Critical Appraisal of medical literature Flashcards
(111 cards)
1
Q
PICO
A
P- population
I- intervention
C- community
O- outcome
2
Q
What is a Randomized Controlled Trial (RCT)?
A
An RCT is a scientific experiment designed to test interventions in an unbiased way, ensuring that the results are reliable and valid.
3
Q
What does “Randomized” mean in the context of an RCT?
A
“Randomized” means that participants are randomly allocated to different treatment groups, which helps eliminate selection bias and ensures that the groups are comparable.
4
Q
What does “Controlled” mean in an RCT?
A
“Controlled” means that there is a control group, which may receive no treatment or the standard of care, allowing for a comparison between the treatment and control conditions.
5
Q
What does “Double-Blinded” mean in an RCT?
A
“Double-Blinded” means that neither the investigator nor the participant knows which group (intervention or control) the participant is in, reducing bias in treatment administration and outcome assessment.
6
Q
Why are these features important in an RCT?
A
These features are crucial because they help:
- Minimize Bias: Randomization and blinding reduce the risk of selection bias, performance bias, and detection bias, leading to more reliable and valid results.
- Ensure Comparability: Random allocation ensures that the treatment and control groups are similar at baseline, making it easier to attribute differences in outcomes to the intervention.
- Enhance Validity: Double-blinding helps ensure that the results are not influenced by the expectations or beliefs of the researchers or participants, enhancing the study’s internal validity.
7
Q
Sources of error
A
Random error
- chance
Bias
- Selection bias
- Information bias
- Cofounding
8
Q
What is bias in the context of research studies?
A
Bias is a systematic error that leads to a statistical overestimation or underestimation of the population parameter being measured.
9
Q
What are the main types of bias in research studies?
A
The main types of bias include:
- Selection bias
- Information bias
- Recall bias
- Confounding
10
Q
What is selection bias?
A
Selection bias occurs when there is poor or no randomization, causing the two groups in a study to be different and not comparable.
11
Q
What is information bias?
A
Information bias occurs when there is misclassification or measurement error, such as using a scale that has not been zeroed, leading to inaccurate data collection
12
Q
What is recall bias and when is it most likely to occur?
A
Recall bias occurs when participants remember past events inaccurately, often seen in retrospective studies where memories may be influenced by personal experiences.
13
Q
What is confounding?
A
Confounding is a form of bias where an outside variable influences both the independent variable and the dependent variable, leading to a spurious association.
14
Q
Can confounding be corrected after the trial is complete?
A
Yes, confounding can generally be corrected for through statistical adjustments after the trial is complete.
15
Q
Can you give an example of information bias?
A
An example of information bias is measurement bias, such as when a scale that has not been zeroed makes everyone appear 5 kg heavier than they really are.
16
Q
What are some other types of bias in research?
A
Other types of bias include:
- Survivorship bias
- Omitted variable bias
- Observer bias
- Funding bias
17
Q
How can bias be minimized in research studies?
A
Bias can be minimized by designing the study well, including proper randomization, blinding, accurate measurement tools, and appropriate statistical adjustments.
18
Q
Cofounding definition
A
Occurs when estimate of association between exposure & disease (outcome) is wholly or partly due to effect of another exposure on the same disease (outcome), & the two exposures are correlated
19
Q
Why should we randomise interventions?
A
- Randomisation unbiased comparison between groups
controls for known and unknown confounding variables - balanced groups with same prognostic variables
- If all groups are prognostically balanced, a difference in outcomes may be attributed to the intervention.
20
Q
What is a placebo?
A
A placebo is an intervention (substance or treatment) that has no intended therapeutic value and does not contain the active substance that affects health.
21
Q
What is the placebo effect?
A
The placebo effect is a beneficial effect produced by a placebo drug or treatment, not attributable to the properties of the placebo itself, but rather due to the patient’s belief in that treatment.
22
Q
What is the standard of care?
A
The standard of care is the treatment agreed upon by experts to be appropriate, acceptable, and widely used for a particular condition.
23
Q
Why is having a control group important in clinical trials?
A
Having a control group is important because it allows for comparison to determine if the intervention has a true effect. The control group may receive a placebo or standard of care to provide a baseline for comparison.
24
Q
Why might it be unethical to use a placebo in some clinical trials?
A
It might be unethical to use a placebo if there is already a proven, effective treatment for a condition. Omitting this treatment would be negligent and could harm patients.
