Design and Analysis of Epidemiological Studies (SBI222)

20 credits

Aim of this module

Descriptive epidemiology helps with hypothesis generation; analytical epidemiology allows the testing of hypotheses and making inference of the results to the population with the use of analytical methods. The overall aim of this module is to provide trainees with the knowledge and skills to design, analyse and interpret observational epidemiological studies used in the public health setting to inform understanding and effect change. The aim of this module is to enable trainees to develop and apply their skills in the design and analysis of epidemiological studies from identifying the public health problem; designing a study to address the problem, and leading the study through to the presentation of the results and their implications for population health, patient care and patient outcomes.

  1. Write a study protocol for an epidemiological study and consider ethical issues if required.
  2. Develop an appropriate quality-assured data collection instrument.
  3. Collate and manage the data and prepare the data set for analysis.
  4. Analyse data according to the data analysis plan; archive data and analysis code as required.
  5. Interpret study results, including a critique of study limitations.
  6. Write a report presenting the results of the analysis and their implications for population health, patient care and patient outcomes.
Number Work-based learning outcome Title Knowledge
1 1

Identify the public health problem and justify the study, considering any ethical issues that may be relevant in order to protect patients and the public.

View
2 1

Conduct a literature review.

View
3 1

Define a research question and study objectives.

View
4 1

Define study hypotheses and translate to statistical hypotheses.

View
5 1

Determine the study population and study design.

View
6 1

Describe sampling strategy and calculate sample size.

View
7 1

Develop an analysis plan, including dummy tables and document the justification for modifications.

View
8 1

Define the strategy for communication of results taking into account diversity in patients and populations.

View
9 1

Consider ethical issues that might be relevant and consult the National Research Ethics Service and apply and gain ethical approval if necessary.

View
10 2

Identify data needed to answer the study question and design an appropriate quality-assured data collection instrument.

View
11 3

Conduct data collection and entry.

View
12 3

Conduct data handling and data management.

View
13 3

Check the data for inconsistencies and missing data and prepare data set for analysis.

View
14 3

Generate summary variables as necessary.

View
15 4

Undertake a descriptive analysis of the data variables.

View
16 4

Conduct bivariate analysis.

View
17 4

Conduct univariate analysis.

View
18 4

Conduct stratified analysis to identify effect modification and confounding.

View
19 4

Conduct multivariable analysis if required to adjust for multiple confounders and effect modifiers.

View
20 4

Archive data and analysis code as required.

View
21 5

Interpret significant and non-significant findings.

View
22 5

Identify limitations and make a final conclusion about the study results.

View
23 6

Draft a report summarising the study findings and their impact on patients and the population.

View
24 6

Share report with appropriate colleagues for comment, collate comments and amend draft following feedback.

View
25 6

Share report for final sign-off and disseminate through appropriate communications channels.

View
26 6

Summarise findings for a lay audience.

View
This module has no work-based assessments.

Important information

The academic parts of this module will be detailed and communicated to you by your university. Please contact them if you have questions regarding this module and its assessments. The module titles in your MSc may not be exactly identical to the work-based modules shown in the e-portfolio. Your modules will be aligned, however, to ensure that your academic and work-based learning are complimentary.

Learning Outcomes

  1. Identify and state the key considerations in the planning and design of epidemiological studies.
  2. Describe the principles and critically evaluate the relative merits of different study designs.
  3. Summarise the key concepts and impact of sampling error, bias and confounding in epidemiological studies and apply strategies to address these.
  4. Select appropriate sampling methods and calculate sample size and power using appropriate software.
  5. Describe the key elements of questionnaire design.
  6. Design data collection instruments.
  7. Design data collection procedures that protect patients, the public and other stakeholders, in line with data protection requirements, e.g. Caldicott requirements.
  8. Perform data cleaning and management on a dummy data set.
  9. Perform data analysis on a dummy data set using statistical software, e.g. R.
  10. Investigate effect modification (interaction) and confounding using both stratified and statistical modelling methods.
  11. Judge the adequacy of evidence supporting causal links between exposure and disease.
  12. Describe the elements and applications of time series analysis.
  13. Describe the applications of infectious disease modelling.

