The use of data Flashcards
1
Q
What percentage of patients are referred to secondary care?
A
3%
2
Q
Compare the factors which are involved in disease vs. illness.
A
Disease = symptoms, signs, diagnosis, bio-medical perspective Illness = ICE, patient perspective
3
Q
Nem 4 factors that affect the uptake of care.
A
- Concept of ‘lay referral’
- Peers, family, internet, TV etc
- Medical factors: new symptoms, disease sevreity
- Non-medical: crisi, peer pressure, age, gender, ethnicity
4
Q
Describe the patterns of uptake in care in males vs. females.
A
Males - High below 4 years then drops from 5-14 then slowly rises up to 75+
Females - not as much of an increase
5
Q
What are the three main aims in epidemiology?
A
- Description
- Explanation
- Disease control
6
Q
Epidemiology compares groups in order to do what?
A
Gain etiological clues
Scope for prevention
Identification of high risk priority groups
7
Q
What is the difference between clinical and epidemiological medicine?
A
Clinical - deals with individual
Epidemiology - deals with population
8
Q
What would be the numerator and denominator in epidemiology? Give an example.
A
Events/ Population at Risk
e.g deaths from IHD/Population at risk
9
Q
What does incidence give information on?
A
Trends in causation and aetiology of disease
10
Q
Define incidence.
A
Number of new cases
11
Q
Define prevalence.
A
Number of people with specific disease at single point of time or time period.
12
Q
What does prevalence give info on?
A
Absence of disease in population
Assess workload of health services
13
Q
In terms of incidence and prevalence describe minor and chronic illnesses.
A
Minor = increased incidence, decreased prevalence (cold) Chronic = reduced incidence and increased prevalence (diabetes)
14
Q
What is relative risk?
A
Strength of an association between suspected risk factors and disease
15
Q
How is relative risk calculated?
A
Incidence of disease in exposed group/ incidence of disease in unexposed groups
16
Q
In terms of risk. what do you need to be able to do?
A
Explain risk to patient
17
Q
Name 8 sources of epidemiological data.
A
Mortality data Hospital statistics Reproductive health statistics NHS expenditure Health and household statistics Cancer statistics GP morbidity Drug misuse database
18
Q
What is health literacy?
A
People having knowledge, skills, understanding and confidence to use health info, active partners in care, navigate systems.
19
Q
Why is health literacy considered a significant health concern?
A
Low health literacy –> poor health outcomes and widens inequality, increases rate of emergency admissions
20
Q
Give an example of poor health literacy which is a significant concern.
A
43% of English working adults struggle to understand instructions for childhood paracetamol dose
21
Q
Describe 5 tips to help when there is poor health literacy.
A
- Teach back
- Chunk and check
- Use pictures
- Use simple language
- Routinely ask people if they require help writing
22
Q
Give two examples of risk scores and what they are used for.
A
CHA2DS2-VASc (stroke risk in AF)
HAS-BLED (bleeding risk score)
23
Q
What are SIGN guidelines based on?
A
Systemic review of scientific literature
24
Q
Name 4 main types of studies.
A
- Descriptive
- Cross-sectional
- Case-control
- Cohort
25
What is a descriptive study?
Describes amount of distribution of disease in given population.
Identifies possible aetiology and emerging public health problems.
Assesses screening programs
26
What is a cross-sectional study?
Looks at disease frequency, survey and prevalence
27
How is a cross-sectional study carried out?
Observations made at a single point in time
28
What is a pro and con of cross-sectional studies?
```
Pro = quick
Con = impossible to infer causation
```
29
Name 3 types of analytical studies.
Cross-sectional
Case-control
Cohort
30
What is a case-control study?
Use two groups (one with disease and one without) and compare.
31
How are results of case-control studies expressed?
Odds ratio or relative risk
| Confidence intervals and p-values
32
What are cohort studies?
Use baseline data from group who don't have disease. The group is then followed through time until sufficient number develop disease to study. They are then split into subgroups to determine incidence of disease.
33
How are results from cohort studies expressed?
Relative risk with p-values
34
What does cohort studies determine?
