Introductory data analysis

2018 ◽  
pp. 168-191
Author(s):  
Alan J. Silman ◽  
Gary J. Macfarlane ◽  
Tatiana Macfarlane
Keyword(s):  
Data Analysis ◽  
Cohort Studies ◽  
Worked Examples ◽  
Epidemiological Studies ◽  
Case Control ◽  
Raw Data ◽  
Easy Task ◽  
Data Files ◽  
Mathematical Calculation ◽  
Study Designs

This chapter builds on the previous one on the analysis of descriptive epidemiological studies and illustrates statistical methods appropriate for analysis of analytical epidemiological studies. It mainly focuses on data obtained from case–control and cohort studies, but also considers other study designs presented in Chapter 6. There are also several practical examples to help with the analysis and interpretation of the results of analytical epidemiological studies. In practice, relatively little mathematical calculation is done without computers. In this chapter, however, formulae are presented for the main measures of effect together with worked examples. Indeed, when data are available in tabulated form, as opposed to raw data files, it is frequently an easy task to calculate the important measures ‘by hand’. The formulae presented will permit the reader, for example, to check or further explore data published by others.

scholarly journals Dietary patterns and CVD: a systematic review and meta-analysis of observational studies

2015 ◽  
Vol 114 (9) ◽  
pp. 1341-1359 ◽  
Author(s):  
Míriam Rodríguez-Monforte ◽  
Gemma Flores-Mateo ◽  
Emília Sánchez
Keyword(s):  
Cohort Studies ◽  
Dietary Patterns ◽  
Observational Studies ◽  
Protective Factor ◽  
Meta Analysis ◽  
Epidemiological Studies ◽  
Case Control ◽  
Case Control Studies ◽  
Chd Risk ◽  
Control Comparison

AbstractEpidemiological studies show that diet is linked to the risk of developing CVD. The objective of this meta-analysis was to estimate the association between empirically derived dietary patterns and CVD. PubMed was searched for observational studies of data-driven dietary patterns that reported outcomes of cardiovascular events. The association between dietary patterns and CVD was estimated using a random-effects meta-analysis with 95 % CI. Totally, twenty-two observational studies met the inclusion criteria. The pooled relative risk (RR) for CVD, CHD and stroke in a comparison of the highest to the lowest category of prudent/healthy dietary patterns in cohort studies was 0·69 (95 % CI 0·60, 0·78; I2=0 %), 0·83 (95 % CI 0·75, 0·92; I2=44·6 %) and 0·86 (95 % CI 0·74, 1·01; I2=59·5 %), respectively. The pooled RR of CHD in a case–control comparison of the highest to the lowest category of prudent/healthy dietary patterns was 0·71 (95 % CI 0·63, 0·80; I2=0 %). The pooled RR for CVD, CHD and stroke in a comparison of the highest to the lowest category of western dietary patterns in cohort studies was 1·14 (95 % CI 0·92, 1·42; I2=56·9 %), 1·03 (95 % CI 0·90, 1·17; I2=59·4 %) and 1·05 (95 % CI 0·91, 1·22; I2=27·6 %), respectively; in case–control studies, there was evidence of increased CHD risk. Our results support the evidence of the prudent/healthy pattern as a protective factor for CVD.

Study designs which can demonstrate and test causation

2017 ◽  
Author(s):  
Mark Elwood
Keyword(s):  
Cohort Studies ◽  
Case Control ◽  
Control Group ◽  
Case Control Studies ◽  
Retrospective Case ◽  
Community Based ◽  
Cross Sectional ◽  
Healthcare Interventions ◽  
Intervention Trials ◽  
Study Designs

This chapter presents study designs which can test and show causation. Cohort and intervention studies compare people exposed to an agent or intervention with those unexposed or less exposed. Case-control studies compare people affected by a disease or outcome with a control group of unaffected people or representing a total population. Surveys select a sample of people, not chosen by exposure or outcome. Cohort studies may be prospective or retrospective; case-control studies are retrospective; surveys are cross-sectional in time, but retrospective or prospective aspects can be added. In part two, strengths, weaknesses and applications of these designs are shown. Intervention trials, ideally randomised, are the prime method of assessing healthcare interventions; special types include crossover trials and community-based trials. Non-randomised trials are noted. The strengths and weaknesses of cohort studies, case-control studies, and surveys are shown.

