DoE (Dept. of Pharmaceutics) PDF | PPT

Save (0)




Dr. V V S Narayana Reddy Karri

Dept. of Pharmaceutics
JSS College of Pharmacy, Ooty

(Constituent College)
JSS academy of higher education and research, Mysuru



Before knowing we should know
first about:

Null hypothesis,
P-Value and ANOVA




µ=0 (Accept null hypothesis)
The hypothesis that there is no significant (<95%) difference between specified populations




µ≠0 (Reject null hypothesis, Accept null hypothesis)
The hypothesis that there is significant (>95%) difference between specified populations


Significant >
95% 5%

95/100=0.95 05/100=0.05
p < 0.05

P Value


Analysis of Variance
(ANOVA): A mathematical
process for separating the
variability of a group of
observations into assignable
causes and setting up various
significance tests.



• At the beginning of the twentieth century, Sir
Ronald Fisher introduced the concept of applying
statistical analysis during the planning stages of
research rather than at the end of experimentation.

• When statistical thinking is applied from the design
phase, it enables to build quality into the product.



• The pharmaceutical industry was late in adopting these
paradigms, compared to other sectors.

• It heavily focused on blockbuster drugs, while
formulation development was mainly performed by One
Factor At a Time (OFAT) studies, rather than
implementing Quality by Design (QbD) and modern
engineering-based manufacturing methodologies.

• E.g., OFAT (Effect of disintegrant on tablet
disintegration time)


DoE exhibits several advantages over OFAT
 Maximize process knowledge, with the minimum use of resources.
 Provide accurate information, in the most efficient way possible.
 Identify factor interactions.
 Characterize the relative significance of each factor.
 Allow for the prediction of the process behavior within the design

 Establish a solid cause and effect relationship between CPPs and

 Allow for multiple response optimization (unlikely OFAT, one

way and two way ANOVA). As pharmaceutical products exhibit
several CQAs, the latter require simultaneous optimization.

 Make the product or process more robust.


DoE will never give an IDEAL formula/batch.

It will only provide a design space (predicted
area) to work with respect to variables.




• Factors/Cause (Independent variable)
– Also called control factors which we can control and may (interaction)

or may not depend on each other factor.
– Input variables that can be changed (excipients and concentrations)

• Response/Effect (Dependent variable)
– Measured output value (Results)
– Depends on independent variables.

• Levels
– Specific values of factors (inputs) (2 or more)
– These high and low levels can be generically coded as +1 and -1.
– For example, a 2 factor experiment will require 4 experimental runs:

• Center Points
– Points at the center value of all factor ranges.

• Replication
– Completely re-run experiment with same input levels
– Used to determine impact of measurement error.
– Replication allows an estimate of the random error independent of any

lack of fit error.


• Randomization
– A schedule for allocating treatment material and for conducting treatment

combinations in a DOE such that the conditions in one run neither depend on the
conditions of the previous run nor predict the conditions in the subsequent runs.

• Uncontrolled Factors/Variables
– Which induce variation under normal operating conditions are referred to as

“Noise Factors”. These factors, such as multiple machines, multiple shifts, raw
materials, humidity, etc., can be built into the experiment so that their variation
doesn’t get lumped into the unexplained, or experiment error.

– A key strength of Designed Experiments is the ability to determine factors and
settings that minimize the effects of the uncontrollable factors.

• Blocking
– We often need to eliminate the influence of extraneous factors when

running an experiment. We do this by “blocking”.
– Typically, a blocking factor is a source of variability that is not of

primary interest to the experimenter.

• Interaction

– Effect of one input factor depends on level of another input factor


Liner (Non-interaction) Vs.
Quadratic (Interactions)


Quality by Design
• Quality by Design (QbD) is defined as a systematic approach to

development that begins with predefined objectives and
emphasizes product and process understanding and process control,
based on sound science and quality risk management.

• Quality by Design was actually suggested by the regulatory
authorities (FDA, EMA) at the beginning of the new millennium,
recognizing that “quality cannot be tested into products, i.e. quality
should be built in by design”.



QBD Example


Quality Target Product Profile (QTPP)
• The QTPP is an essential element of a QbD

approach and forms the basis of design for the
development of the product

• It describes the design criteria for the product, and
should therefore form the basis for development of
the CQAs, CPPs, and Control strategy

• The QTPP provides an understanding of what will
ensure the quality, safety, and efficacy of a specific
product for the patient and is a starting point for
identifying the CQAs.


