Absite: Biostats

• Clinical epidemiology


  • The study of the distribution and determinants of disease in a population
  • Involves the application of traditional epidemiology to patients in a clinical setting
• Two most important types of epidemiologic studies

  • Observational studies
  • Experimental studies
• Observational studies


  Researchers observe the effect of a risk factor or intervention without modifying or controlling the independent variable
• Experimental studies


  Researchers introduce an intervention to study the results
• Descriptive studies


  • Case reports
  • Case series
  • Ecological studies
  • Cross-sectional studies
• Descriptive studies


  • Identify individual characteristics, places, or events in relation to an outcome
  • No clinical intervention is involved, and the independent variable is not manipulated
  • The observations are used to create a hypothesis
  • No comparison group is used
• Analytical studies


  • Cohort studies
  • Case-control studies
  • Cross-sectional studies
• Analytical studies


  • Evaluate the relationship between an exposure and an outcome
  • Always involve a comparison group
  • Used to test a hypothesis
• Case reports


  • Describe disease presentation, treatment, and outcome in a single subject
  • Generally used to describe an unusual diagnosis or clinical treatment algorithm
• Case reports


  • Lack of comparison group or hypothesis
  • Relatively low number of patients
  • Difficult to assess causality
  • Lack of generalizability
  • Selection bias in the patients
• Case series


  • Present data on multiple similar patients
  • Used to present an unusual diagnosis or treatment algorithm
• Case series


  • Similar disadvantages to case reports
• Ecological studies


  • Aim to identify links between an exposure and an outcome (e.g. disease), especially if the outcome is rare
  • Assess aggregated data where at least one variable (e.g. an outcome) is at a population level and not an individual level
• Ecological studies


  • Making inferences about individuals based on group characteristics
  • Uncontrolled confounding variables
• Cross-sectional studies


  • Observe a population at a single point in time or over a specific time interval to identify the prevalence of a disease or disease-associated risk factors
  • Can be either descriptive or analytical
• Cross-sectional studies


  • Cannot establish causality, measure incidence, or measure risk
• Case-control studies


  • Aim to retrospectively study whether an exposure is associated with an outcome
  • Researchers need individuals with and without the disease for comparison
  • The odds ratio is used
• Case-control studies


  • Retrospective design
  • Recall bias
  • Selection bias
  • Cannot be used to determine prevalence and incidence
• Cohort studies


  • Can be retrospective or prospective
  • Examine a large population and separate study groups based on exposure status
  • Aim to study the incidence rate and whether a given exposure is associated with the outcome of interest
• Prospective cohort studies


  • More complete information gathering
  • Limited by time and resources
• Retrospective cohort studies


  • Rely on data previously collected at the start of creating the cohort
• Prospective randomized controlled studies

  • Useful in situations where it is unclear whether a particular treatment provides a benefit over the current standard of care
  • Aim to determine the possible effect of a specific intervention on a given population
• Prospective randomized controlled studies

  • Well-conducted, blinded randomized controlled trials are the gold standard of study design
  • Minimizes bias
  • Can demonstrate causality
  • Cannot be used to evaluate rare diseases
  • Often limited by cost and length of follow-up
• Crossover studies


  • Randomized controlled trials where the participants act as their own controls
  • Observe the effect a series of two or more treatments has on a participant
• Crossover studies


  • Smaller sample size required
  • Every patient receives treatment
  • Need for a treatment washout period
• Phases of clinical trials

  • Phase I: healthy pts, pharmacodynamics of drug
  • Phase II: determines side effects, toxicity and pharmacokinetics of drug
  • Phase III: RCT compare drug to standard tx
  • Phase IV: after market rare side effects
• Systematic reviews


  • Provide a qualitative overview of available literature about a particular topic
  • Address a defined research question and can improve evidence-based clinical decision making
• Systematic reviews


  • Cannot correct issues with the quality of individual studies included
  • Susceptible to publication bias
• Meta-analyses


  • Pool data from multiple studies and then provide statistical analysis of the pooled data
  • Increased statistical power
  • Multivariate analyses are often possible
• Meta-analyses


  • Heterogeneity of the studies that form the basis for the analysis
• Common primary data collection strategies in qualitative research

  • Observation
  • Interview
  • Focus groups
  • Surveys
• Three major types of IRB reviews

  • Exempt reviews
  • Expedited reviews
  • Full reviews
• Exempt reviews


  Reviewed to ensure that a study poses no more than minimal risk to the human subjects involved
• Expedited reviews


  Pose no more than minimal risk to human subjects and fall under one of the federally regulated categories
• Full reviews


  Involve research on human subjects with more than minimal risk
• Information bias


  Critical information is distorted, or compared information is mismatched, leading to study inconsistencies
• Detection bias


  Nonuniform or poorly standardized methods for measuring outcomes
• Ascertainment bias


  Distorted determination of exposure to a factor that is being studied
• Interviewer bias


  Tendency of an interviewer to obtain answers that supports their preconceived notions
• Observer bias (Hawthorne effect)


  Awareness of being under observation alters the way in which a study participant behaves