared to those reporting "no to low" derived stress (P value=.02 and .01, respectively). CONCLUSION: Exposure to racial discrimination is a prevalent psychosocial stressor in African Americans but may not be significantly associated with hypertension prevalence; degree of stress derived from encounters may be an important determinant. More research is needed to clarify the complex relationship between stress-related racial discrimination and hypertension in African Americans.

Adult↗

PubMed1989

Generalized likelihood ratio concept and logistic regression analysis for multiple diagnostic categories.

Albert (Clin Chem 1982;28:1113-9) has proposed estimation of likelihood ratios by logistic regression analysis. The usual likelihood-ratio approach for estimation of post-test probability of disease from sensitivity and specificity data of a diagnostic test has been extended by Birkett (J Clin Epidemiol 1988; 41:491-4) for situations with more than two diagnostic categories. We suggest here a combination of these ideas, demonstrating this by a re-evaluation of previously published data on the validity of neopterin as a tool for differential diagnosis between chronic non-A, non-B hepatitis and fatty liver. Analysis of neopterin data in combination with the ratio between serum concentrations of aspartate aminotransferase and of alanine aminotransferase yielded a good discrimination between three mutually exclusive diagnostic categories, namely, fatty liver and chronic persistent and chronic aggressive non-A, non-B hepatitis. The approach is flexibly applicable to situations with different pre-test probabilities. The sum of estimated post-test probabilities deviates slightly from the sum of pre-test probabilities. This deviation is a function of the coefficients obtained in logistic regression, and an analytical expression for the deviation is given. The generalized likelihood-ratio approach appears promising in complex diagnostic situations when multiple diagnostic tests are available.

Alanine Transaminase↗

PubMed1988

Simulated percentage points for the null distribution of the likelihood ratio test for a mixture of two normals.

We find the percentage points of the likelihood ratio test of the null hypothesis that a sample of n observations is from a normal distribution with unknown mean and variance against the alternative that the sample is from a mixture of two distinct normal distributions, each with unknown mean and unknown (but equal) variance. The mixing proportion pi is also unknown under the alternative hypothesis. For 2,500 samples of sizes n = 15, 20, 25, 40, 50, 70, 75, 80, 100, 150, 250, 500, and 1,000, we calculated the likelihood ratio statistic, and from these values estimated the percentage points of the null distributions. Our algorithm for the calculation of the maximum likelihood estimates of the unknown parameters included precautions against convergence of the maximization algorithm to a local rather than global maximum. Investigations for convergence to an asymptotic distribution indicated that convergence was very slow and that stability was not apparent for samples as large as 1,000. Comparisons of the percentage points to the commonly assumed chi-squared distribution with 2 degrees of freedom indicated that this assumption is too liberal; i.e., one's P-value is greater than that indicated by chi 2(2). We conclude then that one would need what is usually an unfeasibly large sample size (n greater than 1,000) for the use of large-sample approximations to be justified.

Biometry↗

PubMed1987

Parental and sib likelihoods in genealogy reconstruction.

Estimation of genealogical relationships from genetic data provides an approach to answering many questions of population biology. In this estimation problem, previous work has shown that bilateral relatives such as full sibs may be more likely parents than the true parent individuals. This paper investigates the source of this paradox, and the circumstances under which it may arise. Alternative approaches to inferring parentage are proposed; these involve analysis of the distributions of log-likelihood statistics and of the bivariate distribution of sib and parent log-likelihoods. A Bayesian approach may also be superposed on the likelihood analysis, enabling any prior knowledge of the population to be incorporated.

Alleles↗

PubMed1986

The likelihood ratio. An improved measure for reporting and evaluating diagnostic test results.

The principles and use of likelihood ratios are presented as an aid to the reporting and interpretation of diagnostic data. Likelihood ratios can be determined for multiple levels of test results and provide a readily comprehensible and convenient measure for computing the posttest probability of disease. Unlike sensitivity and specificity, they convey an immediate and direct description of a test's ability to revise the initial probability of disease upward or downward. By incorporating all the raw data, the use of likelihood ratios can improve the assessment of the clinical utility of test results. Their application to the diagnostic performance of serum creatine kinase concentrations in predicting patients with suspected myocardial infarction is discussed.

Creatine Kinase↗

PubMed1984

A comparative study of conditional maximum likelihood estimation of a common odds ratio.