25
How does the placebo effect impact the interpretation of clinical trial results?
The placebo effect can cause changes or perceived changes in patients, even if there is no active substance in the placebo. Without a control group, it is difficult to determine if the observed effect is due to the intervention or just a placebo effect.
26
How does comparing the effect of an intervention to a placebo help in clinical trials?
Comparing the effect of an intervention to a placebo helps determine if the intervention has an additional effect beyond the placebo effect, indicating a true therapeutic benefit.
27
What is the endpoint?
The endpoint is the outcome we are looking at
28
Primary endpoint
Main result measured at end of a study to see if a given treatment worked
29
Clinical endpoint
Occurrence of a disease, symptom, or sign that constitutes one of the target outcomes of the trial
30
Secondary endpoint
Other result (s) being measured at different time points of the study to assess impact of intervention
31
Surrogate endpoint
Measure of effect of treatment that may correlate with a real clinical endpoint but does not have a guaranteed relationship, e.g. laboratory marker
32
Composite endpoint
A composite endpoint combines two primary outcomes together. This may be if patients develop vomiting and/or diarrhoea or something like heart failure and/or death.
33
What is the purpose of measures of effect or association in epidemiology?
The purpose is to determine if a disease (outcome) is more common in one (exposed) population compared with another (unexposed) population by comparing the frequency of the disease in both populations.
34
How do we compare the frequency of disease in exposed and unexposed populations?
By calculating and comparing the incidence or prevalence of the disease in the exposed group to the incidence or prevalence in the unexposed group
35
How can the size of the effect or relationship between a disease/outcome and exposure be measured?
The size of the effect can be measured by calculating ratios or differences between the frequencies of the disease in the exposed and unexposed populations.
36
What is the Risk Ratio (Relative Risk) and how is it calculated?
The Risk Ratio (Relative Risk) is a measure of association that compares the risk of a disease/outcome in the exposed group to the risk in the unexposed group. It is calculated as:
Risk Ratio = Incidence in Exposed Group
/ Incidence in Unexposed Group
37
What is the Odds Ratio and how is it used in studies?
The Odds Ratio is a measure of association that compares the odds of a disease/outcome occurring in the exposed group to the odds in the unexposed group. It is commonly used in case-control studies.
38
Absolute measures of effect
Calculate excess risk caused by exposure in the exposed compared to unexposed group.
- Rate difference
- Rate difference percent
- Risk difference
- Risk difference percent
39
Risk (Absolute risk- AR)
Is the probability of occurrence of disease
40
Odds
Ratio of probability of outcome in exposed group to probability in of outcome in unexposed group
41
Odds ration (OR)
Comparison of odds in unexposed and exposed groups
42
Risk ratio (RR)
Comparison of risk in unexposed and exposed groups
43
Hazard ratio (HR)
Comparison between the probability of events in a treatment group, compared to probability of events in a control group
Used to see if patients receiving a treatment progress faster (or slower) than those not receiving treatment.
44
Absolute risk reduction (ARR)
Amount by which therapy/intervention reduces risk of bad outcome
Difference in AR between two groups
45
Number needed to treat (NNT)
): Number needed to treat in order to prevent one additional bad outcome
NNT =1/ ARR
46
Normal distribution
mean= median
47
hypothesis testing in research
1. hypothesis
2. null hypothesis (H0)
3. alternative hypothesis (H1)
48
Hypothesis
A precise, testable statement predicting the outcome of the study
49
Null hypothesis (H0)
There is no difference in outcome estimate between two groups
50
Alternative hypothesis (H1)
There is a difference between two groups
51
What is a p-value?
A p-value is the probability of obtaining the observed result, or something more extreme, if the null hypothesis is true. It indicates how likely it is that the observed difference between treatments occurred by chance.
52
What does a p-value represent if the null hypothesis is true?
If the null hypothesis is true (i.e., treatments A and B are equally effective), the p-value represents the probability of getting a difference as large as the one observed purely by chance.
53
What does a p-value less than 0.05 indicate?
A p-value less than 0.05 indicates strong evidence against the null hypothesis, suggesting that the observed difference is unlikely to have occurred by chance alone. This result is considered statistically significant.
54
What does a p-value greater than 0.05 indicate?
A p-value greater than 0.05 indicates a lack of evidence against the null hypothesis, suggesting that the observed difference could have occurred by chance. This result is not considered statistically significant.
55
What is the range of values for a probability?
Probability can range from 0 to 1, with higher values (closer to 1) indicating a more likely event.