scholarly journals Association Between Infections and Risk of Ankylosing Spondylitis: A Systematic Review and Meta-Analysis

2021 ◽  
Vol 12 ◽  
Author(s):  
Xiao Zhang ◽  
Zhe Sun ◽  
Aihong Zhou ◽  
Lei Tao ◽  
Yingxin Chen ◽  
...  
Keyword(s):  
Ankylosing Spondylitis ◽  
Cohort Studies ◽  
Subgroup Analysis ◽  
Meta Analysis ◽  
Random Effect ◽  
Epidemiological Studies ◽  
Case Control ◽  
Total Risk ◽  
Case Control Studies ◽  
Increased Risk

BackgroundPrevious literature on the association between infections and the risk of developing ankylosing spondylitis (AS) presented controversial results. This meta-analysis aimed to quantitatively investigate the effect of infections on the risk of AS.MethodsWe searched the PubMed, Embase, and Web of Science databases until March 26, 2021 for analytical epidemiological studies on the association between infections and the risk of AS. Fixed or random effect models were used to calculate total risk estimates based on study heterogeneity. Subgroup analysis, and sensitivity analysis were also performed. Publication bias was estimated using funnel plots and Begg’s test.ResultsSix case-control articles (n=1,296,239) and seven cohort articles (n=7,618,524) were incorporated into our meta-analysis. The pooled odds ratio (OR) from these case-control studies showed that infections were associated with an increased risk of AS (OR=1.46, 95% confidence interval [CI], 1.23–1.73), and the pooled relative risk (RR) from the cohort studies showed the same findings (RR=1.35, 95% CI, 1.12–1.63). Subgroup analysis showed that infections in participants with unadjusted comorbidities (OR=1.66, 95% CI, 1.35–2.03), other types of infection (OR=1.40, 95% CI, 1.15–1.70), and infection of the immune system (OR=1.46, 95% CI, 1.42–1.49) were associated with the risk of AS in case-control studies. In cohort studies, infections with adjusted comorbidities (RR=1.39, 95% CI, 1.15–1.68), viral infection (RR=1.43, 95% CI, 1.22–1.66), other types of infection (RR=1.44, 95% CI, 1.12–1.86), and other sites of infection (RR=1.36, 95% CI, 1.11–1.67) were associated with an increased risk of AS.ConclusionsThe findings of this meta-analysis confirm that infections significantly increase the risks of AS. This is helpful in providing an essential basis for the prevention of AS via the avoidance of infections.

Critical Appraisal of Epidemiological Studies and Clinical Trials

2017 ◽  
Author(s):  
Mark Elwood
Keyword(s):  
Cohort Studies ◽  
Statistical Methods ◽  
Critical Appraisal ◽  
Causal Effect ◽  
Internal Validity ◽  
Epidemiological Studies ◽  
Case Control ◽  
Randomised Trials ◽  
Case Control Studies ◽  
Retrospective Cohort Studies

This book presents a system of critical appraisal applicable to clinical, epidemiological and public health studies and to many other fields. It assumes no prior knowledge. The methods are relevant to students, practitioners and policymakers. The book shows how to assess if the results of one study or of many studies show a causal effect. The book discusses study designs: randomised and non-randomised trials, cohort studies, case-control studies, and surveys, showing the presentation of results including person-time and survival analysis, and issues in the selection of subjects. The system shows how to describe a study, how to detect and assess selection biases, observation bias, confounding, and chance variation, and how to assess internal validity and external validity (generalisability). Statistical methods are presented assuming no previous knowledge, and showing applications to each study design. Positive features of causation including strength, dose-response, and consistency are discussed. The book shows how to do systematic reviews and meta-analyses, and discusses publication bias. Systems of assessing all evidence are shown, leading to a general method of critical appraisal based on 20 key questions in five groups, which can be applied to any type of study or any topic. Six chapters show the application of this method to randomised trials, prospective and retrospective cohort studies, and case-control studies. An appendix summarises key statistical methods, each with a worked example. Each main chapter has self-test questions, with answers provided.

