Search PubMed⌕ Search

Biomedical subjects

D Böhning

Publications and source records attributed to D Böhning.

At least 19 recordsLinked to original sources
PubMed2005

Meta-analysis: a unifying meta-likelihood approach framing unobserved heterogeneity, study covariates, publication bias, and study quality.

OBJECTIVES: This contribution provides a unifying concept for meta-analysis integrating the handling of unobserved heterogeneity, study covariates, publication bias and study quality. It is important to consider these issues simultaneously to avoid the occurrence of artifacts, and a method for doing so is suggested here. METHODS: The approach is based upon the meta-likelihood in combination with a general linear nonparametric mixed model, which lays the ground for all inferential conclusions suggested here. RESULTS: The concept is illustrated at hand of a meta-analysis investigating the relationship of hormone replacement therapy and breast cancer. The phenomenon of interest has been investigated in many studies for a considerable time and different results were reported. In 1992 a meta-analysis by Sillero-Arenas et al. concluded a small, but significant overall effect of 1.06 on the relative risk scale. Using the meta-likelihood approach it is demonstrated here that this meta-analysis is due to considerable unobserved heterogeneity. Furthermore, it is shown that new methods are available to model this heterogeneity successfully. It is argued further to include available study covariates to explain this heterogeneity in the meta-analysis at hand. CONCLUSIONS: The topic of HRT and breast cancer has again very recently become an issue of public debate, when results of a large trial investigating the health effects of hormone replacement therapy were published indicating an increased risk for breast cancer (risk ratio of 1.26). Using an adequate regression model in the previously published meta-analysis an adjusted estimate of effect of 1.14 can be given which is considerably higher than the one published in the meta-analysis of Sillero-Arenas et al. In summary, it is hoped that the method suggested here contributes further to a good meta-analytic practice in public health and clinical disciplines.

Confidence Intervals↗
PubMed2002

[Comparison of expertises between long-term care insurance (SGB XI) and the law on ambulatory care for disabled individuals in Berlin (HPG)].

A long-term care insurance (SGB XI) was established in Germany in 1995. Previously, care for predominantly disabled and disabled elderly people in Berlin was regulated by a law enforced by the Berlin government (HPG). The objective of this study was to investigate the differences in age, sex, diagnosis and care of the disabled individuals at home. In a cross-sectional study, the social medicine certificates of 3.916 disabled individuals were evaluated. The certificate assesses the grade of disability and the care needed for a person entitled to benefit from either the HPG or later from SGB XI. Disabled females benefiting from the HPG and SGB XI regulations are in the majority. The male: female ratio for HPG was 1:2.215 and for SGB XI 1:2.759. The differences between male and female as well as between HPG and SGB XI are statistically significant. The even higher proportion of females for those benefiting from SGB XI is due to the demographic change over time. The results of the study also show that the certificates differ in diagnosis, in the care needed at home, and in the grades of disability. By logistic regression analysis, the univariate method shows the amount of care needed at home to be greater for the HPG than for the long-term care insurance. The collective of disabled persons benefiting from HPG were derived from the ambulatory and the stationary sector. In this study, those individuals who were investigated and supported by SGB XI originated from the ambulatory sector, although the law also supports those under hospital care. By calculating analysis of multivariate logistic regression with grades of disability taken as dependent and disability of a given person as independent variables related to the grades of disability when applying SGB XI were compared to HPG. The independent variables have different hierarchies for both laws and need to be divided into more specific subgroups. Interestingly, age and sex did not influence the grading of disability. Diseases leading to disability differ between male and females. Males suffer mostly from neurologic and psychiatric and females from orthopaedic diseases. The criteria for the particular grades are too broad and a better specification is suggested, in that the grades should be increased from three to five. For exceptional cases, a more flexible procedure should be allowed, which is possible under the present SGB XI regulations.

Adolescent↗
PubMed2000

Estimating risk difference in multicenter studies under baseline-risk heterogeneity.

In this paper, we consider the case of efficient estimation of the risk difference in a multicenter study allowing for baseline heterogeneity. We consider the optimally weighted estimator for the common risk difference and show that this estimator has considerable bias when the true weights (which are inversely proportional to the variances of the center-specific risk difference estimates) are replaced by their sample estimates. In addition, we propose a new estimator for this situation of the Mantel-Haenszel type that is unbiased and, in addition, has a smaller variance for small sample sizes within the study centers. Simulations illustrate these findings.

Bias↗
PubMed1999

The relationship between alcohol consumption, health indicators and mortality in the German population.

