Module 1 Trial Exam

Study Design  à review different study designs à chart

Panel

-          one group studied

-          exposed to intervention

-          Group measured before / after treatment

 

Quasi Panel

 

-          similar to panel design

-          but two separate groups used

-          one before treatment / other after treatment

-          used to avoid difficulty of keeping track of participants over time

 

Steps in Survey Research  à are there other lists?

 

-          set specific measurable objectives

-          select research design

-          choose population and sample

-          develop reliable / valid survey instruments

-          administer survey

-          analyse survey data

-          interpret / report

 

Sampling Methods à review all

 

Systematic sampling

 

-          probability sampling method

-          every nth

-          requires good sampling frame

-          not suitable for f2f where respondents over large area

-          systematic bias if periodicity in sampling frame exists

-          simpler than other probability methods

 

Availability sampling

 

-          non probability

-          individuals who are easily available and willing to participate

-          may not be rep

-          large samples

-          pilot test / exploratory

 

Sample Size / Statistical Accuracy

 

-          what happens to statistical accuracy if small sample size used

o   statistical accuracy ( + / - ) will be larger, meaning less accurate.

 

 

Potential Problems With Questions

 

-          term not defined

-          ambiguous question

 

 

Types of Questions  à review list

 

-          behavioural

o   measure what people do / what they have done

-          leading

o   question structure / wording makes respondents provide answer they would not have given had question been asked in more neutral way.

§  Do you agree that

§  Does this seem like a good idea to you.

 

Exhaustiveness Principle à list other principles

 

-          response alternatives provide sufficient range of responses to cover all responses

 

 

Advantages of face to face interviews over paper surveys  à list other types

 

-          questionnaires can be reasonably complex

-          accurate recording of responses

-          misunderstandings can be clarified

-          body language etc can be used to encourage participation

-          probe

-          be certain right person has answered.

 

 

Disadvantages of face to face over paper surveys

 

-          expense

-          possible interviewer bias

-          tiredness / impatience

-          social desirability considerations

-          interviewer contamination

 

 

Coding à can be more than 1 code per answer

 

List wise deletion

-          any case that has missing data on any of the set of variables deleted

Pair Wise

-          use only cases with complete responses for each calculation

 

 

Validity à look at associated definitions

-          how well question / scale measures what it is intended to measure

 

 

alpha =

 

0.95        - very high internal consistency

 

0.55        - reliability is unacceptable

 

0.86        - high internal consistency

 

0.63        - low internal consistency

 

 

 

 

Topic 10: Mortality : Revision Exercises

 

Arguments that time trends are real, not artifactual

-          trend steadily downwards

-          decline is dramatic

-          decline occurred for both groups

 

Improved Birth Registration

 

-          increase denominator

-          decrease infant mortality rate

 

Improved Death Registration

 

-          increase numerator

-          increase infant mortality rate

 

Reasons For Decline in Infant Mortality – early Period

 

-          elimination / reduction in infectious diseases

-          improved sanitation / sterile techniques

-          vaccinations / antibiotics

 

Reasons for Decline in Infant Mortality – later Period

 

-          hospital based technology

 

Hypothesis That Might Explain Differences in Infant Mortality B/ W

 

-          genetics

-          less access to medical care

-          lower socio economic

-          differences in recoding births / deaths

 

Crude Rate

 

-           percentage *  rate

 

 

Crude Mortality rate

 

Sex Specific Mortality Rate

 

Cause Specific Mortality rate  à  death from disease / total pop

 

Case fatality ratio à death from disease / total suffering from disease

 

Proportionate mortality rate  death from disease / total deaths

 

Standardised Mortality rate

-          observed total deaths in population, divided by expected deaths in pop, multiplied by 100.

 

standardized rate à includes indirect & direct.

 

 

Two Way Contingency Tables – Chapter 2 : An Introduction to Categorical Data Analysis – Alan Agresti

 

-          Association between two categorical  variables

-          In many contingency tables, one variable is a response variable and the other an explanatory variable.

o   Then informative to construct separate probaility distribution for Y at each level of x

§  Conditional probabilities for Y, given the level of X

·         Called : conditional distribution.

-          Independence

o   Statistically independent if conditional distributions of Y are identical at each level of X

§  Independent : probability of any particular column is same for each row.

§  Statistical independence: property that all joint probabilities equal the product of their marginal probabilities

·         Joint

·         Marginal

-          When rows of contengency table refer to different groups

o   Sample sizes for groups often fixed by sampling design

o   When marginal totals fixed rather than random

§  Joint distribution for x and y is no longer meaningful

·         But conditional distributions for Y at each level of X are

-          Difference of Proportions

o   Compares the success probabilities in the two rows

o   Difference falls between -1 and 1  / equals zero when two probabilities equal

o   See formula for SE   à calculate confidence interval.

-          Aspirin and heart attack example

o   Two rows à independent bimonial sample

-          Relative Risk

o   Difference between two proportions of certain fixed size may have greater importance when both proportions are near 0 or 1.

o   RR = ratio of success probabilities.

o   Any non-negative number

o   Complicated CI formula

-          Odds Ratio

o   Odds = prob of success / prob of failure

o   Odds non-negative

-          Odds Ratio for Aspirin Study

-          Relationship betwen odds ratio and relative risk

o   When proportion of successes is close to zero à fraction in last term is approx zero

o   OR and RR then take similiar values.

o   For some data sets calculation of RR is not possible

§  Case control study à where marginal distribution is fixed by sampling design

·         Two controls for each case

·         Might wish to compare ever smokers with non smokers in terms of proportions who suffered a disease

o   These proportions refer to conditional distribution of disease, given smoking status

