Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “multilevel analysis”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 271 records · Page 15Linked to original sources
PubMed2004

Team climate at Antarctic research stations 1996-2000: leadership matters.

INTRODUCTION: The popular assumption is that extreme environments induce a climate of hostility, incompatibility, and tension by intensifying differences and disagreements among team members. Team members' perceptions of team climate are likely to change over time in an extreme environment, and thus team climate should be considered as a dynamic outcome variable resulting from multiple factors. In order to explore team climate as a dynamic outcome, we explored whether variables at multiple levels of analysis contributed to team climate over time for teams living and working in Antarctica. METHOD: Data for this study were collected from volunteers involved in Australian National Antarctic Research Expeditions conducted from 1996 to 2000. Multilevel analysis was used to partition and estimate the variance in team climate and to explore factors explaining variance at the group/team, individual, and weekly levels. RESULTS: Most of the variance in perceptions of team climate was at the individual level (57%). Team climate had less variance at the group level (16%) and at the weekly level (26%). Results indicated that perceived leadership effectiveness was significantly related to team climate. Perceived leadership effectiveness accounted for an estimated 77% of the group level variance, which equated to 14% of the overall variance in team climate. DISCUSSION: Our results suggest that exploring the characteristics and behaviors that constitute effective leadership would contribute to a more complete and useful picture of team climate, as well as guide selection research.

Adult↗
PubMed2002

Prevalence of common infections among employees in different work schedules.

This study examined the prevalence of common infections among employees in different work schedules. Self-administered questionnaire data from the Maastricht Cohort Study on "Fatigue at Work" (n = 12,140) were used. Job title was used as a matching variable between day and shift workers to control for their different work environment. We used a multilevel analysis of a two-level structure, in which the individual employees (level 1) were nested within job titles (level 2), adjusted for demographics, longstanding disease, health behavior, work-related factors, fatigue and sleep quality. Results from the multilevel analyses showed that, compared to day work, shift work was associated with a higher risk for common infections, with the highest risk in three-shift workers. Compared to day work, shift work was further associated with differences in health, health behavior, sleep, fatigue and perceived job characteristics, factors that may influence the occurrence of infections and should be taken into account in future studies as well.

Adult↗
PubMed2001

Assessing the effects of a school-based intervention on unscheduled school transfers during elementary school.

Raising Healthy Children is a cluster-randomized study of a school-based intervention aimed at preventing problem behaviors among children recruited into the project in the first or second grade of elementary school. Multilevel analysis was used to compare students in intervention and control schools with respect to whether they transferred out of their original schools. Students in intervention schools were less likely to transfer within the first 5 years of the project. A multilevel discrete-time survival model that included both time-varying and contextual variables revealed that the difference in hazard of transfer was greatest in the earlier years of the project.

Child↗
PubMed2003

[Income inequality and self rated health: an analysis from a contextual perspective in Chile].

BACKGROUND: The correlation between income inequality and life expectancy was demonstrated 10 years ago, but later, several studies suggested that the negative impact of a low economic income on the health status was disappearing. AIM: To assess the independent effects of community income inequality on self rated health in Chile. MATERIAL AND METHODS: Multilevel analysis of the 2000 National Socio Economic Characterization Survey (CASEN) data from Chile. Individual level information included self rated health, age, sex, ethnicity, marital status, education, income, type of health insurance and residential setting (urban/rural). Community level variables included the Gini coefficient and median income. The main outcome measure was dichotomized self rated health (0 if excellent, very good or good; 1 if fair or poor). RESULTS: 101,374 individuals (at level 1) aged 18 and above, nested within 285 communities (at level 2) and 13 regions (at level 3) were studied. Controlling for a range of individual level predictors, a significant gradient was observed between income and poor self rated health, with very poor most likely to report poor health (10.5%) followed by poor (9.5%) low (9%) middle (7%), high (6%) and very high (4.5%) income earners. Controlling for individual and community effects of income, a significant non linear effect of community income inequality was observed, with the most unequal communities being associated with approximately 5% higher likelihood of reporting poor health compared to the most equal communities. CONCLUSIONS: Individual income does not explain any of the between community differences and neither does it wash the significant effects of income inequality on poor self rated health. The contextual effect of inequality is almost as large as the differential observed in poor health comparing the very poor to the very rich individual income categories.