BACKGROUND: The patterns of total alcohol, beer and wine consumption were evaluated in the German National Health Surveys. The impact of these habits on cardiovascular and all-cause mortality as well as cardiovascular risk factors and liver disease parameters was estimated. METHODS: Independent representative samples of the German population (15,400 people), and regional samples of the Berlin-Spandau population (2,370 in total), aged 25-69 years, were analysed. The amount and frequency of alcohol consumption was assessed with standardized questionnaires. Biochemical analyses included serum lipids and gamma-glutamyl-transpeptidase (Gamma GT). Multiple analyses of variance were used to determine the relationship between alcohol intake and biochemical parameters. A mortality follow-up of about 7 years was conducted for the Berlin-Spandau population. Proportional hazard models were used to estimate hazard ratios (HR) for all-cause and cardiovascular mortality. RESULTS: Over 80% of men and 55% of women in Germany drink alcohol on a regular base. The majority of the consumers (65% of men, 87% of women) are light (1-20 g/day) or moderate (21-40 g/day) drinkers. Higher serum high density lipoprotein (HDL)-cholesterol and Gamma GT levels were observed with increasing alcohol intake. In light and moderate drinkers no significant relationship was seen with non-HDL-cholesterol, triglycerides, blood pressure and body mass index, compared to teetotallers. Men who consumed 1-20 g alcohol/day had a significantly lower all-cause and cardiovascular mortality. As compared to nondrinkers, the risk was almost 50% lower. CONCLUSION: The results suggest that light (and possibly moderate) alcohol consumption reduces the risk of cardiovascular and total mortality risk and is favourably related to HDL-cholesterol.

Adult↗
PubMed1999

Analysis of geographical heterogeneity in live-birth ratio in Thailand.

BACKGROUND: Live-birth (male-female) ratios are a standard measure used in demography. Recently, live-birth ratios have been considered as a potential indicator for various environmental hazards. In this paper, mixture modeling is applied to analyse the geographic heterogeneity of live-births in their composition of male and female proportions (live-birth ratio) in the Kingdom of Thailand. METHODS: Live-birth data are taken from the 1990 census of the Kingdom of Thailand. The level of aggregation is the province, of which there are 73 in Thailand. The analysis is based on the simple observation that a logical equivalent to the live-birth ratio is available, namely the proportion of male live-births. Based on this measure a simple and exact statistical model is easily derived: conditional on the number of live-births, the number of male live-births forms a binomial distribution, with parameter lambda. If there is homogeneity in the proportion of male live-births, then all provinces can be described by means of a single binomial distribution. However, if there is heterogeneity in the proportion parameter lambda, then a mixture of binomial distributions will occur. RESULTS: For the 1990 census data, three groups could be identified: a majority group containing 84% of the provinces and a proportion parameter of lambda = 0.513, a group of five provinces having a reduced proportion parameter of lambda = 0.500 (fewer male live-births), and a group of four provinces having an increased proportion parameter of lambda = 0.525. CONCLUSIONS: It is unclear how this can be explained, although some potential explanations are offered. The stability of these groups in time should be confirmed and regularly monitored.

Bias↗
PubMed1998

Prevalence estimation under heterogeneity in the example of bovine trypanosomosis in Uganda.

We examine variance estimators of a binomial parameter established under cluster sampling using data from a cross-sectional study of bovine trypanosomosis in Mukono County, Uganda. Fifty farms (referred to as clusters), were sampled with a total sample size of 487 cattle. Trypanosomes were found in 17.9% (87/487) of the total sample. The cluster-level (CL) prevalences were not homogeneously distributed. According to maximum-likelihood parameters established by mixture-distribution analysis, 18% of the cluster had 0% prevalence whereas 48% and 34% of the clusters could be allocated to subpopulations of clusters with mean prevalences 11.6% and 31.9% respectively. We show that this form of heterogeneity invalidates the applicability of the Beta distribution as a model for the distribution of CL prevalences. Furthermore, we provide empirical evidence for a variance inflation due to heterogeneity (inflation factor 2.07) that exceeds the design-based variance inflation due to clustering alone (inflation factor 1.82). The variance inflation due to heterogeneity is given in a closed form so that the approach can be conveniently applied to survey data that involve cluster sampling under heterogeneity.

Animals↗
PubMed1998

Recent developments in computer-assisted analysis of mixtures.

