Health Pop Flashcards
What is a census?
simultaneous recording of demographic data to ALL persons in a defined area
What is a census used for?
- allocation of resources
- projection of populations (estimate for future)
- trends e.g. ethnicity or age
What is crude birth rate (CBR?)
the number of LIVE births per 1000 population
What is general fertility rate (GFR)?
number of LIVE births per 1000 FERTILE women between 15-44 (accurate but not always possible)
What is total period fertility rate (TPFR)?
average number of children born to a hypothetical women in her lifetime - sum of age specific boundaries
What is crude death rate (CDR)?
number of deaths per 1000 population
What is age specific mortality rate?
number of deaths per 1000 in a specific age group
What is incidence rate?
number of new cases of the disease per 1000 people per YEAR
What is prevalence?
amount of people who currently have the disease in a set population (no time frame)
no follow up, NOT rate
What is incidence rate ratio (IRR)?
incidence rates of 2 SEPARATE populations with VARYING exposure compared to see if exposure CAUSES certain diseases
How do you work out incidence rates and prevalence?
e^2√(1/d)
d is events observed in a population
What’s another formula to work out incidence rate?
new cases exposed / (population x year)
How do you work out incidence rate ratio (IRR)?
(new cases exposed / population x year) ÷ (new cases UNexposed / population x year)
What does it mean if IRR = 2?
2 x more likely to have the disease in an exposed region compared to an unexposed region
What’s another way to work out IRR?
e^2 √ ((1÷d1) + (1÷d2))
d1 & d2 = cases in each population respectively
What are confounding factors?
something that is associated with both outcome and exposure of interest, but is not on the casual pathway between exposure and outcome
What is disadvantage about confounding factors?
distorts results and give misleading results
they can show potential casual links which are actually unfounded
What is standardised mortality rate (SMR)?
takes into account confiding factors to provide summative figure compared to general population
compares observed VS expected number of deaths, takes into account age-sex distribution
How do you work out SMR? what do the figures mean?
O/E x 100
age & sex most commonly accounted for confounding factors
SMR > 100 suggest EXCESS mortality
SMR
How else can SMR be calculated and how come other numbers are not required?
e^2 (1/O)
O = observed number of events in a population
no need for other values as SMR is compared to the standard population
What is incidence?
measurement of population’s average risk of disease
What is variation?
occurs in an epidemiological study whereby there is a difference between observed and expected value
What is carried out to allow for variation?
error factor is produced and from that confidence intervals are worked out
What is the confidence interval?
a range of values that we can say (with 95% confidence) that the actual value will lie in between this range
How do you work out confidence interval?
lower bound = observed value ÷ error factor
upper bound = observed value x error factor
What is biasing?
deviation of the results from the truth via certain processes
What is selection bias?
error due to SYSTEMATIC differences in ways in which the two groups were collected
information bias
error due to systematic misclassification of subjects in the group
recall / publication bias
studies with statistically significant or favourable results are more likely to be published
RIP HS(U)A
Recall bias Information bias Publish bias Healthy worked effect Selection bias Allocation bias
What are cohort studies?
recruiting disease FREE individuals and classifying them according to EXPOSURE status
then followed up for extended periods, disease progress monitored and incidence rates (IR) are calculated
What are cohort studies good for?
rare exposure or diseases which take a long time to develop
Prospective studies
disease free individuals recruited and followed up
‘look to the future’
Retrospective studies
disease free individuals recruited then exposure status calculated from historical documentation and followed up
can be prone to recall bias
use IR
Internal comparisons
when you have sub cohorts within your original group then compare exposed and unexposed within your within the cohort
use IRR
External comparison
exposed population compared against reference population e.g. general population
use SMR to remove cofounders
Case control studies, how are they carried out?
recruiting disease free individuals (control) and diseased individuals (control) and their exposure status determined (recorded)
then word out the odds ratio
(a x d) / (b x c)
What does null hypothesis assume?
2 things are equal / no difference
What does p
strong suggestion that the null hypothesis is false (statistically significant)
What does p>0.05 mean?
observed findings consistent with null hypothesis (2 things are equal) (NOT true, just consistent, so can’t reject)
If the CI includes IRR = 1 then what does that mean?
p>0.05, can’t reject null hypothesis
What value is the null hypothesis normally given? what does it mean?
null hypothesis = 1
being exposed will give the same odds for developing the disease as unexposed
(being exposed will give the odds of developing the disease for the odds calculated)
When is case control studies used?
for rare diseases
number of controls usually 5x amount of cases, controls easier to find - minimising EF
(more control, less error - more representative of the population of the disease being rare)
What are pros and cons of cohort study?
pros: good for rare EXPOSURES
cons: expensive, time-consuming (esp. if disease has long latent period - takes long time to develop)
but can calculate absolute risk
opportunity to look for different potential outcomes at once from varying exposures
What are the pros and cons of case-control studies?
good for rare DISEASE
opportunity to look for different EXPOSURES
cheap and quick (pt already have the disease)
can’t obtain absolute risks
heavily affected by recall + selection bias
What types of bias are in case control bias?