Adolescent↗
PubMed2000

Do patients matter? Contribution of patient and care provider characteristics to the adherence of general practitioners and midwives to the Dutch national guidelines on imminent miscarriage.

OBJECTIVE: To assess the relative contribution of patient and care provider characteristics to the adherence of general practitioners (GPs) and midwives to two specific recommendations in the Dutch national guidelines on imminent miscarriage. The study focused on performing physical examinations at the first contact and making a follow up appointment after 10 days because these are essential recommendations and there was much variation in adherence between different groups of providers. DESIGN: Prospective recording by GPs and midwives of care provided for patients with symptoms of imminent miscarriage. SETTING: General practices and midwifery practices in the Netherlands. SUBJECTS: 73 GPs and 38 midwives who agreed to adhere to the guidelines; 391 patients were recorded during a period of 12 months. MAIN MEASURES: Adherence to physical examinations and making a follow up appointment were measured as part of a larger prospective recording study on adherence to the guidelines on imminent miscarriage. Patient and care provider characteristics were obtained from case recordings and interviews, respectively. Multilevel analysis was performed to assess the contribution of several care provider and patient characteristics to adherence to two selected recommendations: the number of recommended physical examinations at the first contact and the number of days before a follow up appointment took place. RESULTS: In the multilevel model explaining variance in adherence to physical examinations, the care provider's acceptance of the recommendations was the most important factor. Severity of symptoms and referral to an obstetrician were significant factors at the patient level. In the model for follow up appointments the characteristics of the care provider were less important. Referral to an obstetrician and probability diagnosis were significant factors at the patient level. CONCLUSIONS: The study showed that characteristics of both the patient and care provider contribute to the variability in adherence. Furthermore, the contribution of the characteristics differed per recommendation. It is therefore advised that the contribution of both patient and care provider characteristics per recommendation should be carefully examined. If implementation is to be successful, strategies should be developed to address these specific contributions.

Abortion, Spontaneous↗
PubMed2006

Influence of individual- and area-level measures of socioeconomic status on obesity, unhealthy eating, and physical inactivity in Canadian adolescents.

BACKGROUND: Low socioeconomic status (SES) is a risk factor for obesity. However, few studies have used a multilevel analysis to determine the influence of both individual- and area-level determinants of SES on obesity, and these studies have been limited to adults. OBJECTIVE: The primary objective was to examine associations between individual- and area-level measures of SES and obesity among adolescents by using a multilevel analytic approach. A secondary objective was to examine associations between individual- and area-level measures of SES with unhealthy eating and physical inactivity. DESIGN: The study sample consisted of 6684 youth in grades 6-10 from 169 schools across Canada. Individual-level SES exposures included material wealth and perceived family wealth. Area-level SES exposures included unemployment rate, percentage of adult residents with less than a high school education, and average employment income from head of household. Associations between SES and the outcome measures were examined by using multilevel logistic regression procedures that modeled students (individual level) nested within schools (area level). RESULTS: Both individual-level and all 3 area-level SES measures were inversely associated with obesity. The odds for unhealthy eating were increased for those living in an area with a low percentage of residents with a high school education. The odds of being physically inactive increased with decreasing levels of material wealth and perception of family wealth. CONCLUSIONS: Individual- and area-level SES measures were independently related to obesity, which suggests that both individual and environmental approaches may be required to curtail adolescent obesity.

Adolescent↗
PubMed2003

Measuring [(18)F]FDG uptake in breast cancer during chemotherapy: comparison of analytical methods.