This paper reviews recent developments in the area of computer-assisted analysis of mixture distributions (C.A.MAN). Given a biometric situation of interest in which, under homogeneity assumptions, a certain parametric density occurs, such as the Poisson, the binomial, the geometric, the normal, and so forth, then it is argued that this situation can easily be enlarged to allow a variation of the scalar parameter in the population. This situation is called unobserved heterogeneity. This naturally leads to a specific form of nonparametric mixture distribution that can then be assumed to be the standard model in the biometric application of interest (since it also incorporates the homogeneous situations as a special case). Besides developments in theory and algorithms, the work focuses on developments in biometric applications such as meta-analysis, fertility studies, estimation of prevalence under clustering, and estimation of the distribution function of survival time under interval censoring. The approach is nonparametric for the mixing distribution, including leaving the number of components (subpopulations) of the mixing distribution unknown.

Animals↗
PubMed1997

A case-control study of non-fatal traffic accidents on hospital patients in Bangkok metropolis.

Morbidity and mortality from road traffic accidents in Bangkok metropolis is high with a trend strictly increasing in time, thus establishing one of the major public health problems and targets of the Kingdom of Thailand. An unmatched case-control study was undertaken to evaluate various risk patterns connected with the occurrence of traffic accidents. 350 cases were sampled from five major hospitals distributed over the Bangkok Metropolitan area and 350 controls were included from out-patients of Rajawithi hospital. The major interest in this case-control analysis is to estimate relative risk (through odds ratio) and measure statistical significance of association between road traffic accident and the following risk factors: the demographic risk factors are age (age between 15 and 19 years has odds ratio of 8.25 [1.01, 67.72]; age above 40 is baseline), marital status (single has odds ratio of 2.74 [1.26, 5.95]; baseline is not-single), sex (males have an odds ratio of 2.05 [0.51, 8.24], occupation (salespersons have an odds ratio of 4.27 [1.47, 12.46]; baseline are office workers); the driving related exposure factors type of vehicle (motorbikers have an odds ratio of 6.22 [2.82, 13.78]; baseline are other vehicle drivers), driving time per day (odds ratio 1.35 [0.88, 2.09]) and the number of accidents in the last three years (odds ratio of 4.36 [2.51, 7.58]); the behavioral factors one hour before driving such as the consumption of alcohol (odds ratio 3.01 [1.53, 5.94]; baseline in no alcohol) and the consumption of a lot of food (odds ratio of 0.07 [0.03, 0.16]; baseline is not eating a lot before driving). The odds ratios for traffic accident when being under allergic medication and the consumption of tranquilizer were not significant. The results are consistent with current results in traffic accident research and point to target groups for potential prevention programmes in traffic accident prevention.

Accidents, Traffic↗
PubMed1997

Impact of biological factors on the interpretation of bovine trypanosomosis serology.

A total of 457 cattle from dairy farms in Mukono County, Uganda, were investigated for Trypanosoma antibodies by ELISA. The objective of the study was to identify explanatory covariate factors for seropositivity among nine farm-specific and four animal-specific variables. We used logistic regression models for parasitological and serological outcome variables and then compared the adjusted odds ratios for explanatory factors between the models. Age is positively correlated with seropositivity but not with the detection of the parasite. Therefore, age group-specific cut-off values were established using mixture-distribution analysis. This procedure, as well as a mixture-distribution-derived cut-off value for the total sample, resulted in a greater relative efficiency of the ELISA as compared to conventional interpretation (cut-off value defined using non-exposed negative controls). The relevance of age and other biological factors for the serological status is briefly discussed.

Age Factors↗
PubMed1997

Risk factors for traffic accidents in Bangkok Metropolis: a case-reference study.

It was aimed to study injures from road traffic accidents in Bangkok Metropolis and identify patients' characteristics as well as to search for risk factors for traffic accidents leading to hospitalization. The study included 346 in-patient cases suffering injuries from road traffic accidents in Bangkok Metropolis. The patients were recruited during a period of 4 months of the year 1992 from five hospitals in various areas of Bangkok which were judged to be representative for Bangkok Metropolis. Using the method of case-reference, relative risk could be estimated for various exposure factors. Most of the patients drove a motorcycle, had their license for only a short period, and drove more than 5 hours a day. About one third of the patients were under the influence of alcohol. The traffic accident characteristics were that they occurred mainly at night time with the peak between 21.00 and 24.00 hours. About 90% of all traffic accidents occurred during the rainy season and most of them occurred near to road junctions. Reference data was available for some variables and the following risk group could be identified: RR (male-age 20-24) = 17.06 (8.8-33.9), RR (single-marital status) = 2.25 (1.7-3.1), RR (primary-education) = 6.2 (2.9-12.6), RR (unskilled labourer-occupation) = 3.91 (2.7-5.9), RR (salesperson-occupation) = 3.34 (2.2-5.0).

