ST 520 Statistical Principles of Clinical Trials - NCSU Statistics ...
ST 520 Statistical Principles of Clinical Trials - NCSU Statistics ...
ST 520 Statistical Principles of Clinical Trials - NCSU Statistics ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
CHAPTER 4 <strong>ST</strong> <strong>520</strong>, A. TSIATIS and D. Zhang<br />
Disadvantages<br />
• The major disadvantage is that the number <strong>of</strong> patients assigned to the different treatments<br />
are random. Therefore, the possibility exists <strong>of</strong> severe treatment imbalance.<br />
– leads to less efficiency<br />
– appears awkward and may lead to loss <strong>of</strong> credibility in the results <strong>of</strong> the trial<br />
For example, with n = 20, an imbalance <strong>of</strong> 12:8 or worse can occur by chance with 50% probability<br />
even though π = .5. The problem is not as severe with larger samples. For instance if n = 100,<br />
then a 60:40 split or worse will occur by chance with 5% probability.<br />
4.2.2 Permuted block randomization<br />
One way to address the problem <strong>of</strong> imbalance is to use blocked randomization or, more<br />
precisely, permuted block randomization.<br />
Before continuing, we must keep in mind that patients enter sequentially over time as they become<br />
available and eligible for a study. This is referred to as staggered entry. Also, we must realize that<br />
even with the best intentions to recruit a certain fixed number <strong>of</strong> patients, the actual number<br />
that end up in a clinical trial may deviate from the intended sample size. With these constraints<br />
in mind, the permuted block design is <strong>of</strong>ten used to achieve balance in treatment assignment. In<br />
such a design, as patients enter the study, we define a block consisting <strong>of</strong> a certain number and<br />
proportion <strong>of</strong> treatment assignments. Within each block, the order <strong>of</strong> treatments is randomly<br />
permuted.<br />
For illustration, suppose we have two treatments, A and ⎛B,<br />
and we choose a block size <strong>of</strong> 4,<br />
⎜<br />
with two A’s and two B’s. For each block <strong>of</strong> 4 there are ⎝ 4<br />
⎞<br />
⎟<br />
⎠ or 6 possible combinations <strong>of</strong><br />
2<br />
treatment assignments.<br />
PAGE 57