Phase II design with sequential testing of hypotheses within each stage

Stavroula Poulopoulou, Dimitris Karlis, Constantin Yiannoutsos, Urania Dafni

Research output: Contribution to journalArticle

Abstract

The main goal of a Phase II clinical trial is to decide, whether a particular therapeutic regimen is effective enough to warrant further study. The hypothesis tested by Flemings Phase II design (Fleming, 1982) is H0: p ≤ p0 versus HA: p > p0, with level and α with a power β at p = pA, where p0 is chosen to represent the response probability achievable with standard treatment and pA is chosen such that the difference pA ? p0 represents a targeted improvement with the new treatment. This hypothesis creates a misinterpretation mainly among clinicians that rejection of the null hypothesis is tantamount to accepting the alternative, and vice versa. As mentioned by Storer (1992), this introduces ambiguity in the evaluation of type I and II errors and the choice of the appropriate decision at the end of the study. Instead of testing this hypothesis, an alternative class of designs is proposed in which two hypotheses are tested sequentially. The hypothesis H 01: p≤ p0 versus HA1: p > p0 is tested first. If this null hypothesis is rejected, the hypothesis H 02: p ≥ pA versus HA2: p < pA is tested next, in order to examine whether the therapy is effective enough to consider further testing in a Phase III study. For the derivation of the proposed design the exact binomial distribution is used to calculate the decision cut-points. The optimal design parameters are chosen, so as to minimize the average sample number (ASN) under specific upper bounds for error levels. The optimal values for the design were found using a simulated annealing method.

Original languageEnglish
Pages (from-to)768-784
Number of pages17
JournalJournal of Biopharmaceutical Statistics
Volume24
Issue number4
DOIs
StatePublished - Jul 4 2014

Fingerprint

Sequential Testing
Binomial Distribution
Phase II Clinical Trials
Null hypothesis
Therapeutics
Average Sample number
Type II error
Testing
Binomial distribution
Type I error
Alternatives
Exact Distribution
Rejection
Simulated Annealing
Clinical Trials
Therapy
Design
Upper bound
Minimise
Calculate

Keywords

  • Clinical trials
  • Multistage
  • Phase II trials
  • Sequential testing

ASJC Scopus subject areas

  • Pharmacology (medical)
  • Pharmacology
  • Statistics and Probability
  • Medicine(all)

Cite this

Phase II design with sequential testing of hypotheses within each stage. / Poulopoulou, Stavroula; Karlis, Dimitris; Yiannoutsos, Constantin; Dafni, Urania.

In: Journal of Biopharmaceutical Statistics, Vol. 24, No. 4, 04.07.2014, p. 768-784.

Research output: Contribution to journalArticle

Poulopoulou, Stavroula ; Karlis, Dimitris ; Yiannoutsos, Constantin ; Dafni, Urania. / Phase II design with sequential testing of hypotheses within each stage. In: Journal of Biopharmaceutical Statistics. 2014 ; Vol. 24, No. 4. pp. 768-784.
@article{d2612cc02d04456b9d44bcfe9af10d23,
title = "Phase II design with sequential testing of hypotheses within each stage",
abstract = "The main goal of a Phase II clinical trial is to decide, whether a particular therapeutic regimen is effective enough to warrant further study. The hypothesis tested by Flemings Phase II design (Fleming, 1982) is H0: p ≤ p0 versus HA: p > p0, with level and α with a power β at p = pA, where p0 is chosen to represent the response probability achievable with standard treatment and pA is chosen such that the difference pA ? p0 represents a targeted improvement with the new treatment. This hypothesis creates a misinterpretation mainly among clinicians that rejection of the null hypothesis is tantamount to accepting the alternative, and vice versa. As mentioned by Storer (1992), this introduces ambiguity in the evaluation of type I and II errors and the choice of the appropriate decision at the end of the study. Instead of testing this hypothesis, an alternative class of designs is proposed in which two hypotheses are tested sequentially. The hypothesis H 01: p≤ p0 versus HA1: p > p0 is tested first. If this null hypothesis is rejected, the hypothesis H 02: p ≥ pA versus HA2: p < pA is tested next, in order to examine whether the therapy is effective enough to consider further testing in a Phase III study. For the derivation of the proposed design the exact binomial distribution is used to calculate the decision cut-points. The optimal design parameters are chosen, so as to minimize the average sample number (ASN) under specific upper bounds for error levels. The optimal values for the design were found using a simulated annealing method.",
keywords = "Clinical trials, Multistage, Phase II trials, Sequential testing",
author = "Stavroula Poulopoulou and Dimitris Karlis and Constantin Yiannoutsos and Urania Dafni",
year = "2014",
month = "7",
day = "4",
doi = "10.1080/10543406.2014.900784",
language = "English",
volume = "24",
pages = "768--784",
journal = "Journal of Biopharmaceutical Statistics",
issn = "1054-3406",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