The finite-sample properties of various point estimators of a common odds ratio from multiple 2 X 2 tables have been considered in a number of simulation studies. However, the conditional maximum likelihood estimator has received only limited attention. That omission is partially rectified here for cases of relatively small numbers of tables and moderate to large within-table sample sizes. The conditional maximum likelihood estimator is found to be superior to the unconditional maximum likelihood estimator, and equal or superior to the Mantel-Haenszel estimator in both bias and precision.

Biometry↗

PubMed1984

Usefulness of synovial fluid analysis in the evaluation of joint effusions. Use of threshold analysis and likelihood ratios to assess a diagnostic test.

This study applied threshold analysis and likelihood ratios to determine the usefulness of a diagnostic test. Eleven staff rheumatologists or rheumatology fellows provided probability estimates for the most likely diagnoses both before and after synovial fluid analyses were performed on 180 patients with joint effusions. They also indicated whether the planned therapy was altered by the test results. The therapeutic thresholds and log likelihood ratios were derived for the six most frequent diagnoses. Synovial fluid analysis was most useful for patients likely to have gout, pseudogout, or infectious arthritis. The derived therapeutic thresholds were consistent with recommended medical practice, for example, with a lower threshold for possible septic arthritis (20%) than for possible gout (65%). This study demonstrates that threshold analysis and likelihood ratios can be used to assess the clinical contribution of diagnostic tests.

Arthritis, Infectious↗

PubMed1993

Phylogenetic inference: linear invariants and maximum likelihood.

We develop a new statistical method for inferring phylogenies, based on a likelihood ratio test. This method does not require parameter constraints but does require identical evolutionary processes in the sites considered. Another method of phylogenetic inference is the method of linear invariants, described by Cavender (1989, Molecular Biology and Evolution 6, 301-316), based on a notion of Lake (1987, Molecular Biology and Evolution 4, 167-191). We describe a sound mathematical basis for the use of linear invariants. We show that the validity of the method requires parameter constraints, but does not require that the evolutionary processes in differing sites be identical. We show that the method of linear invariants is asymptotically equivalent to a less powerful version of our likelihood ratio test, and is thus essentially a maximum likelihood technique.

Animals↗

PubMedDate not supplied

[Determination of likelihood of the development of pleural empyema in patients with cancer after pneumonectomy].

Management of 150 patients with lung cancer after pneumonectomy has shown that the basic factors causing the onset of postoperative pleural empyema are chronic nonspecific pulmonary diseases, pronounced obstructive syndrome and right heart overload, tumor histology, local tumor size and cancer complications, duration and techniques of surgery, duration of controlled lung ventilation, as well as the onset of postoperative suppurative tracheobronchitis and surgical wound suppuration. The likelihood of empyema development may be assessed only upon analysis of the combination of risk factors with reference to their low informative value, preoperative patients' condition and intraoperative factors which have an equal impact on the onset of this complication. The efficacy of the algorithm elaborated is as follows: if the likelihood of empyema development exceeds 50%, it occurs in 96.3% of cases, if the likelihood of the complication is less than 10%, the postoperative period is uncomplicated in 87.8% of patients.

Algorithms↗

PubMed1994

On the use of the quasi-likelihood method in teratological experiments.

This paper concerns the impact of litter effects on the inference of dose-response relationships in teratological experiments with binary response. Kupper et al. (1986, Biometrics 42, 85-98) concluded that when intra-litter correlations are dose-dependent, the use of a common intra-litter correlation likelihood based on the beta-binomial distribution could lead to severe bias in estimation of the parameter beta, which characterizes the dose-response relationship. We show here that the problems of bias and coverage probability for beta could still be substantial when one uses the likelihood with heterogeneous intra-litter correlations. We then examine through a simulation study the performance of the quasi-likelihood method (Wedderburn, 1974, Biometrika 61, 439-447) and recommend that this method with a common intra-litter correlation parameter be used when the number of litters is small or modest.

Analysis of Variance↗

PubMed1998

Does a visible retinal embolus increase the likelihood of hemodynamically significant carotid artery stenosis in patients with acute retinal arterial occlusion?