Incidence of disease
35
What is the purpose of trials?
Test ideas about aetiology or evaluate interventions.
36
What is the definitive trial method?
Randomised control trial
37
What is a randomised control trial?
2 groups, alterations to intervention group, gain data on subsequent outcomes of both groups to calculate relative risk.
38
what does randomised control trials look at.
Whether modification of a factor alters incidence
39
What factors should you consider when interpreting results? Name 6.
```
Standardisation
Standardised mortality ratio
Quality of data
Case definition
Coding and classification
Ascertainment
```
40
What is 'case definition' for interpreting results?
Deciding whether individual has condition of interest or not.
41
What is the standardised mortality ratio?
Standard death rate converted into ratio for easy comparison
42
What is bias?
any trend in collection, analysis, interpretation, publication or review of data that can lead to conclusions different from truth.
43
What are the 4 types of bias?
Selection
Information
Follow-up
Systematic error
44
What are confounding factors?
Associated with disease and exposure. May be true causal factor.
45
What make up cofounding factors?
Risk factor A
Disease B
Cofounder C
46
Give examples of cofounding factors and what would happen if you adjust for these?
Age
Sex
Social class
--> stratify results
47
What is the criteria for causality?
```
Strength of association
Consistency
Specificity
Temporality
Biological gradient
Biological plausibility
Coherence
Analogy
Experiment
```
48
How is strength of association measured?
Odds ratio, relative risk
49
How is consistency achieved?
Repeated observation
50
What is the only absolute criterion?
Temporality
51
What is temporality?
Ensure exposure comes before disease
52
What is the difference between biological gradient and analogy?
Biological gradient = dose-response relationship
| Analogy = exposure-disease relationship
53
What are the main considerations in designing an audit?
What to assess?
How?
How to compare figures with other practices?
How to compare figures with national figures?
List possible suggestions for disparities.
54
What is the purpose of an audit?
Assess, Evaluate and Improve patient care (A,E,I)
55
How do you design and audit?
Need SET CRITERIA and STANDARDS to measure
Can utilise other guideline available e.g those based on systematic reviews (SIGN guidelines)
NICE produce audit toolkits
Carry out intervention before repeating the audit
56
What is the main objective of the audit?
Measure current practice against defined standard
57
What is involved in a full cycle of an audit?
Initial audit, change implemented, re-audit
58
What are the main sections of an audit?
- TITLE
- REASON for choice
- DATES of 1st and 2nd sets of data collection
- CRITERIA/ clinical condition/process of care to be audited and standards set with justification
- RESULTS of FIRST data collection and comparison to standards
- SUMMARY of subsequent discussion and plan of change agreed
- CHANGES IMPLEMENTED
- RESULTS of SECOND and comparison to standards set
- Quality improvement achieved
- Reflection in terms of GMP
59
What are the 5 main steps of an audit?
1. Set standards
2. Measure current practice
3. Compare results of practice to standards
4. Reflect, plan change and implement
5. Re-audit
60
What are the 4 main steps when designing an audit?
1. Preparation - choose topic (against guideline) and identify resources
2. Select criteria - define criteria (statement) and standard (usually target %)
3. Measure level of performance - collect and analyse data
4. Making improvements - present and discuss with relevant teams. Then repeat.
61
What are the top three causes of mortality in low income countries?
LRTI
CHD
Diarrhoeal disease
62
What are the top three causes of mortality in high income countries?
CHD
Stroke
CVD
63
What does population health look at?
Health risks and burden of disease
64
What are leading global risks of mortality worldwide?
Hypertension, tobacco, diabetes, overweight, physical inability, unsafe water and sanitation
65
What does disease surveillance look at?
Disease progression
66
What is relative risk?
Risk of event relative to exposure
67
What do screening tests show?
Smaller populations who need further testing
68
On what levels can screening test be carried out giving examples.
Individual - diabetes
| National - cervical cancer
69
Wilson and Junger (1968) created a criteria to decide if screening programmes were worthwhile, What were the 4 main points?