Over the years several analytical methods have been proposed for the measurement of glucose metabolism using fluorine-18 fluorodeoxyglucose ([(18)F]FDG) and positron emission tomography (PET). The purpose of this study was to evaluate which of these (often simplified) methods could potentially be used for clinical response monitoring studies in breast cancer. Prior to chemotherapy, dynamic [(18)F]FDG scans were performed in 20 women with locally advanced ( n=10) or metastasised ( n=10) breast cancer. Additional PET scans were acquired after 8 days ( n=8), and after one, three and six courses of chemotherapy ( n=18, 10 and 6, respectively). Non-linear regression (NLR) with the standard two tissue compartment model was used as the gold standard for measurement of [(18)F]FDG uptake and was compared with the following methods: Patlak graphical analysis, simplified kinetic method (SKM), SUV-based net influx constant ("Sadato" method), standard uptake value [normalised for weight, lean body mass (LBM) and body surface area (BSA), with and without corrections for glucose (g)], tumour to non-tumour ratio (TNT), 6P model and total lesion evaluation (TLE). Correlation coefficients between each analytical method and NLR were calculated using multilevel analysis. In addition, for the most promising methods (Patlak, SKM, SUV(LBMg) and SUV(BSAg)) it was explored whether correlation with NLR changed with different time points after the start of therapy. Three methods showed excellent correlation ( r>0.95) with NLR for the baseline scan: Patlak10-60 and Patlak10-45 ( r=0.98 and 0.97, respectively), SKM40-60 ( r=0.96) and SUV(LBMg) ( r=0.96). Good correlation was found between NLR and SUV-based net influx constant, TLE and SUV(BSAg) (0.90< r<0.95). The 6P model and TNT had the lowest correlation ( r<or=0.84). SUV was least accurate in predicting changes in [(18)F]FDG uptake over time during therapy. For all methods, correlation with NLR was significantly lower for bone metastases than for other (primary or metastatic) tumour lesions ( P<0.05). In conclusion, three methods with different degrees of complexity appear to be promising alternatives to NLR for measuring glucose metabolism in breast cancer: Patlak, SKM and SUV (normalised for LBM and with a correction for plasma glucose).

Adult↗
PubMedDate not supplied

Interest of investigating p53 status in breast cancer by four different methods.

Alterations in the p53 gene are among the most common molecular changes in breast cancer although its potential role as a prognostic and predictive factor of tumor response is controversial. All abnormalities of this tumor marker need to be detected with efficacy and practical technologies. In the present study, our aim was to compare four different analyses of p53 in order to determine how many and which techniques were necessary and sufficient to reflect the p53 status in breast cancer. We analyzed p53 at the molecular level with sequencing of total cDNA and allelic loss. Furthermore, we detected p53 protein accumulation by immunohistochemistry and presence of p53 antibodies by serological analysis. Nine patients (17%) had p53 gene alterations, 33% had allelic loss, 19% had positive p53 nuclear staining and p53 antibodies were detected in 12% of patients. We found p53 gene alterations associated with p53 nuclear protein expression (p=0.001). The single significant statistical correlation in this study suggested the need to define p53 status with more than one abnormality. The general view of a simultaneous multilevel analysis would make it possible to encompass the consequences of p53 abnormalities in breast cancer.

Adenocarcinoma, Mucinous↗
PubMed2003

Prophylactic effects of two selective dry cow strategies accounting for interdependence of quarter.

Infusion of a long-acting antibiotic preparation at drying off in dairy cows as a prophylactic therapy is usually recommended for all quarters where it is in use. Studying the effectiveness of such treatment using quarter as the unit of analysis assumes that each quarter within a cow has a risk of being infected independent of the other quarters of the cow. Failure to account for interdependence of quarters within a cow may lead to inaccurate variance estimates and errors in assessing treatment effects. Data from two trials assessing different dry-cow strategies were examined for interdependence of infection between quarters. Logistic regression with a variance inflation factor or a multilevel analysis was used to assess the effect of antibiotic and internal teat-sealant dry cow strategies. Parity and infection status at drying off were covariates in the analysis. Interdependence of the risk of quarter infections within control-group cows was demonstrated in both dry-cow antibiotic and teat-seal trials. However, cows that received either of these treatments did not demonstrate interdependence. Treated quarters in both trials were 3.0 times less likely to acquire a new infection at calving compared with the untreated controls. Quarters in cows of parity 3 or greater were also at an increased risk in the antibiotic treatment trial. In both trials, quarters with either Corynebacterium spp. or coagulase-negative staphylococci infections at drying off had an increased risk of a new intramammary infection at calving. This study has demonstrated the beneficial and comparable effects of antibiotic and teat seal dry cow strategies; both decreased the risk of intramammary infection at calving. The application of dry-cow strategies at the cow level and not the quarter level is also supported.