OBJECTIVE: To determine the value of visible retinal emboli as a diagnostic "test" for the detection of hemodynamically significant carotid artery stenosis in the setting of acute retinal artery occlusion. METHODS: A cross-sectional diagnostic accuracy study was performed in a tertiary North American center, with the results of the dichotomous diagnostic test (the presence or absence of visible retinal emboli) being placed against the dichotomous outcome of the presence or absence of hemodynamically significant carotid artery stenosis (defined as > or = 60%, or < 60%, carotid artery stenosis on either side). RESULTS: Forty-eight (18.7%) of our 256 patients had hemodynamically significant carotid artery stenosis. The sensitivity and specificity of retinal emboli for the detection of hemodynamically significant carotid artery stenosis were 39% and 68%, respectively. The presence of a visible retinal embolus generated a likelihood ratio of 1.24 (95% confidence interval, 0.84-1.86). This value corresponds to a patient with a pretest probability of 50% having a posttest probability of 55.3%. The absence of a visible retinal embolus generated a likelihood ratio of 0.88 (95% confidence interval, 0.68-1.15). CONCLUSIONS: The presence of a visible retinal embolus is a poor diagnostic test for the detection of hemodynamically significant carotid artery stenosis in the setting of acute retinal artery occlusion. Accordingly, the presence of an embolus should not influence the decision to perform carotid Doppler ultrasonography in patients with acute retinal arterial occlusion.

Adult↗

PubMed1998

Stature estimation and calibration: Bayesian and maximum likelihood perspectives in physical anthropology.

Many applied problems in physical anthropology involve estimation of an unobservable quantity (such as age at death or stature) from quantities that are observable. Two of the more disparate subdisciplines of our discipline, paleoanthropology and forensic anthropology, routinely make use of various estimation methods on a case-by-case basis. We discuss the rationales for making estimations on isolated cases, taking stature estimation from femoral and humerus lengths as an example. We show that the entirety of our discussion can be placed within the context of calibration problems, where a large calibration sample is used to estimate an unobservable quantity for a single skeleton. Taking a calibration approach to the problem highlights the essentially Bayesian versus maximum likelihood nature of the question of stature estimation. On the basis of both theoretical arguments and practical examples, we show that inverse calibration (regression of stature on bone length) is generally preferred when the stature distribution for a reference sample forms a reasonable prior, while classical calibration (regression of bone length on stature followed by solving for stature) is preferred when there is reason to suspect that the estimated stature will be an extrapolation beyond the useful limits of the reference sample statures. The choice between these two approaches amounts to the decision to use either a Bayesian or a maximum likelihood method.

Anthropology, Physical↗

PubMedDate not supplied

The relationship between hot-deck multiple imputation and weighted likelihood.

Hot-deck imputation is an intuitively simple and popular method of accommodating incomplete data. Users of the method will often use the usual multiple imputation variance estimator which is not appropriate in this case. However, no variance expression has yet been derived for this easily implemented method applied to missing covariates in regression models. The simple hot-deck method is in fact asymptotically equivalent to the mean-score method for the estimation of a regression model parameter, so that hot-deck can be understood in the context of likelihood methods. Both of these methods accommodate data where missingness may depend on the observed variables but not on the unobserved value of the incomplete covariate, that is, missing at random (MAR). The asymptotic properties of hot-deck are derived here for the case where the fully observed variables are categorical, though the incomplete covariate(s) may be continuous. Simulation studies indicate that the two methods compare well in small samples and for small numbers of imputations. Current users of hot-deck may now conduct their analysis using mean-score, which is a weighted likelihood method and can thus be implemented by a single pass through the data using any standard package which accommodates weighted regression models. Valid inference is now straightforward using the variance expression provided here. The equivalence of mean-score and hot-deck is illustrated using three clinical data sets where an important covariate is missing for a large number of study subjects.

Angioplasty, Balloon, Coronary↗

PubMed1997

Variation of sensitivity, specificity, likelihood ratios and predictive values with disease prevalence.

The sensitivity, specificity and likelihood ratios of binary diagnostic tests are often thought of as being independent of disease prevalence. Empirical studies, however, have frequently revealed substantial variation of these measures for the same diagnostic test in different populations. One reason for this discrepancy is related to the fact that only few diagnostic tests are inherently dichotomous. The majority of tests are based on categorization of individuals according to one or several underlying continuous traits. For these tests, the magnitude of diagnostic misclassification depends not only on the magnitude of the measurement or perception error of the underlying trait(s), but also on the distribution of the underlying trait(s) in the population relative to the diagnostic cutpoint. Since this distribution also determines prevalence of the disease in the population, diagnostic misclassification and disease prevalence are related for this type of test. We assess the variation of various measures of validity of diagnostic tests with disease prevalence for simple models of the distribution of the underlying trait(s) and the measurement or perception error. We illustrate that variation with disease prevalence is typically strong for sensitivity and specificity, and even more so for the likelihood ratios. Although positive and negative predictive values also strongly vary with disease prevalence, this variation is usually less pronounced than one would expect if sensitivity and specificity were independent of disease prevalence.