Indicative Content

  • Identify and state the key considerations in the planning and design of epidemiological studies
    • Ethics
    • Cost
    • Human resource
    • Time
    • The underpinning evidence base to:
      • determind what study needs doing (what is the right uqestion)
      • inform how it migh tbest be conducted
    • The involvement of patient and the public
  • Describe the principles and critically evaluate the relative merits of different study designs
    • Cross-sectional studies
      • Prevalence and prevalence ratios
    • Prospective and retrospective cohort studies
      • Attack rates, risk ratios and rate ratios
      • Attributable fraction
    • Case-control studies
      • Odds and odds ratios
    • Case-case studies
    • Case-crossover studies
    • Case and control selection

 

  • Summarise the key concepts and impact of sampling error, bias and confounding in epidemiological studies and apply strategies to address these
    • Information bias
    • Selection bias
    • Prevention of bias in data collection
    • Correction of bias in data analysis
  • Select appropriate sampling methods and calculate sample size and power using appropriate software
    • Non-probability sampling: convenience sampling, judgement sampling, quota sampling and snowball sampling
    • Probability sampling: simple random, systemic, stratified, multi-stage, cluster
    • Selection of participants
      • Cohort
      • Cases selection
      • Control selection
      • Sampling techniques
    • Impact of response to the study
    • Sample size calculation
      • Open Epi, R
      • Methods for estimating the desired value of the measure of effect
      • Power, significance
  • Describe the key elements of questionnaire design
    • Information required
    • Define target respondents
    • Choose the method(s) of reaching target respondents
    • Question content: demographic information, establish rapport, information required for the study
    • Question wording: closed, open-ended and open response-option questions
    • Meaningful order and format
    • Length of the questionnaire
    • Layout
    • Piloting
  • Design data collection instruments
    • Web-based, e.g. SelectSurvey
    • EpiData Entry
    • Evolving technologies, e.g. social media, wearable technologies
  • Design data collection procedures which protect patients and the public, in line with data protection requirements e.g. Caldicott requirements
    • Questionnaire: paper or electronic, self-administered versus interviewer, mail v email v web
    • Confidentiality
    • Secure data storage
    • Data entry: double entry, data entry checks
  • Perform data cleaning and management on a dummy data set using statistical software, e.g. R
    • Software commands and syntax
    • Reading data
    • Data types
    • Generating and recoding variables
    • Sorting data
    • Saving the data
    • Automate analysis, e.g. by generating a ‘.R’ file
    • Missing data
      • Types of missing data
      • Missing-data mechanism
      • Imputation and sensitivity analyses
  • Perform data analysis on a dummy dataset using statistical software. e.g. R
    • Measurement error
      • Description
      • Monitoring measurement error
      • Measurement error in the outcome variable
      • Impact of measurement error
  • Compare the outcome of two or more groups using appropriate statistical techniques
    • Tests for comparing means
    • Tests for comparing proportions
    • Non-parametric methods
    • Stratified analysis
    • Regression models
    • Survival analysis
  • Describe the association between variables
    • Appropriate use of graphical techniques for displaying and describing univariate and multivariate data
    • Meaning of correlation and regression analysis
    • Meaning of scatter plots
    • Non-independent measurements
    • Clustered data: description, problems with analyses, methods of analysis including generalised estimating equations and multilevel models and examination of contextual effects
  • Investigate effect modification (interaction) and confounding using both stratified and statistical modelling methods
    • Causal diagrams
    • Stratified analysis
    • Regression methods for case-control studies: unconditional and conditional logistic regression
    • Regression methods for cohort studies and survival analysis: stratifying on time, poisson regression, cox regression
    • Non-linear relationships
  • Judge the adequacy of evidence supporting causal links between exposure and disease
    • Bradford-Hill criteria
  • Describe the elements and applications of time series analysis
    • Elements:
      • assessment of trend
      • periodicity
      • smoothing and differencing
      • outliers
    • Applications:
      • compact description of data
      • interpretation
      • forecasting
      • control
      • hypothesis testing
      • simulation
  • Describe the applications of infectious disease modelling
    • Describe a compartmental model
    • Describe course of outbreak (retrospective and real-time)
    • Predict impact of control strategies
    • Describe infectious disease dynamics (e.g. Ro)