**2.2 Study design**

The main object of the experimental design is to minimize the experimental variables and to avoid a bias [14]. *In vivo* bioavailability study is determined by taking into consideration of the following points:

1.The nature of the reference drug and the dosage form to be tested

2.Benefit risk ratio considerations in regard to testing in humans

### 3.The availability of analytical methods

#### 4.What is the scientific questions to be answered

Bioavailability studies are influenced by various factors such as age, sex, disease state, food habits, physical and mental health condition, body weight human volunteer, experimental design, time of administration, time of sampling, analytical method used and compartment model used in estimating pharmacokinetic parameters or bioavailability that contribute to the observed blood concentration time profile. Therefore, it is necessary to consider all these important factors in a study design.

The bioavailability study should be designed in such a way that the formulation effect can be distinguished from other effects. If two formulations are to be compared, a two-period, two-sequence crossover design is the design of choice which should ideally be equal to or more than five half-lives that have to be measured. Alternative study designs include the parallel design for very long half-life substances with highly variable disposition [15].

In the following sections, various factors are discussed keeping the bioequivalence study also in mind. However, they are valid for simple bioavailability studies also.

#### *2.2.1 Parallel design*

In a parallel design, two formulations are administered to two groups of volunteers. To avoid a bias, formulations may be administered randomly to the volunteers. The major disadvantage of this design is that the intersubject variation is not being corrected. It has been proved beyond doubt that most of the times intersubject variation is greater than the variation between any formulation. Therefore, a cross over design is preferred in bioavailability or bioequivalence trails to avoid influence of a intersubject variation. This design is used mainly for drug, and its metabolites have long elimination half-life. The carryover effects or dropouts were less in parallel studies compared to crossover studies.

#### *2.2.2 Crossover design*

As recommended by the USFDA [7], in most bioequivalence studies, a test drug is compared with the standard reference drug in a group of normal healthy subjects of age 18–55 years, each receives both the treatments alternately, in a crossover fashion (two-period, two-treatment crossover design), with the two phases of treatment separated by a washout period of generally a week's duration and it mainly depends on the half-life of the drug [16]. If elimination half-life of the drug increases, the washout period also increases. The drug formulation either test or reference is given to each human volunteer randomly but an equal number of subjects receives each treatment in each period, as given in **Table 1**. In case of two treatments, groups 1 and 2, one group receives the treatment in the order A and B, and the second group receives in the reverse order B and A. A similar allocation is done in case of a three-treatment crossover design (three-period, three-treatment crossover design). Intersubject variability is observed for several drugs in clearance. The intrasubject coefficient of variation (approximately 15%) is usually substantially smaller than that between subjects (approximately 30%), and therefore crossover designs are generally recommended for bioequivalence studies.

In crossover design, the treatments are compared on the same human subject, and the intersubject variability is reduced. Both the designs depend on the three fundamental statistical concepts of study design, and these are randomization, replication, and error control. Randomization means allocation of treatments to the

**69**

*Bioavailability and Bioequivalence Studies DOI: http://dx.doi.org/10.5772/intechopen.85145*

Two-way crossover

Three-way crossover

Four-way crossover

**Table 1.**

*Latin square design.*

subjects without bias. Replication involves the application of more than one experimental subject for reliable estimates than a single observation and also provides a more precise measurement of treatment effects. The number of replicates required mainly depends upon the degree of differences to be detected and inherent variability of the data. Commonly used cross over designs in bioavailability trails are Latin

1. 1, 2, 3, 4, 5, 6 A B C D 2. 7, 8, 9, 10, 11, 12 B D A C 3. 13, 14, 15, 16, 17, 18 C A D B 4. 19, 20, 21, 22, 23, 24 D C B A

**Group No. Subject in group Treatment for period No.**

1. 1, 2, 3, 4, 5, 6 A B 2. 7, 8, 9, 10, 11, 12 B A

1. 1, 2, 3, 4, 5, 6 A C B 2. 7, 8, 9, 10, 11, 12 B A C 3. 13, 14, 15, 16, 17, 18 C B A

I II

I II III

I II III IV

A standard approach for conducting a comparative bioavailability study to use a randomized, balanced, cross over design called Latin square or complete cross over design is as shown in **Table 1**. Incomplete block design (BIBD) eliminates many of the difficulties encountered with the Latin square design. In this, each subject receives not more than two formulations, each formulation is administered the same number of times and each pair of formulations occurs together in the same number of subjects. **Table 2** shows BIBD four formulations A, B, C, and D. In this design, as discussed above, each subject receives two formulations, each formulation is administered six times and each pair of formulations occurs together in two subjects (the