56
What does a p-value of 0.5 mean?
A p-value of 0.5 means that there is a 50% probability that the observed result occurred by chance if the null hypothesis is true. This is quite high, indicating no strong evidence against the null hypothesis.
57
What does a p-value of 0.05 mean?
A p-value of 0.05 means there is a 5% probability that the observed result occurred by chance if the null hypothesis is true. This is relatively low, providing some evidence against the null hypothesis.
58
When is a result considered statistically significant?
A result is considered statistically significant if the p-value is less than 0.05, indicating that the observed difference is unlikely to have occurred by chance alone.
59
How does a smaller p-value affect the null hypothesis?
A smaller p-value provides stronger evidence to reject the null hypothesis, suggesting a true difference between treatments.
60
What does a larger p-value indicate about the evidence against the null hypothesis?
A larger p-value indicates a lack of evidence to reject the null hypothesis, suggesting the observed result may have occurred by chance in populations with no difference between treatments.
61
What is a confidence interval?
A confidence interval is a range within which the true population mean is expected to lie, based on sample data.
62
What does a 95% confidence interval imply?
A 95% confidence interval implies that if we repeat the experiment 100 times, on 95 occasions, the true population estimate will lie within this interval.
63
What does a narrow confidence interval indicate?
A narrow confidence interval indicates more precision in the estimate of the population mean.
64
What does a wider confidence interval indicate?
A wider confidence interval indicates less precision in the estimate of the population mean.
65
What factors influence the width of a confidence interval?
The width of a confidence interval is influenced by the variability of the data (standard deviation) and the sample size.
66
Why do we use samples in statistics?
We use samples to gather data and infer something about the population because it is often impractical to investigate the entire population.
67
Why might a sample not represent the underlying population well?
A sample might not represent the population well due to chance inclusion of outliers or bias in selection and measurement.
68
How do confidence intervals help in statistical inference?
Confidence intervals construct a range of values around the sample mean within which we can be confident the true population mean will lie.
69
How do we calculate the 95% confidence interval using sample data?
Using the sample data and a margin of error (based on sample size and variability), we calculate the 95% confidence interval, which tells us the range within which the true population mean is expected to lie.
70
What does a 95% confidence interval tell us about the population mean?
A 95% confidence interval tells us that we are 95% confident that the population mean lies within this interval.
71
How does sample size affect the width of confidence intervals?
Larger sample sizes usually result in narrower confidence intervals, making the estimate more precise.
72
How does data variability affect the width of confidence intervals?
Higher variability in data (larger standard deviation) leads to wider confidence intervals, indicating less precision.
73
What is the difference between confidence intervals and standard deviations?
Confidence intervals apply to the population parameter, indicating where the true mean lies, while standard deviations describe the spread of sample data values.
74
What is a common misinterpretation of a 95% confidence interval?
A common misinterpretation is thinking that 95% of the sample values lie within the confidence interval, whereas it actually means we are 95% confident that the population parameter is within that range.
75
What does it mean to correctly reject the null hypothesis?
It means our study shows a statistically significant result because such an effect truly exists in the population (true positive).
76
What is the power of a study?
The power of a study is the probability of correctly rejecting the null hypothesis when it is false.
77
What does it mean to correctly fail to reject the null hypothesis?
It means our study shows no difference or effect associated with an intervention because no effect truly exists (true negative).
78
What is a Type 1 error in hypothesis testing?
A Type 1 error occurs when we incorrectly reject the null hypothesis even though it is true.
79
What is the probability of making a Type 1 error?
The probability of making a Type 1 error is called alpha (α), which is equal to the p-value threshold set. For example, if we set a p-value of less than 0.05 as significant, the risk of incorrectly rejecting the null hypothesis is less than 5 in 100.
80
What is a Type 2 error?
A Type 2 error occurs when our study shows no effect associated with an intervention, even though an effect truly exists in the population, leading us to incorrectly fail to reject the null hypothesis when it is false.
81
What is the probability of making a Type 2 error?
The probability of making a Type 2 error is denoted by beta (β).
82
What does the alpha level represent in hypothesis testing?
The alpha level represents the threshold for significance, the probability of rejecting the null hypothesis when it is true (Type 1 error). It is commonly set at 0.05.
83
What does the beta level represent in hypothesis testing?
The beta level represents the probability of failing to reject the null hypothesis when it is false (Type 2 error).
84
How is the alpha level related to the power of a study?
