Meta-analysis Flashcards
Meta-analysis
GV Glass (1976): Primary, secondary, and meta-analysis of research. Edu Researcher 5:3-8.
→ “Meta-analysis refers to the analysis of analyses … The statistical analysis of a large collection of an analysis results from individual studies for the purpose of integrating findings. It connotes a rigorous alternative to the casual, narrative discussions of research studies which typify our attempts to make sense of the rapidly expanding literature …”
BMJ Book (2001)
→ “Statistical analysis of the results of independent studies, which generally aims to produce a single effect estimate”
How does meta-analysis work using published studies?
- Combine results from 2 or more studies
- Increase statistical power
- Provides a single numerical
value of overall treatment effect - Forest plots
Real world data (RWD)
Refers to data that is collected outside of the controlled environment of a clinical trial or experimental setting. This data is typically derived from various sources such as electronic health records (EHRs), claims databases, registries, wearable devices, and patient-reported outcomes.
Randomised control trials (RCT)
- a type of scientific study that is considered the gold standard for evaluating the effectiveness of medical interventions.
- In an RCT, participants are randomly assigned to receive either the treatment being studied (the experimental group) or a comparison or control group, which may receive a placebo, standard treatment, or no treatment.
RCTs strengths and limitations
Strengths:
- excellent internal validity
- provides precise measures of efficacy and acute toxicity of new therapies under ideal conditions
- because of randomisation, measurement of effect size is less prone to bias
- allows exploratory measures of secondary endpoints, including patient-reported outcomes and aspects of correlative biology
- can evaluate prognostic and predictive properties of new biomarkers and cancer therapies
-provide a mechanism whereby new (and potentially toxic) treatments can be carefully studied in centres of excellence
Limitations:
- limited external validity
- provide evidence of efficacy (drug effect under ideal circumstances), but not about effectiveness (i.e true benefit to patients in routine practice)
- Applicability to clinical practice can be limited
(i) because patients and practitioners in RCTs are different from those in routine practice
(ii) elderly and patients in comorbidity are under-represented in RCTs
(iii) often powered to detect a clinically modest effect size that may not apply to less selected patients
(iv) may use a surrogate primary endpoint that it not a valid measure of patient benefit
RWD strengths and limitations (aka population-based observational studies)
Strengths:
- good external validity
- provide insight delivery care in routine practice to all patients, including elderly and those with comorbidity
-provide info to guide future knowledge translation
-can provide evidence of effectiveness of new therapies in the general population
-large samples provide the opportunity to study rare disease for which RCTs are not possible
-can provide insight into short and long term toxicity in routine practice
-can address questions that have not and will not be evaluated in an RCT
Limitations:
- limited internal validity: may be difficult to separate effects of a new treatment from other factors
- population- level databases often do not include detail regarding comorbidity, performance status and specific treatment plan identification of comparative benefit in these studies is prone to multiple biases, including confounding by indication for a given treatment and/or concurrent changes in practice and/ or disease biology.
Process of meta-analysis example
1 research proposal and grant secured
2 one study coordinator employed to homogenise definitions of variables across five countries
3 methodological protocol finalise - including all ICD codes and read-codes for exposures, outcomes and other variables
4 Data extracted in all five countries using the specific codes and algorithms in the methodological protocol
5 Ethical issues - only results shared with study coordinator
6 Countries provide study coordinator with relevant information for the meta-analysis
Heterogeneity in Meta analytical model
Heterogeneity:
refers to the variability or diversity among the results of different studies included in a meta-analysis or systematic review.
In the context of the I² statistic, it quantifies the proportion of total variation across studies that is due to heterogeneity rather than chance.
A high I² value indicates substantial heterogeneity among the study results, suggesting that the true effect size may vary across studies.
Cox Proportional Hazards Model
Univariate analysis:
- each predictor variable is examined individually in relation to the outcome (in this case, cardiovascular disease or CVD), without considering other variables. Age is used as the timescale, which means that the analysis considers how the hazard of developing CVD changes over time as individuals age.
Multivariate analysis:
- multiple predictor variables are included simultaneously in the model to assess their independent effects on the outcome.
- The first multivariate analysis includes only the ‘history of CVD indicator’ variable, while the second multivariate analysis includes both the ‘history of CVD indicator’ and the ‘number of previous CVD events’ variables. These analyses allow researchers to determine whether these variables have significant associations with the risk of developing CVD when accounting for each other.
Stratified analysis:
- involves dividing the data into subgroups based on a particular characteristic (in this case, the ‘history of CVD indicator’) and then conducting separate analyses within each subgroup. This allows researchers to examine whether the relationship between predictor variables and the outcome differs across different levels of the stratifying variable.
Crude meta-analysis
Typically refers to a basic or preliminary analysis that aggregates the results of multiple studies without considering potential sources of variation or adjusting for confounding factors.
This is where we combine number of exposed vs unexposed (cases vs controls) in a meta-analysis.
Stratification in Meta-analytical model
Involves conducting separate meta-analyses within different age groups.
By stratifying the analysis by age, researchers can explore whether the effect of a certain intervention or exposure varies among different age groups.
This approach allows for a more nuanced understanding of how age influences the relationship between variables and the outcome of interest.
Adjusted meta-analysis
researchers take into account potential confounding factors or sources of variation that could influence the association between the exposure and outcome of interest. Unlike crude meta-analyses, which combine raw data from individual studies without adjusting for confounders, adjusted meta-analyses aim to provide more accurate estimates of the true effect size by controlling for these factors.
This is where we combine hazard ratios and confidence intervals from adjusted analyses (i.e. country-specific analyses in this case)
