To use or not to use the odds ratio in epidemiologic analyses?

@article{Nurminen2005ToUO,
  title={To use or not to use the odds ratio in epidemiologic analyses?},
  author={Markku Mikael Nurminen},
  journal={European Journal of Epidemiology},
  year={2005},
  volume={11},
  pages={365-371}
}
This paper argues that the use of the odds ratio parameter in epidemiology needs to be considered with a view to the specific study design and the types of exposure and disease data at hand. Frequently, the odds ratio measure is being used instead of the risk ratio or the incidence-proportion ratio in cohort studies or as an estimate for the incidence-density ratio in case-referent studies. Therefore, the analyses of epidemiologic data have produced biased estimates and the presentation of… 
Statistical Issues in Estimation of Adjusted Risk Ratio in Prospective Studies.
TLDR
The potential misinterpretation of odds ratios should be considered by researchers when researchers want to estimate the effect of exposure or intervention by controlling potential covariates using regression models that can estimate risk ratios instead of logistic regression.
Prevalence odds ratio versus prevalence ratio: choice comes with consequences.
TLDR
This article provides an example of a cross-sectional study wherein prevalence ratio was chosen over (Prevalence) odds ratio and demonstrates the analytic implications of the choice of category to be modeled and choice of reference level for independent variables.
Prevalence odds ratio versus prevalence ratio: choice comes with consequences.
TLDR
This article provides an example of a cross-sectional study wherein prevalence ratio was chosen over (Prevalence) odds ratio and demonstrates the analytic implications of the choice of category to be modeled and choice of reference level for independent variables.
A practical guide for multivariate analysis of dichotomous outcomes.
TLDR
3 alternative multivariate statistical models to replace Logistic Regression for the analysis of data from cross-sectional and time-to-event studies are reviewed, viz, Modified Cox Proportional Hazard Regression Model, Log-Binomialregression Model and Poisson Regression model incorporating the Robust Sandwich Variance are reviewed.
Exact inference for the risk ratio with an imperfect diagnostic test
TLDR
Procedures that control for sensitivity and specificity of the diagnostic test are proposed, given the risks are measured by proportions, as it is in cross-sectional studies or studies with fixed follow-up times, for the true risk ratio.
Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio
TLDR
Cox or Poisson regression with robust variance and log-binomial regression provide correct estimates and are a better alternative for the analysis of cross-sectional studies with binary outcomes than logistic regression, since the prevalence ratio is more interpretable and easier to communicate to non-specialists than the odds ratio.
What is so odd about odds?
TLDR
It is shown that the OR is often presented and misinterpreted as equivalent to the RR, and using OR uncritically as an estimate of RR will strongly bias inferences about treatment effect or cause of disease by amplifying the apparent strength of an association between an exposure and an outcome.
Constraint approaches to the estimation of relative risk
TLDR
This article proposes an alternative constraint optimization approach for estimating relative risk that can reach similar or better performance than Bayesian approach in terms of bias, root mean square error, coverage rate, and predictive probabilities.
Detecting Potential Safety Issues in Large Clinical or Observational Trials by Bayesian Screening When Event Counts Arise from Poisson Distributions
  • A. Gould
  • Mathematics, Medicine
    Journal of biopharmaceutical statistics
  • 2013
TLDR
A Bayesian screening approach is described that does not test hypotheses, is self-adjusting for multiplicity, provides a direct assessment of the likelihood of no material drug–event association, and quantifies the strength of the observed association.
A general binomial regression model to estimate standardized risk differences from binary response data.
TLDR
A constrained maximum likelihood estimation algorithm is presented that ensures the feasibility of risk estimates of the LEXPIT model and procedures for defining the feasible region of the parameter space, judging convergence, and evaluating boundary cases are described.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 46 REFERENCES
What does the odds ratio estimate in a case-control study?
  • N. Pearce
  • Medicine
    International journal of epidemiology
  • 1993
TLDR
The use of the term 'odds ratio' in reporting the findings of case-control studies is technically correct, but is often misleading, and authors should be encouraged to not only specify the manner in which controls have been selected but also the corresponding effect measure which is being estimated by means of calculating the odds ratio in the subjects actually studied.
Estimators of relative risk for case-control studies.
TLDR
The authors present an approach to hypothesis testing for crude and stratified data from a case-exposure study and suggest such studies may be preferable to cohort studies and to cumulative-incidence case-control studies.
On the need for the rare disease assumption in case-control studies.
TLDR
The conditions under which matched and unmatched odds ratios are consistent estimators of the incidence-density ratio in case-control studies are examined and the odds ratio obtained under "incidence-density" sampling will in general provide a better approximation to the risk ratio.
Use of the prevalence ratio v the prevalence odds ratio as a measure of risk in cross sectional studies.
The letter by Lee and Chia (1993;50: 861-2) is a welcome discussion of the debatable but increasing use of the prevalence odds ratio as a quantified measure of association in cross sectional studies
Interpretation and choice of effect measures in epidemiologic analyses.
The concept of the odds ratio is now wellestablished in epidemiology, largely because it serves as a link between results obtainable from follow-up studies and those obtainable from case-control
Analysis of epidemiologic case-base studies for binary data.
TLDR
This paper modifies and advances a consistent likelihood-based procedure-analogous to Miettinen and Nurminen's proposal for a full cohort design--for interval estimation (and also point estimation and significance testing) in the context of binary case-base data and extends the analysis to encompass stratified data.
Estimability and estimation in case-referent studies.
  • O. Miettinen
  • Mathematics, Medicine
    American journal of epidemiology
  • 1976
The concepts that case-referent studies provide for the estimation of "relative risk" only if the illness is "rare", and that the rates and risks themselves are inestimable, are overly superficial
The rare-disease assumption revisited. A critique of "estimators of relative risk for case-control studies".
TLDR
One should be wary of methods of studying incidence that involve the use of prevalent cases and any etiologic study employing prevalent cases may be biased by such factors, as each case-control design has certain practical implications for selection and interviewing.
Risk ratio and rate ratio estimation in case-cohort designs: hypertension and cardiovascular mortality.
TLDR
New pseudo-likelihood methods are developed for the case-cohort design, which are a good, sometimes even advantageous alternative to the nested case-control design, in studying a disease that is not very rare.
Assessment of excess risks in case-base studies.
TLDR
In this paper, likelihood-based statistics are derived which can be used for interval estimation (and also point estimation and significance testing) of risk differences and the unified approach generalizes to stratified analysis.
...
1
2
3
4
5
...