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 109 records · Page 6Linked to original sources
PubMed2003

International differences in in-hospital revascularization and outcomes following acute myocardial infarction: a multilevel analysis of patients in ASSENT-2.

BACKGROUND: Revascularization rates vary substantially between countries in patients with acute ST-elevation myocardial infarction (STEMI). The impact of early revascularization on clinical outcomes in such patients remains uncertain. The ASSENT-2 fibrinolytic trial provides the opportunity to compare revascularization rates following STEMI in patients across 29 countries, and to explore the relationship between revascularization and clinical outcome. METHODS: Countries participating in ASSENT-2 were grouped into tertiles according to their in-hospital revascularization rates (<15%, 15-39%, >39%). Baseline characteristics, medication and procedure use, and clinical outcomes of the 16949 patients enrolled were compared. Multiple Cox regressions were used to assess the relationship between the tertiles and 30-day mortality, the primary endpoint of the ASSENT-2 trial. Multilevel logistic regression models were developed to validate and further extend the findings from the single-level analyses. RESULTS: Patients in highest tertile countries were younger, heavier, and more often diabetic or hypertensive. They were more likely to have had a previous myocardial infarction or revascularization procedure. Time to treatment and hospital length of stay were shorter in the highest tertile, and beta-blocker use was more frequent. Stroke rates were low and similar across tertiles, with no statistically significant difference in rates of intracranial haemorrhage. Recurrent ischaemia and reinfarction were less common in the highest tertile. Mortality rates at 30 days were lower for countries with the highest revascularization rates (5.1% vs 6.9% vs 6.5% for the lower two tertiles, P<0.001). At 1 year, mortality remained significantly lower in the highest tertile countries (8.4% vs 10.6% vs 9.9%, P=0.001). Following adjustment for baseline patient characteristics, Cox regression analysis confirmed an excess of 30-day and 1-year mortality in the lowest and intermediate tertiles compared to the highest tertile. The multilevel analyses validated these findings, and demonstrated that a country's life expectancy and the hospital volume were inversely related to both 30-day and 1-year mortality. CONCLUSIONS: The highest rate of in-hospital revascularization following fibrinolytic therapy for acute myocardial infarction in this international study was associated with a reduction in recurrent ischaemia, reinfarction, and improved survival at both 30 days and at 1 year. The optimal rates of revascularization in this setting remain to be determined.

Aged↗
PubMed2004

Racial/ethnic differences in cigarette smoking initiation and progression to daily smoking: a multilevel analysis.

OBJECTIVES: We sought to identify individual and contextual predictors of adolescent smoking initiation and progression to daily smoking by race/ethnicity. METHODS: We used data from the National Longitudinal Study of Adolescent Health to estimate the effects of individual (adolescent, family, peer) and contextual (school and state) factors on smoking onset among nonsmokers (n = 5374) and progression to daily smoking among smokers (n = 4474) with multilevel regression models. RESULTS: Individual factors were more important predictors of smoking behaviors than were contextual factors. Predictors of smoking behaviors were mostly common across racial/ethnic groups. CONCLUSIONS: The few identified racial/ethnic differences in predictors of smoking behavior suggest that universal prevention and intervention efforts could reach most adolescents regardless of race/ethnicity. With 2 exceptions, important contextual factors remain to be identified.

Adolescent↗
PubMed2002

Individual and neighbourhood determinants of social participation and social capital: a multilevel analysis of the city of Malmö, Sweden.

The aim of this study was to analyse the impact of neighbourhood on individual social capital (measured as social participation). The study population consisted of 14,390 individuals aged 45-73 that participated in the Malmö diet and cancer study in 1992-1994, residing in 90 neighbourhoods of Malmö, Sweden (population 250,000). A multilevel logistic regression model, with individuals at the first level and neighbourhoods at the second level, was performed. The study analysed the effect (intra-area correlation and cross-level modification) of the neighbourhood on individual social capital after adjustment for compositional factors (e.g. age, sex, educational level, occupational status, disability pension, living alone, sick leave, unemployment) and, finally, one contextual migration factor. The prevalence of low social participation varied from 23.0% to 39.7% in the first and third neighbourhood quartiles, respectively. Neighbourhood factors accounted for 6.3% of the total variance in social participation, and this effect was reduced but not eliminated when adjusting for all studied variables (-73%), especially the occupational composition of the neighbourhoods (-58%). The contextual migration variable further reduced the variance in social participation at the neighbourhood level to some extent. Our study supports Putnam's notion that social capital, which is suggested to be an important factor for population health and possibly for health equity, is an aspect that is partly contextual in its nature.

