Excess Mortality Flashcards
What is Excess Mortality?
Excess mortality is the number of deaths recorded which are greater than the numbers expected in a given year.
What is epidememlogical study?
Epidemiological studies measure the risk of illness or death in an exposed population compared to that risk in an identical, unexposed population (for example, a population the same age, sex, race and social status as the exposed population).
Describe Scotland’s life expectancy data throughout history to contempoary times compared to other European countries.
Scotland’s life expectancy from the mid 1800s to mid 2000s has steadily increased. They started off with an average life expectancy relative to other european countries but as time progressed, they now have one of the worst life expectancies of Europe.
What is deindustrialisation?
Deindustrialisation is the reduction of industrial activity or capacity in a region or economy.
What is the male life expectancy of Scotland and the male life expectancy of the rest of the UK?
The life expectancy in Scotland is 77.1 years old and the life expectancy in the rest of the UK is 79.2 years old.
What is the female life expectancy of Scotland compared to the female life expectancy in the rest of the UK?
The female life expectancy of Scotland is 81.1 years old and the female life expectancy of the rest of the UK is 82.9 years old.
Within Scotland, what council area has the worst life expectancy in males?
Glasgow city has the worst life expectancy in males.
What contributes to Scotland/Glasgow having the poorest health?
The reasons for Glasgow’s/Scotland’s poor health are:
- Cardiovascular disease/Stroke
- Type 2 Diabetes
- Obesity
- Alcohol/Drugs/Suicide
These health issues usually arise from the current socioeconomic deprivation and poverty.
What is the Determinants of Health model by Dahlgren and Whitehead sometimes referred to as?
The Determinants of Health model by Dahlgren and Whitehead is sometimes referred to as the ‘onion model’.
Determinants of Health Model by Dahlgren and Whitehead.
What was the onion model designed to show?
The onion model displays the levels of impact/influence on ones health and is used by people who develop public health policies and strategies.
Give an ordinal scale the biggest killers in today’s scottish men.
- Ischaemic Heart Disease
- External Causes
- Cerebovascular Disease
- Chronic Liver Disease
- Oesophageal Cancer
How do we know if there’s a causal relationship/an associative relationship between certain factors and the associated outcome?
You’d go to the avilable evidence to find out if it’s a known causative factor or if it’s an associated factor.
What do we need to keep in mind when investigating if two factors are associated?
When investigating if two factors are associated, we need to keep in mind if:
- X causes Y
- Y causes X
- X and Y are caused by Z
- Z would be known as a cofounder.
- X and Y are asssociated only by artefact
- X and Y are associated by chance.
What does the phrase “X and Y are only associated by artefact” mean?
The phrase “X and Y are only associated by artefact” means that the association between the two variables are a result of the investigative procedure and does not occur naturally.
What is a Risk Factor?
A risk factor is something that increases your likelihood of getting disease?
What is a modifiable risk factor vs non-modifiable risk factor?
Risk factors are either modifiable, meaning you can take measures to change them, or non-modifiable, which means they cannot be changed.
What are examples of modifiable risk factors?
Examples of modifiable risk factors are:
- Smoking
- Diet
- Physical activity
- Blood pressure
What are examples of non-modifiable risk factors?
Examples of non-modifiable risk factors are:
- Age
- Family history
- Ethnicity
What does research design do?
A research design orovides a framework for the collection and analaysis of data. It freflects the aims of the research including:
- Describing what is going on
- Expressing causal connections between variables of interest
- Understanding health risk/outcomes/behaviour over time
- Generalising beyond participants in the investigation.
What is the Research Design Framework?
WHat is the hypothesis you want to investogate?
- What is the aim/objective?
- Is it descriptive (what is going on?)
- Explanatory (why - causation?)
- Is the aim researchable - can apply and have appropriate research methods to assess the outcomes of interest.
What are the two research methods?
The two research methods are:
- Quantative (e.g. anthropometric measurements, blood tests, objective physical activity, fitness, self-report questionnaires)
- Qualitative (interviews, focus groups, observations)
What are anthropometric measurements?
Anthropometric measurements are a series of quantitative measurements of the muscle, bone, and adipose tissue used to assess the composition of the body.
What are the two main tyoes of research design?
The 2 main types of research design are:
- Causal (explanatory) research
- Experimental (RCT) design
- Quasi-experimental design
- Descriptive research
Whagt is a powered outcome, what’s the powered aim of FFIT programme?
A powered outcom is the main aim of the research determined by the number of the participants for reliability and the powered aim of FFIT it is weight loss at 12 months.
What is a non powered outcome?
A non powered outcome is one which does not dictate the sample size of the sample population.
How does a One group before-and-after design work?
A Longitudinal experiment works as so:
- Measure outcome in exposed group (A) before they have recieved intervention x (T1)
- Then apply intervention x to Group A
- Measure outcome in group A after they have recieved intervention x (T2)
- Estimate of the effect of intervention x is the change in outcome (T2 - T1)
What is the most common causal evaluation design?
Before-and-after research is the most common as it is easy and inexpensive.
Can the observed change from intervention x in a before and after experiment be explained soley by intervention x, and why?
No because there are external factors which may influence the T2 of the group and learning effects.
How is a Two group before and after experiment conducted?
Have two groups, one exposed to intevention (A), and control group (B).
Measure outcome in both groups at baseline (T1)
Apply inteverntion x to (A)
Measure outcome in both groups at follow up
Estimate effect of x by comparing change (T2 - T1) between groups.
What is selection bias?
