Article
On the distinction between inclusion and exclusion criteria
Search Medline for
Authors
-
Reinhard Vonthein
- Universität zu Lübeck, Lübeck, Germany
-
Felix Matthews
- Universität zu Lübeck, Lübeck, Germany
| Published: | August 19, 2022 |
|---|
Outline
Text
Purpose: While the CONSORT statement [1] requires eligibility criteria and reasons for each of them, the distinction between inclusion and exclusion criteria may be unnecessary [2], or it could have several valid reasons: statistical, legal, ethical, organizational, linguistical or there could be no rule at all.
Methods: A random sample of clinical trials published in the New England Journal of Medicine 2018 was piloted, statistically planned for McNemar tests, and analyzed by generalized mixed models. Fitting rules for all selection criteria, the criteria’s length and number, and the proportion of randomized participants among the screened persons were extracted from all articles and supplemental trial protocols.
Results: The 62 randomly selected trials had 482 inclusion and 1073 exclusion criteria. All but 4 were compatible with the rule, that inclusion criteria define the population for generalization, while exclusion criteria are the compromises needed to make a trial feasible. Second and third most plausible rules were: no rule applies (90%), negations are avoided (82%). Proportions were similar when studies were considered as clusters. The possible rule that inclusion criteria are assessed before formal consent and exclusion criteria afterwards would be followed by commercial sponsors in another frequency than by academia. The more selection criteria there were, the less plausible was the possible rule that inclusion criteria are short and exclusion criteria are long.
Conclusion: While there are hints, that NEJM authors try to avoid negation and consider practicalities of screening for RCT, which produces some exemptions to the rule, the rule that should be taught in courses on regulatory affairs is: Inclusion criteria define the population of statistical inference, while exclusion criteria are the compromises needed to make the trial feasible legally, ethically, organizationally, but not statistically.
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- Moher D, Hopewell S, Schulz KF, Montori V, Goetzsche PC, Devereaux PJ, Elbourne D, Egger M, Altman DG. CONSORT 2010 Explanation and elaboration: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c869. DOI: 10.1136/bmj.c869
- 2.
- Fuks A, Weijer C, Freedman B, Shapiro S, Skrutkowska M, Riaz A. A study in contrasts: Eligibility criteria in a twenty-year sample of NSABP and POG clinical trials. J Clin Epidemiol. 1998;51:69-79.
