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 235 records · Page 13Linked to original sources
PubMed2005

Whose health is affected by income inequality? A multilevel interaction analysis of contemporaneous and lagged effects of state income inequality on individual self-rated health in the United States.

The empirical relationship between income inequality and health has been much debated and discussed. Recent reviews suggest that the current evidence is mixed, with the relationship between state income inequality and health in the United States (US) being perhaps the most robust. In this paper, we examine the multilevel interactions between state income inequality, individual poor self-rated health, and a range of individual demographic and socioeconomic markers in the US. We use the pooled data from the 1995 and 1997 Current Population Surveys, and the data on state income inequality (represented using Gini coefficient) from the 1990, 1980, and 1970 US Censuses. Utilizing a cross-sectional multilevel design of 201,221 adults nested within 50 US states we calibrated two-level binomial hierarchical mixed models (with states specified as a random effect). Our analyses suggest that for a 0.05 change in the state income inequality, the odds ratio (OR) of reporting poor health was 1.30 (95% CI: 1.17-1.45) in a conditional model that included individual age, sex, race, marital status, education, income, and health insurance coverage as well as state median income. With few exceptions, we did not find strong statistical support for differential effects of state income inequality across different population groups. For instance, the relationship between state income inequality and poor health was steeper for whites compared to blacks (OR=1.34; 95% CI: 1.20-1.48) and for individuals with incomes greater than $75,000 compared to less affluent individuals (OR=1.65; 95% CI: 1.26-2.15). Our findings, however, primarily suggests an overall (as opposed to differential) contextual effect of state income inequality on individual self-rated poor health. To the extent that contemporaneous state income inequality differentially affects population sub-groups, our analyses suggest that the adverse impact of inequality is somewhat stronger for the relatively advantaged socioeconomic groups. This pattern was found to be consistent regardless of whether we consider contemporaneous or lagged effects of state income inequality on health. At the same time, the contemporaneous main effect of state income inequality remained statistically significant even when conditioned for past levels of income inequality and median income of states.

Adult↗
PubMed2006

Recent increase of neighborhood socioeconomic effects on ischemic heart disease mortality: a multilevel survival analysis of two large Swedish cohorts.

Studies have shown that the decrease in ischemic heart disease mortality over the past decades was paralleled by an increase in socioeconomic disparities. Using two large Swedish cohorts defined in 1986 and 1996, the authors examined whether the effect of neighborhood socioeconomic position on ischemic heart disease mortality strengthened over the period and whether the relative contribution of individual and neighborhood socioeconomic effects changed over time. Multilevel survival models adjusted for individual factors indicated that neighborhood socioeconomic effects on ischemic heart disease mortality increased markedly between the two periods (hazard ratios for residing in the most vs. least deprived neighborhoods were 1.60 (95% credible interval: 1.36, 1.89) for the 1986 cohort and 2.54 (95% credible interval: 1.99, 3.21) for the 1996 cohort). Comparing the neighborhood socioeconomic effect with the strongly predictive effect of 15-year individual income indicated that the neighborhood effect was two times weaker than the individual effect in the 1986 cohort (-48%, 95% credible interval: -22%, -68%) but of comparable magnitude in the 1996 cohort (-11%, 95% credible interval: -42%, 29%). This increase in the contribution of neighborhood factors to the socioeconomic gradient in ischemic heart disease urges investigation into the exact mechanisms between the residential context and coronary health.

Coronary Artery Disease↗
PubMed2004

Characteristics of long-term-care facility residents associated with receipt of influenza and pneumococcal vaccinations.

