Dr. Mehrnaz Nikouyeh Emergency physician

  The study of the distribution and determinants of health-related states or events in specified populations and the application of this study to the control of health problems Definitions of Research

  To describe the frequency and extent of health conditions and disease.  To determine the burden of disease in a community, including socioeconomic impact of disease occurrence in specific populations.  To identify the causes and risk factors of specific diseases. This is the basis of disease prevention.  To evaluate medical interventions, including both preventive and therapeutic measures, and evaluate the delivery of these measures in health care settings.  To study the natural history and prognosis of disease.  To provide the foundation for developing public policy and regulatory decisions relating to health. PURPOSE OF EPIDEMIOLOGY

  A good clinical study starts with  a good question based on good hypothesis that is based on good and comprehensive review of the available evidence from pre-clinical and clinical data  Type of design depends on the question to be answered Where to Start?

  Focused and specific  Supported by available data  Not a replication of already established evidence  Ethical  Answerable  Methods, resources ….etc. Formulating a Research Question

  Specific aims  Clear and detailed  End point(s)  Primary  The main answer to the research question  Secondary  Answer other related questions Objectives

 Describe Analyze Your question Retrospective Prospective Your resources Acceptance of research Observational Interventional Community Study Design

 Descriptive Case repotCase seriesSurvey Analytic Observational Cross sectinal Case control Cohort Experimental RCT Clinical Study Types

 Types of primary studies  Descriptive studies  describe occurrence of outcome  Analytic studies  describe association between exposure and outcome

10 Observational study  Clinical trial exposed non exposed outcome Clinical Trial observational study describe as occurring in nature allocate randomly Ethics!

 Hierarchy of Epidemiologic Study Design

  Validity: Truth  External Validity:  Can the study be generalized to the population  Internal Validity:  Results will not be due to chance, bias or confounding factors  Symmetry Principle: Groups are similar Important issues in Study Design

 13  Confounding: distortion of the effect of one risk factor by the presence of another  Bias: Any effect from design, execution, & interpretation that shifts or influences results  Confounding bias: failure to account for the effect of one or more variables that are not distributed equally  Measurement bias: measurement methods differ between groups  Sampling (selection) bias: design and execution errors in sampling Important issues in Study Design

 Descriptive studies  Cannot establish causal relationships  Still play an important role in describing trends and generating hypotheses about novel association  Describing a novel phenomena

  Attempt to establish a causal link between a predictor/risk factor and an outcome.  You are doing an analytic study if you have any of the following words in your research question:  greater than, less than, causes, leads to, compared with, more likely than, associated with, related to, similar to, correlated with Analytic Studies

 Timeframe of Studies  Prospective Study - looks forward, looks to the future, examines future events, follows a condition, concern or disease into the future time Study begins here

 Timeframe of Studies  Retrospective Study - “to look back”, looks back in time to study events that have already occurred time Study begins here

 Case Reports  Detailed presentation of a single case or handful of cases  Generally report a new or unique finding  e.g. previous undescribed disease  e.g. unexpected link between diseases  e.g. unexpected new therapeutic effect  e.g. adverse events

 Case Series  Experience of a group of patients with a similar diagnosis  Assesses prevalent disease  Cases may be identified from a single or multiple sources  Generally report on new/unique condition  May be only realistic design for rare disorders

 Case Series  Advantages  Useful for hypothesis generation  Informative for very rare disease with few established risk factors  Characterizes averages for disorder  Disadvantages  Cannot study cause and effect relationships  Cannot assess disease frequency

Case Report Case Series Descriptive Epidemiology Study One case of unusual findings Multiple cases of findings Population-based cases with denominator

 Observational Studies  non-experimental  observational because there is no individual intervention  treatment and exposures occur in a “non-controlled” environment  individuals can be observed prospectively, retrospectively, or currently  Possibility of confounding  No control over study units  need to clearly describe study individuals  Can study risk factors that have serious consequences  Study individuals in their natural environment (>> extrapolation)

  cohort  Case control  Cross sectional Types of observational studies

  Evaluate the effect of a suspected risk factor (exposure) on an outcome (e.g. disease)  define “exposure” and “disease”  Describe the impact of the risk factor on the frequency of disease in a population Aims of observational studies

