PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Conditional logit”

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 73 records · Page 4Linked to original sources

Analysis of matched case-control data with multiple ordered disease states: possible choices and comparisons.

In an individually matched case-control study, effects of potential risk factors are ascertained through conditional logistic regression (CLR). Extension of CLR to situations with multiple disease or reference categories has been made through polychotomous CLR and is shown to be more efficient than carrying out separate CLRs for each subgroup. In this paper, we consider matched case-control studies where there is one control group, but there are multiple disease states with a natural ordering among themselves. This scenario can be observed when the cases can be further classified in terms of the seriousness or progression of the disease, for example, according to different stages of cancer. We explore several popular models for ordered categorical data in this context. We first adopt a cumulative logit or equivalently, a proportional-odds model to account for the ordinal nature of the data. The important distinction of this model from a stratified dichotomous and polychotomous logistic regression model is that the stratum-specific nuisance parameters cannot be eliminated in this model via the conditional-likelihood approach. We discuss a Mantel-Haenszel approach for analysing such data. We point out possible difficulties with standard likelihood-based approaches with the cumulative logit model when applied to case-control data. We then consider an alternative conditional adjacent-category logit model. We illustrate the methods by analysing data from a matched case-control study on low birthweight in newborns where infants are classified according to low and very low birthweight and a child with normal birthweight serves as a control. A simulation study compares the different ordinal methods with methods ignoring sub-classification of the ordered disease states.

Case-Control Studies↗

Distribution-free fitting of logit models with random effects for repeated categorical responses.

This article discusses random effects models for within-subject comparisons of repeated responses on the same categorical scale. The models account for the correlation that normally occurs between repeated responses. The standard way of fitting such models maximizes the marginal likelihood after integrating with respect to a distribution for the random effect. An alternative non-parametric approach does not assume a distributional form for the random effects. Recent literature shows that for certain simple logit models, this approach yields essentially the same model parameter estimates as conditional maximum likelihood. Moreover, these estimates also result from fitting corresponding quasi-symmetric log-linear models. For simple data sets in which primary interest relates to subject-specific comparisons of the repeated responses, one can easily obtain the estimates with standard software for log-linear models. Examples include data from crossover designs and from comparisons of treatment and control groups regarding the change between baseline and follow-up observations.

Clinical Trials as Topic↗

The increasing proportion of men with low earnings in the United States.

This study has provided an examination of recent changes in the lower tail of the male earnings distribution. Data from the CPS for 1967 through 1978 were used to analyze the increasing proportion of male workers with annual and weekly earnings below a fixed low earnings threshold. Our central purpose was to assess the extent to which the growth in the probability of low earnings could be explained by the more salient changes in the structure of the male labor force over this period. To this end, logit analysis was used to examine the roles of education, experience, cyclical conditions, and cohort size in explaining variations in the probability of subthreshold earnings, conditional on experience and education. The estimates generally yielded the expected effects. However, the most important findings from our analysis concern the trends estimated net of education, experience, unemployment, and cohort size. These variables appear to explain satisfactorily the recent growth in the proportion of men with low earnings among those with at least sixteen years of education. For all other educational categories, our independent variables were unable to account for a major portion of the growth in the probability of low earnings. Our results supplement previous findings of positive trends for mean annual and weekly earnings net of a similar set of independent variables. Hence, we have provided substantial evidence of stagnation in the lower tail of the male earnings distribution--a stagnation not shared by the average worker nor fully explicable by education, experience, aggregate unemployment, or the entrance of the baby boom cohort into the labor market. Investigation of alternative explanations for this phenomenon, such as changes in female labor supply or the structure of labor demand, is clearly warranted.

Adolescent↗

Discussing patient management online: the impact of roles on knowledge construction for students interning at the paediatric ward.

