Semester 1 - Lecture Revision Flashcards
Definition of research?
“a detailed study of a subject, especially in order to discover (new) information or reach a (new) understanding”
Difference between positivism and interpretivism?
Positivism, at its simplest, holds that so-called ‘positive’ knowledge is derived from empirical data - quantitative
Interpretivist research is concerned explicitly with generating understanding of the world - qualitative
Definition of evidence-based medicine?
“The conscientious, explicit, and judicious development and use of current best evidence in making decisions about the care of individual patients.”
What are the different ways health research may be used - think different levels of society?
What are elements that one considers when making a clinical decision?
Different factors that go into make a plan/decision for patient treatment - not only scientific research - other pragmatical factors need to be considered.
What are some limitations to health research?
Some questions/areas more likely to be researched
1. Is the condition druggable?
2. Is the condition profitable?
3. Is it trendy?
4. Is it a political choice to focus on a particular condition?
In areas where there is evidence:
1. Study population may not match patient in front of you
2. Individuals may respond differently within study populations
3. Many patients have multiple conditions
4. Single studies may be over-interpreted
5. Adverse effects may take years to emerge
6. Statistical versus clinical significance - is statistical significance sufficient to warrant clinical significance
What questions are asked when conducting a population based health study?
- What conditions cause most morbidity?
- How many people have these conditions, who are they?
- What are the outcomes in a population?
- How good are our services?
- What services and interventions improve outcomes?
- How much does it all cost?
- What needs to change?
What are the three components of public health?
Three components of public health
1. Health improvement - Addressing preventable risk factors for disease
2. Health Protection - Addressing communicable diseases (primarily)
3. Healthcare improvement - Working with health services for population health
All of which is underpinned by health intelligence - research
What types of research can inform population health?
- Biomedical research
- Epidemiology
- Health services research
- Studies of behavioural, social and economic factors that affect health
Definition of evidence?
“one or more reasons for supporting a proposition about the natural or social world”
Pragmatically - what counts as evidence relates to what kinds of problems you want to address or questions you want to answer.
What should we keep in mind when thinking about when examining evidence?
Research is shaped by the researchers previous beliefs and views (power structers) - shapes the experimental design + the results that they are looking for
What evidence based tool did Edinburgh develop for delirium detection?
4 ‘A’s Test (4AT)
- Alertness
- 4AMT - recalling age, DOB, place and current year
- Attention
- Acute changes/fluctuating changes
Provies a score - allowing us to assess level of delirium
Tool improved delirium diagnosis
What are the limitations of evidence based medicine?
- How evidence is used - will it always be relevant to the case in front of you, how have the numbers/conclusions been generated and does the evidence come from a large or small study?
- Overstandardisation and control over clinical practice - limits clinicians ability to deal with patient idiosyncrasies/perspectives
- Difficulty getting evidence into clinical practice and policy - long arduous process
- How different values come to shape how and what forms of evidence are, or are not, taken up in the policymaking process.
What are two barriers for people to access scientific articles/journals?
- Language - primary language is english
- Financial - subscriptions cost money
What is peer-review? What are the different ways that it can be done?
Peer-review - experts review
Note - not perfect as there are lots of examples/levels where bias can be introduced
Types
Double blind – authors don’t know reviewers (visa-versa)
Single blind – one group do not know about the other
Open – everyone knows everything
What are pre-print repositories?
Pre-print repositories – articles are released before peer-review/journal acceptance
What are predatory journals?
Predatory Journals - accept papers for a fee
What does the critical appraisal of articles entail?
“Critical appraisal is the process of carefully and systematically examining research to judge its trustworthiness, and its value and relevance in a particular context”.
Not about being negative - aim is to look for strengths and weaknesses and make a balanced decision about its usefulness.
Why is critical appraisal important?
- Lots of articles/reports on your subject of interest – you need to be able to decide which ones are most relevant and best quality.
- Not all studies are good quality, some have biases that affect interpretation of the findings.
- Basing decisions on evidence that is not reliable or trustworthy can have serious implications
When critiquing qualitative studies, what should you keep in mind?
Qualitative studies
Looking for detail on: sample diversity, recruitment approach (any bias) and analysis, limitations.
When critiquing quantitative studies, what should you keep in mind?
Quantitative
Looking for detail on: sample type (probability - i.e. random, or non-probability – i.e. quota), sample size (is it representative or not, any bias), analysis, limitations
Provide a brief outline for the development of clinical guidelines.
- Guidelines construction - Different forms of evidence used to construct guidelines - Meta-analysis, systematic reviews and RCTs rated highly
- Note - Sometimes some areas lack any evidence – clinical experience and recommendations are used when it is lacking - Publication of guidelines and dissemination of information
- Guidelines under review - data collected and changes are implemented where appropriate.