Selection bias is the bias introduced by the selection of individuals, groups or data for analysis in such a way that proper randomization is not achieved, thereby ensuring that the sample obtained is not representative of the population intended to be analyzed.
How does selection bias effect two group before and after experiments?
Selection bias effects two group before and after experiments as it leads to one group being fundamentally different to the other groups, rendering results effectively unreliable.
How did selection bias effect the Lankarshire milk experiment and what was the problem with the experiment?
10k kids had to drink 3/4 pint of milk a day
Another 10k to get none
teachers tended to put weaker, less well nourished kids in interventuion group
study found that control group were taller and heavier than intervention group
milk had not led to smaller children but the two groups were fundamentally disimilar due to the initial selection bias.
How do we reduce selection bias?
To reduce selection bias one must use randomisation control trial (RCT) experimental design.
In this exoerimental design randomisation is used to allocate participanrts in group A or B at T1.
Then apply intevention x to group A only.
Measure outcome in both groups at follow up (T2)
Estimate the effect of x by comparing T2 -T1 between groups.
What is Cross sectional design research?
In research, a cross-sectional study is a type of observational study that analyzes data from a population, or a representative subset, at a specific point in time—that is, cross-sectional data.
Coined snapshot study as each participant is analysed at one point in time.
Allows researchers to look at many factors at once (age, income, gender as well as health risk factors and outcomes).
Designed to determine current state - prevelance
Give 3 sampling methods.
- Random Sampling
- Most representative - every member of the population has an equal chance of being picked.
- Stratified Sampling
- Separate groups e.g.ethnic minority vs majority, then sample systematically (e.g. every 20th postcode) to ensure you have adequate samples of each group.
- Convenience Sampling
- Take those people who are available however the representativeness is questionable.
What is the Longitudinal design?
Descriptive research design involves measuring participants at more than one point in time
Observing change in these participants gives a better basis for causal inferences than a cross sectional design.
What is rekiability?
Reliability refers to the consistency of an instrument/ measurements
- Determines if the results are repeatable
What are the three common forms of reliability?
The three common forms of reliability are:
- Inter-rater reliability - when 2 or more people are taking the measurement or applying the instrument of measurement.
- Test-retest reliability - when you repeat the measure on one or more occasions and expect the result to be similar.
- Internal consistency reliability - Applies to questionnaires. Looks at the extent to which the items on the scale are all telling you the same thing.
What is validity?
Validity refers to the accuracy of an instrument/measurement
The degree to which you’re measuring what you’re claiming to measure.
There are different types:
Construct validity: Does the test measure the concept that it’s intended to measure?
Internal validity: The extent to which the dependent variables can be attributed to changeds in independent vairable.
What studies have the highest internal validity?
Studies with control and randomisation like an RCT have the highest internal validity.
What’s an RCT?
An RCT is a randomised control trial
What are the different types of validity?
The different types are:
Construct validity: Does the test measure the concept that it’s intended to measure?
Internal validity: The extent to which the dependent variables can be attributed to changeds in independent vairable.
External validity: The extent to which the results can be generalised.
Where does tension arise between internal validity and external validity?
The problem is, the more srictly controlled the study is, the higher the internal validity will be. However the lower the external validity will be.
What research design did the FFIT (Football Fans in Training) trial use?
The FFIT was a RCT (randomised control trial) to design the effectiveness of a weight managent programme delivered through a football regime to help men between 35 and 65 lose weight and retain the weight loss after 12 months.
Describe the two FFIT groups (control vs intervention group).
747 men in study, half randomly allocated to intervention group for 12 weeks. While other half were the control and did not participate in the intervention at all.
What happened after the 12 week programme in FFIT and why?
Both groups had measurements taken and were followed up after 12 months to find out if they had retained the changes over a longer term.
What were some of the challenges faced by researchers in the FFIT programme?
Between group contamination. It’s important that the control groups don’t know anything about the intervention group. They had to keep the groups seperate.
Hard to do amongst football fans. E.g. man in intervention group may be a fan of man in comparison group as they support the same team.
If the control group member found out about intervention may decide to do it as well and the results for the control will be skewered.
What is a solution to the FFIT group-group contamination issue?
A solution would be to do a cluster randomised control trial - randomisation of football clubs, not the men. Randomising the collective, not the individuals.
What solution did FFIT use to combat the contamination issue?
The groups were measured at different times at 12 weeks but not at baseline or 12 months
What is blinding?
Blinding is where researchers and participants do not know what group the participants are in.
Why is blinding important?
Blinding is important because researchers and/or participants may influence outcomes of the experiment.
Why did FFIT struggle with blinding initially and how did they overcome it?
FFIT struggled with blinding because the participants know if they are in the intervention or not and may of the fieldworkers knew the participants, becoming friendly and participants would tell researchers intentionally or non intentionally about their intervetion.
FFIT managed to overcome this issue by hiring new field workers for primary outcome (weight at 12 months). As they have no recollection of any of the participants groups.
Why was retention a problem for FFIT?
Lack of retention in a study makes results less reliable as there is an attrition bias due to the powered nature of the primary aim.
- This was caused by the fact that only those with good outcomes want to be followed up.
Also if the loss:follow up ratio is higher in one group in an RCT the groups are no longer similar, leading to the same problem found with selection bias.
How did FFIT maximise retention?
FFIT offered home visits to maximise retention.
What were some of the challenges of home visits for FFIT?
Challenges faced included:
- Distance
- SATNAV locating issues
- Communicare link was down (system they were using)
- Very costly to get researchers to participants homes depending on where the participant was.