The objectives of this study are to explore the use of asynchronous discussion groups during medical students' clinical rotation in paediatrics. In particular, the impact of role assignment on the level of knowledge construction through social negotiation is studied. Case-based asynchronous discussion groups were introduced to enhance reflection and critical thinking on patient management and treatment, and to offer an exercise in evidence-based medical practice. Groups of approximately 4-5 students were asked to discuss 4 authentic cases during clinical rotation in paediatrics. 49 students interning at the paediatric ward participated in this study. With respect to role assignment, differences between groups (1) with a student or an instructor as moderator and (2) with or without a developer of alternatives for patient management were explored. A content analysis was performed to explore the different levels of social construction of knowledge. The results of multilevel logit analyses show a significant difference in knowledge construction through social negotiation between conditions with a student moderator and conditions where the instructor is moderating, but only when a developer of alternatives is involved. No significant difference was revealed between student-moderated and instructor-moderator groups without a developer of alternatives. It can be concluded that when both the moderator and developer role are assigned to students, their contributions are more likely to reflect a high level of knowledge construction.

Adult↗

Correlates of accuracy in the assessment of psychiatric inpatients' risk of violence.

OBJECTIVE: The authors evaluated characteristics of patients whom clinicians accurately assessed as being at high or low risk for violence and patients for whom clinicians overestimated or underestimated the risk. METHOD: At admission, physicians estimated the probability that each of 226 psychiatric inpatients would physically attack someone during the first week of hospitalization. Nurses rated assaultive behavior in the hospital with the Overt Aggression Scale. Acute symptoms were rated with the Brief Psychiatric Rating Scale. RESULTS: For the group as a whole, assessed levels of risk were substantially related to later physical aggression (sensitivity = 67%, specificity = 69%). Multinomial logit analysis showed that patients with psychotic disorders such as schizophrenia, organic psychotic conditions, and mania were more likely to be accurately assessed by clinicians as being at high risk (true positives) than to be true negatives or false positives. A recent history of violence was associated with higher estimated risk but did not distinguish true positives from false positives. An admission mental status characterized by low levels of hostility, uncooperativeness, and suspiciousness and high levels of depression, guilt, and anxiety differentiated true negative patients from others, but symptom profiles did not differ among true positives, false positives, and false negatives. Clinical judgments emphasizing gender and race/ethnicity were associated with predictive errors: nonwhite and male patients tended to be false positives. CONCLUSIONS: While clinicians can accurately classify the potential for violence in the majority of patients at admission, systematic errors characterize inaccurate assessments of the risk. Awareness of these patterns may help improve assessment of the risk of violence in clinical practice.

Adult↗

Causal explanations for class inequality in health--an empirical analysis.

One of the most important issues for research on social class inequalities in health are the causes behind such differences. So far, the debate on class inequalities in health has mainly been centred around hypotheses on artefactual and selectional processes. Although most contributors to this branch of research have argued in favour of causal explanations, these have gained very little systematic scrutiny. In this article, several possible causal factors are singled out for empirical testing. The effect of these factors on class differences in physical and mental illness is studied by means of logit regressions. On the basis of these analyses, it is shown that physical working conditions are the prime source of class inequality in physical illness, although economic hardship during upbringing and health related behaviours also contribute. For class inequality in mental illness these three factors plus weak social network are important. In sum, a large part of the class differences in physical as well as mental illness can be understood as a result of systematic differences between classes in living conditions, primarily differences in working conditions.

Adolescent↗

Multivariate analysis for matched case-control studies.

A multivariate method based on the linear logistic model is presented for the analysis of case-control studies with pairwise matching. This technique enables one to investigate the effect of several variables simultaneously in the analysis while allowing for the matched design. The odds ratio is used as the basic measure of risk. One is able to control for variables which are not matching variables while investigating the odds ratio for a particular factor, and to estimate the change in the odds ratio as the level of one or more interval variables changes. The computing methods used for obtaining maximum conditional likelihood estimates of the parameters of interest are modifications of standard programs for logit regression.

Epidemiologic Methods↗

Long-term benzodiazepine use by elderly people living in the community.