In a Latin square cross over design, each subject receives each formulation, and even in BIBD, each subject receives two formulations at different occasions. The time interval between the two treatments is called "washout period." Washout period is required for the elimination of the administered dose of a drug so as to avoid the carryover. For most of drugs in crossover design, at least 10 half-lives should be allowed between treatments. This should ensure an elimination of 99.9% of the administered dose and a maximum carryover of less than 0.1% from first treatment. The number of washout period is a function of the half-life and the dose of the drug administered. The number of washout periods in a study depends upon the type of crossover design used and the number of formulations to be evaluated.

square cross over design and balanced incomplete block design.

pairs are AB, AC, AD, BC, BD, and CD).

**2.3 Washout period**

*Bioavailability and Bioequivalence Studies DOI: http://dx.doi.org/10.5772/intechopen.85145*


**Table 1.**

*Pharmaceutical Formulation Design - Recent Practices*

3.The availability of analytical methods

stances with highly variable disposition [15].

less in parallel studies compared to crossover studies.

*2.2.1 Parallel design*

*2.2.2 Crossover design*

4.What is the scientific questions to be answered

Bioavailability studies are influenced by various factors such as age, sex, disease state, food habits, physical and mental health condition, body weight human volunteer, experimental design, time of administration, time of sampling, analytical method used and compartment model used in estimating pharmacokinetic parameters or bioavailability that contribute to the observed blood concentration time profile. Therefore, it is necessary to consider all these important factors in a study design.

The bioavailability study should be designed in such a way that the formulation effect can be distinguished from other effects. If two formulations are to be compared, a two-period, two-sequence crossover design is the design of choice which should ideally be equal to or more than five half-lives that have to be measured. Alternative study designs include the parallel design for very long half-life sub-

In the following sections, various factors are discussed keeping the bioequivalence

In a parallel design, two formulations are administered to two groups of volunteers. To avoid a bias, formulations may be administered randomly to the volunteers. The major disadvantage of this design is that the intersubject variation is not being corrected. It has been proved beyond doubt that most of the times intersubject variation is greater than the variation between any formulation. Therefore, a cross over design is preferred in bioavailability or bioequivalence trails to avoid influence of a intersubject variation. This design is used mainly for drug, and its metabolites have long elimination half-life. The carryover effects or dropouts were

As recommended by the USFDA [7], in most bioequivalence studies, a test drug is compared with the standard reference drug in a group of normal healthy subjects of age 18–55 years, each receives both the treatments alternately, in a crossover fashion (two-period, two-treatment crossover design), with the two phases of treatment separated by a washout period of generally a week's duration and it mainly depends on the half-life of the drug [16]. If elimination half-life of the drug increases, the washout period also increases. The drug formulation either test or reference is given to each human volunteer randomly but an equal number of subjects receives each treatment in each period, as given in **Table 1**. In case of two treatments, groups 1 and 2, one group receives the treatment in the order A and B, and the second group receives in the reverse order B and A. A similar allocation is done in case of a three-treatment crossover design (three-period, three-treatment crossover design). Intersubject variability is observed for several drugs in clearance. The intrasubject coefficient of variation (approximately 15%) is usually substantially smaller than that between subjects (approximately 30%), and therefore crossover designs are generally recommended for bioequivalence studies.

In crossover design, the treatments are compared on the same human subject, and the intersubject variability is reduced. Both the designs depend on the three fundamental statistical concepts of study design, and these are randomization, replication, and error control. Randomization means allocation of treatments to the

study also in mind. However, they are valid for simple bioavailability studies also.

**68**

*Latin square design.*

subjects without bias. Replication involves the application of more than one experimental subject for reliable estimates than a single observation and also provides a more precise measurement of treatment effects. The number of replicates required mainly depends upon the degree of differences to be detected and inherent variability of the data. Commonly used cross over designs in bioavailability trails are Latin square cross over design and balanced incomplete block design.

A standard approach for conducting a comparative bioavailability study to use a randomized, balanced, cross over design called Latin square or complete cross over design is as shown in **Table 1**. Incomplete block design (BIBD) eliminates many of the difficulties encountered with the Latin square design. In this, each subject receives not more than two formulations, each formulation is administered the same number of times and each pair of formulations occurs together in the same number of subjects. **Table 2** shows BIBD four formulations A, B, C, and D. In this design, as discussed above, each subject receives two formulations, each formulation is administered six times and each pair of formulations occurs together in two subjects (the pairs are AB, AC, AD, BC, BD, and CD).