Animals↗
PubMed2005

Neighborhood influences on physical activity in middle-aged and older adults: a multilevel perspective.

Over the past few years, attention has been drawn to the importance of neighborhood influences on physical activity behavior and the need to consider a multilevel analysis involving not only individual-level variables but also social- and physical-environment variables at the neighborhood level in explaining individual differences in physical activity outcomes. This new paradigm raises a series of issues concerning systems of influence observed at different hierarchical levels (e.g., individuals, neighborhoods) and variables that can be defined at each level. This article reviews research literature and discusses substantive, operational, and statistical issues in studies involving multilevel influences on middle-aged and older adults' physical activity. To encourage multilevel research, the authors propose a model that focuses attention on multiple levels of influence and the interaction among variables characterizing individuals, among variables characterizing neighborhoods, and across both levels. They conclude that a multilevel perspective is needed to increase understanding of the multiple influences on physical activity.

Aged↗
PubMedDate not supplied

The impact of hospital nursing characteristics on 30-day mortality.

BACKGROUND: Evidence indicates that hospital nursing characteristics such as staffing contribute to patient outcomes. Less attention has been given to other hospital nursing characteristics central to optimal professional practice, namely nurse education and skill mix, continuity of care, and quality of the work environment. OBJECTIVE: To assess the relative effects and importance of nurse education and skill mix, continuity of care, and quality of work environment in predicting 30-day mortality after adjusting for institutional factors and individual patients characteristics. METHOD: A cross-sectional analysis of outcome data for 18,142 patients discharged from 49 acute care hospitals in Alberta, Canada, for diagnoses of acute myocardial infarction, congestive heart failure, chronic obstructive pulmonary disease, pneumonia, or stroke between April 1, 1998, and March 31, 1999, was done. Mortality data were linked to patient demographic and comorbidity factors, institutional characteristics, and hospital nursing characteristics derived from a survey of all registered nurses working in acute care hospitals. RESULTS: Using multilevel analysis, it was determined that the log-odds for 30-day mortality varied significantly across hospitals (variance = .044, p < .001). Patient comorbidities and age explained 44.2% of the variance in 30-day mortality. After adjustment for patient comorbidities and demographic factors, and the size, teaching, and urban status of the study hospitals in a fixed-effects model, the odds ratios (95% confidence interval) of the significant hospital nursing characteristics that predict 30-day mortality were as follows: 0.81 (0.68-0.96) for higher nurse education level, 0.83 (0.73-0.96) for richer nurse skill mix, 1.26 (1.09-1.47) for higher proportion of casual or temporary positions, and 0.74 (0.60-0.91) for greater nurse-physician relationships. The institutional and hospital nursing characteristics explained an additional 36.9%. DISCUSSION: Hospital nursing characteristics are an important consideration in efforts to reduce the risk of 30-day mortality of patients.

Acute Disease↗
PubMed2001

Interviewers' and respondents' effects on self-reported alcohol consumption in a Swiss health survey.

OBJECTIVE: Characteristics of interviewers and respondents have been shown to influence the quality of data from survey research in various domains. There is little evidence for such effects in alcohol research, however. The purpose of the study reported here was to examine effects of gender and age of interviewers and respondents simultaneously This was done using hierarchical linear modeling, the advantage of which is that it can account for the clustering effects of respondents being nested within interviewers. METHOD: Data were obtained from the first wave of an ongoing randomized longitudinal study on changes in alcohol consumption in Switzerland. The response rate was 77.9%. Analyses were based on 2,746 (1,749 male) subjects with an average of at least monthly consumption in the 6 months before the telephone interview. Consumption was assessed by means of a graduated frequency measure. Five different hierarchical linear models of increasing complexity were used to test several hypotheses of interviewer and respondent effects. Findings from hierarchical linear modeling were compared with those from "classical" analysis of variance. RESULTS: A theoretical design effect of 1.89 attributable to interviewers was found. Both analyses of variance and hierarchical linear modeling provide support for a structure with a main effect for gender of respondents, as well as a main effect for age of respondents and an interaction effect between interviewers' and respondents' ages. CONCLUSIONS: Interviewer effects affect the estimation of statistics in survey research and must be adjusted for either by means of multilevel analysis or by the use of specialized sample survey software.