OBJECTIVE: To investigate the prevalence of long-term benzodiazepine use in an elderly community sample, and factors associated with such use. METHOD: Data came from the Sydney Older Persons Study, a longitudinal study of people aged 75 or over. There were 337 subjects who were interviewed in 1991-93, and subsequently followed up after three and 4.5 years. At the first interview, subjects were assessed for socio-demographic characteristics, physical and mental health, and use of health services. At the first and subsequent interviews, subjects were asked about use of medications, including benzodiazepines. RESULTS: There were 16.6% who were using benzodiazepines at the time of all three interviews, while a further 19.6% were using them at one or two interviews. In a multivariate ordered logit regression model, long-term benzodiazepine use was associated with treatment for nervous conditions, restless sleep, being female, being divorced and greater contact with medical services. CONCLUSIONS: The prevalence of benzodiazepine use in the elderly is high and much of this use is long term. The high prevalence of benzodiazepine use stands in contrast to the findings from national surveys that the elderly living in the community tend to have better mental health than younger age groups. IMPLICATIONS: Efforts are needed to reduce the number of elderly people becoming long-term users. The use of benzodiazepines in this age group is of particular concern, because they may be a risk factor for falls and for cognitive impairment in the elderly.

Aged↗

Small sample properties of some parametric bioassay estimators of the LD90.

Bioassay experiments are frequently conducted with small sample sizes, but the literature offers little comparative information for small sample sizes among the analytical procedures. A computer simulation study was performed to compare several traditional analytical procedures for estimating the LD90. These estimation procedures include probit analysis with maximum likelihood estimation, logit analysis with maximum likelihood, and minimum chi 2 estimation methods. The stimulation results indicate the conditions under which each analytical method appears most useful. Recommendations are tentatively made for minimally adequate designs for bioassay studies where an LD90 is to be estimated.

Animals↗

Pressure ulcer risk assessment: application of logistic analysis.

AIM: The aim of this study was to investigate the relative importance of key factors affecting the likelihood of hospital acquired pressure ulcers, thus offering the groundwork for the development of an empirically-based risk assessment scale. It also evaluated the predictive performance of the underlying empirical model and compared its classification ability with the empirically observed ability of the Waterlow risk assessment scale. BACKGROUND: Pressure ulcers impose a significant burden on patients and carers and demand substantial resources from the health care system. There is, however, a lack of systematic empirical research on which to base the development of risk assessment measurement tools. METHODS: Multivariate statistical methods were applied to data derived from the records of a cross-sectional sample of around 500 randomly selected inpatient episodes drawn from the population of admissions to a single unit in a large acute hospital during a 2-year period (October 1996 to September 1998). Double-checking of a randomly selected 25% sample of the original records and careful screening out of records with incomplete information or inconsistent values was carried out to ensure a high quality sample. Logit analysis was used to investigate the relative contribution of risk factors, such as continence, skin condition, mobility and inter-hospital transfer to the risk of hospital-acquired pressure ulcer occurrence, whilst cross-validation techniques were employed to check the predictive performance of the model. RESULTS: The results suggest that a simplified version of the Waterlow risk assessment tool has satisfactory predictive ability and the potential for further development. CONCLUSIONS: Two main conclusions emerged from this study. First, it is both feasible and worthwhile to pursue improvement in the development of risk assessment tools using statistical methods. Second, locally-determined risk factors will need to be incorporated into the construction of future risk assessment scales.

Aged↗

Analysis of failure time hierarchical data in the presence of competing risks with application to oral contraceptive pill use in Egypt.

Problems of practical interest in the analysis of data on contraceptive use, from Demographic and Health Surveys (DHS), include the estimation of the cause-specific probability of discontinuation by time t (the cumulative incidence function), in the presence of other competing causes and the evaluation of the effect of covariates on the cause-specific hazards of discontinuation. Methods of analysis of failure time data with competing risks are by now fairly well developed in the case of a simple random sample. However, the data from the DHS are clustered by geographical areas and include multiple episodes per woman. For a marginal (population average) approach, we propose using methods developed for simple random samples with standard errors calculated using a double bootstrap to take account of the clustered hierarchical nature of the data. In the conditional approach, the cause-specific hazards are modelled as log-linear functions of the covariates conditional on random effects of clusters and women, using a three-level multinomial discrete-time logit model. The methods are applied to data from Egypt 1992 DHS on the oral contraceptive pill use.