Quality Target Product Profile (QTPP)
Considerations for QTPP include:
• Intended use in clinical setting, route of administration, dosage form, delivery

• Dosage strength(s)
• Container closure system
• Therapeutic moiety release or delivery and attributes affecting pharmacokinetic

characteristics (e.g., dissolution, aerodynamic performance) appropriate to the drug
product dosage form being developed;

• Drug product quality criteria (e.g., sterility, purity, stability and drug release)
appropriate for the intended marketed product.


QTPP Example
• Particle size is critical to the dissolution of a

solid oral product, then QTPP should include
dissolution (product profile), but not particle

• Particle size will be a critical material attribute



QTPP Examples


General discriptions

2Cosmetic Appearance
3Rigidity Interface

Chemical analysis
5Assay of doxycycline
6Assay of NO
7Assay of curcumin
9CU of NO

10CU of curcumin
Pharmaceutical characterization

12Buffering capacity
15Adhesion strength
16Gel strength

In vitro release testing using synthetic
In vitro release testing using skin
In vitro tests

22Microbial efficacy studies
23Cell viability/Toxicity
24Anti-inflamatory study

In vivo testing
Tissue regeneration
Ulcer area measuremnt


Critical Quality Attributes (CQAs)
• Any attribute of dosage form which relates to safety and efficacy.
• A CQA is a physical, chemical, biological, or microbiological

property or characteristic that should be within an appropriate
limit, range, or distribution to ensure the desired product quality.
CQAs are generally associated with the drug substance, excipients,
intermediates (in-process materials) and drug product.

• CQAs of solid oral dosage forms are typically those aspects
affecting product purity, strength, drug release and stability. CQAs
for other delivery systems can additionally include more product
specific aspects, such as aerodynamic properties for inhaled
products, sterility for parenterals, and adhesion properties for
transdermal patches.

• For drug substances, raw materials and intermediates, the CQAs
can additionally include those properties (e.g., particle size
dTiasbtlreit bfruiatbiiolinty, abnud ldkisdsoeluntsiointy) that affect drug product CQAs.


MR Tablet


Critical Process Parameter (CPP):
• A process parameter whose variability has an

impact on a critical quality attribute (CQAs)
and therefore should be monitored or
controlled to ensure the process produces the
desired quality.


• The most influential factors are defined in ICH Q8 as
Critical Process Parameters (CPPs), i.e. process parameters
whose variability has an impact on a critical quality
attribute of the product.

• Therefore, they should be monitored and controlled to
ensure that the process produces the desired quality. The
latter is the sum of the product’s characteristics, which
should consistently comply with predefined ranges, widely
known as specifications. Within this context, ICH Q8
defines Critical Quality Attributes (CQAs) as physical,
chemical, biological, or microbiological properties or
characteristics that should be within an appropriate limit,
i.e. range or distribution to ensure the desired product


Over view of attributes

(Excipients) (Procedure) (Responses, (Safety and

Dissolution) efficacy)


Design of Experiments


How much DoE is important.?
A more comprehensive review on Scopus revealed

more than 5200 results for the keywords
“Design of Experiments” and “pharmaceutical”

since 1978.
It is worth noting that a steadily increasing trend is

observed, with only two documents in 1978,
10 in 1990,
29 in 2000,
24 in 2005,

281 in 2010, and
more than 800 in 2016.


• “Design of Experiments” (DoE) is a structured, organized method

for determining the relationships between factors affecting a
process and the output of that process.

• In other words, achieving process knowledge, through the
establishment of mathematical relationships between process inputs
and its outputs.

• Design of Experiments (DoE) is the main component of the
statistical toolbox to deploy (Spread/distribute) Quality by Design
in both research and industrial settings.


• The impact of the controlled factors on the quality

characteristics of the final product varies according to a
general pattern known as the Pareto principle, indicating
that a relatively small number of factors are responsible for
the substantial percentage of the effect. This is also known as
the 20:80 rule, stating that 20% of the causes (factors) are
responsible for 80% of the results (responses).


• DoE is an approach where the controlled input factors of the
process are systematically and purposefully varied in order to
determine their effects on the responses.

