The research activities at the Division of Biostatistics focus on the development and application of statistical methods that may help improve medical and public health research from designing new studies, to analyzing data, interpreting results, and disseminating scientific findings. The Division of Biostatistics has a multidisciplinary approach to research and fosters collaboration with researchers in epidemiology, clinical research, and public health, biological, social, and basic sciences.
Theory and Methods
The following are some of the current topics of our research in statistical theory and methods.
Artificial effect size magnification may occur in underpowered studies, where effects are only reported because they or their associated p-value have passed some threshold.
Event Probabilities Risks and Regression Methods
Methods for the analysis of the probability of occurrence of events, such as death or cancer diagnosis.
Inferences and Conjectures
Separating inferences from conjectures is essential when interpreting conditional and average effects in randomized and observational studies.
A regression method for the analysis of percentiles of survival time like, for example, median survival.
Quantile Regression Methods with Dependent Data
Regression methods for the analysis of percentiles of a variable with dependent data, as those that may arise from longitudinal, cluster, and matched case-control studies.
Logistic Quantile Regression
A regression method for the analysis of variables bounded within a specified range like, for example, visual analog scales, psychological scales, and percentages.
Optimal Correlation-Based Prediction
Best predictors of values of an outcome variable based on the observed values of predictor variables.
Parametric Quantile Models
Parametric modeling of the conditional quantile function of an outcome variable given covariates, applicable to censored and truncated data.
Singular Information Matrix
Non-regular scenario in likelihood-based inference when the Fisher information matrix is singular at some points in the parameter space.
Summary measures of probability distributions
Novel statistical summary measures to extract maximum information from biomedical data.
An imputation method of the analysis of data with missing observations.
The following are some of the current topics of our research applications.
Methods for assessing the spatio-temporal relationship between traffic-related air pollutants and respiratory health.
Assessment of the relationships between dietary intake of trace elements and adverse outcomes of chronic diseases.
Assessment of the perinatal exposure to phthalates, heavy metals and insecticides on mental and physical health in children.
The following are online calculators for diagnosis and classification of different diseases and syndroms.
European League Against Rheumatism/American College of Rheumatology classification criteria for adult and juvenile idiopathic inflammatory myopathies and their major subgroups
Progression-free probability after radical prostatectomy for localized prostate cancer
Risk of metachronous peritoneal carcinomatosis from colorectal cancer
Three-month risk of psychological morbidity in ICU survivors
Division of Biostatistics, Nobels väg 13, karolinska Institutet, 17177 Stockholm, Sweden