Adolescent↗

Seat belt usage rates: a test of Peltzman's hypothesis.

Despite the universally accepted belief that the use of seat belts would have a significant impact upon the number of vehicular fatalities, current evidence indicates that relatively few drivers employ their seat belts. Various hypotheses have been offered to explain this phenomenon, many of which conclude that a driver's decision to use a seat belt is independent of the risk experienced in his trip making. This paper develops an economic model which focuses upon the relationship between driver use of seat belts and the travel conditions under which trips are made. Using data obtained from a national survey of households, a binary logit model is developed to test the hypothesis that seat belt usage is influenced by the level of risk experienced in one's trip making. The estimation results were consistent with the underlying hypothesis that individuals travelling in more risky environments are more likely to use their seat belts. The econometric results are then employed to examine various policy issues, including the predicted use of seat belts for population sub-groups, driver response to the introduction of a small urban car, and the impact upon the probability of a fatality resulting from vehicle fleet downsizing.

Accidents, Traffic↗

Development of a radioimmunoassay for papain.

Various well-tried radioimmunoassay (RIA) techniques were compared for the quantitation of papain. The evaluation of individual assays was performed by logit-log analysis. The most compatible analytical steps were combined in order to obtain the optimal analytical conditions of the assay. The preferred RIA involves papain labelling with lactoperoxidase, a double antibody method as the separation step and a 24 h incubation period at 2 degrees C. It permits the detection of 16 ng of papain per tube. In contrast, a method using immobilized antibody was satisfactory for rapid quantitation of papain with quite acceptable accuracy.

Antibody Affinity↗

Variation in severity of respiratory syncytial virus infections with subtype.

Two major subtypes of respiratory syncytial virus have been identified. This study assessed the hypothesis that A-subtype infections were more severe than B-subtype infections among the 157 infants hospitalized in two hospitals in Rochester, N.Y., during two winters. Severity was measured both by specific clinical observations and by a severity index that was derived empirically. Among all subjects, several clinical observations suggested that A-subtype infections were more severe. For example, mechanical ventilation was required in 12.6% of those with A-subtype compared with 1.6% of those with B-subtype infection (relative risk = 7.88; p = 0.01). Among high-risk infants (infants with underlying conditions or age 3 months or less at admission), carbon dioxide tension greater than 45 mm Hg was found in 37.0% of those with A-subtype compared with 12.0% of those with B-subtype infection (relative risk = 3.08; p = 0.04). In discrete multivariate (logit) analysis, effects of subtype (odds ratio = 6.59; p less than 0.01) on severity remained after adjustment for other statistically significant effects of age less than 3 months, underlying condition, and premature birth. The finding that A-subtype infections were more severe might have important implications for vaccine development, studies of the virulence of respiratory syncytial virus, clinical management (e.g., selection for antiviral therapy), and long-term prognosis.

Age Factors↗

A multistate Markov chain model for longitudinal, categorical quality-of-life data subject to non-ignorable missingness.

Quality-of-life (QOL) is an important outcome in clinical research, particularly in cancer clinical trials. Typically, data are collected longitudinally from patients during treatment and subsequent follow-up. Missing data are a common problem, and missingness may arise in a non-ignorable fashion. In particular, the probability that a patient misses an assessment may depend on the patient's QOL at the time of the scheduled assessment. We propose a Markov chain model for the analysis of categorical outcomes derived from QOL measures. Our model assumes that transitions between QOL states depend on covariates through generalized logit models or proportional odds models. To account for non-ignorable missingness, we incorporate logistic regression models for the conditional probabilities of observing measurements, given their actual values. The model can accommodate time-dependent covariates. Estimation is by maximum likelihood, summing over all possible values of the missing measurements. We describe options for selecting parsimonious models, and we study the finite-sample properties of the estimators by simulation. We apply the techniques to data from a breast cancer clinical trial in which QOL assessments were made longitudinally, and in which missing data frequently arose.