scholarly journals The Microbiome

2016 ◽  
Vol 45 (1) ◽  
pp. 190-194 ◽  
Author(s):  
Ellen K. Silbergeld
Keyword(s):  
Data Analysis ◽  
Human Development ◽  
Epidemiological Studies ◽  
Critical Component ◽  
Health And Disease ◽  
Study Designs

The microbiome is increasingly recognized as a critical component in human development, health, and disease. Its relevance to toxicology and pharmacology involves challenges to current concepts related to absorption, metabolism, gene:environment, and pathways of response. Framing testable hypotheses for experimental and epidemiological studies will require attention to study designs, biosampling, data analysis, and attention to confounders.

Consumption of Artificially-sweetened Soft Drinks and Risk of Gastrointestinal Cancer: A Meta-analysis of Observational Studies

2021 ◽  
pp. 1-50
Author(s):  
Alfred Jatho ◽  
Jansen Marcos Cambia ◽  
Seung-Kwon Myung ◽  
Keyword(s):  
Cohort Studies ◽  
Observational Studies ◽  
Meta Analysis ◽  
Epidemiological Studies ◽  
Case Control ◽  
Soft Drinks ◽  
Case Control Studies ◽  
Gi Cancer ◽  
Increased Risk ◽  
Meta Analyses

Abstract Objective: There remain inconclusive findings from previous observational epidemiological studies on whether consumption of artificially-sweetened soft drinks (ASSDs) increases the risk of gastrointestinal (GI) cancer. We investigated the associations between the consumption of ASSDs and the risk of GI cancer using a meta-analysis. Design: Systematic review and meta-analysis. Setting: PubMed and EMBASE were searched using keywords until May 2020 to identify observational epidemiological studies on the association between the consumption of ASSDs and the risk of GI cancer. Subjects: Twenty-one case-control studies and 17 cohort studies with 12,397 cancer cases and 2,474,452 controls. Results: In the random-effects meta-analysis of all the studies, consumption of ASSDs was not significantly associated with the risk of overall GI cancer (odds ratio (OR)/relative risk (RR), 1.02; 95% CI, 0.92-1.14). There was no significant association between the consumption of ASSDs and the risk of overall GI cancer in the subgroup meta-analyses by study design (case-control studies: OR, 0.95; 95% CI, 0.82-1.11; cohort studies: RR, 1.14; 95% CI, 0.97-1.33). In the subgroup meta-analysis by type of cancer, consumption of ASSDs was significantly associated with the increased risk of liver cancer (OR/RR, 1.28; 95% CI,1.03-1.58). Conclusions: The current meta-analysis of observational epidemiological studies suggests that overall, there is no significant association between the consumption of ASSDs and the risk of GI cancer.

scholarly journals Research made Easy: Answering Important Questions with Valid Designs

2012 ◽  
Vol 46 (1) ◽  
pp. 8-11
Author(s):  
Harman Chaudhry ◽  
Mohit Bhandari
Keyword(s):  
Clinical Study ◽  
Cohort Studies ◽  
Case Reports ◽  
Research Question ◽  
Case Control ◽  
Controlled Trials ◽  
Case Control Studies ◽  
Randomized Controlled ◽  
Observational Cohort ◽  
Study Designs

ABSTRACT Clinical research fundamentally involves finding answers to questions. Next to asking important questions, determining what type of study design to use is arguably the most pivotal step for a researcher. In this article, we provide an overview of various clinical study designs, including case reports and series, case-control studies, observational cohort studies, randomized controlled trials and systematic reviews. We aim to elucidate the utility, advantages and drawbacks of these study designs in order to assist researchers in selecting the most valid design for their research question. How to cite this article Chaudhry H, Bhandari M. Research made Easy: Answering Important Questions with Valid Designs. J Postgrad Med Edu Res 2012;46(1):8-11