Data Interpretation, Statistical↗

PubMed2000

Comparing diagnostic tests: a simple graphic using likelihood ratios.

The diagnostic abilities of two or more diagnostic tests are traditionally compared by their respective sensitivities and specificities, either separately or using a summary of them such as Youden's index. Several authors have argued that the likelihood ratios provide a more appropriate, if in practice a less intuitive, comparison. We present a simple graphic which incorporates all these measures and admits easily interpreted comparison of two or more diagnostic tests. We show, using likelihood ratios and this graphic, that a test can be superior to a competitor in terms of predictive values while having either sensitivity or specificity smaller. A decision theoretic basis for the interpretation of the graph is given by relating it to the tent graph of Hilden and Glasziou (Statistics in Medicine, 1996). Finally, a brief example comparing two serodiagnostic tests for Lyme disease is presented. Published in 2000 by John Wiley & Sons, Ltd.

Diagnostic Tests, Routine↗

PubMed1999

Likelihood-based approach to estimating twin concordance for dichotomous traits.

Genetic epidemiologists are well aware that the casewise and pairwise twin concordances are two different measures. In determining appropriate estimators for each of these measures, the method of ascertainment must be considered. Here, we derive expressions for the concordance estimators and their asymptotic variances appropriate to different twin ascertainment schemes using a likelihood framework, and apply these formulas to existing data. We emphasize the distinction between concordance measures (i.e., the parameters of interest) and the concordance estimators based on the number of pairs observed. Under random or complete ascertainment the casewise estimator is asymptotically unbiased for the casewise concordance, and the pairwise estimator is asymptotically unbiased for the pairwise concordance. Under incomplete ascertainment, the casewise estimator is biased for the casewise concordance, the pairwise estimator is biased for the pairwise concordance, but the probandwise estimator is asymptotically unbiased for the casewise concordance. One can extend the likelihood equations presented here to allow the concordance parameter of interest to depend on zygosity and, if measured, other factors such as cohabitation status and similarity for genetic markers, while concurrently allowing the disease prevalence to depend on measured covariates.

Humans↗

PubMedDate not supplied

Non-parametric maximum likelihood estimators for disease mapping.

A Non-Parametric Maximum Likelihood approach to the estimation of relative risks in the context of disease mapping is discussed and a NPML approximation to conditional autoregressive models is proposed. NPML estimates have been compared to other proposed solutions (Maximum Likelihood via Monte Carlo Scoring, Hierarchical Bayesian models) using real examples. Overall, the NPML autoregressive estimates (with weighted term) were closer to the Bayesian estimates. The exchangeable NPML model ranked immediately after, even if it implied a greater shrinkage, while the truncated auto-Poisson showed inadequate for disease mapping. The coefficients of the autoregressive term for the different mixtures have clear interpretations: in the breast cancer example, the larger cities in the region showed high rates and very low correlation with the neighbouring areas, while the less populated rural areas with low rates were strongly positively correlated each other. This pattern is expected since breast cancer is strongly correlated with parity and age at first birth, and the female population of the rural areas experienced a decline in fertility much later than those living in the larger cities. The leukemia example highlighted the failure of the Poisson-Gamma model and other general overdispersion tests to detect high risk areas under specific conditions. The NPML approach in Aitkin is very general, simple and flexible. However the user should be warned against the possibility of local maxima and the difficulty in detecting the optimal number of components. Special software (such as CAMAN or DismapWin) had been developed and should be recommended mainly to not experienced users.

Algorithms↗

PubMed2000

Maximum likelihood estimation for longitudinal data with truncated observations.

We obtain maximum likelihood estimates of the parameters when the observations on the response variable in a repeated measures design are truncated above a cutpoint. The maximum likelihood equations are solved iteratively using an EM-like procedure. It is observed that these estimates have smaller mean squared error than recently proposed iterative weighted least-squares estimates. The results are applied to data arising from a study of dioxin elimination in Air Force veterans. Published in 2000 by John Wiley & Sons, Ltd.

Aerospace Medicine↗