Introduction to Epidemiology 1 PDF

**[Core Principles of Mental Health Research]** **[Introduction to Epidemiology 1]** [Learning Outcomes:] - Be familiar with the history of epidemiology and psychiatry epidemiology - Understand the core principles of epidemiology research and differences between key measures of disea...

**[Core Principles of Mental Health Research]** **[Introduction to Epidemiology 1]** [Learning Outcomes:] - Be familiar with the history of epidemiology and psychiatry epidemiology - Understand the core principles of epidemiology research and differences between key measures of disease frequency - Understand how to critique a research study along three key lines of investigation -- chance, bias (selection & measurement), confounding and reverse causality - How to discuss whether associations are casual [Investigating causes:] - Hypothesis: Physical activity reduces the likelihood of depression - Outcome: Depression - Exposure: Physical activity - Study Design: Cross Sectional survey Come from: - Lab studies - Disease surveillance - Case reports - Theoretical speculation - Clinical observation Outcome happens after exposure [Epidemiological Reasoning:] Study Design: - Identify the question + - Systematic collection of data + - Systematic analysis of data + Interpretation - Assess threats to validity - - Casual interference -- [Making Inferences from associations:] - Association does not mean causation - When deciding if a relationship is casual, these issues must be considered: - Chance (statistics) - Bias (selection bias, measurement bias) - Confounding - Reverse causality (cause opposite to what we assume -- e.g., depression causes less exercise) - Must then weigh up likelihood that association found is actually causal (i.e., causal inference) [Making inferences from epidemiological observations:] - Estimated from the sample are not necessarily the same as the (true) value from the population Why does chance matter? - Can't test everyone - Since we must sample, we are faced with uncertainty -- we need to quantify this uncertainty using statistics (confidence intervals & probability) - Conversely, to save us studying everyone, we can use adequately sized samples to make statistically robust inferences about the population [Chance:] - Need to consider if the association results by chance, as a result of random error - Statistical inference (p-values and confidence intervals) help make a judgement about whether sampling variation can explain the result - Type 1 error: "statistically significant" result occurs by chance (p-value) accept null when its false - Type 2 error: real result not "statistically significant" because of chance (1-P) reject null when its true - Confidence intervals -- give you information about likely values and whether your study has excluded an important result [95% Confidence Intervals:] - The CI and p-value are both derived from the size of the difference between groups and its standard error, so they are closely related - The standard error decreases as sample size increases so the width of the CI and size of p are as dependent on sample size as they are on the size of the difference - The larger the sample, the narrower the CI and the smaller the p-value (more precise) - in confidence intervals mid-point is odds ratio value - If confidence interval ranges between 0.3 -- 1.5 (aka crosses 1) means we have failed to reject the null hypothesis (results found by chance) [Bias: ] - Systematic error introduced during the design or conduct of a study - If there is bias your estimate of association is not the true value as observed in the population, you are studying - Bias can reduce and increase the size of association - It is difficult or impossible to quantify bias -- but you can identify likely sources and discuss possible impact [Selection Bias:] - Bias that arises from the procedures used to select individuals into the study or analysis - It is about the people in your study (or not in your study) - The sample of participants selected differs on characteristics than the target population with respect to factors that are associated both with the exposure and the outcome [Examples:] - Low response rate in a survey - Participants in a clinical trial drop out of follow up - Case-control studies are particularly sensitive to selection bias [Measurement Bias:] - Bias that results from the measurement of exposure, outcome or other information obtained on participants -- measurement bias is sometimes called "information bias" [Many types including:] - Observer bias: error introduced by observer expectations - Recall bias: if the illness (outcome) affects memory for past events [Confounding:] - Alternative explanation of the association between an exposure and outcome - A confounder is a "third" variable that is associated with the exposure and outcome and results in a change in the observed association - Confounding can lead to spurious associations or eliminate real ones [Methods for dealing with confounding:] - You have to collect data on potential confounders when you carry out a study so need to think about them at the design stage [Main methods for taking account of confounding are:] 1. Randomisation: the best way if you can but often not possible 2. Adjusting using multivariable methods: after adjustment for a confounder the association between exposure and outcome will change Residual Confounding: you can almost never take account of confounding perfectly and there are unknown confounders, so there is always the possibility of confounding even after adjustment [Confounding or Bias?] - Confounding still occurs even if the study is perfectly designed (which is impossible) - Confounding is an alternative explanation for the association that is present in the study population - Bias is introduced by the investigator (not intentionally) as a result of carrying out the study [Reverse Causality:] - Main way of addressing this is to carry out a longitudinal study where you measure the exposure the exposure in people without the outcome and then follow them up to see who becomes ill [Dealing with chance, bias & confounding] - Chance - Sample Size - Confidence Intervals - Bias: - Sampling Strategy - Measures Used - Design of study - Retrospective vs Prospective - Confounding: - Restriction (Design) - Matching (Design) - Adjustment (Analysis) [Models of Causation:] - "Neither necessary nor sufficient": disease can occur without the cause - E.g., elderly woman who had smoked thousands of cigarettes in their life being fine [Bradford-Hill Considerations on causality:] - Temporality: exposure precedes outcome - Strength: stronger associations are more likely to be casual -- usually taken to be \> Relative Risk of 1.5 to 2 - Dose-Response: as the level of exposure increases so does the risk of the outcome - Consistency: different studies in different populations find the association - Specificity: exposure produces a specific effect - Coherence: triangulation across different sources of evidence - Plausibility: aligns with known biological or other processes - Analogy: similar exposures lead to the outcome - Experimental evidence: animal work or experimental medicine studies support the hypothesis [Study Designs:] - Observational vs Experimental (choice of study design will depend on whether the investigator assigned the exposure [Descriptive Epidemiology: Cross-sectional studies] - Conducted on individuals - Exposure & outcome status ascertained at a "cross-sectional" point in time - Often survey based i.e., questionnaire - Chiefly concerned with assessing frequency of disease occurrence - i.e., prevalence - and distribution i.e., by sex, age, race, social class - less useful for determinants of health [Strengths & Limitations] [Descriptive Epidemiology: Ecological Studies] - Conducted on populations - Disease frequencies compared between different groups in same period - Or, in the same population at different times - Often use routine data - Can reveal patterns at the population level - As before, limited in terms of determinants [Strengths & Limitations] [Analytical Epidemiology: Case-Control Studies ] - Conducted at the individual level - Starting point is case status - Identification of people with the disease outcome i.e., cases - A sample of controls (without the disease) are recruited & compared with respect to exposure of interest [Strengths & Limitations] [Analytical Epidemiology: Cohort Studies] - A population is identified without the disease outcome - Risk factors measured at points during follow up - Followed over time to see who develops disease [Strengths & Limitations] [Intervention Studies: Randomised Controlled Trials ] - A form of prospective cohort study - People identified based on exposure status & followed to see who experiences outcome - What is the major strength of a randomised trial? - Random allocation to exposure will, on average, ensure balance of other factors which could affect disease risk [Summary:] - No study design is optimal, all have strengths & weaknesses - Is some circumstances only certain designs will be possible? - In terms of casual inference: - RCT Very Strong - Cohort Study Moderate/Quite Strong - Case-Control Moderate - Cross-Sectional Weak - Ecological Very Weak