Biomedical Engineering Reference
In-Depth Information
substance such as saline, which will have no pharmacological or other effect); (c) the therapy most
commonly used at that time to combat the target disease/condition.
The size of the trial will be limited by a number of factors, including:
•
economic considerations (level of supporting fi nancial resources);
•
size of population with target condition;
•
size of population with target condition after additional trial criteria have been imposed (e.g.
specifi c age bracket, lack of complicating medical conditions, etc.);
•
size of eligible population willing to participate in the trial.
Whereas a comprehensive phase III trial would normally require at least several hundred pa-
tients, smaller trials would suffi ce if, for example:
•
the target disease is very serious/fatal;
•
there are no existing acceptable alternative treatments;
•
the target disease population is quite small;
•
the new drug is clearly effective and exhibits little toxicity.
Choosing the study population is obviously critical to adequate trial design. The specifi c criteria
of patient eligibility should be clearly predefi ned as part of the primary question the trial strives
to answer.
A number of trial design types may be used (Table 4.5), each having its own unique advantages
and disadvantages. The most scientifi cally pure is a randomized, double-blind trial (Box 4.3).
However, in many instances, alternative trial designs are chosen based on ethical or other grounds.
In most cases, two groups are considered: control and test. However, these designs can be adapted
to facilitate more complex subgrouping. Clinical trial design is a subject whose scope is too broad
to be undertaken in this text. The interested reader is referred to the 'Further reading' section at
the end of this chapter.
Table 4.5
Some clinical trial design types. Additional information may be
sourced from appropriate references provided in the 'Further reading' section
Randomized control studies (blinded or unblinded)
Historical control studies (unblinded)
Non-randomized concurrent studies (unblinded)
Cross-over trial design
Factorial design
Hybrid design
Large simple clinical trials
Search WWH ::
Custom Search