Abstract
Stopping rules—rules dictating when to stop accumulating data and start analyzing it for the purposes inferring from the experiment—divide Bayesians, Likelihoodists and classical statistical approaches to inference. Although the relationship between Bayesian philosophy of science and stopping rules can be complex (cf. Steel 2003), in general, Bayesians regard stopping rules as irrelevant to what inference should be drawn from the data. This position clashes with classical statistical accounts. For orthodox statistics, stopping rules do matter to what inference should be drawn from the data. “The dispute over stopping rule is far from being a marginal quibble, but is instead a striking illustration of the divergence of fundamental aims and standards separating Bayesians and advocates of orthodox statistical methods” (Steel 2004, 195).
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
The likelihood principle says that if P(x|θ) = cP(x′|θ), where c is some positive constant and x and x′ similar data from different experiments testing the same hypotheses H about θ, then both data have identical evidential import. E.g. see Mayo (1996).
- 2.
Steel is right that, although cumbersome for Bayesians, they can make stopping rules matter. The problem however is not so much one of possibility; instead, the key issue—as I see it—is whether or not such rules should matter for an account of statistical evidence. The fact that Bayesians can make them matter, does not imply they would. For RCTs, so I argue, the answer seems to me clear: not only stopping rules matter in practice but that they should.
- 3.
See Chapter 3, Section 3.2 for further details.
- 4.
This is the probability at the start of the trial of achieving a statistically significant result at a pre-specified significance level and a pre-specified alternative treatment size.
- 5.
Computed according to two-proportion z-test.
- 6.
Similar to the original trial where the TE event rates for experimental and standard treatments were both low, i.e. 5.5 and 5.3% respectively.
- 7.
See Chapter 3 (Section 3.2 Conditional Power for Futility) for details pp. 45–52.
- 8.
dd is the cost of having a patient switch treatments, after drug acceptance; while cc is the loss for assigning a new drug that is non-superior to a patient.
References
Ellenberg, S. 2003. Are all monitoring boundaries equally ethical? Controlled Clinical Trials 24: 585–588.
Ellenberg, S. et al. 2003. Data monitoring committees in clinical trials. New York: Wiley.
Jacobson, M.A. et al. 1994. Primary prophylaxis with pyrimethamine for toxoplasmic encephalitis in patients with advanced human immunodeficiency virus disease: Results of a randomized trial. Journal of Infectious Diseases 169: 384–394.
Mayo, D. 1996. Error and the growth of experimental knowledge. Chicago: University of Chicago Press.
Mayo, D., and M. Kruse. 2001. Principles of inference and their consequences. In Foundations of Bayesianism, eds. D. Corfield and J. Williamson, 381–403. Dordrecht: Kluwer Academic.
Montori, V.M. et al. 2005. Randomized trials stopped early for benefit: A systematic review. JAMA 294: 2203–2209.
Neaton, J. et al. 2006. Data monitoring experience in the AIDS toxoplasmic encephalitis study. In Data monitoring in clinical trials, eds. D. DeMets, C. Furberg, and L. Friedman, 320–329. New York: Springer.
Pocock, S.J. 1993. Statistical and ethical issues in monitoring clinical trials. Statistics in Medicine 12: 1459–1469.
Proschan, M. et al. 2006. Statistical monitoring of clinical trials. New York: Springer.
Steel, D. 2003. A Bayesian way to make stopping rules matter. Erkenntnis 58: 213–227.
Steel, D. 2004. The facts of the matter: A discussion of Norton’s material theory of induction. Philosophy of Science 72: 188–197.
Acknowledgements
I am grateful to Paul Bartha for his supervision, helpful discussion and feedback. I am also grateful to two anonymous reviewers for their comments and criticisms, and the audience at EPSA 2009 in Amsterdam. Earlier version of this work was presented at the PSX in the Center for Philosophy of Science at the University of Pittsburgh.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer Science+Business Media B.V.
About this paper
Cite this paper
Stanev, R. (2012). Stopping Rules and Data Monitoring in Clinical Trials. In: de Regt, H., Hartmann, S., Okasha, S. (eds) EPSA Philosophy of Science: Amsterdam 2009. The European Philosophy of Science Association Proceedings, vol 1. Springer, Dordrecht. https://doi.org/10.1007/978-94-007-2404-4_31
Download citation
DOI: https://doi.org/10.1007/978-94-007-2404-4_31
Published:
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-007-2403-7
Online ISBN: 978-94-007-2404-4
eBook Packages: Humanities, Social Sciences and LawPhilosophy and Religion (R0)