BACKGROUND: Studies have found residency in long-term-care facilities (LTCFs) a risk factor for influenza and pneumonia and have demonstrated that vaccinations against these diseases reduce the risk of disease. However, rates are below Healthy People 2010 goals of 90% for LTCFs. During 1999-2002, a multi-state demonstration project was conducted in LTCFs to implement standing orders programs for immunizations. OBJECTIVE: Identify nursing home resident-specific characteristics associated with vaccination coverage at baseline. METHODS: Facility-level data were collected from self-reported surveys of selected nursing homes in 14 states and from the On-line Survey and Certification Reporting System. Resident-level data, including demographics and physical functioning, were obtained from the Centers for Medicare & Medicaid Services' Minimum Data Set; 2000-2001 vaccination status was obtained by chart review. Influenza vaccination status reflected a single season, whereas pneumococcal vaccination status reflected vaccination in the past. Multilevel analysis was used to control for facility-level variation. RESULTS: Of 22,188 residents sampled in 249 LTCFs, complete data were obtained for 20,516 (92%). The average coverage for immunizations was 58.5% +/- 0.7% for influenza and 34.6% +/- 0.3% for pneumococcal. On bivariate analyses, residents with cognitive, psychiatric, or neurologic problems were more likely to be vaccinated; those with accidental injuries, unstable conditions, or cancer were less likely to receive either vaccine. On multilevel analysis, the strongest resident characteristics associated with receipt of immunizations, controlling facility variation, were cognitive deficits and psychiatric illness. CONCLUSION: The variation in baseline vaccination coverage associated with LTCF resident characteristics supports the need for strategies to increase vaccination coverage in LTCFs.

Aged↗
PubMedDate not supplied

[Multilevel models or the importance of ranking].

Many researchers in Public Health have data bases with a hierarchical structure. The studied patients (level 1) can be nested in groups, i.e., district, doctor, hospital, etc. (level 2). It is possible that patients in the same group be similar, so traditional regression models can not be used because the hypothesis of independent observations is not satisfied. A Multilevel Analysis, using hierarchical models, can be a solution for this problem; these models take into account the distribution of the data at different levels to estimate two types of variability: one due to individuals in the study and another due to the groups in which patients are nested. These types of models were applied in education in the last decade, however they have been recently applied in Health Research. This paper is a review about multilevel analysis. A discussion about hierarchichal models versus traditional regression models is presented and some applications in Epidemiology and Health Research are showed.

Health Services Research↗
PubMed2003

Role of organizational factors in poor blood pressure control in patients with type 2 diabetes: the QuED Study Group--quality of care and outcomes in type 2 diabetes.

BACKGROUND: A large body of evidence supports the need for reducing the cardiovascular burden of diabetes. Only indirect and occasional data describe the adequacy of routine management of hypertension in patients with diabetes. The aim of this study was to explore the interplay of some potential key determinants of quality of antihypertensive care, including the settings, physicians' beliefs about blood pressure (BP) control, and patient-related factors. METHODS: We evaluated physicians' beliefs about BP control using questionnaire responses at study entry. A sample of 3449 patients with type 2 diabetes mellitus, of whom 1782 (52%) were considered to have hypertension, was recruited by 212 physicians practicing in 125 diabetes outpatients clinics (DOCs) and 106 general practitioners (GPs). We evaluated the type and number of antihypertensive agents used and the BP values at study entry and after 1 year of follow-up. We used multilevel analysis to investigate correlates of poor BP control (> or =160/90 mm Hg). RESULTS: Only 16% of GPs and 14% of DOC physicians targeted BP values of less than 130/85 mm Hg. At study entry, 6% of the patients had values below 130/85 mm Hg, whereas 52% showed values of 160/90 mm Hg or greater. Only 12% of subjects were treated with more than 2 drugs at study entry, compared with 16% at the 1-year follow-up (P =.001). Multilevel analysis showed that patients attending DOCs had a more than 2-fold increased risk for inadequate BP control, compared with those treated by GPs. The risk for poor BP control was 2 times higher for patients treated by male physicians compared with those treated by female physicians, and it was halved when the physician responsible for the diabetes care specialized in diabetology or endocrinology. CONCLUSION: In a model situation of comorbidity, the overall quality of care depends on structural and organizational factors, which are likely to be more influential than existing guidelines.

Antihypertensive Agents↗
PubMed1994

Design and analysis of multilevel analytic studies with applications to a study of air pollution.