• The overall scope is the connection of the CPPs (x1, x2,.xi)
with the CQAs (yi) through mathematical functions y¼f(x),
polynomial equation.

• Such relationships enable the determination of the most
influential factors (CPPs) and identification of optimum
factor settings leading to enhanced product performance and
assuring CQAs.


Execution of an experimental
design (Seven distinct phases)


Execution of an experimental design..
1. Setting solid objectives: during this step the Quality

Target Product Profile (QTPP) should be clearly
defined, using the knowledge base deriving from
scientific literature and technical experience, critically
evaluated under the Quality Risk Management approach
(Ex: Desired dissolution and disintegration time).

2. Selection of process variables (factors) and responses
(CQAs): This choice is justified based on the DoE
objective and should assure that the whole design is
executable and appropriate for satisfying its scope. This
means that factor types either numeric (Temperature) or
categorical (color) and their levels, i.e. their values
within the design, are selected accordingly.


Execution of an experimental design..
3. Selection of an experimental design: based on

DoE objectives, screening, characterization or
optimization of process and formulation,
several types of designs can be chosen, e.g.
factorial or Response. Decisions on
randomization, blocking, and replications
should be also addressed.

4. Execution of the design: The generated design
matrix should be accurately executed, assuring
that parameters not included in the design are
identified and kept constant.


Execution of an experimental design..
5. Checking that the data are consistent with the

experimental assumptions: Related to our hypothesis and
scientific logics.

6. Analyzing the results: ANOVA and graphical tools are very
helpful in identifying main effects of the significant factors
and interactions thereof. The results are evaluated not only on
a statistical basis but also versus the established theories and
practically lead to process knowledge.

7. Use and interpretation of the results: at this step, the
evaluation of the DoE outcome can support scientifically
sound decision making on following actions, such as
executing confirmatory runs, augmenting designs, or further
proceeding with scale up and technology transfer activities.


Types of DoE
Types of DoE

Factorial Response

Box behnken Central
(BB) composite



Factorial designs

• They refer to parameters that can be adjusted
independently of each other, such as
compaction force, temperature, and spraying
rate. In this case, the responses are functions of
factor levels as described in Equation


Factorial designs
• Factorial designs are mainly used for screening

of factors.
• Factorial design screens unnecessary factors

to include in BB and CCD, since they have
more runs and repetabilities.

• Process characterization focuses
on discriminating the critical
process factors from the useful
many, using screening designs,
which typically include two
level factorial designs, either full
or fractional.


Response surface designs
• Once screening is completed, the selected significant factors are

further studied using more comprehensive designs aiming at
process optimization, which refers to setting the most influential
factors at levels that enhance all product CQAs simultaneously.

• Such designs typically include at least three factor levels and can
support quadratic or higher order effects. Response Surface
Methodology (RSM) techniques constitute the main tool for such

• More repeats are done to get experimental error.
• These designs are most effective when there are less than 5

• Quadratic models are used for response surface designs and at

least three levels of every factor are needed in the design.


• Four corners of the square represent the factorial

(+/- 1) design points
• Four star points represent the axial (+/- alpha)

design points
• Replicated center point (Usually 06)
What is the effect of Alpha (distance from center point)


CCD Concept


Calculating α Value (Coded)..


Calculating α Value (Actual value)



Box-Behnken Designs (BBD)
• They are very useful in the same setting as the central

composite designs.
• Their primary advantage is in addressing the issue of where

the experimental boundaries should be, and in particular to
avoid treatment combinations that are extreme.

• By extreme, we are thinking of the corner points and the
star points, which are extreme points in terms of region in
which we are doing our experiment.

• The Box-Behnken design avoids all the corner points, and
the star points.

• One way to think about this is that in the central composite
design we have a ball where all of the corner points lie on
the surface of the ball. In the Box-Behnken design the ball
is now located inside the box defined by a ‘wire frame’ that
is composed of the edges of the box.