- selection bias - participants not representing general population e.g. diseased and control from the same ward (heart disease), SIMILAR risk factors / exposures
- recall bias - exposure status incorrectly determined - due to retrospective (looking back to history to determine exposure)
What can be removed using case control studies?
confounders
matching up same band controls with similar details
What does randomised controlled trial involve?
identify a source of eligible pts
allocate participants to treatment fairly - randomisation
follow up participants in identical ways
minimise losses to follow up (try and follow up often) and maximised compliance with treatment (follow the treatment plan given)
analyse data obtained and results
Why use randomisation?
used to remove any confounders that may be present in the study, known or unknown
What is double blinding?
neither the patient nor the doctor now which treatment they are on - removing selection bias (doctors are more likely to place healthy patients in new drug groups and show more positive results for the new drug to produce statistically significant results)
When are Placebo’s used?
if no current treatment already in place for disease in question
to remove placebo effect from the trial
what does loosing compliers lead to?
loss of randomisation as the patients who remain in the trail are normally ones who are predisposed to or have the disease, so want a cure
How are outcomes measure in randomised controlled trial?
intention to treat analysis
ignores noncompliance, withdrawal, and anything that happens after randomisation
What is the bradford hill criteria used for?
to determine whether causal-effect relationship has been established (once confounders, bias and chance have been removed)
What are the association features in the bradford hill criteria?
- STRENGTH of association: stronger association, more likely to be causal
- SPECIFICITY of association: outcome associated with specific factor (mesothelioma caused by asbestos)
- CONSISTENCY of association: association occurs in other studies too
What are the exposure / outcome factors of the Bradford Hill criteria?
- temporal sequence: causative factor precedes outcome (cause before outcome)
- Dose response: increasing exposure increases outcome
- Reversibility: removing causative factor reduces risk of outcome
What are the other evidence of the bradford hill criteria?
- Coherence of theory: observed observation confirms current scientific thinking
- Biological plausibility: biological mechanism support theory
- Analogy: study more likely if there is a link with another study
How do you minimise losses in randomised controlled trial?
follow up participants at appropriate times
no coercion or inducements (don’t force anything!)
honest to patients
Ho do you maximise compliance in a randomised controlled trial?
How can compliance be measured?
- simplify instructions
- patient allowed to ask any questions they have
- made simple and accessible for patients
measuring blood test urine samples etc.
What are ethics of randomised control trial?
- clinical equipoise - reasonable, certain which drug is better for patient, not subjecting patient to less effective treatment
- scientifically robust - in search for good of the general population
- ethical recruitment - recruitment for region where drug will take affect (all factors taken in - side effects etc.)
- valid consent - participants given sufficient knowledge cooling off period
What is systematic review?
an overview of primary studies that used explicit and reproducible method
large amount of studies identified and narrowed down with the most relevant and credible
studies should be transparent, reproducible and explicit (stated clearly)
What is meta analysis?
quantitive synthesis of primary studies within systematic review (continuos outcome e.g. long term effects)
provides overall value with associated confidence intervals
results shown via forest plots
What are the 2 types of systematic reviews?
- fixed effects model
2. random effects model
What is fixed effect model?
assumes that the studies used are homogenous and any variation between data comes from within - study variation
(study estimating SAME effect size)
What is random effects model?
assumes the studies are heterogeneous and variation between the data comes from between - study variation and in between study variation
(study estimating SIMILAR effect size)
What does systematic review rely on?
quality of primary studies AND publication bias
should use published AND unpublished trial
what is publication bias?
a study is more likely to be published if the results are statistically significant or has a large sample size - results will be biased
How can publication bias be determined?
using funnel plot
well shaped and balanced systematic review - funnel shape in its studies when plotted
a biased systematic review will vary in shape
Advantages of systematic review
explicit methods can reduce bias and exclusion of poor quality studies
meta-analysis provides overall figure for the studies
large amounts of information can be assimilated quickly by healthcare professionals
reduction in time between research discovery and implementation of clinical use
uses evidence based practice guidelines
(quick, large amount of information assimilated, short time period between discovery and use of treatment)
What is Henle-Koch’s Postulates?
- Necessary: cause BEFORE disease
- Specific: cause ABSENT in other diseases
- Sufficient: cause ALONE can cause disease
What is epidemiology?
the study of the distribution and determinants of health-related states or even in specified populations, and the application of this study to the control of health problems
What is clinical trail?
any form of planned experiment which involves patients and is designed to elucidate the most appropriate method of treatment of future patients
what is the purpose of a clinical trial?
reproducible controlled - comparison of interventions fair - unbiased without confounding autonomy - patient has a say pt able to withdraw anytime
What is the placebo effect?
even if the therapy is irrelevant to the patient’s condition, the patient’s attitude to his or her illness and indeed the illness itself, ay be improved by a feeling that something is being done about it