Aged↗
PubMed2003

Social participation, social capital and daily tobacco smoking: a population-based multilevel analysis in Malmö, Sweden.

AIMS: The aim of this study was to investigate the influence of contextual and individual factors on daily tobacco smoking. METHODS: The public-health survey in Malmö 1994 is a cross-sectional study. A total of 5600 individuals aged 20-80 years were invited to answer a postal questionnaire. The participation rate was 71%. A multilevel logistic regression model, with individuals at the first level and neighbourhoods at the second, was performed. We analysed the effect (intra-area correlation, cross-level modification and odds ratios) of individual and neighbourhood factors on smoking after adjustment for individual factors. RESULTS: Neighbourhood factors accounted for 2.5% of the crude total variance in daily tobacco smoking. This effect was significantly reduced when the individual factors such as education were included in the model. However, individual social capital, measured by social participation, only marginally affected the total neighbourhood variance in daily tobacco smoking. In fact, no significant variance in daily tobacco smoking remained after the introduction of the individual factors other than individual social capital in the model. CONCLUSION: In Malmö, the neighbourhood variance in daily tobacco smoking is mainly affected by individual factors other than individual social capital, especially socioeconomic status measured as level of education.

Adult↗
PubMed2003

Social capital and leisure time physical activity: a population based multilevel analysis in Malmö, Sweden.

OBJECTIVE: To investigate the influence of social capital and individual factors on the level of leisure time physical inactivity in the neighbourhoods. METHODS: The public health survey in Malmö 1994 is a cross sectional study. A total of 5600 people aged 20-80 years were invited to answer a postal questionnaire. The participation rate was 71%. A multilevel logistic regression model, with individuals at the first level and neighbourhoods at the second, was performed. The effect (intra-area correlation, cross level modification, and odds ratios) was analysed of individual and neighbourhood (the 1993 migration out of an area as a proxy for social capital) factors on leisure time physical inactivity after adjustment for individual factors. RESULTS: Neighbourhood factors accounted for 5.0% of the crude total variance in physical inactivity. This effect was significantly reduced when the individual factors, especially country of origin, education, and social participation, were included in the model. In contrast, it was not reduced by the introduction of the contextual social capital variable. CONCLUSION: This study suggests that in the neighbourhoods of Malmö leisure time physical inactivity is mainly affected by individual factors.

Adult↗
PubMed2001

Survival after initial hospitalisation for heart failure: a multilevel analysis of patients in Swedish acute care hospitals.

STUDY OBJECTIVE: Although national variation in short-term prognosis (that is, 30 day mortality) after a patient's first hospitalisation for heart failure may depend on individual differences between patients, dissimilarities in hospital practices may also influence prognosis. This study, therefore, sought to disentangle patient determinants from institutional factors that might explain such variation. DESIGN: A multilevel logistic regression modelling was performed with patients (1st level) nested in hospitals (2nd level). Institutional effects (that is, 2nd level variance and intra-hospital correlation) were calculated unadjusted and adjusted for specific patient (that is, age and previous diseases) and institutional (that is, size of hospital) characteristics. Patients were followed up until death or 30 days from hospital admission. SETTING: Hospitals in Sweden. PATIENTS: The study identified all the 20420 men and 17923 women (ages 65 to 85) admitted to the 90 acute care hospitals in Sweden during the period 1992-1995 for their first hospitalisation attributable to heart failure. MAIN RESULTS: Patient age and previous diseases (particularly senile dementia) were major determinants of impaired prognosis. Institutional factors explained only 1.6% and 2.3% of the total variation in 30 day mortality in men and women, respectively. These modest institutional effects remained after adjusting for patient age and previous diseases, but were in part explained by hospital size. CONCLUSIONS: National variation in short-term prognosis after an initial hospitalisation for heart failure was mainly explained by differences between patients, with hospital factors playing a minor part. Of the latter, hospital size seemed to emerge as one determinant (that is, the greater the number of patients, the better the individual prognosis).

Acute Disease↗
PubMed2005

Area-level determinants of specialty care utilization in France: a multilevel analysis.

