Independence and statistical inference in clinical trial designs: a tutorial review - PubMed (original) (raw)
Review
Independence and statistical inference in clinical trial designs: a tutorial review
S Bolton. J Clin Pharmacol. 1998 May.
Abstract
The requirements for statistical approaches to the design, analysis, and interpretation of experimental data are now accepted by the scientific community. This is of particular importance in medical studies where public health consequences are of concern. Investigators in the clinical sciences should be cognizant of statistical principles in general, but should always be wary of the pursuing their own analyses and engage statisticians for data analysis whenever possible. Examples of circumstances that require statistical evaluation not found in textbooks and not always obvious to the lay person are pervasive. Incorrect statistical evaluation and analyses in such situations will result in erroneous and potentially serious misleading interpretation of clinical data. Although a statistician may not be responsible for any misinterpretations in such unfortunate circumstances, the quote often cited about statisticians and "damned liars" may appear to be more truth than fable. This article is a tutorial review and describes a common misuse of clinical data resulting in an apparently large sample size derived from a small number of patients. This mistake is a consequence of ignoring the dependency of results, treating multiple observations from a single patient as independent observations.
Similar articles
- Invited review: Reproducible research from noisy data: Revisiting key statistical principles for the animal sciences.
Bello NM, Renter DG. Bello NM, et al. J Dairy Sci. 2018 Jul;101(7):5679-5701. doi: 10.3168/jds.2017-13978. Epub 2018 May 3. J Dairy Sci. 2018. PMID: 29729923 Review. - Data sharing and the evolving role of statisticians.
Manamley N, Mallett S, Sydes MR, Hollis S, Scrimgeour A, Burger HU, Urban HJ. Manamley N, et al. BMC Med Res Methodol. 2016 Jul 8;16 Suppl 1(Suppl 1):75. doi: 10.1186/s12874-016-0172-9. BMC Med Res Methodol. 2016. PMID: 27410483 Free PMC article. - Repeated Measures Designs and Analysis of Longitudinal Data: If at First You Do Not Succeed-Try, Try Again.
Schober P, Vetter TR. Schober P, et al. Anesth Analg. 2018 Aug;127(2):569-575. doi: 10.1213/ANE.0000000000003511. Anesth Analg. 2018. PMID: 29905618 Free PMC article. - Statistics in physiology and pharmacology: a slow and erratic learning curve.
Ludbrook J. Ludbrook J. Clin Exp Pharmacol Physiol. 2001 May-Jun;28(5-6):488-92. doi: 10.1046/j.1440-1681.2001.03474.x. Clin Exp Pharmacol Physiol. 2001. PMID: 11428385 Review. - Clustered data in sports research.
Hayen A. Hayen A. J Sci Med Sport. 2006 May;9(1-2):165-8. doi: 10.1016/j.jsams.2006.02.003. Epub 2006 Mar 30. J Sci Med Sport. 2006. PMID: 16574483
Cited by
- Donald Mainland: anatomist, educator, thinker, medical statistician, trialist, rheumatologist.
Altman D. Altman D. J R Soc Med. 2020 Jan;113(1):28-38. doi: 10.1177/0141076819896274. J R Soc Med. 2020. PMID: 31922445 Free PMC article. No abstract available. - CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials.
Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, Elbourne D, Egger M, Altman DG. Moher D, et al. BMJ. 2010 Mar 23;340:c869. doi: 10.1136/bmj.c869. BMJ. 2010. PMID: 20332511 Free PMC article. No abstract available.
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources