Terms in this set (43)
Information about a defined population collected in a consistent manner for administrative reasons.
Such data may be used to describe the needs of and services provided to different population groups.
Almost all countries produce vital statistics, the most important of which are life expectancy and infant mortality.
Routine data need to be reliable, valid and complete.
The number of children per 1000 live births who die in their first year of life.
Strengths of vital statistics
- Cheap and readily available
- Almost complete data
- Can be used by ecological studies to develop hypotheses
- Recorded at regular intervals, so can be used for monitoring trends.
Weaknesses of vital statistics
- Not 100% complete
- Potential for bias - postmortem inflation of social status; underreporting of stigmatised disease.
- Become out of date - census data only every 10 years.
Ways to improve the quality of routine data
1. Computerised data collation and analysis.
2. Feedback - better feedback to providers helps maintain their interest and leads to improvement in data quality.
3. Presentation - in ways meaningful to different stakeholders.
4. Training - improved training of clinical coders and those responsible for data entry.
crude mortality rate = deaths/1000 pop per year.
age-specific mortality rate = deaths aged x/1000 population aged x per year
child mortality < 5 years
infant mortality < 1 year
neonatal < 28 days
post-neonatal 4-52 weeks
perinatal moratlity rate - stillbirths and deaths aged < 7 days
The population used in the denominatory must be defined using the same period of time as that used to accumulate the numerator using the concept of person-years at risk. The population estimate used is typically the mid-year (July 1) population count estimate. Rates may also be expressed per 100,000.
Crude birth rate = no live births / 1,000 population per year
- poor fertility indicator as denominator includes men, children and post-menopausal women.
General fertility rate = no live births / 1,000 women aged 15 to 44 per year.
age-specific fertility rate = no live births to women aged X / 1,000 women aged X per year.
- more precise because it takes into account difference in fertility at different ages.
Total period fertility rate = sum of the age-specific fertility rates.
This indicates the average number of children that would be born to a woman during her lifetime, assuming that she experienced the current age-specific fertility rates and that she survived to the end of her reproductive life.
Prevalence : sources and uses
Use: measures the burden of disease in a population, therefore enables policymakers and planners to judge what services are required for a particular population.
Although sometimes individuals say prevalence rate, prevalence is a proportion not a rate.
Usually used when calculating a relative measure of exposure effect.
A variation on prevalence (or point prevalence). The number of cases at any (short) time period / the total number in the population during the time period.
Incidence vs prevalence
The number of cases in a population at a particular point in time depends not only on the frequency with which new cases occur and are diagnosed, but on the disease duration.
For this reason, the prevalence of a disease will vary between populations due to differing durations of disease, which makes this less than ideal in establishing the determinants of disease.
Incidence looks at the number of new cases in a population at risk during a specified time interval. There are three measures of incidence:
- Cumulative incidence (risk)
- Incidence odds
- Incidence rate
Time at risk and cohort studies
Time at risk is particularly important in the analysis of cohort studies. For example:
- Some may joint the study after data collection has commenced
- Others may develop the disease of interest and thereby stop being at risk.
- Still others may leave the at-risk population before the event of interest has occurred (lost to follow-up or death by another cause): this phenomenon is known as right censoring.
- Some may join the study occasionally after the event of interest has already occurred - this is left censoring and usually these individuals are excluded before entering.
Person-time at risk
Sum of all the individual times at risk
Number of new cases /
total person-time at risk
The incidence rate of a disease is defined as the number of new cases of a disease that occurring during a specified period of time in a population at risk of developing the disease.
Because the incidence rate is a measure of events (transition from non-diseased to diseased), it is a measure of risk.
In contrast to prevalence, incidence is not affected by survival.
For example, in a food-borne disease outbreak, an attack rate is defined as the number of people exposed to a suspect food who became ill, divided by the number of people who were exposed to that food.
In this case time is specified implicitly rather than explicitly.
Otherwise known as RISK or absolute risk.
Proportion of a population who become disease in a defined time period. The cumulative incidence is a measure of the risk/probability that an individual will become diseased during a defined time period. (e.g. the attack rate in an epidemic).
Relationship between prevalence and incidence rate
A population in which the numbers of people with and without the disease remain stable is known as a steady-state population. In such (theoretical) circumstances, the point prevalence of disease is approximately equal to the product of the incidence rate and the mean duration of disease (i.e. length of time from diagnosis to recovery or death), providing that the prevalence is less than about 10%.
Although true steady-state conditions are never met in practice, the above relationship provides a useful `rule of thumb' for making rough estimates of prevalence in real populations when no dramatic changes in the incidence or duration of disease have occurred.
Relative measures of exposure effect, collectively known as relative risk
Absolute measures of effect
Attributable risk explanation
If it is possible to assume that the relationship between the exposure and disease is causal, then these absolute measures of effect can be re-interpreted as another measure, the attributable risk.
The attributable risk is especially useful in evaluating the impact of eliminating a risk factor.
Its value indicates the number of cases of the disease among the exposed group that could be prevented if the exposure were completely eliminated. Or, alternatively, what is the risk of disease attributable to the exposure in the exposed group, after taking into account the risk of the disease which would have occurred anyway.
Attributable risk percent
Attributable risk percent interpretation.
With an AR of 50%, for example, 50% of the disease in the exposed group would be removed if they had not been exposed.
Numbers needed to treat
Population attributable risk defintion
The excess rate of disease in the whole study population (of exposed and non-exposed people) that is attributable to the exposure.
Population attributable risk formula.
Population attributable risk example.
Mortality in whole population = 55 per 100,000
Mortality in non-smokers = 16 per 100,000
PAR = rate in population - rate in non-smokers
PAR = 55-16 = 39 deaths per 100,000
Population attributable fraction
This statistic estimates the proportion of disease in the study population that is attributable to the exposure (and thus the proportion of diseases that could be eliminated if the exposure were eliminated).
Also known as the preventable fraction or PAR percent when expressed as a percentage.
It is a key statistic for prioritising population interventions.
Population attributable fraction calculation
PAF = PAR / risk in whole population = (Rt-R0)/Rt
PAF = [p(RR-1)] / [ p(RR-1) +1] or (RR-1)/ (RR +1/p - 1) where p is the proportion of the total population that is exposed.
Establishing causation: other possible reasons to account for an association between an exposure and an outcome
1. Chance - random error.
4. Actual causation
5. Reverse causation
When the association between an exposure and an outcome in truth reflects the outcome altering the exposure.
E.g. the association between illicit drug use and psychological and social harm may, in part, occur because people suffering from psychological and social distress are more likely to take illicit drugs.
Bradford Hill criteria for determining causality
1. Strength of association
2. Biological credibility / plausibility
3. Consistency of findings
4. Temporal sequence
A systematic error that leads to differences between the comparison groups in how they are chosen, treated, measured or interpreted.
This error leads to an incorrect estimate of the association between the exposure and the risk of disease.
Unlike confounding and the role of chance, the magnitude of bias cannot be quantified.
Where the associations seen at a population level are wrongly assumed to hold at the individual level.
General outline / steps of a RCT
1. Specify hypothesis.
2. Define outcomes.
3. Define reference population.
4. Select study population.
5. Select suitable participants.
6. Obtain informed consent.
7. Collect baseline data.
8. Randomise to trial arms.
10. Assess outcomes continuously/intermittently
11. Analyse results
12. Interpret findings
13. Explore alternative explanations.
14. Policy implications
15. Feedback to participants.
16. Communicate key results.
Type I error
Reject the null hypothesis when there is no difference.
Known as : alpha
Type II error
Fail to reject the null hypothesis when there is a difference.
Known as : beta
Results from Hypothesis Testing
Calculating Power & Sample Sizes
1. Control the precision.
2. Control the power.
The probability of rejecting the null hypothesis.
1 - beta.
Advantages and disadvantages of primary prevention strategies
5 reasons why governments and clinical practitioners tend to favour prevention strategies targeting high-risk individuals.
- it appears to be more cost-effective, in that interventions are directed at those whose health will improve most dramatically if the intervention achieves its objective;
- people who recognise that they are at increased risk may be more likely and more motivated to act to change their behaviour;
- the focus is on individuals rather than society - this message fits in with the ethos and organisation of medical care and is easier for doctors and health care providers to promote;
- the risk associated with the adverse exposure (e.g. alcohol and traffic 'accidents') is generally apparent;
- society (especially the politicians and media) tends to sanction concern which focuses on needy individuals rather than on changing societal structures to achieve health gains.