OBJECTIVES: We investigated the effects of the density of specialists and of the area-level percentage of highly educated individuals on the odds of consulting a specialist, and examined whether these variables could explain the observed urban/rural contrast in utilization of specialty care. STUDY DESIGN: The study sample, representative of the French population aged 18-75 years in 1999, comprised 12,435 individuals. METHODS: Multilevel logistic models allowed us to investigate predictors of the odds of consulting a specialist occasionally, regularly and frequently over the previous 12 months. RESULTS: We observed a modest but significant clustering within areas of the utilization of specialty care, with higher levels of clustering for behaviours representing heavy consumption of care. After adjustment for individual factors, the odds of consulting a specialist were higher in larger cities compared with rural areas, but most of this effect was attributable to other area-level variables. These area-level effects were different in magnitude and nature among males and females. Among males, the odds of consulting a specialist increased with the area-level density of specialists. Among females, such an effect was not significant, but the odds of consulting a specialist increased with the area-level percentage of highly educated individuals. CONCLUSIONS: Further investigation is required to better understand the processes operating at the area level that were shown to affect healthcare utilization in a different way for males and females. Policies may be needed to address problems of geographical access to specialty care, as well as situations of overuse of specialty care without regular recourse to primary care.

Adolescent↗
PubMed2006

Social capital and administrative contextual determinants of lack of access to a regular doctor: a multilevel analysis in southern Sweden.

BACKGROUND: The aim of this study was to investigate the influence of contextual (social capital and administrative/neo-materialist) and individual factors on lack of access to a regular doctor. METHODS: The 2000 public health survey in Scania is a cross-sectional study. A total of 13,715 persons answered a postal questionnaire, which is 59% of the random sample. A multilevel logistic regression model, with individuals at the first level and municipalities at the second, was performed. The effect (intra-class correlations, cross-level modification and odds ratios) of individual and municipality (social capital and health care district) factors on lack of access to a regular doctor was analysed using simulation method. The Deviance Information Criterion (DIC) was used as information criterion for the models. RESULTS: The second level municipality variance in lack of access to a regular doctor is substantial even in the final models with all individual and contextual variables included. The model that results in the largest reduction in DIC is the model including age, sex and individual social participation (which is a network aspect of social capital), but the models which include administrative and social capital second level factors also reduced the DIC values. CONCLUSIONS: This study suggests that both administrative health care district and social capital may partly explain the individual's self reported lack of access to a regular doctor.

Adolescent↗
PubMed2003

Social capital and sense of insecurity in the neighbourhood: a population-based multilevel analysis in Malmö, Sweden.

The aim of this study was to investigate the influence of social capital on self-reported sense of insecurity in the neighbourhood. The public health survey in Malmö, Sweden in 1994 was a cross-sectional study. A total of 5600 individuals aged 20-80 years were asked to answer a postal questionnaire. The participation rate was 71%. A multilevel logistic regression model, with individuals at the first level and neighbourhoods at the second, was performed. We analysed the effect (intra-area correlation, cross-level modification and odds ratios) of individual (social participation) and neighbourhood social capital (electoral participation in the 1994 municipal election) on sense of insecurity after adjustment for compositional factors. Neighbourhood factors accounted for 7.2% of the total variance in individual insecurity. This effect was marginally reduced when the individual factors were included in the model. In contrast, it was reduced by 70% by the introduction of the contextual variable. This study suggests that social capital, measured as electoral participation, may partly explain the individual's sense of insecurity in the neighbourhood.

Adult↗
PubMed1999

Neighborhood environment and self-reported health status: a multilevel analysis.

OBJECTIVES: This study examined whether neighborhood socioeconomic environment helps to explain the proportion of community members with self-reported poor health status. METHODS: A random sample of 9240 persons aged 25 to 74 years were interviewed during 1988 and 1989. The socioeconomic environment of each respondent's neighborhood was measured with the Care Need Index (CNI) and the Townsend score. The data were analyzed with a multilevel model adjusted for the independent variables. The second-level variables were the 2 neighborhood scores. RESULTS: There was a clear gradient for poor health and education within every CNI interval so that with an increasing CNI (indicating more deprivation), the prevalence of poor health increased in all 3 education groups (P = .001). In the full model, decreasing educational level, obesity, length and frequency of smoking, physical inactivity, and increasing CNI were associated with poor health. Persons living in the most deprived neighborhoods had a prevalence ratio of 1.69 (95% confidence interval = 1.44, 1.98) for poor health compared with those living in the most affluent areas. CONCLUSIONS: Both neighborhood socioeconomic environment and individual educational status are associated with self-reported poor health.