§  Cannot estimate such proportions for this data set

§  Study matched each case with two controls

·         We can compute proportions in reverse direction

o   Conditional distribution of smoking status, given disease status

o   Use odds ratio

§  Odds ratio takes same value when it is defined using the condtional distribution of x given y as it does using the distribution of y given x  à treats variables symetrically

§  Or – conditional distribution in either direction

 

Topic 6: Case Control Studies: Revision Questions

 

Why are case control studies called "retrospective studies"

-          study starts with end point / disease

-          goes backward to identify risk factors

-          no follow up of cases

 

Meaning of Odds Ratio

 

-          approximation of the relative risk of disease in those exposed to risk factor compared to those who were not exposed to risk factor

-          odds of exposed person having disease / odds of unexposed person having disease

 

Selection of Cases & Control

               

-          controls à sourced from hospital catchment area / patients attending same hospital for treatment of other diseases

-          match for age / sex

 

 

Data Collection

               

-          pre-designed questionnaire à preferred as medical records are often incomplete

 

 

Differential Medical Surveillance

 

-          biased because sub clinical cases would have gone undetected in community

 

 

Differential Classification

 

-          questions about exposure from clinician depend on the disease being treated

 

 

Relative Risk and Odds Ratio

 

-          RR – measures likelihood of developing disease in exposed group relative to those who are unexposed.

-          OR – compares odds of exposure to risk factor among cases with that among controls.

 

Absolute & Relative Risk

 

-          absolute risk – risk of having disease

-          RR – risk of disease in exposed group relative to risk of disease in unexposed group.

o   RR = 3.0  à risk of outcome in exposed group is 3 times that of unexposed group.

 

 

Incidence rate & Case Control Study

 

-          case control study does not allow calculation of incidence

o   subjects not carried fwd in time

o   proportion of cases controlled by study design

 

 

Exposure Measurement

 

-          breast feeding history obtained through interview

o   recall bias

 

 

 

OR

 

-          OR of 0.72  à   women with a history of breastfeeding have a 28% lower risk of developing cancer than women without a history of breastfeeding.

 

-          breast feeding has a protective effect

 

 

 

Case Control

 

-          Before it is accepted that breast feeding provides protection ….   results of study would need to be reproduced by other larger studies in a range of population groups

 

Revision Exercises: Module 2 : Topic 5 : Randomised Controls and Cohort Studies

RCT

-          therapeutic à trial for laser eye surgery

-          intervention à pap smears to detect early signs of cervical cancer

-          preventative à oestrogen treatment to prevent osteoporosis

 

-          double blinding à observer

-          what is triple blinding

 

Aetiological factor	Risk factor	Disaese
low dietary calcium	no dairy products in diet	osteoporosis
salmonella	unhygienic food preparation	food poisoning
carcinogens in smoke	cigarette smoke	lung cancer

 

 

No statistically significant relationship

-          no relationship exists

-          there is a relationship, but trial unable to detect (eg, low power)

 

Strategy to assess subject compliance

-          testing of blood etc

 

Keeping subjects enrolled

-          incentive or inducement

-          altruism

-          promise if assigned to control group, that they will be offered treatment after trial

 

Baseline characteristics of study groups differ

-          realloc subjects randomly to groups after stratifying on key variables

-          randomly discard subjects with certain characteristics

-          include variables with imbalance as covariates

 

Withdrawal of subjects in trial

-          decrease power

-          increase likelihood of type 2 error

 

Randomization

-          only then can outcome between groups  be ascribed to treatment

-          if randomization omitted à differences in outcome may be due to differing characteristics of groups

-          unbiased distribution of risk factors

 

Incidence / Prevalence

-          incidence à rate of appearance of new cases

-          prevalence à proportion of cases in population

 

Duration = Prevalence / Incidence

 

Chi Square degrees of freedom

-          (c – 1 ) ( r – 1 )

 

 

 

 

Topic 6: Case Control Studies: Revision Questions

 

Why are case control studies called "retrospective studies"

-          study starts with end point / disease

-          goes backward to identify risk factors

-          no follow up of cases

 

Meaning of Odds Ratio

 

-          approximation of the relative risk of disease in those exposed to risk factor compared to those who were not exposed to risk factor

-          odds of exposed person having disease / odds of unexposed person having disease

 

Selection of Cases & Control

               

-          controls à sourced from hospital catchment area / patients attending same hospital for treatment of other diseases

-          match for age / sex

 

 

Data Collection

               

-          pre-designed questionnaire à preferred as medical records are often incomplete

 

 

Differential Medical Surveillance

 

-          biased because sub clinical cases would have gone undetected in community

 

 

Differential Classification

 

-          questions about exposure from clinician depend on the disease being treated

 

 

Relative Risk and Odds Ratio

 

-          RR – measures likelihood of developing disease in exposed group relative to those who are unexposed.

-          OR – compares odds of exposure to risk factor among cases with that among controls.

 

Absolute & Relative Risk

 

-          absolute risk – risk of having disease

-          RR – risk of disease in exposed group relative to risk of disease in unexposed group.

o   RR = 3.0  à risk of outcome in exposed group is 3 times that of unexposed group.

 

 

Incidence rate & Case Control Study

 

-          case control study does not allow calculation of incidence

o   subjects not carried fwd in time

o   proportion of cases controlled by study design

 

 

Exposure Measurement

 

-          breast feeding history obtained through interview

o   recall bias

 

 

 

OR

 

-          OR of 0.72  à   women with a history of breastfeeding have a 28% lower risk of developing cancer than women without a history of breastfeeding.

 

-          breast feeding has a protective effect

 

 

 

Case Control

 

-          Before it is accepted that breast feeding provides protection ….   results of study would need to be reproduced by other larger studies in a range of population groups