We discuss a hybrid epidemiologic design that aims to combine two approaches to studying exposure-disease associations. The analytic approach is based on comparisons between individuals, e.g., case-control and cohort studies, and the ecologic approach is based on comparisons between groups. The analytic approach generally provides a stronger basis for inference, in part because of freedom from between-group confounding and better quality data, but the ecologic approach is less susceptible to attenuation bias from measurement error and may provide greater variability in exposure. The design we propose entails selection of a number of groups and enrollment of individuals within each group. Exposures, outcomes, confounders, and modifiers would be assessed on each individual; but additional exposure data might be available on the groups. The analysis would then combine the individual-level and the group-level comparisons, with appropriate adjustments for exposure measurement errors, and would test for compatibility between the two levels of analysis, e.g., to determine whether the associations at the individual level can account for the differences in disease rates between groups. Trade-offs between numbers of groups, numbers of individuals, and the extent of the individual and group measurement protocols are discussed in terms of design efficiency. These issues are illustrated in the context of an on-going study of the health effects of air pollution in southern California, in which 12 communities with different levels and types of pollution have been selected and 3500 school children are being enrolled in a ten-year cohort study.(ABSTRACT TRUNCATED AT 250 WORDS)

Adolescent↗
PubMed2005

The quality of the quality indicator of pain derived from the minimum data set.

OBJECTIVE: To examine facility variation in data quality of the level of pain documented in the minimum data set (MDS) as a function of level of hospice enrollment in nursing homes (NHs). DATA SOURCE: Clinical assessments on 3,469 nonhospice residents from 178 NHs were merged with On-line Survey Certification and Reporting data of 2000, Medicare Claims data of 2000 and the MDS of 2000-2002. STUDY DESIGN: Using the same assessment protocol, NH staff and study nurses independently assessed 3,469 nonhospice residents. Study nurses' assessments being gold standard, we quantified and compared quality of NH staff's pain rating across NHs with high, medium, or low hospice use. Multilevel models were built to assess the effect of NH hospice use levels on the occurrence of false positive (FP) and false negative (FN) errors in NH-rated "severe pain." PRINCIPAL FINDINGS: Of 178 NHs, 25 had medium and 41 high hospice use. NHs with higher hospice use had lower sensitivities. In multilevel analysis, we found a significant facility-level variation in the probability of FP and FN errors in facility-rated "severe pain." Resident characteristics only explained 4 and 0 percent of the facility variation in FP and FN, respectively; characteristics and locations (state) of NHs further explained 53 and 52 percent of the variance. After controlling for resident and NH characteristics, staff in NHs with medium or high hospice use were less likely to have FP or FN errors in their MDS documentation of pain than were staff in NHs with low or no hospice use. CONCLUSIONS: The examination of data quality of pooled MDS data from multiple NHs is insufficient. Multilevel analysis is needed to elucidate sources of heterogeneity in the quality of MDS data across NHs. Facility characteristics, e.g., hospice use or NH location, are systematically associated with overrated/underrated pain and may bias pain quality indicator (QI) comparisons. To ensure the integrity of QI comparison in the NH setting, the government may need to institute regular audits of MDS data quality.

Aged↗
PubMed1998

The efficient organization of blood donation.

This paper models the costs of collecting whole blood in the north of Scotland in order to investigate strategies whereby the annual collection target can be met at lower cost. Data on the costs of the individual sessions held in 1993-1995 are analyzed using multilevel analysis. A new technique, namely the conditioned iterative generalized least squares (CIGLS) estimator is applied. Then the feasibility of collecting increased volumes from particular panels and areas is assessed by examining which factors determine the number of blood donors at a session. Results show that fixed cost and marginal cost vary across panels but marginal cost does not vary by volume. This implies that the cost-minimizing policy is to equalize marginal costs and collect higher volume at fewer panels (those with lower fixed costs). The level of donations can be increased by increasing the number of opportunities to donate and/or increasing the average length of a session. The latter policy is shown to be more cost-effective. Multilevel analysis proves not only to be appropriate but also particularly useful.

Blood Banks↗
PubMed2003

Biomechanical analysis of multilevel cervical corpectomy and plate constructs.