Adult↗
PubMed2005

Public health application comparing multilevel analysis with logistic regression: immunization coverage among long-term care facility residents.

PURPOSE: Public health studies often sample populations using nested sampling plans. When the variance of the residual errors is correlated between individual observations as a result of these nested structures, traditional logistic regression is inappropriate. We used nested nursing home patient data to show that one-level logistic regression and hierarchical multilevel regression can yield different results. METHODS: We performed logistic and multilevel regression to determine nursing home resident characteristics associated with receiving pneumococcal immunizations. Nursing home characteristics such as type of ownership, immunization program type, and certification were collected from a sample of 249 nursing homes in 14 selected states. Nursing home resident data including demographics, receipt of immunizations, cognitive patterns, and physical functioning were collected on 100 randomly selected residents from each facility. RESULTS: Factors associated with receipt of pneumococcal vaccination using logistic regression were similar to those found using multilevel regression model with some exceptions. Predictors using logistic regression that were not significant using multilevel regression included race, speech problems, infections, renal failure, legal responsibility for oneself, and affiliation with a chain. Unstable health conditions were significant only in the multilevel model. CONCLUSIONS: When correlation of resident outcomes within nursing home facilities was not considered, statistically significant associations were likely due to residual correlation effects. To control the probability of type I error, epidemiologists evaluating public health data on nested populations should use methods that account for correlation among observations.

Aged↗
PubMed2005

Social capital and neo-materialist contextual determinants of sense of insecurity in the neighbourhood: a multilevel analysis in Southern Sweden.

BACKGROUND: The aim of this study was to investigate the influence of contextual (social capital and neo-materialist) and individual factors on sense of insecurity in the neighbourhood. METHODS: The 2000 public health survey in Scania is a cross-sectional study. A total of 13,715 persons answered a postal questionnaire, which is 59% of the random sample. A multilevel logistic regression model, with individuals at the first level and municipalities at the second, was performed. The effect (median odds ratios, intra-class correlation, cross-level modification and odds ratios) of individual and municipality/city quarter (social capital and police district) factors on sense of insecurity was analysed. RESULTS: The crude variance between municipalities/city quarters was not affected by individual factors. The introduction of administrative police district in the model reduced the municipality variance, although some of the significant variance between municipalities remained. The introduction of social capital did not affect the municipality variance. CONCLUSIONS: This study suggests that the neo-materialist factor administrative police district may partly explain the individual's sense of insecurity in the neighbourhood.

Adolescent↗
PubMed2004

Social support and social strain among husbands and wives: a multilevel analysis.

In response to recent calls in the literature for within-person examinations of social support processes over time, this study explores the relationships of spousal support, spousal strain, and well-being among husbands and wives, both within the same day and across days. Eighty-three couples were interviewed and completed a structured diary twice daily for 1 week. The results of multilevel hierarchical modeling suggest that both spousal support and spousal strain made significant, independent contributions to concurrent negative affect, although only spousal support was a significant predictor of next-day negative affect. Spousal strain interacted with spousal support to predict next-day negative affect. Direct and moderating effects of perceived marital adjustment on negative affect were discussed.

Adaptation, Psychological↗
PubMedDate not supplied

A multilevel analysis examining the relationship between social influences for smoking and smoking onset.

OBJECTIVES: To examine how older smoking peers at school and the smoking behavior of friends and family members are related to youth smoking. METHODS: The School Smoking Profile was used to collect data on tobacco use and determinants of tobacco use from 22,091 students from 29 secondary schools in Ontario, Canada. Correlates of occasional and regular smoking were examined using multilevel logistic regression analyses. RESULTS: Students are at increased risk for smoking if they (a) have smoking friends, (b) have smoking family members, and (c) attend a school with a relatively high senior-student smoking rate. CONCLUSION: These findings suggest that prevention programs should target both at-risk schools and at-risk students.

Adolescent↗
PubMed2006

Time and place: putting individual health into context. A multilevel analysis of the British household panel survey, 1991-2001.

