Medicine & Surgery
- Number of new cases of a disease in a given time period
- (New cases / population at risk) * 1000
- Total number of cases in a population at one point in time
- (All cases / population at risk) * 1000
- Or what question says - per 1000 or per 100, etc.
- Alternate explanation for the association between exposure & outcome
- Standardisation: Match people into groups with similar characteristics
- Variation in health & disease for different generations
- Application of a statistic at a population level to the individual
- Availability of good healthcare varies inversely with the need for it in the population served.
- E.g. In areas with most sickness and death GPs have more work, larger lists and less support so the care is worse despite a greater need.
Used in decision: Should a cancer patient have chemo or not?
- Quality-adjusted life years: Looks at remaining years they have
- Remaining life left * QoL in those years = QALY
- Disability-adjusted life years: Looks at what they've lost
- Years of life lost (YLL) + Years lost to disability (YLD) = DALY
How do you calculate the following:
- (No. of people with a disease / No. of people at risk of disease)
- This is a proportion not rate
- (No. of new events in a year / No. of people exposed to risk in that year) * 1000
- (No. of deaths in age group in year / Avg. population in age group in year) * 100000
- (No. of incidence deaths from disease in time period / Incidence in the time period) * 100
- (No. of live births in a year / Avg. population in same year) * 1000
- (No. of live births in a year / Avg. mid year population of women of childbearing age) * 1000
- 15-44 years
- (No. of deaths in a specified period / No. of people at risk of dying in that period) * 1000
What are the types of studies within the following categories?
- Point prevalence
- 'Snapshot' of frequency of disease'
- 'Single point in time'
- 'On a certain day'
- Case series
- More than one person has a condition.
- Describing trends and patterns
- Case report
- One study of a particular person with a condition.
- Rare diseases.
- Cohort: Prospective
- Take 2 groups, one exposed to a RF and one not and compare their outcomes - How many develop the disease?
- Pros: ↓ Recall & selection bias
- Cons: Expensive & time consuming, lose patients along the way
- Relative risk: Ratio of the probability of an outcome in an exposed group to the probability of an outcome in an unexposed group
- (0.17 / 0.2) * 100 = 85%
- Case-control: Retrospective
- Look at diseased and non-diseased people and compare their past exposures
- Pros: Quick, easy, cheap
- Cons: Recall & selection bias (Match groups to eliminate confounders)
- Odds ratio: Measure of association between an exposure and an outcome
- (10/50) / (20/80) = 0.2 / 0.25 = 0.8
- Group B are less likely to get the disease than group A
- There is increased odds of getting disease with that exposure
- Randomised controlled trial (RCT)
- Participants randomised into treatment & control group and outcomes compared
- Double blinding: Neither the patient nor the doctors know who has the drug
- Unethical to give diseased patients a placebo that is guaranteed not to work
- Control group is given 'usual care' to combat this
- Number needed to treat (NNT):
- The average number of patients who need to be treated to prevent one additional bad outcome
- Absolute risk ratio = Control event rate - experimental event rate
- Participants are not assigned by chance, they can choose which group they want to be in OR they may be assigned to the group by researchers.
- Can be considered more ethical if the participants choose to be assigned to the control group
- Participants knowledge that they're in the treatment group may affect how they feel their treatment is going.
- There is no correlation between A & B
- Probability that the results are due to chance, i.e. probability of making a type I error
- Optimist: The null hypothesis is rejected when it should have been accepted
- Robust study design
- Pessimist: The null hypothesis is accepted when it should have been rejected
- Large sample size
- Probability of rejecting H₀ when it is false.
- High power = solid study
- Calculation: 1 - β
Which of the following represent a type I error, type II error and the power of a study.
Study Accepts H₀
Study Accepts H₁
- Type II error (β)
- Type I error (ɑ)
- Variability of the values about the mean
- Correlation coefficient
- -1 ≤ x ≤ 1
- Regression: "If this is changed, how much will this change?"
- Relying on memory to report things that happened in the past
- A disease is detected at an early stage, giving the impression that the survival time is longer when it isn't
- Screening tests
- Not randomly selecting patients for a trial to get a desired outcome
- Loss of participants due to loss to follow up.
Based on the following table, which letter represents a true positive, false positive, true negative and false negative?
- True positive (TP)
- False positive (FP)
- False negative (FN)
- True negative (TN)
What are the following and how do you calculate them:
- Proportion of patients that test positive and have the disease
- Probability that those with the disease will be detected
- TP / (TP + FN)
- Number of patients that test negative and don't have the disease
- Probability that those without the disease won't be detected
- TN / (TN + FP)
- Probability that if you test positive you have the disease
- TP / (TP + FP)
- Probability that if you test negative you don't have the disease
- TN / (TN + FN)
- How good a test is at getting the correct answer
- How good the test is at getting the same result if repeated
- PPV & NPV
- Wilson-Jungner criteria
- Proving that correlation is causation between 2 variables
- Does the cause precede the effect?
- If you remove it does the effect go away?
- If you increase the dose does the effect increase?
- Biological plausibility
- Study design
- It is a Healthcare evaluation framework. Split into the following:
- Efficiency: Cost/benefit
- Validity of a screening programme. There are 10 targets that all screening programmes should meet.
- Natural history understood
- Detectable at an early stage
- Treatment early stage better than later
- Suitable test for early stage
- Acceptable test
- Intervals for repeating test determined
- Adequate health service provision
- Risk < benefits
- Cost balances benefits
- Health education
- Health protection
- Fiscal & legal policies to protect health
- Disease prevention- Leavell and Clark's levels
- Primary: ↓ incidence
- Secondary: ↓ prevalence
- You must treat thousands to benefit a few people, the rest of them did it for no benefit directly to themselves
- E.g. Vaccinations - You may have never got that disease anyway
- Impairment: Loss of psychological, physical or social function (Leg amputation)
- Disability: Inability to complete a task (Can't drive your car)
- Handicap: Inability to fulfil a role (Can't drive son to football practice)
- Maslow's theory states that our actions are motivated by certain physiological and psychological needs that progress from basic to complex.
- Individual perceptions of variations from normal health.
- Felt need turned into action. Vocalising their need by seeking help.
- Based on professaional judgements of relative needs of a group.
- Based on professional judgement defined by experts.
- States that most ill people aren't seen by healthcare professionals
- 1/4 have no symptoms
- 1/4 self medicate
- 1/4 do nothing
- 1/4 go to the GP
- Notify the health protection officer within 3 days
A limited list…(you don’t need to know all of them)
- Hepatitis A
- Yellow fever
- Endemic: Disease always present in the population
- Epidemic: Temporary but clear increase in the incidence of a disease in a population
- Pandemic: An epidemic that travels worldwide
- Communicable disease is transmissible between people, infectious disease encompasses all disease transmissible between anything, therefore communicable disease is classed as a type of infectious disease
- Sentinel: Using key areas throughout the U.K to look at trends across the country.
- Enhanced: Report → Throat swab → etc. (Algorithm to follow for the individual disease).
- Laboratory data: Not very reliable due to clinical iceberg & not testing everyone.
- Syndromic: Track how many people have presented to NHS with certain symptoms & signs.