OBJECT: The authors compared the biomechanical stability of two multilevel cervical constructs involving the placement of equal size anterior cervical plates (ACPs) after decompressive surgery: the first is placed after three-level corpectomy with strut graft and the second after two-level corpectomy and aggressive discectomy with strut graft. In addition, both constructs were evaluated with and without the application of a screw attaching the ACP to the strut graft to determine whether the additional screw enhanced stability in any mode of loading. METHODS: Nondestructive repeated-measures in vitro flexibility tests were performed in human cadaveric cervical spines. Nonconstraining pure moments of up to 1.5 Nm were applied while recording three-dimensional angular motion stereophotogrammetrically at each level from C4-5 to C7-T1. Nine specimens underwent the three-level corpectomy/strut graft procedure and eight specimens the two-level corpectomy/discectomy strut graft procedure. Failures during testing eliminated two of the former specimens and three of the latter specimens from analysis. The construct applied after the two-level procedure allowed a significantly smaller normalized neutral zone during flexion-extension than the three-level construct (p = 0.04). Normalized elastic zone and range of motion were consistently smaller in the two- than in the three-level construct, but the differences were not significant. Addition of a screw to the strut graft significantly reduced motion in the three-level procedure-treated specimens during flexion and lateral bending but had no effect on two-level corpectomy-treated specimens. CONCLUSIONS: The construct associated with the two-level corpectomy/discectomy provided better immediate postoperative stability than that associated with the three-level corpectomy. The addition of a screw to the strut graft conferred stability on the three-level construct but not the two-level construct.

Adult↗
PubMed2006

Competing definitions of contextual environments.

BACKGROUND: The growing interest in the effects of contextual environments on health outcomes has focused attention on the strengths and weaknesses of alternate contextual unit definitions for use in multilevel analysis. The present research examined three methods to define contextual units for a sample of children already enrolled in a respiratory health study. The Inclusive Equal Weights Method (M1) and Inclusive Sample Weighted Method (M2) defined communities using the boundaries of the census blocks that incorporated the residences of the CHS participants, except that the former estimated socio-demographic variables by averaging the census block data within each community, while the latter used weighted proportion of CHS participants per block. The Minimum Bounding Rectangle Method (M3) generated minimum bounding rectangles that included 95% of the CHS participants and produced estimates of census variables using the weighted proportion of each block within these rectangles. GIS was used to map the locations of study participants, define the boundaries of the communities where study participants reside, and compute estimates of socio-demographic variables. The sensitivity of census variable estimates to the choice of community boundaries and weights was assessed using standard tests of significance. RESULTS: The estimates of contextual variables vary significantly depending on the choice of neighborhood boundaries and weights. The choice of boundaries therefore shapes the community profile and the relationships between its components (variables). CONCLUSION: Multilevel analysis concerned with the effects of contextual environments on health requires careful consideration of what constitutes a contextual unit for a given study sample, because the alternate definitions may have differential impact on the results. The three alternative methods used in this research all carry some subjectivity, which is embedded in the decision as to what constitutes the boundaries of the communities. The Minimum Bounding Rectangle was preferred because it focused attention on the most frequently used spaces and it controlled potential aggregation problems. There is a need to further examine the validity of different methods proposed here. Given that no method is likely to capture the full complexity of human-environment interactions, we would need baseline data describing people's daily activity patterns along with expert knowledge of the area to evaluate our neighborhood units.

Adolescent↗
PubMed2026

Systematic Identification of Therapeutic Targets and Repurposed Drugs for Stroke: From Genome Causal Analysis to Multilevel Validation.

BACKGROUND: Stroke is a severe cerebrovascular disease characterized by narrow time windows and complications. This study aimed to identify novel drug targets and repurposed drugs for stroke. METHODS: This study used expression quantitative trait loci data from druggable genes in brain and blood as instrumental variables. Mendelian randomization, colocalization, and phenome-wide Mendelian randomization were applied to evaluate causal relationships and potential side effects, with stroke and ischemic stroke as primary outcomes. Preclinical validation used oxygen-glucose deprivation/reperfusion and middle cerebral artery occlusion/reperfusion models. Pharmacological and behavioral assessments evaluated the therapeutic potential of candidate targets and drugs. Additionally, proteomic sequencing was performed following GGCX (γ-glutamyl carboxylase) overexpression to explore its biological functions. RESULTS: Elevated GGCX expression in brain and blood was potentially causally associated with reduced risk of stroke and ischemic stroke, supported by colocalization evidence, although potential cardiovascular risks could not be excluded. Drug repositioning identified ifenprodil as a candidate agent that reduced infarction volume, improved motor and cognitive functions, and reversed GGCX downregulation in mice. Ifenprodil treatment and GGCX overexpression alleviated oxygen-glucose deprivation/reperfusion-induced injury and upregulated GGCX expression. Mechanistically, GGCX conferred neuroprotection by regulating protein homeostasis, suppressing inflammation, promoting metabolic recovery, and modulating nuclear transcriptional regulation. CONCLUSIONS: This study established a potential causal link between GGCX and stroke risk, particularly ischemic stroke. GGCX represents a promising therapeutic target for ischemic stroke. Targeted GGCX expression upregulation and drug repurposing, particularly ifenprodil, may offer novel therapeutic avenues. Further validation is warranted to assess clinical efficacy and safety.

