2. Contents
• Choice of study design.
• Qualitative methods
• Quantitative methods
• The research question, PICO (again!).
• Testing the Hypothesis.
• Elements of study design.
• Types of studies.
• Chance, bias and confounding.
3. Choice of study design
• Which type is suitable for my research?
• The study type is determined by the research question to be
answered.
• Are you going to observe or experiment?
• Also depends on:
• Researcher Beliefs and Values.
• Researcher Skills.
• Time and Funds.
4. Q1. What was the aim of the study?
1. To simply describe a population.
(PO questions) descriptive
1. To quantify the relationship between factors. (PICO
questions) analytic
5. Quantitative vs qualitative
Qualitative
• Understanding.
• Interview/observation.
• Textual (words).
• Theory generating.
• Quality of informant more
important than sample
size.
• Subjective.
• Embedded knowledge.
Quantitative
• Prediction.
• Survey/questionnaires.
• Numerical.
• Theory testing (experimental).
• Sample size core issue in
reliability of data.
• Objective.
• Public.
6. Quantitative vs qualitative (II)
Quantitative Methods
• Observational
• Experimental
• Mixed
• Sampling: Random
(simple, stratified,
cluster, etc) or
purposive
Qualitative Methods
• Focus Groups
• Interviews
• Surveys
• Self-reports
• Observations
• Document analysis
• Sampling: Purposive
7. Quantitative methods
• Observational: studies that do not involve any
intervention or experiment.
• Experimental: studies that entail manipulation of
the study factor (exposure)
8. Qualitative methods
• Participant observation (field notes).
• Interviews /Focus group discussions with key
informants.
• Video / Text and Image analysis.
• Surveys.
• User testing.
9. Hypothesis testing
• The research hypothesis is a predicative statement that
relates an independent variable to dependent variable.
• When the purpose of research is to test a research
hypothesis, it is termed as hypothesis testing research.
16. Important terms…
• The main purpose of an epidemiologic study is to
investigate, whether there is an effect of exposure
on outcome.
• Exposure: environmental or clinical factor, that could
potentially influence incidence in diseases or
mortality.
• smoking, alcohol drinking, low birth weight,
antidepressant use, vaccination against diphtheria.
• Outcome: disease (cancer, cardiovascular diseases)
or death.
17. Cohort Study
• Identify a large group of individuals, often a well
defined population.
• Identify exposed subjects and not exposed
subjects.
• Follow over a specific time.
• Record the fraction in each group who develop
the health outcome.
• Compare these fractions using RR, AR or OR.
19. Toxicity in Cancer Patients
Explanatory variable = generic drug use (generic: 1 = yes, 2 = no)
Response variable = cerebellar toxicity (tox: 1 = yes, 2 = no)
Data are
observational,
individual-level,
longitudinal, with all
individuals followed
over time. Thus, data
are cohort.
Comment: This is a retrospective cohort based on data
abstracted data from medical records.
20. Prospective Cohort Study
Non smoker
Regular smoker
No Lung
cancer
Lung
cancer
Lung
cancer
No Lung
cancer
Present Future
Cohort
21.
22. Pros and cons of cohort study
• Advantages:
• ethically safe.
• can establish timing and directionality of events.
• suitable for incidence estimation.
• Disadvantages:
• controls may be difficult to identify.
• exposure may be linked to a hidden confounder.
• blinding is difficult.
• for rare disease, large sample sizes necessary.
• expensive; time-consuming.
23. case-control studies
• Participants selected on the basis of whether or
not they are DISEASED (remember in a cohort
study participants are selected based on
exposure status).
• Those who are diseased are called CASES.
• Those who are not diseased are called
CONTROLS.
• A relationship of a risk factor to the disease is
examined by comparing the diseased and non-
diseased with regard to how frequently the risk
factor is present.
24. The logic of Case-Control Studies
• Cases differ from controls only in having the disease.
• If exposure does not predispose to having the disease,
then exposure should be equally distributed between the
cases and controls.
• The extent of greater previous exposure among the
cases reflects the increased risk.
26. Esophageal Cancer and Alcohol Consumption
Data are
observational,
individual-level, with
study of all
population cases but
only a sample of non-
cases. Thus, data are
case-control.
Explanatory var = alcohol consumption (alc2: 1 = high, 2 = low)
Response var = esophageal cancer (case: 1 = case, 2 = control)
29. Pros and cons of case-control studies
• Advantages:
• quick and cheap as fewer people needed than cross-
sectional studies.
• Disadvantages:
• representativeness of cases and controls is a major
concern.
• can not distinguish between causes & associated
factors
• confounders.