Accidents, Traffic↗
PubMed1994

Construction of an intrinsic cut-off value for the sero-epidemiological study of Trypanosoma evansi infections in a canine population in Brazil: a new approach towards an unbiased estimation of prevalence.

In the serodiagnosis of tropical infectious diseases, cut-off values are often established by using sera from individuals living under moderate climatic conditions, not exposed to the risk of infection (non-endemic controls). This approach guarantees the disease-free status of the individuals within that control group but leads to an assembly of samples which are not representative for the disease-free individuals of the target population (selection bias). Using data from an epidemiological study of Trypanosoma evansi infection in dogs, two alternative methods to construct cut-off values for a T. evansi antibody ELISA are described which are solely based on a distribution analysis of the data from the endemic animals. By cluster analysis these data could be divided into 'high', 'intermediate' and 'low responders'. High responders could also be identified by using the computer-assisted analysis of mixtures (C.A.MAN). Conventional cut-offs were calculated from a group of non-endemic individuals. A receiver operating characteristic (ROC) analysis was performed to demonstrate the impact of the choice of cut-offs on the test specificity and on the estimated seroprevalence among the endemic population. The data indicate that distribution analysis, especially the mixture analysis (C.A.MAN), are valuable tools for the unbiased estimation of seroprevalence when representative negative controls are not available.

Animals↗
PubMed1993

Mixture models and disease mapping.

The analysis and recognition of disease clustering in space and its representation on a map is one of the oldest problems in epidemiology. Some traditional methods of constructing such a map are presented. An alternative approach using mixture models to identify population heterogeneity and map construction within an empirical Bayes framework is described. For hepatitis B data from Berlin in 1989, a map is presented and the different methods are evaluated using a parametric bootstrap approach.

Bayes Theorem↗
PubMed1992

The area between curves (ABC)--measure in nutritional anthropometry.

This paper considers a statistic--recently suggested by Mora--for the deviation of a sample distribution from a reference distribution which typically arises in anthropometry when using the nutritional indicators height/age, weight/age or weight/height. The statistic measures the area between curves (ABC) and stands for the mass of the sample distribution which is not covered by the reference distribution. The paper provides a statistical framework for the ABC and includes some minor corrections of Mora's original paper. For the normal distribution situation with common or different variances, formulae are derived which include a partition of ABC into parts corresponding to malnourished and well-nourished groups. However, the main result is a non-parametric generalization of the ABC, motivated by the fact that the nutritional indicators often have skewed distributions with heavier left tails. Non-parametric statistical inference is provided by linking the ABC to the Kolmogorov-Smirnov statistic.

Analysis of Variance↗
PubMed1992

Computer-assisted analysis of mixtures (C.A.MAM): statistical algorithms.

This paper presents various algorithmic approaches for computing the maximum likelihood estimator of the mixing distribution of a one-parameter family of densities and provides a unifying computer-oriented concept for the statistical analysis of unobserved heterogeneity (i.e., observations stemming from different subpopulations) in a univariate sample. The case with unknown number of population subgroups as well as the case with known number of population subgroups, with emphasis on the first, is considered in the computer package C.A.MAN (Computer Assisted Mixture Analysis). It includes an algorithmic menu with choices of the EM algorithm, the vertex exchange algorithm, a combination of both, as well as the vertex direction method. To ensure reliable convergence, a step-length menu is provided for the three latter methods, each achieving monotonicity for the direction of choice. C.A.MAN has the option to work with restricted support size-that is, the case when the number of components is known a priori. In the latter case, the EM algorithm is used. Applications of mixture modelling in medical problems are discussed.

Algorithms↗
PubMed1990

Seasonal variation of wasting and stunting in preschool children during a three-year community-based nutritional intervention study in northeast Thailand.

A nutritional intervention project was undertaken for three years in four villages located in a relatively prosperous irrigation area in northeast Thailand where sub-clinical undernutrition is still prevalent. Data were collected at each measuring pont from approximately 270 preschool children in four intervention villages. Approximately 120 children in three other villages in the same area served as controls for each measuring point. Various activities such as teaching nutrition, introducing supplementary feeding, and using and understanding weight charts, as well as home visits, were carried out in the intervention villages. An obvious improvement in the nutritional status of children in the intervention villages could be observed only after two years and nine months. In order that immediate action could be taken when a child was found to be undernourished or at risk of becoming undernourished in the near future, the nutritional indicator weight-for-height was used to assess the children's nutritional status. However, the best indicator determining the impact of the project was height-for-age. Intervention measures were successful in diminishing marked seasonal variations in the nutritional status observed in the study area. Statistical methods are given to evaluate a prospective study based on an incomplete repeated measure method.

Age Factors↗