Antineoplastic Agents↗

Veterans and functional status transitions in older Americans.

OBJECTIVES: This research examines the relationship of veteran status with functional status transitions in older Americans. METHODS: Data for this study come from the Survey of Asset and Health Dynamics among the Oldest Old. We use a structural multinomial logit model to decompose the effect of older veterans into the direct effect and the indirect effects via physical health conditions and mental disorders on functional status transitions. RESULTS: Although there is no distinct association among those functionally independent at baseline, veteran status significantly impacts age-dependent transitions from functional dependence to other statuses. At age 85, the excess mortality and the lower level of functional resolution among functionally dependent veterans are considerable. CONCLUSIONS: Physical health is more important than mental health in transmitting the effect of veteran status on functional status transitions in functionally dependent persons.

Activities of Daily Living↗

Quality and provider choice: a multinomial logit-least-squares model with selectivity.

A Federal Trade Commission survey of contact lens wearers is used to estimate a multinomial logit-least-squares model of the joint determination of provider choice and quality of care in the contact lens industry. The effect of personal and industry characteristics on a consumer's choice among three types of providers--opticians, ophthalmologists, and optometrists--is estimated via multinomial logit. The regression model of the quality of care has two features that distinguish it from previous work in the area. First, it uses an outcome rather than a structural or process measure of quality. Quality is measured as an index of the presence of seven potentially pathological eye conditions caused by poorly fitted lenses. Second, the model controls for possible selection bias that may arise from the fact that the sample observations on quality are generated by consumers' nonrandom choices of providers. The multinomial logit estimates of provider choice indicate that professional regulations limiting the commercial practices of optometrists shift demand for contact lens services away from optometrists toward ophthalmologists. Further, consumers are more likely to have their lenses fitted by opticians in states that require the licensing of opticians. The regression analysis of variations in quality across provider types shows a strong positive selection bias in the estimate of the quality of care received by consumers of ophthalmologists' services. Failure to control for this selection bias results in an overestimate of the quality of care provided by ophthalmologists.

Adolescent↗

Looking into the demography of an iron age population in the western Mediterranean. I. Mortality.

In this paper, we attempt to reconstruct the mortality pattern of the population buried in S'Illot des Porros (Majorca), an Iron Age necropolis in the western Mediterranean, by means of paleodemographic analysis. The skeletal sample consists of 285 individuals, 93 subadults (under 20 years old) and 192 adults. The aim of this study is twofold: first, to identify and to evaluate the structural anomalies of the skeletal sample, and second, to obtain a possible and realistic description of the biological dynamics of this population, with special reference to its mortality pattern. The study uses current demographic methodology and several demographic models (for comparison). An abridged life table was built to estimate the mortality parameters. To evaluate the likelihood of the estimated data, an indirect analysis, which consisted of a comparison of our results with different population models (Weiss [1973] American Antiquity 38; Coale and Demeny [1996] Regional Model Life Tables and Stable Populations. Princeton: Princeton University Press; Ledermann [1969] Nouvelles tables-types de mortalité. Paris: Presses Universitaires de France), was carried out. An important bias was identified in the case of children, mostly affecting infants but also children between the ages of 1 and 5. This was interpreted as a census error due to taphonomic reasons and to an excluding differential funeral rite. A life expectancy at birth of approximately 28 years was estimated from the observed data. When this bias was removed, the estimated life expectancy at birth dropped to 23 years. The use of the Brass logit system allowed us to sketch a possible mortality profile for this population: low life expectancy, high infant mortality and hard life conditions, which were the cause of the low levels of survivorship in old ages. Am J Phys Anthropol 110:285-301, 1999.

Adolescent↗