Week 2 Flashcards
What are the three ways for searching for evidence?
- informally, recreationally
- focused, looking for answers
- surveying existing literature
Informally searching for evidence
google searches, social media
Focused, looking for answers
- usually related to questions that have occurred in a clinical setting
- synthesizing information
Surveying existing literature
clarifying gaps in knowledge, clinical audit
What is primary research?
a single research study
What is secondary research?
synthesized findings, takes form of a literature review
What is bias?
a systematic error or deviation from the truth in resulsts of inferences
What are the 6 types of study designs?
- intervention
- diagnostic
- aetiological
- prognostic
- screening intervention
What are the types of bias found in intervention studies?
- allocation
- detection
- performed
- attrition
- measurement
Allocation bias
differences between treatment and control conditions at the start
Detection
differences in how treatment and control groups are assessed or measured
Performed
events other than intended occur during experiment
Attrition
some participants leave the study, or unwanted differences between groups
Measurement
outcomes measured inaccurately
What is levels of evidence?
to what degree, based on study design, can information be trusted
What has occurred in secondary research?
- examine multiple primary sources
- summarised research papers
- combined findings to provide strong evidence
What are Boolean operators?
words/connectors placed between search terms to narrow or expand a search
What are common Boolean operators?
or, and, not
What is truncation?
identifies variations of a word without the need to individually enter each word
What are wildcards?
enables entering one search term to account for many different ways a word can be spelt, uses a question mark
What is a federated search engine?
allows the simultaneous search of multiple searchable resources - make a single query that is distributed to search engines in the federation
What is the National Health and Medical Research Council?
gvt organisation offering expert advice about health and medical research, ethical healthcare and health research
What is an intervention study?
a sample of patients are treated with the disorder and changes to the patient’s condition is measured
this is then compared to patients who were not treated
What is a diagnotic study?
measured how well an assessment identifies people with a specific disease
What is an aetiological study?
measures association between people’s exposure to a suspected hazard and rates of disease among the same people
What is a prognostic study?
records what happens to people who have a specific health condition, especially in cases where the condition is untreated/
What is a screening intervention?
a large number of people without signs are tested. Where the patient may have the condition, measure whether follow-up procedures lead to a benefit or harm
Systematic reviews of RCTs
-level 1
+more reliable, less bias
-expensive, time consuming, not good for new studies
RCTs
-level 2
-participants are randomly selected into control and intervention groups
+less bias, more reliable
-issues with ethics
Quasi
-non-randomised RCT
+can use specific groups, doesn’t take much time
-more biased
Cohort
-level 4
-observational study on a group of people over time
-longitudinal, exposure, follow-up
+specific
-people not completing the follow-up (waste money)
Case control
-comparative study with control and case groups
-cause group has condition, other doesn’t
+allows researchers to look at specific cases
-biased, can be hard to find people with a rare disease
Cross section
-snapshot of a population at a given time
-eg census
+lots of information
-reporting bias
Anecdotes
-stories
+takes into account people’s experiences
-lots of bias
What is an experimental study?
- quasi and RCT
- able to find a cause because two groups are compared
- intervention
What is an observational study?
- systematic review, cohort, case control, cross-sectional, anecdotes
- cannot say one thing causes another as there are confounding variable s
What are the types of biases involved in systematic reviews?
- publication
- time lag
- duplicated publication
- outcome reporting
- citation
- database inclusion
- language
- reviewer’s personal biases
Publication bias
studies researching an unpopular topic don’t get published –> evidence is unavailable
Time lag bias
delays in publication prevent practitioners using research
Duplicated publication bias
same results from the same study are repeatedly published, suggesting there is more evidence than there really is
Outcome reporting bias
desirable, expected or statistically significant resulsts are published, even though other results are equally valid and informative
Citation bias
reviewers are more likely to find a study cited by lots of authors
Database inclusion bias
studies are more easily found if they are available from online databases
Language bias
researchers will have a preference for studies published in their own langauge
Reviewer’s personal bias
reviews may unfairly exclude an article because they don’t like the topic or result, even though the results are valid and relevant
According to the NHMRC, what should be evaluated when choosing evidence?
- amount of evidence
- quality of evidence
- level of evidence
- consistency of evidence
- clinical importance
- generalisabiliy
- applicability
What is a clinical practice guideline?
educational document with recommendations for patient care for a named health condition
-doesn’t tell health practitioners what to do, but gives an evidence-based suggestion