 Temporal relations of observational studies

 Cross-sectional studies  An “observational” design that surveys exposures and disease status at a single point in time (a cross-section of the population) time Study only exists at this point in time

 Cross-sectional Design time Study only exists at this point in time Study population No Disease Disease factor present factor absent factor present factor absent

 Cross-sectional Studies  Often used to study conditions that are relatively frequent with long duration of expression (nonfatal, chronic conditions)  It measures prevalence, not incidence of disease. Comparison of prevalence among exposed and non- exposed.  Not suitable for studying rare or highly fatal diseases or a disease with short duration of expression

 Cross - Sectional Study  Random sample from population  i.e. results reflect reference population  Estimates the frequencies of both exposure and outcome in the population  Measuring both exposure and outcome at one point in time  Typically a survey

 Cross - Sectional Study advantages  Can study several exposure factors and outcomes simultaneously  Determines disease prevalence  Helpful in public health administration & planning  Quick  Low cost (e.g. mail survey)  Very useful for public health planning  Disease etiology. Conduct this by obtaining data on risk factors for a disease.  Hypothesis generating

 Cross-sectional studies disadvantages  Weakest observational design, (it measures prevalence, not incidence of disease). Prevalent cases are survivors  The temporal sequence of exposure and effect may be difficult or impossible to determine  Usually don’t know when disease occurred  Rare events a problem. Quickly emerging diseases a problem  Does not determine causal relationship

 Cross-Sectional: Pediatrician-to-Child Ratio Greg et al. (2001) Pediatrics.107(2):e18

 Case-Control Study Design Cases Controls Exposed Unexposed Exposed Unexposed Time Data collection Direction of inquiry Q: What happened?

 Case control studies  Determines the strength of the association between each predictor variable and the presence or absence of disease  Cannot yield estimates of incidence or prevalence of disease in the population (why?)  Odds Ratio is statistics

 Case-Control Study  Retrospective  Can use hospital or health register data  First identify cases  Then identify suitable controls  Hardest part: who is suitable ??  Critical that the exposure in the controls is representative of the exposure in the population  Ideal controls would have same/similar characteristics as the cases  Matching cases to controls  Then inquire or retrieve previous exposure  By interview  By databases (e.g. hospital, health insurance)

 Case-Control Study  Strengths  Less expensive and time consuming  Efficient for studying rare diseases  Can study many risk factors at the same time  Smaller sample  Multiple etiologic factors evaluated for single disease  Good for diseases with long latency

  Limitations  Inappropriate when disease outcome for a specific exposure is not known at start of study  Exposure measurements taken after disease occurrence  Disease status can influence selection of subjects  Confounding likely  difficulties in selection of controls  ascertainment of disease & exposure status  inefficient for rare exposures unless attributable risk is high Case-Control Study

 Case-control study-minuses - Causality still difficult to establish - Selection bias (appropriate controls) - Caffeine and Pancreatic cancer in the GI clinic - Recall bias: sampling (retrospective) - Abortion and risk of breast cancer in Sweden

Case Selection Define source population Cases –incident/prevalent –diagnostic criteria (sensitivity + specificity) Controls –selected from same population as cases –select independent of exposure status

Control Selection Random selection from source population Hospital based controls: –convenient selection –controls from variety of diagnostic groups other than case diagnosis –avoid selection of diagnoses related to particular risk factors –limit number of diagnoses in individuals

 Case-Crossover Studies  Study of “triggers” within an individual  ”Case" and "control" component, but information of both components will come from the same individual  ”Case component" = hazard period which is the time period right before the disease or event onset  ”Control component" = control period which is a specified time interval other than the hazard period

 Prospective Cohort Study without outcome Cohort with outcome without outcome Exposed Unexposed Time Onset of study Direction of inquiry Q: What will happen?

