Study Design - Lecture notes 1 PDF

Preview

Summary

Lecture notes...

Description

Principles of Epidemiology According to the World Health Organisation (WHO), epidemiology is “the study of the distribution and determinants of health-related states or events, and the application of this study to the control of diseases and other health problems” Study base- is a defined population whose disease experience during some period of time is the source of the study data Person time- Person-time is an estimate of the actual time-at-risk – in years, months, or days – that all participants contributed to a study. A subject is eligible to contribute person-time to the study only so long as that person does not yet have the health outcome under study and, therefore, is still at risk of developing the health outcome of interest. Time will differ by people depending on the endpoint Person-time= (Number of patients in study group) * (Time) Exposure A risk factor being observed or measured that is hypothesised to influence an event or manifestation Outcome Disease, disease progression, death, comorbidity (one or more disease happening at the same time), questionnaire data are all examples of outcome Prevalence Proportion of a population found to have a disease at a specific point in time Incidence Probability of occurrence of a given medical condition in a population within a specified period of time Risk The probability of the disease developing in an individual in a specific time period

L2 Study Design

1. Research question Firstly, you need to think of what kind of study you are carrying out, is it quantitative or qualitative? Within quantitative data deciding whether the study is observational or experimental.

Observational studies

Experimental studies

Researcher does not control the exposure

Researcher controls the exposure

Don’t want to determine if something is casual

Want to determine if something is casual

Treatment and exposure occur in a 'non-controlled' Treatment occurs in a controlled environment environment Individuals can be observed prospectively, retrospectively, or currently

Planned research designs

Observational because there is no individual intervention

Are stronger in determining the etiology of disease than descriptive studies

Observational study

Descriptive

Analytical

Used to formulate certain hypothesis e.g. case studies, cross-sectional studies, ecological studies

Used to test hypothesis, e.g. case-control, cohorts

Ex: What is the prevalence and trends in obesity?

Ex: How much exercise is necessary to reduce the risk of specific diseases

Cross-sectional An observational design that surveys exposure and disease status at a single point in time. Often used to study conditions that are relatively frequent with long duration of expression. It measures the prevalence not the incidence of disease. Not suitable for studying rare or highly fatal diseases with short duration of expression. The process •

Select a sample (representing the population of interest)

•

Measure exposure and outcome variables at the same time

•

Determine prevalence

Measure of association: odds ratios

Strengths •

fast and inexpensive

•

immediate answers – no follow-up time

•

no loss to follow-up (but can have non-responders)

Disadvantage: weakest observational design as it only measures prevalence, over time changes in exposure and effect may be difficult to determine, usually don’t know when the disease occurred, not good for rare events or outcomes. Potential bias- measurement bias and survivor bias Example: Community surveys

Case-control studies The process   

Select a sample of ‘cases’ (i.e. people who have the condition/disease) Select a sample of ‘controls’ (i.e. people without the disease but who have the same chance of having the disease) Measure (past) exposure to risk factors of interest (their past exposure to suspected causal factors is compared with that of controls)

Limitations: not appropriate when you don’t know the specific exposure for the disease outcome Selecting study subjects Results may be biased if exposures are different due to the selection process CASES – all those who develop a disease (or random sample)   

Cases from a hospital Cases on a population registry (i.e. cancer registry) Best to be new cases

CONTROLS must come from a source population with similar chance of being exposed to the risk factor of interest  

Hospital-based controls with different disease Population controls (good if cases are from a registry)

How many controls? Since case-control studies have a high statistical power, sample sizes are lower. 1:1 ratio is okay, 1:2 is more common, increasing it further makes no difference (not much power gained) Matching Controls can be matched with cases to ensure they are comparable with respect to other influencing factors (confounders)    

age and sex matching is common - strongly linked to many diseases and exposures sometimes geographical area individual matching frequency matching (overall proportions are the same)

Matching is not necessary but can be useful to control for strong confounders Overmatching can ‘hide’ true effects, and can limit the analysis that can be done Measurement Bias

Information about exposures collected after the outcome is known can introduce bias:   

different level of recall/reporting of risk factors between cases and controls more details about risk factors in cases (because of the disease) researchers look harder for evidence of exposure in cases

Possible solutions    

use data collected before outcome was known researcher blinded to outcome (not always possible) subjects asked about multiple possible risk factors pick controls have a disease that is linked to similar risk factors

Strengths and Weaknesses Strengths  

Useful for rare outcomes (e.g. specific rare cancers) Efficient/less costly than cohort study: smaller sample size; no follow-up required

Weaknesses    

Biases if cases and controls come from different populations Biases due to measuring exposure after the outcome Confounding due to other influential factors (not measured) Can only study one outcome

Cohort studies

Comparisons between individuals with a known risk factor or exposure with those without The Process    

Start with the POPULATION of interest Identify or assemble a cohort Measure risk factor(s) and potential confounders Measure the outcome over the follow-up period

Can be Prospective: Start with assembling a cohort, measure risk factors then follow over time to measure outcomes or Retrospective (historical): Identify a suitable cohort (from the past), collect risk factor data measured in the past, collect subsequent outcome data

Ecological studies Studies that investigate risk factors of health outcomes in which the unit of analysis is at the group level rather than the individual. Group measures (exposure and or outcome) can include: •

summary measures of a group (mean, average rate)

•

environmental factors (air pollution, hours of sun-light, fast-food shops)

i.e. something that is not measured at the individual level Examples: •

Time trends, geographic comparisons

Advantages:   

Easy to do No individual data necessary Good to generate ideas about potential associations

Disadvantages:  

No information on the individual level Not able to account for other factors that might explain the association  ecological fallacy: false claims

Potential errors and biases in observational studies Selection bias: study population differs from broader population in terms of the relationship between exposure and outcome Measurement (information) bias: Measurement of exposure or outcome differs between groups Confounding: Other factors associated with both exposure and outcome can distort the main effect if not considered Random error: Due to natural variation in the population / less precise measurements

Summary Research question determines the study design Source population & study population should reflect the target population Valid and reliable measures of exposure, outcome and confounders Hard to establish causation in observational studies due to bias/confounding/temporality...