Adolescent↗
PubMed2004

Location, location, location: contextual and compositional health effects of social capital in British Columbia, Canada.

After decades of epidemiological exploration into individual-level risk factors for ill health, a recent surge of interest in the health effects of socially patterned attributes of geographically defined 'places' has given the structural side of the agency-structure debate new prominence in population health research. Utilizing two original data sets, one pertaining to features of communities in British Columbia, Canada and the other to characteristics of individuals living in them, this article distinguishes the health effects of socially patterned attributes of communities, including the social capital of communities, from the health effects of characteristics of residents that contribute to social capital, e.g., trust and participation in voluntary associations. Results from multilevel analysis demonstrated that, of three different individual-level measures of health and well-being (and including measures of long-term limiting illness and self-rated health), only a measure of depressive symptoms had variability that could be reasonably attributed to the level of the community. The social capital of communities in the form of the availability of public spaces explained some of this variability, but in the direction contrary to expectations. Overall, location (community of residence) did little to explicate health inequalities in this context. The strongest predictors of health in multivariate and multilevel models were characteristics of individual survey respondents, namely, income, trust in politicians and governments, and trust in other members of the community. Breadth of participation in networks of voluntary association was not significantly related to health in multivariate models.

Adult↗
PubMed1993

Prevention of torture and care of survivors. An integrated approach.

An analysis of the 1992 Amnesty International report demonstrates the nature and global distribution of reports of human rights violations. Systematic torture was reported in 93 of 204 countries. Reports of torture were more common from regions affected by political unrest, including mass demonstrations, riots, outbreaks of violence, killings, coup attempts, civil war, armed tribal conflict, rebellions, and conflicts with various opposition groups demanding social and political reform. These observations suggest that effective measures against torture require a multilevel analysis of underlying social, political, cultural, and psychological factors. Effective care of survivors of organized violence and torture is an indispensable component of the human rights struggle and needs more scientific attention. An integrated approach involving both preventive efforts and care of survivors can promote the human rights cause.

Global Health↗
PubMed2005

An examination of word relevance in a modified Stroop task in patients with chronic low back pain.

The present study examined the role of personal relevance of sensory pain-related words in selective attentional processing measured with a modified Stroop task administered to 30 patients with chronic low back pain. A related aim of this study was to introduce the application of multilevel analysis to test the influence of personal relevance on selective attentional processing in this sample. Patients completed the modified Stroop task, as well as a set of self-report measures aimed to assess Fear of Pain, Trait Anxiety, Catastrophizing, Pain Vigilance, and Pain Intensity. The modified Stroop task comprised 33 sensory pain-related words for which the personal relevance towards current concerns was rated afterwards by each participant on a 7-point Likert-type scale. The multilevel analyses did not support the hypothesis that personal relevance of sensory pain-related words interacted with Fear of Pain scores of patients in accounting for reaction times in naming the color of sensory pain-related words. None of the other self-report measures accounted for reaction times in isolation or in interaction with personal relevance. The modified Stroop task does not appear to be a robust measure of selective attentional processing in patients with chronic low back pain. The usefulness of other paradigms, such as the visual dot-probe task, should be explored in examining selective attentional processing in this population.

Adult↗
PubMed1998

Clusters of marijuana use in the United States.