Box-Behnken Designs (BBD)


Comparison of factorial, BBD and CCD
(For 3 Factors)

Run/Exp Starch HPMC Avicel Starch HPMC Avicel Starch HPMC Avicel
1 + + + + + + – – 0
2 + + – + + – + – 0
3 + – – + – – – + 0
4 – – + – – + + + 0
5 – + + – + + – 0 –
6 – + – – + – + 0 –
7 + – + + – + – 0 +
8 – – –
9 – – – + 0 +
10 0 0 0 0 – –
11 0 0 0 0 + –
12 0 0 0 0 – +
13 0 0 0 0 + +
14 0 0 0 0 0 0
15 0 0 0 0 0 0
16 -α 0 0 0 0 0
17 +α 0 0 0 0 0
18 0 -α 0 0 0 0
19 0 +α 0
20 0 0 -α

0 0 +α

Factorial CCD Box behnken


Types of graphs in DoE



Contour plots
• Contour: Outline or boundary line/shape
• Contour plots display the 3-dimensional

relationship in two dimensions, with x- and y-
factors (predictors) plotted on the x- and y-scales
and response values represented by contours.

• The darker regions identify higher z (response)

Effect of lipid and surfactant concentrations
on entrapment efficiency

Lipid Conc (mg)
Note: The body and line colors can be changed

Surfactant conc (mg)


3D Plots
• 3D surface and 3D wireframe plots are graphs that you can use to

explore the potential relationship between two variables. The
predictor variables are displayed on the x- and y-scales, and the
response (z) variable is represented by a smooth surface (3D
surface plot) or a grid (3D wireframe plot).

3D surface plots can provide a clearer concept of the response
surface than contour plots.


Overlaid contour plot
• Used to visually identify an area where the

predicted means of one or more response
variables are in an acceptable range.

• Used to visually identify the feasible variables
for multiple responses for a model.

• The plot emphasizes the region (if any) where
all responses are within their bounds. This is
known as the feasible region.


Polynomial equation
In mathematics, a polynomial is an expression consisting of variables (also
called indeterminates) and coefficients, that involves only the operations of
addition, subtraction, multiplication, and non-negative integer exponents of


Polynomial equation


Slope Intercep


F, ANOVA relation with
polynomial equation

p value


Model Summary
• Suitable mathematical models of the mixture

design such as linear, quadratic and special
cubic models were analyzed by the software.

• Significance of the model was determined by
comparisons of statistical parameters such as
SD, R2, adjusted R2, predicted R2 and
predicted residual error sum of squares
(PRESS). The best model was decided on the
basis of higher values of adjusted R2 and
predicted R2.


Model Summary/fit summary


predicted residual error sum of
squares (PRESS) Value

• PRESS value should be small for the best

• PRESS demonstrates the excellence of model
fitting. The optimized formulation was
prepared for further evaluation.


Power transformations
In statistics, a power transform is a family of functions that are applied to
create a monotonic transformation of data using power functions and power
parameter λ.



Model selection

Near value to 1


Model significance and lack of fit

FIT Lack of FIT
(Significant) (Non-Significant)


Experimental vs. predicted
• Exp indicates experimental values.

Experimental results of first to last runs (total
runs) as mean ± S.D. (n = 3).

• Pred represents predicted values predicted
through Equations of respective polynomial



Practical DoE


Solid lipid nanoparticles (SLNs) formulation of

diallyl disulfide drug
A solid lipid nanoparticle is typically spherical with an average
diameter between 10 and 1000 nanometers. Solid lipid
nanoparticles possess a solid lipid core matrix that can
solubilize lipophilic molecules. The lipid core is stabilized by
surfactants (emulsifiers)

Solid lipid
Vehicle (water, ethanol)


Method: solvent diffusion method
1. Palmitic acid and drug were dissolved

completely in ethanol in a water bath at 70 C.
2. The resultant organic solution was quickly
3. dispersed into 50 mL of an aqueous phase

containing surfactant under continuous
mechanical agitation at 400 rpm in a water bath
at 70 C for 5 min.

4. The obtained pre-emulsion (melted lipid droplet)
was subsequently transferred into an ice bath to
solidify the lipid droplets then cooled to room
temperature till SLN dispersion was obtained.


Diagrammatic representation


17-run, 3-factor, 3-level Box–

design was employed to optimize
the SLN formulation

Factors Responses
Surfactant (X1), Particle size
amount of lipid (X2) Entrapment efficiency
volume of solvent (X3)





• Dr. K. Gowthamarajan and Dr. Rajkumar

• The Principal and Staff, Krupanidhi college of

• The organizing team