- Note - Guidelines are shaped by public involvement
What is the role of guidelines and who is normally responsible for creating them?
Guidelines help with diagnosis, management and treatment
Formed by governments or major professional organisations
Note - Guidelines inform clinical decisions – should not be used as set rules
What are the moral principles governing governing ethical research?
- Ensuring respect for people and doing no harm.
- Recognising that participants are not a means to an end.
- Ensuring research is socially and/or medically valuable for participants and/or as part of a wider contribution to knowledge.
- Ensuring participants are cared for, particularly those who are marginalised, vulnerable, disempowered or disadvantaged.
What is the difference between research ethics and integrity?
Responsible conduct of research - overarching
What are the 3Rs for reducing animal suffering in research?
Can research be conducted without an hypothesis?
Yes, when you have an open-ended question which you don’t have information available to predict/create a hypothesis - Research question is exploratory, so one does not predict the outcome
Inductive approach (data is collected to build a theory) rather than the deductive approach
Typically associated with quanlitative studies
Examples…
1. Research project exploring patients’ understandings and experiences of cancer genetic testing.
2. Research project exploring men’s experiences of chronic illness.
What is the difference between a descriptive and analytical apporach in epidemiological research?
Descriptive - Usually focus on issues related to “person”, “time” and “place” - describing/identifying patterns
For example…
1. Person: who is affected?
2. Time: when did the disease event occur?
3. Place: where did the disease event occur?
Analytic - trying to understand something about the underlying relationship between one or more “exposures” or “risk factors” and a specific outcome
Does exposure/risk factor ‘X’, increase the liklihood of ‘Y’ occuring.
Exposure - can be fixed/non-modifiable factors, modifiable risk factors or interventions (health education, drug, etc.)
Outcome - can be negative or positive - death, disease status, degree of disability, length of survival or survival, risk reduction, satisfaction, quality of life
What are the two broad categories of study designs used by epidemiologists?
What is the difference between incidence and prevalence?
Ways of measuring disease frequency
Prevalence - proportion of a population with disease at one time point - the number of cases divided by the size of the population
Incidence - new cases in a population over a specified time period.
Note - incidence is reported as number of new cases per person-years at risk
Person-years at risk is calculated by multiplying length of time (specified time period) by the current population.
What are the following epidemiological study types and what are they used for?
a) cross-sectional study
b) Prospective/Co-hort study
c) Retrospective/case-control study
- Cross-sectional study- looking at the prevalance at a specific time period - taking a snapshot - e.g. relationship between levels of physical activity (PA) and diabetes in scotland in 2010 or 2019
-
Prospective/Co-hort study - follow a sample of people selected and examine whether the exposure ‘X’ influences the likelihood of ‘Y’ over this time period
Example - follow a sample of people in 2010 with known PA levels all the way through to 2019, at which time we would check to see whether they had been diagnosed with diabetes over that follow-up period. -
Retrospective/case-control study - sample a group of participants with a conditon and a group without, and establish the levels of the ‘exposure’ they encountered/experienced over a specified time period. Then by comparing the two groups, we can evaluate the impact of the exposure/risk factor
Example - one group of people known to have diabetes plus a second sample of individuals known to be diabetes-free (controls). We would then look backwards in time to establish what their PA levels had been during the previous decade.
What is relative risk and odds ratio? What are they used for?
In both cohort and case-control studies, we would usually wish to have some form of numerical measure of the association between the exposure and outcome of interest
Cohort studies - we typically measure association via the relative risk, which is a ratio of the incidence of the outcome in one exposure group relative to the incidence of outcome in the other exposure group.
Case-control study - use the Odds Ratio, which compares the odds of an outcome in one exposure group relative to the other
What are three sources of error in epidemiological studies?
-
Chance - Random Variation - due to sampling variation and measurement error
- Address this issue with statistical interference -
Bias (systematic error) - induced by study design or conduct - caused by selection bias (not representative cohort), measurment bias (different groupss treated in different ways) and recall bias (can’t remeber or exaggerating?)
- Bias is systematic error (or variation) in results away from the true population relationship - Confounding - detection of an association between exposure and outcome that is an artefact of a third variable (called a confounder) being strongly associated individually with both of the exposure and outcome variables - removes casaulity
Confounding problems can be dealth with at the design stage (e.g. randomising participants to interventions) but is more commonly dealth with at the analysis stage (stratifying groups or introducing cofounder in the multifactoral statistical model).
What is a cohort study?
Study whereby a designated group of individuals are followed or traced over a period of time.
Concept
1. Define a population of interest
2. Sample from population of interest
3. Group people in terms of whether they have been exposed to the exposure of interest or not.
4. Follow these groups over time to see whether our outcome of interest develops
What are the differen stages for developing and testing complex interventions?