Characteristics of disease (e.g must have clinically detectable pre-symptomatic phase), test (e.g must be safe), treatment (e.g safe and proven effectiveness)
Consideration of organisation and costs
70
The WHO recently updated the criteria for screening programmes in 2008 to include what?
```
Respond to recognised need
Defined target population
Scientific evidence of effectiveness
Quality assurance
Promote equity
Evaluation planned
Benefits should outweigh harm
```
71
What is tertiary prevention?
Reduce consequence of disease and disability and prevent deterioration to increase QoL
72
The PICO framework is useful for formulating searchable questions. What does PICO stand for?
People/patients
Intervention
Comparison
Outcome
73
Compare qualitative and quantitative research.
```
Qualitative = generate language data (for exploration)
Quantitative = generate numerical data (test hypothesis)
```
74
What type of research tests hypotheses?
Quantitative
75
Name 5 disadvantages of randomised control trials (the gold standard).
```
Costly
Difficult to set-up
Time consuming
Selection bias
Can't address all research questions
```
76
Is a prospective or case-control study more powerful and why?
Prospective - can more easily control issues relating to time and confounding factors
77
What is a disadvantage of a prospective cohort?
Greater chance of losing subjects in follow-up
| Costly
78
What studies have the most bias?
RCT -> Cohort -> Case-control -> Case reports -> Opinion
79
What is a nested cohort study?
Variation of case-control study
80
Give an example of a nested cohort study.
7000 woman - 3500 flagged for cancer and death who had a hysterectomy. 3500 were flagged and didn't have an operation
81
What is an observational study?
Uses questionnaires for example
82
What is an individual patient meta-analysis study?
Compare relevant studies identified by computer-assisted searches fro example.
83
What type of study investigates aetiology?
Case-control
84
How is the odds ratio calculated?
2 x 2 contingency table = (a/c)/(b/d) = ad/bc
85
What is relative risk a measure of?
Relative incidence
86
What does the odds ratio do?
Approximate relative risk
87
How would you design a case control study?
1. Decide on CASE DEFINITION
2. Calculate number of cases and controls that need to be include to have sufficient STATISTICAL POWER
3. IDENTIFY CASES from appropriate source
4. IDENTIFY CONTROLS from preferable the same source (e.g from community)
5. Decide on EXPOSURE DATA
6. Collect the data in an UNBIASED manner
7. ANALAYSE the data
8. PRESENT results
88
What else do you need to ensure when designing a car control study?
Those selected are representative of the underlying population of "all people" without the disease
89
Describe matched vs unmatched controls.
In an unmatched study, sizes of two groups need not be identical
90
What can improve the power of the study?
More than once control is recruited
| Match factors such as age and sex
91
What are matched factors present in?
In case and control groups in same proportions
92
What does it mean if there are cases with no control data in a fully matched study?
They are discarded
93
Describe a form of information bias.
Recall bias - where cases or their relatives have "mulled over" the possible reasons for developing the disease and "improved" their recollection of possible risk factors or they "fail to remember"
94
Social class IV/V "causes" lung cancer? Why is this an example of confounding?
```
Distorted because higher prevalence of smoking in social classes IV/V than in other social classes.
There are independent associations between smoking and being in social class IV/V and between smoking and lung cancer.
```
95
What type of study would you design to make more reliable assessment of one of the theories from your case control study?
Prospective or cohort study
96
What does cohort studies test?
Associations between exposure(s) and disease outcome
97
What is the basis hypothesis for cohort studies?
If (risk factor A) is aetiologically associated with development of (disease X) then a group of people with higher levels of exposure to (risk factor A) "exposed" will have a higher proportion of individuals who will go on to develop (disease X) than a group of people with lower levels of exposure to (risk factor A) "unexpected/less expected"
98
What is the gold standard test to measure relative risk?
Cohort study
99
What are disadvantages of a cohort study?
Expensive
Time lag between exposure and development of disease
Difficulty of ease of follow-up
100
A mother with a young child has questions about MMR. Heard about autism risk and never seen measles anyway, feels it's not something she wants her child to have. Are you better placed to advise? What kind of info would you require and have at hand and how would you present it to the mother?
Yes, have statistics on risks of not having MMR.