Observational Studies

2018 ◽  
pp. 324-361
Author(s):  
Minou Djannatian ◽  
Clarissa Valim ◽  
Andre Brunoni ◽  
Felipe Fregni
Keyword(s):  
Cohort Studies ◽  
Observational Studies ◽  
Real Life ◽  
Case Control ◽  
Sample Size Determination ◽  
Size Determination ◽  
Cross Sectional ◽  
Life Conditions ◽  
Study Designs

This chapter on observational studies provides an understanding of the main concepts in epidemiology, introduces common study designs, such as cross-sectional, case-control, and cohort studies, and outlines their importance for clinical research. The hallmark of epidemiological research is that it observes unexposed and exposed individuals under “real-life conditions” without intervening itself. The chapter emphasizes the important role of bias and confounding in interpreting results from such studies and explains how bias and confounding can be controlled. It furthermore discusses specific aspects of sample size determination that are relevant to observational studies. The chapter concludes with a brief review of the special nature of surgical research.

Introduction

2017 ◽  
Author(s):  
Mark Elwood
Keyword(s):  
Cohort Studies ◽  
Case Control ◽  
Randomised Trials ◽  
Case Control Studies ◽  
Causal Explanations ◽  
Bradford Hill ◽  
Definition Of ◽  
Meta Analyses ◽  
Observation Bias ◽  
Study Designs

This chapter shows the plan of the book. Later chapters will cover the definition of causation, study designs that can demonstrate causation, how results are presented, the interpretation of studies stressing the non-causal explanations of observation bias, confounding, and chance variation. Then come positive aspects of causation, the Bradford-Hill principles, systematic reviews and meta-analyses, and an overall scheme for assessing studies and diagnosing causation. Further chapters present appraisals of six published studies: randomised trials, cohort studies, and case-control studies. An appendix presents statistical methods with examples.

scholarly journals Towards non-conventional methods of designing register-based epidemiological studies: An application to pediatric research

2017 ◽  
Vol 45 (17_suppl) ◽  
pp. 30-35 ◽  
Author(s):  
Tong Gong ◽  
Bronwyn Brew ◽  
Arvid Sjölander ◽  
Catarina Almqvist
Keyword(s):  
Birth Weight ◽  
Low Birth Weight ◽  
Epidemiological Studies ◽  
Case Control ◽  
Observational Research ◽  
Advantages And Disadvantages ◽  
Study Population ◽  
Family Based ◽  
Study Designs ◽  
Residual Confounding

Aims: Various epidemiological designs have been applied to investigate the causes and consequences of fetal growth restriction in register-based observational studies. This review seeks to provide an overview of several conventional designs, including cohort, case-control and more recently applied non-conventional designs such as family-based designs. We also discuss some practical points regarding the application and interpretation of family-based designs. Methods: Definitions of each design, the study population, the exposure and the outcome measures are briefly summarised. Examples of study designs are taken from the field of low birth-weight research for illustrative purposes. Also examined are relative advantages and disadvantages of each design in terms of assumptions, potential selection and information bias, confounding and generalisability. Kinship data linkage, statistical models and result interpretation are discussed specific to family-based designs. Results: When all information is retrieved from registers, there is no evident preference of the case-control design over the cohort design to estimate odds ratios. All conventional designs included in the review are prone to bias, particularly due to residual confounding. Family-based designs are able to reduce such bias and strengthen causal inference. In the field of low birth-weight research, family-based designs have been able to confirm a negative association not confounded by genetic or shared environmental factors between low birth weight and the risk of asthma. Conclusions: We conclude that there is a broader need for family-based design in observational research as evidenced by the meaningful contributions to the understanding of the potential causal association between low birth weight and subsequent outcomes.

Sign in / Sign up

Export Citation Format

Share Document