Lowering the alpha level reduces the probability of a Type 1 error but can also reduce the power of the study, making it harder to detect a true effect.
85
How is the beta level related to the power of a study?
The power of a study is 1 - beta. Reducing the beta level increases the power of the study, improving the ability to detect a true effect.
86
How does sample size affect the significance level and power of a study?
Larger sample sizes generally increase the power of a study and decrease the margin of error, making it easier to detect true effects and reduce the likelihood of Type 1 and Type 2 errors.
87
What is the power of a study?
The power of a study is the probability of finding an effect in a sample if such an effect truly exists in the population.
88
How do you critically appraise the power of a study?
To critically appraise the power of a study, assess if the study recruited enough participants to detect an effect, if one exists.
89
What is the formula for calculating study power?
Power = 1 - β, where β is the probability of making a Type 2 error.
90
How does study power relate to Type 2 error?
Lowering the probability of a Type 2 error (β) increases the power of the study
91
How can study power be increased?
Study power can be increased by:
- Increasing the sample size
- Increasing the effect size
- Reducing variability in the data
92
How does sample size affect study power?
A larger sample size increases the study power by reducing the margin of error and making it easier to detect a true effect.
93
How does effect size affect study power?
A larger effect size increases the study power by making it easier to detect a significant difference if one exists.
94
How does variability affect study power?
Lower variability in the data increases the study power by making it easier to detect a true effect.
95
Why is understanding study power important for research?
Understanding study power is important to ensure that a study is adequately designed to detect a true effect and avoid Type 2 errors.
96
What is Intention-to-Treat (ITT) analysis?
ITT analysis includes all participants who were randomized, regardless of the treatment they actually received, and analyzes them according to their original group assignment.
97
What are the advantages of ITT analysis?
ITT analysis preserves randomization, maintains sample size, eliminates bias, and provides information about the effectiveness of an intervention.
98
Why is ITT analysis considered the gold standard?
ITT analysis is considered the gold standard because it maintains the benefits of randomization, avoids biases that can occur with other methods, and provides a realistic estimate of the effectiveness of the treatment.
99
What is Per-Protocol (PP) analysis?
PP analysis includes only participants who completed or complied with the study protocol and analyzes them based on the treatment they actually received.
100
What are the advantages of PP analysis?
PP analysis provides information about the efficacy of the intervention, focusing on those who adhered to the protocol.
101
What are the disadvantages of PP analysis?
PP analysis can create selection bias, undermine randomization, and may overestimate the effect size by excluding non-compliant participants.
102
How do ITT and PP analyses differ in terms of bias and randomization?
ITT analysis minimizes bias and maintains randomization by including all randomized participants, while PP analysis may introduce selection bias and undermine the benefits of randomization by only including compliant participants.
103
What do ITT and PP analyses reveal about treatment?
ITT analysis provides insights into the effectiveness of an intervention in real-world conditions, while PP analysis offers information about the efficacy of an intervention under ideal conditions.
104
What is generalisability in research?
Generalisability is the applicability of research findings and conclusions from the sample population to the larger population.
105
How can you assess the generalisability of a study?
Assess generalisability by considering whether the study population is representative of the general population, examining how participants were selected, and checking for any potential biases in the selection process.
106
What factors might affect the generalisability of a study's findings?
Factors affecting generalisability include the selection criteria for participants, the characteristics of the sample compared to the broader population, and any biases introduced during the study.
107
What is clinical significance?
Clinical significance refers to a change in a patient’s life or disease status that is considered important or meaningful in a clinical context.
108
How does clinical significance differ from statistical significance?
Clinical significance focuses on the real-world importance of a treatment effect, while statistical significance refers to whether an effect is unlikely to have occurred by chance. An effect can be statistically significant but not clinically significant if it doesn’t lead to a meaningful change in patient outcomes.
109
Can you provide examples of clinical significance?
Examples of clinical significance include a medication that significantly improves a patient’s quality of life, a treatment that reduces symptoms to a degree that changes daily functioning, or an intervention that has a meaningful impact on disease progression.
110
How can you evaluate the clinical significance of study results?
Evaluate clinical significance by considering the magnitude of the effect, its impact on patient well-being, how it compares to existing treatments, and whether it leads to a meaningful improvement in health outcomes
111
Why is generalisability important in clinical practice?
Generalisability is important because it determines whether study findings can be applied to broader patient populations and influences the relevance and applicability of research results to everyday clinical settings.