 Prospective Cohort study Measure exposure and confounder variables Exposed Non-exposed Outcome Baseline time Study begins here

 Retrospective Cohort study Measure exposure and confounder variables Exposed Non-exposed Outcome Baseline time Study begins here

 Cohort studies  Follow-up studies; subjects selected on presence or absence of exposure & absence of disease at one point in time. Disease is then assessed for all subjects at another point in time.  Typically prospective but can be retrospective, depending on temporal relationship between study initiation & occurrence of disease.

 Cohort studies 1. Prospective cohort study – outcome event of interest occur after the study is initiated 2. Retrospective cohort study – outcome events of interest have already occurred at study initiation  Can be regarded as a reconstruction of a cohort study which has already taken place  Generally cheaper and faster than a prospective study  Requires good historical exposure data on subjects

 Cohort studies  More clearly established temporal sequence between exposure & disease  Can establish population-based incidence  Accurate relative risk (risk ratio) estimation  Can examine rare exposures (asbestos > lung cancer)  Can be used where randomization is not possible  Allows direct measurement of incidence  Examines multiple effects of a single exposure  - Magnitude of a risk factor’s effect can be quantified  Selection and information biases are decreased

 Cohort Study  Strengths  Exposure status determined before disease detection  Subjects selected before disease detection  Can study several outcomes for each exposure  Directly measure incidence of a disease outcome  Limitations  Expensive and time-consuming  Inefficient for rare diseases or diseases with long latency  Loss to follow-up and unavailability of data  potential confounding factors  Unexpected environmental changes may influence the association

 Basic Question in Analytic Epidemiology  Are exposure and disease linked? ExposureDisease

 Experimental Studies  In an experiment, we are interested in the consequences of some treatment on some outcome.  The subjects in the study who actually receive the treatment of interest are called the treatment group.  The subjects in the study who receive no treatment or a different treatment are called the comparison group.

 Epidemiologic Study Designs  Randomized Controlled Trials (RCTs)  a design with subjects randomly assigned to “treatment” and “comparison” groups  provides most convincing evidence of relationship between exposure and effect  not possible to use RCTs to test effects of exposures that are expected to be harmful, for ethical reasons

  Uncontrolled  Controlled  Before/after (cross-over)  Historical  Concurrent, not randomized  Randomized Clinical Trial: Study Design

  May Be Appropriate  Early studies of new and untried therapies  Uncontrolled early phase studies where the standard is relatively ineffective  Investigations which cannot be done within the current climate of controversy  Truly dramatic response Non-randomized Trials

 Randomized controled trial time Study begins here (baseline point) Study population Intervention Control outcome no outcome outcome no outcome baseline future RANDOMIZATION

 Epidemiologic Study Designs  Randomized Controlled Trials (RCTs)  the “gold standard” of research designs  provides most convincing evidence of relationship between exposure and effect  trials of hormone replacement therapy in menopausal women found no protection for heart disease, contradicting findings of prior observational studies

 62 Advantages of Randomized Control Clinical Trial  advantages  Randomization tends to produce comparable groups  Assure causal relationship  Randomization produces valid statistical tests

 63 Disadvantages of Randomized Control Clinical Trial 1.Generalizable Results?  Participants studied may not represent general study population. 2.Recruitment  Hard 3.Acceptability of Randomization Process  Some physicians will refuse  Some participants will refuse 4.Administrative Complexity 5.Very expensive

64 Clinical Trials-Phases Phase I - Does it hurt the Patient? Phase I - Does it hurt the Patient? Usually in normal volunteers, small groups for safety testing Usually in normal volunteers, small groups for safety testing Phase II - Does it help the Patient? Phase II - Does it help the Patient? On patients to confirm the effectiveness of the drug On patients to confirm the effectiveness of the drug Phase III - Is it any better? Phase III - Is it any better? Large groups of patients for statistical confirmation of effect and incidence of side-effects Large groups of patients for statistical confirmation of effect and incidence of side-effects Phase IV - Does it work in the community? Phase IV - Does it work in the community? Post marketing studies. Fine tuning and new rare findings from a very large population Post marketing studies. Fine tuning and new rare findings from a very large population