Quantitative Research Design Flashcards
State 5 types of biases
Detection, reporting, attrition, performance, selection
Explain detection bias and provide ONE strategy to avoid such bias
Systematic differences between groups in how outcomes are determined (e.g. confirmation bias).
Blinding investigators may reduce the risk that knowledge of which intervention was received affects outcome measurement.
Explain performance bias and provide ONE strategy to avoid such bias
Systematic differences between groups in the care provided, or in exposure to confounding factors.
Matching both groups demographically and ensuring both groups receive equal treatment apart from the actual intervention.
Explain selection bias and provide ONE strategy to avoid such bias
Systematic differences between baseline characteristics of groups that are compared. E.g. self- selection / self-exclusion / sampling / clinician bias
Randomisation helps reduce sampling bias; clinician referrals can help reduce self-selection and ensure you have rigorous exclusion and inclusion categories.
Explain attrition bias and provide ONE strategy to avoid such bias
Systematic differences between groups in withdrawals from a study. E.g. adverse events may cause more people in the treatment group to withdraw than in the placebo group.
Utilise correct statistical strategies for missing data points.
Explain reporting bias and provide ONE strategy to avoid such bias
Systematic differences between reported and unreported findings. E.g. statistically significant differences are more likely to be published than non-significant ones.
Need a culture-shift towards reporting null results!
What questions would you ask when critically appraising scientific literature
Was the assignment to groups randomised? How?
Were the group’s demographics matched?
Were the participants masked? The researchers? Was the masking measured?
Were the groups treated the same apart from the intervention?
How large and precise was the treatment effect?
Were adverse events (AEs) considered?
Was patient safety a priority?
Sample size calculations?
Explain the characteristics of a case report study, and discuss the advantages and disadvantages
Case report studies are typically qualitative and are in-depth studies on a single individual or group. They are generally used in specific scenarios of abnormal behaviour or development, unexplained adverse events, emerging disease, or new treatments.
Advantages - quick, cheap, can detect novelties, help generate hypotheses, provide detailed information
Disadvantages - cannot establish causality, over-interpret results (applying happenings of one person to a population, non-randomised, usually a moment of time - no follow up
What needs to be considered when critically appraising a case report?
Was there clear reporting of demographics and clinical information?
Were the objectives clearly outlined?
Were the outcomes clearly reported?
Were appropriate diagnostic tests performed?
Were valid methods used to determine who should be included in a series?
* clear inclusion criteria?
Was it made clear that conclusions about relationships between variables
Explain the characteristics of a cross-sectional study, and discuss the advantages and disadvantages
Cross-sectional studies look at a population at a single point in time to demonstrate the prevalence or distribution of variables across that population. They are typically used to find the prevalence of disease or adverse events of treatment.
Advantages - quick, cheap, examine a large number of variables simultaneously, find an association between variables, easy recruitment, investigates harmful situations without breaking ethics
Disadvantages - cannot establish causality, selection bias, may rely on patient recall, limited ability to predict future trends
What needs to be considered when critically appraising a cross-sectional study?
o Were the recruitment methods valid?
o What was the response rate?
o Was the sample size justified?
o Were the variables measured appropriately?
o Were the data collection methods reliable?
o Were confounding variables appropriately adjusted for?
o Were all the results presented?
Explain the characteristics of a case-control study, and discuss the advantages and disadvantages.
A case-control study is a retrospective study that compares a group with certain outcomes to a control group to find risk factors and the rate of exposure to such risk factors.
Advantages - quick, cheap, examines multiple exposures in the past, can utilise exiting medical records for efficiency, useful for rare disease
Disadvantages - cannot establish causality, recall bias, finding matching control group is difficult, only assesses one outcome/disease
What needs to be considered when critically appraising a case-control study?
o Case recruitment – clearly defined inclusion criteria?
o Were all pertinent cases included?
o Were the controls matched / representative of the demographic?
o Was there a confounding factor about the controls?
o How accurate were the measurements of the exposure?
o Were they reliable/valid? Prescriptions? Dispensing? Patient reports?
o Was the sample size adequate?
Explain the characteristics of a cohort study, and discuss the advantages and disadvantages.
A cohort study is a longitudinal study that samples a cohort, performing a cross-section at intervals through time. They can be prospective, recruiting a group of people without an outcome of interest, or retrospective, using data collected for other purposes.
Advantages - temporal, can pick up on rare events, provide rich data, relatively safe ethically.
Disadvantages - expensive, time-consuming, loss to follow-up, nature of being a trial may make the sample less like the population, not appropriate for some chronic diseases.
What needs to be considered when critically appraising a cohort study?
o Representative sample?
o Were exclusion and inclusion criteria appropriate?
o Have they defined the exposure and treatment well?
o Have all the confounds been outlined?
o Was there a sufficient follow-up to capture the outcome of interest?
o Were the data collection methods reliable and valid?
o Was there an appropriate adjustment for confounding factors?
o Were participants unable to move between exposure categories in their everyday life?
o Was the sample size adequate?
o Was there appropriate handling of missing data?
o Are the results generalizable to the rest of the population?