In this study, the authors test for and estimate the clustering of marijuana use within United States neighborhoods, making use of data from annual nationally representative household sample surveys conducted during the period 1990-1995. A recently developed statistical method, alternating logistic regression, was used to quantify the clustering of marijuana users in neighborhoods. The resulting estimates of pairwise odds ratios ranged from 1.3 (95% confidence interval 1.22-1.42) for the lifetime history of marijuana use to 2.0 (95% confidence interval 1.6-2.6) for recent sharing of marijuana from one person to another. Exploratory analysis showed a slight decrease of clustering effects after adjustment for individual-level covariates: age, sex, race, education, annual family income, and history of tobacco use. Nevertheless, the main factors that account for clustering remain to be determined. Alternating logistic regression provided useful estimates of marijuana use clustering and can be used to estimate clustering of the other drug-related behavior, including sharing of needle injection equipment and other human immunodeficiency virus risk behaviors. As a form of multilevel analysis, the alternating logistic regression can accommodate shared, community-level characteristics that might influence drug taking (e.g., collective efficacy), as well as individual-level covariates, such as age and sex.

Adolescent↗
PubMed1994

Quantitative and qualitative evaluation of patterns of cerebral language dominance. An amobarbital study.

As a part of presurgical evaluation, 173 patients received bilateral intracarotid amobarbital tests for determination of cerebral language dominance. Language testing during intracarotid amobarbital procedures (IAP) consisted of the following tasks: automatic speech, sentence comprehension, body commands, naming, repetition, reading, and spontaneous speech. Patterns of cerebral language dominance were evaluated and discussed on five levels of analysis: (1) quantification of language dominance on the basis of a lateralization index derived from the total language scores in each IAP; (2) determination of five dominance subpatterns (left or right dominant, strongly bilateral, and incomplete left or right dominant) according to quantification performed on level (1) and clinical judgement; (3) qualitative differentiation of three kinds of bilaterality (positive, negative, and general) according to total language performance in left and right IAP; (4) analysis of grouped linguistic subfunctions extracted from performance in specific IAP subtests; (5) extraordinary individual case histories. The distribution of lateralization indices revealed only partially continuous degrees of lateralization, especially between the left-dominant and bilateral subgroups. As for the clinically oriented classification, incomplete left dominance is frequent (16.2%), while incomplete right dominance does not occur at all. Atypical dominance patterns are mostly correlated to bilateral and/or extratemporal foci. Concerning grouped subfunctions, a rotated factor matrix statistic yields an analysis of clusters of IAP subtests, where functions involving expressive language capacities are separated from those that are purely receptive. Further analyses of bilaterality subpatterns suggest that there are mainly four bilaterality phenomena, namely interhemispheric dissociation, double representation, unilateral representation of subfunctions, and partial representation of subfunctions in either hemisphere. Application of these differentiations to individual cases yields additional evidence that can be used in patient selection for operation in order to avoid postoperative neuropsychological deficits, especially in candidates for extratemporal surgery. In conclusion, a multilevel analysis of IAP language data is recommended since it permits a detailed account of varieties of language dominance patterns and contributes to more adequate presurgical decision-making in planned operations in cognitively relevant brain areas.

Adolescent↗
PubMedDate not supplied

Spatial analysis of the air pollution-mortality relationship in the context of ecologic confounders.

Lack of control for confounding by ecological covariates that may relate to sulfate air pollution and mortality was a key criticism of the two studies that were the focus of the Particle Reanalysis Project. To assess the validity of this criticism, we address the question: "Does sulfate air pollution exert health effects when the impact of other individual and ecologic variables thought to influence health is taken into account?" A related question arises from the possibility of autocorrelation in the mortality risks and ecologic covariates. Failure to control for autocorrelation can lead to false positive significance tests and may indicate bias resulting from a missing variable or group of variables. We control for more than 25 individual risk factors and for 20 ecologic variables representing environmental, socioeconomic, demographic, health- care, and lifestyle determinants of health in a two-stage multilevel analysis. Four modeling strategies are used to control for spatial autocorrelation. Of the 20 ecologic variables tested, only sulfate and sulfur dioxide are significant in models that incorporate spatial autocorrelation. Accounting for autocorrelation also reduces the size and certainty of the sulfate effect on mortality when compared to results generated from Cox models where independent observations are assumed. Confidence limits for the sulfate relative risk include unity in models that simultaneously control for sulfur dioxide and autocorrelation.

Air Pollution↗