Animals↗
PubMed2006

A general approach for two-stage analysis of multilevel clustered non-Gaussian data.

In this article, we propose a two-stage approach to modeling multilevel clustered non-Gaussian data with sufficiently large numbers of continuous measures per cluster. Such data are common in biological and medical studies utilizing monitoring or image-processing equipment. We consider a general class of hierarchical models that generalizes the model in the global two-stage (GTS) method for nonlinear mixed effects models by using any square-root-n-consistent and asymptotically normal estimators from stage 1 as pseudodata in the stage 2 model, and by extending the stage 2 model to accommodate random effects from multiple levels of clustering. The second-stage model is a standard linear mixed effects model with normal random effects, but the cluster-specific distributions, conditional on random effects, can be non-Gaussian. This methodology provides a flexible framework for modeling not only a location parameter but also other characteristics of conditional distributions that may be of specific interest. For estimation of the population parameters, we propose a conditional restricted maximum likelihood (CREML) approach and establish the asymptotic properties of the CREML estimators. The proposed general approach is illustrated using quartiles as cluster-specific parameters estimated in the first stage, and applied to the data example from a collagen fibril development study. We demonstrate using simulations that in samples with small numbers of independent clusters, the CREML estimators may perform better than conditional maximum likelihood estimators, which are a direct extension of the estimators from the GTS method.

Animals↗
PubMed2024

Geospatial Analysis of Multilevel Socioenvironmental Factors Impacting the Campylobacter Burden among Infants in Rural Eastern Ethiopia: A One Health Perspective.

Increasing attention has focused on health outcomes of Campylobacter infections among children younger than 5 years in low-resource settings. Recent evidence suggests that colonization by Campylobacter species contributes to environmental enteric dysfunction, malnutrition, and growth faltering in young children. Campylobacter species are zoonotic, and factors from humans, animals, and the environment are involved in transmission. Few studies have assessed geospatial effects of environmental factors along with human and animal factors on Campylobacter infections. Here, we leveraged Campylobacter Genomics and Environmental Enteric Dysfunction project data to model multiple socioenvironmental factors on Campylobacter burden among infants in eastern Ethiopia. Stool samples from 106 infants were collected monthly from birth through the first year of life (December 2020-June 2022). Genus-specific TaqMan real-time polymerase chain reaction was performed to detect and quantify Campylobacter spp. and calculate cumulative Campylobacter burden for each child as the outcome variable. Thirteen regional environmental covariates describing topography, climate, vegetation, soil, and human population density were combined with household demographics, livelihoods/wealth, livestock ownership, and child-animal interactions as explanatory variables. We dichotomized continuous outcome and explanatory variables and built logistic regression models for the first and second halves of the infant's first year of life. Infants being female, living in households with cattle, reported to have physical contact with animals, or reported to have mouthed soil or animal feces had increased odds of higher cumulative Campylobacter burden. Future interventions should focus on infant-specific transmission pathways and create adequate separation of domestic animals from humans to prevent potential fecal exposures.

Humans↗
PubMed2002

The development and application of a multilevel decision analysis model for the remediation of contaminated groundwater under uncertainty.