Examination of the extent to which time and place affect people's health has been constrained by the resources available to answer this question. A British longitudinal, nationally representative survey of 8301 adults aged 16 years and older living in private households was used to consider the influence of household membership, area of residence and time using multilevel logistic regression. Self-rated health was assessed by general health and limiting illness during periods characterized by economic decline (1992), economic improvement (1996) and prosperity (2000). There was modest evidence of clustering of poor general health within areas and stronger support for within household similarities in general health which increased over time. Individual, household and area level deprivation accounted for almost all the area-level variability but had little effect on household variance. There was greater evidence of clustering of limiting illness within areas: deprivation did not account for this to any great extent. Area differences in general health reduced as the economy improved but time trends in differences in limiting illness lagged behind the timing of economic recovery. Both time and place are shown to affect self-rated health although the processes may differ depending on the health outcome.

Adolescent↗
PubMed2003

A multilevel analysis of the relation of socioeconomic status to adolescent depressive symptoms: does school context matter?

OBJECTIVE: To determine whether the socioeconomic context of the school environment is associated with adolescent depressive symptoms independent of individual household income. Study design Data were drawn from a 1995 nationally representative study of 7th to 12th grade students. Multivariable linear regression at the school and individual levels assessed the relation between income and depressive symptoms. Multilevel modeling techniques were then used to understand how these factors are jointly associated with adolescent depressive symptoms. PARTICIPANTS: Adolescents (n=13,235) in grades 7 through 12 from 132 schools whose parent provided income information. RESULTS: Linear regression analyses indicated that lower household income, average school income, and increasing school-level income inequality were significantly (P<.001) associated with depressive symptoms. Further examination of these relations through multilevel modeling indicated that both household income (P<.01) and average school income (P<.05) were significantly related to depressive symptoms after adjusting for covariates, with evidence for an interaction between the two. The impact of lower household income on depressive symptoms was approximately 2-fold greater for students attending a poor versus a rich school. CONCLUSIONS: School context is associated with adolescents' depressive symptoms, even after adjusting for individual-level factors. The school environment may partially buffer the adverse influence of lower household income on adolescent depressive symptoms.

Adolescent↗
PubMed2004

A multilevel analysis of race, community disadvantage, and body mass index among adults in the US.

This study examined the contributions of both individual socioeconomic status (SES) and community disadvantage in explaining the higher body mass index (BMI) of black adults in the US. Data from a national survey of adults (1986 American's Changing Lives Study) were combined with tract-level community data from the 1980 census. Results of multilevel regression analyses showed that black women had an age-adjusted BMI score three points higher than non-black women. Individual SES (income, education, assets) was negatively associated with BMI in women, but it only reduced the association between race and BMI from 2.99 to 2.50. Adding community socioeconomic disadvantage index further reduced the race coefficient slightly from 2.50 to 2.21. Nevertheless, living in communities with higher socioeconomic disadvantage was associated with higher BMI net of age, race, individual SES, smoking, physical activity, stress, and social support. Community income inequality (Gini) had an independent positive association with BMI, but did not substantially reduce racial differences among women. Community percent black was not associated with BMI. Results for men demonstrated no statistically significant racial differences in BMI, and no association between BMI and either individual SES or community disadvantage. Although individual SES and community socioeconomic disadvantage each partly explained the higher average BMI among black women, clear racial disparities persisted. Moreover, race, individual SES, community socioeconomic disadvantage, and individual health behaviors were each independent predictors of BMI among women. Unexplained within- and between-community variance in BMI remained among both women and men, with most unexplained variation due to within-community variance. Because our evidence for women suggests that the determinants of obesity are multiple and multilevel, attempts to address this growing social problem will similarly require a multi-faceted and multilevel approach.

Adult↗
PubMed2006

Multilevel analysis of workplace and individual risk factors for long-term sickness absence.

OBJECTIVE: The objective of this study was to examine if psychosocial and physical work-environment factors predict long-term sickness absence (>8 weeks) at both the individual and the workplace level. MATERIAL AND METHODS: Data were collected in a prospective study in 52 Danish workplaces. Psychosocial factors were aggregated as workplace means. We used multilevel logistic regression models with psychosocial factors as predictors of long-term sickness absence over 5 years based on data from a national absence register. RESULTS: Long-term sickness absence was predicted by physical work-environment factors at the individual level and psychosocial work environment factors at the workplace level. Interaction between the individual physical and workplace-level psychosocial risk factors was found. CONCLUSION: Workplace-based absence reduction interventions can be enhanced by concurrently addressing the individual and the workplace level.

Absenteeism↗