TY - JOUR

T1 - Phase II design with sequential testing of hypotheses within each stage

AU - Poulopoulou, Stavroula

AU - Karlis, Dimitris

AU - Yiannoutsos, Constantin

AU - Dafni, Urania

PY - 2014/7/4

Y1 - 2014/7/4

N2 - The main goal of a Phase II clinical trial is to decide, whether a particular therapeutic regimen is effective enough to warrant further study. The hypothesis tested by Flemings Phase II design (Fleming, 1982) is H0: p ≤ p0 versus HA: p > p0, with level and α with a power β at p = pA, where p0 is chosen to represent the response probability achievable with standard treatment and pA is chosen such that the difference pA ? p0 represents a targeted improvement with the new treatment. This hypothesis creates a misinterpretation mainly among clinicians that rejection of the null hypothesis is tantamount to accepting the alternative, and vice versa. As mentioned by Storer (1992), this introduces ambiguity in the evaluation of type I and II errors and the choice of the appropriate decision at the end of the study. Instead of testing this hypothesis, an alternative class of designs is proposed in which two hypotheses are tested sequentially. The hypothesis H 01: p≤ p0 versus HA1: p > p0 is tested first. If this null hypothesis is rejected, the hypothesis H 02: p ≥ pA versus HA2: p < pA is tested next, in order to examine whether the therapy is effective enough to consider further testing in a Phase III study. For the derivation of the proposed design the exact binomial distribution is used to calculate the decision cut-points. The optimal design parameters are chosen, so as to minimize the average sample number (ASN) under specific upper bounds for error levels. The optimal values for the design were found using a simulated annealing method.

AB - The main goal of a Phase II clinical trial is to decide, whether a particular therapeutic regimen is effective enough to warrant further study. The hypothesis tested by Flemings Phase II design (Fleming, 1982) is H0: p ≤ p0 versus HA: p > p0, with level and α with a power β at p = pA, where p0 is chosen to represent the response probability achievable with standard treatment and pA is chosen such that the difference pA ? p0 represents a targeted improvement with the new treatment. This hypothesis creates a misinterpretation mainly among clinicians that rejection of the null hypothesis is tantamount to accepting the alternative, and vice versa. As mentioned by Storer (1992), this introduces ambiguity in the evaluation of type I and II errors and the choice of the appropriate decision at the end of the study. Instead of testing this hypothesis, an alternative class of designs is proposed in which two hypotheses are tested sequentially. The hypothesis H 01: p≤ p0 versus HA1: p > p0 is tested first. If this null hypothesis is rejected, the hypothesis H 02: p ≥ pA versus HA2: p < pA is tested next, in order to examine whether the therapy is effective enough to consider further testing in a Phase III study. For the derivation of the proposed design the exact binomial distribution is used to calculate the decision cut-points. The optimal design parameters are chosen, so as to minimize the average sample number (ASN) under specific upper bounds for error levels. The optimal values for the design were found using a simulated annealing method.

KW - Clinical trials

KW - Multistage

KW - Phase II trials

KW - Sequential testing

UR - http://www.scopus.com/inward/record.url?scp=84901374240&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901374240&partnerID=8YFLogxK

U2 - 10.1080/10543406.2014.900784

DO - 10.1080/10543406.2014.900784

M3 - Article

C2 - 24697358

AN - SCOPUS:84901374240

VL - 24

SP - 768

EP - 784

JO - Journal of Biopharmaceutical Statistics

JF - Journal of Biopharmaceutical Statistics

SN - 1054-3406

IS - 4

ER -