A study was initiated which combined elements of stochastic hydrology, risk assessment, simulation modeling, cost analysis and decision making to define the optimum remediation choice(s) for a Superfund site in the southern United States. The effort focused upon the premise that groundwater remediation is inherently complex due to uncertainties in the geological matrix as well as in contaminant concentrations at points of compliance and/or exposure. The technical analyst should supply the decision maker with estimates of these uncertainties as well as the cost penalties required to reduce them to manageable levels. Monte Carlo transport modeling was employed to define the probability of contaminant excursions from the site, while geostatistical simulation identified a joint plume configuration and its attendant probability. Bayesian modeling was used to define the worth of additional data. These individual components were combined within a Decision Model to identify optimum remediation configurations for a given levels of risk tolerance which could be supplied by the decision maker or affected community. Sensitivity analyses were conducted to define ranges over which the decision would not be affected by variation in the respective decision parameter.

Decision Making↗
PubMed2006

US state- and county-level social capital in relation to obesity and physical inactivity: a multilevel, multivariable analysis.

Although social capital has been linked to a variety of health outcomes, its association with obesity has yet to be elucidated. This study explored the relations between social capital measured at the US state and county levels and individual obesity and leisure-time physical inactivity. Individual-level data were drawn from the 2001 Behavioral Risk Factor Surveillance System survey, while data from other surveys and administrative sources were used to construct contextual measures. Two state-level social capital scales were derived from 10 indicators and two county-level scales from five indicators. In 2-level analyses of over 167,000 adults nested within 48 states plus the District of Columbia, residence in a state above the median on one or both state social capital scales (vs. neither scale) was associated with lower relative odds of obesity and physical inactivity, controlling for individual-level covariables and state-level estimates of mean household income, the Gini coefficient, and the percentage of Black residents. In 3-level analyses, the adjusted odds ratio (OR) for physical inactivity associated with residence in a county above the median on one or both county social capital scales was significantly below 1, while the association with obesity was not significant. Significantly weaker inverse ORs for the relations between state- and county-level social capital and obesity were observed among American Indians/Alaska Natives compared to Whites. Meanwhile, little support was found for mediation by social capital of the associations of urban sprawl and income inequality with obesity or physical inactivity. Overall, this study provides some evidence for the promotion of social capital as a potential strategy for addressing the burgeoning obesity epidemic.

Adult↗
PubMed2006

Modeling heterogeneity in social interaction processes using multilevel survival analysis.

More than 15 years ago, survival or hazard regression analyses were introduced to psychology (W. Gardner & W. A. Griffin, 1989; W. A. Griffin & W. Gardner, 1989) as powerful methodological tools for studying real time social interaction processes among dyads. Almost no additional published applications have appeared, although such data are commonly collected and the applicable questions are central to many important theoretical perspectives. To revisit the basic methods, the authors use an example from emotion regulation theory in which the level of child antisocial behavior is hypothesized to be positively associated with the hazard rate of angry emotions and negatively associated with sad, fearful emotions in the face of parental negative behavior (scolding). The authors discuss the limitations of traditional approaches to the analysis of social interaction and demonstrate improvements in the ability to model individual differences now available in existing software.

Antisocial Personality Disorder↗
PubMed2005

Influence of short stature on the change in pulse pressure, systolic and diastolic blood pressure from age 36 to 53 years: an analysis using multilevel models.

BACKGROUND: Previous cross-sectional analyses of this cohort have shown that short height and leg length are associated with higher pulse pressure and systolic blood pressure in middle age. It is unclear how these adult measures of childhood growth influence the change in blood pressure as it increases with age. METHODS: Multilevel models were fitted to investigate associations between components of height and the change in blood pressure between 36, 43, and 53 years in a prospective national cohort of 1472 men and 1563 women followed-up since birth in 1946. RESULTS: Shorter height and leg length, but not trunk length, were associated with higher blood pressure, similarly in men and women. Longitudinal analyses showed that the effects of both height and leg length on pulse pressure and systolic blood pressure became significantly stronger with age. For example, the change in systolic blood pressure was found to be -0.021 mm Hg (95% confidence interval -0.029 to -0.013) per year lower for every centimetre increase in leg length (P < or = 0.001). In other words, the increase in systolic blood pressure over a 10 year period of a participant whose legs were 10 centimetres shorter was 2.1 mm Hg higher (P < or = 0.001), compared with a taller participant. Associations were independent of a number of potential confounders. CONCLUSIONS: These results support the hypothesis that short people may be more susceptible to the effects of ageing on the arterial tree. Childhood growth may contribute to the tracking of cardiovascular risk throughout life.

Adult↗