• potential bias: recall, selection.
30. Cross sectional studies
• Examines the relationship between
1) diseases/other health related characteristics and
2) other variables of interest
as they exist in a defined population at one time.
• Exposure and outcomes both measured at the same
time (Snapshot in time).
• Quantifies prevalence, risk, or diagnostic test accuracy.
31. HIV in a Women’s Prison
Explanatory var = IV drug use (1 = users, 2 = non-user)
Response var = HIV serology (1 = positive, 2 = negative)
Data are observational on the
individual-level. But onset data
cannot be unraveled. Thus,
data are cross-sectional
32. Cross-sectional Study
Sample of Population
Regular
smoker
Non
smoker
Prevalence of
lung cancer
Prevalence of
Lung cancer
Time Frame = Present
Defined Population
33.
34.
35. Pros and cons of cross sectional study
• Advantages:
• cheap and simple.
• ethically safe.
• Disadvantages:
• establishes association at most, not causality.
• recall bias, researcher’s bias.
• group sizes may be unequal.
• confounders.
36. Experimental studies
• Identify subjects,
• Place in common context,
• Intervene, then
• Observe/evaluate effects of intervention.
• An experimental comparison study where
participants are allocated to treatment
(intervention) or control (placebo )groups.
• Best for studying the effect of an intervention.
37. Experimental studies….Why?
• Provide stronger evidence of the effect (outcome)
compared to observational designs, with maximum
confidence and assurance.
• Yield more valid results, as variation is minimized and bias
controlled.
• Determine whether experimental treatments are safe and
effective under “controlled environments” (as opposed to
“natural settings” in observational designs).
38. Weight Gain on Different Diets
Explanatory variable = diet group (1=standard, 2=junk, 3=health)
Response variable = weight gain (grams)
Data are experimental
because the investigator
assigned the explanatory
variable
39. Randomised controlled trials
• A clinical trial is a comparative, prospective experiment
conducted in human subjects.
• “The RCT trial is the “gold standard of all research designs
because it provides the strongest evidence for concluding
causation.”
40. time
Study begins here (baseline point)
Study
population
Intervention
Control
outcome
no outcome
outcome
no outcome
future
RANDOMIZATION
41. Types of trials
Blinded Not blinded
Randomised Not randomised
Controlled Not controlled
Trial
42.
43. Pros and cons of the RCT
• Advantages:
• no bias (using blinding).
• controlling for possible confounders.
• comparable groups (using randomization).
• Disadvantages:
• expensive: time and money.
• volunteer bias.
• ethically problematic at times.
• subjects may not be representative
44. Blinding
• Factors that can affect the evaluation of outcome should
not be permitted to influence the evaluation process
• Double-blind design
• Neither patient nor outcome evaluator knows Rx to which patient
was assigned
• Single-blind
• Patient or evaluator is blinded as to Rx, but not both
• Triple-blind
• Patient, Physician, and Data analyst are blinded as to Rx identity
45. Meta-analysis
• Combining the results from many studies dealing with the
same topic.
• Statistically combines results of existing research to
estimate overall size of relation between variables.
• Helps in
• Developing theory,
• Identifying research needs,
• Establishing validity.
• Can replace large-scale research studies.
46.
47. Best Type of Study for Your Question
Type of Question Appropriate Study Design
Burden of illness
Prevalence
Incidence
Field Surveys
Cross Sectional Survey
Cohort studies
Cross Sectional Survey Case Control Study,
Cohort study, RCT
Treatment Efficacy Randomized Controlled
study
Diagnostic Test Evaluation Randomized Controlled
study
48. Chance: definition of p-value
The probability that an effect at least as extreme as
that observed in a particular study could have
occurred by chance alone, given that there is truly no
relationship between exposure and disease.
49. Chance: Definition of P-value
• If p-value > 0.05 - by convention:
• We generally consider that chance cannot be
excluded as a likely explanation.
• Findings are stated to be not statistically
significant at that level.
50. Bias
• Any systematic error in the design, conduct, or analysis of
a study which acts to make the observed results non-
representative of the true effect of therapy/risk.
• Examples:
• healthier patients given treatment A, sicker patients given treatment
B
• treatment A is “new and exciting” so both the physician and the
patient expect better results on A
51. Confounding: another type of bias
• A variable that:
• Is causally related to (or at least associated with) the
disease under study, and
• Is associated with the exposure under study in the study
population but is not a consequence of this exposure.
52. Next session…
• How to do the maths?
• Sample size.
• Statistical test.
• P-value and confidence intervals
• Odds ratio, Relative risk and Hazard ratio.
• Sensitivity and specificity.