Guide to Blend Uniformity (IVP)

1. 46 Journal of GXP Compliance
A Guide to Blend Uniformity
By Trudy Yin
OVERVIEW OF BLENDING
PROCESS AND EQUIPMENT
The process of blending is commonly defined in the
pharmaceutical industry as: the mixing of various bulk
powder ingredients (e.g., drugs, excipients), targeted for
a homogeneity and uniform distribution of the mixture.
Each pharmaceutical product has its unique blend of
raw materials; therefore, a careful selection of manufac-
turing process equipment and methods is inevitable.
The quality of the resulting blend depends highly on the
types of blenders or mixers used; the “flowability” (flow
behavior) of powder during the blend cycle is factored
into the finished product quality. The mechanics of
blending, which involves the principles of shear, convec-
tion, and diffusion (or dispersion), is achieved by the
use of industrial blenders or mixers. The equipment
employed to accomplish the process of blending can be
categorized as either rotating or fixed shell blenders.
Rotating shell blenders (i.e., drum, cross-flow, double
cone, and twin-shell) accomplish the mixing process by
rotating the blender shell around a fixed axis and by
relying upon a sliding or rolling motion of the powder. To
aid in this mixing process, many systems will utilize
internal baffles. Fixed shell blenders (i.e., ribbon, screw,
and impeller mixers), on the other hand, rotate internal
blender parts, such as an impeller or paddle, and pro-
duce a continuous cutting and shuffling motion. Shear-
ing force that breaks apart large conglomerates of pow-
der is, therefore, developed by this kind of cutting and
shuffling motion. Pony, planetary, and high shear mixers
are commonly used in conjunction with blenders.
There are a number of process and product vari-
ables and characteristics that influence the quality of
ABSTRACT
The powder blending process has been historically
identified as a challenging operation in Oral Solid
Dosage (OSD) form manufacturing. Insufficient
blending results in poor active ingredient mixing with
excipients. Excessive blending could adversely
impact the distribution of drug content (content
uniformity) in the final product. Companies suffer
financial setbacks due to rework of poor quality
product, and more severely, legal action being taken
because of non-compliance. With harmonizing
efforts, government agencies, such as the U.S. Food
and Drug Administration (FDA), join with industry
experts and scientists to develop corresponding
principles and guidelines in order to assist drug
manufacturers. According to the FDA's current Good
Manufacturing Practice (cGMP) Code of Federal
Regulations (CFR), Title 21, Part 211-Current Good
Manufacturing Practice for Finished Pharmaceuticals
standard; particularly section 211.110, Sampling and
testing of in-process materials and drug products
of Subpart F states, written procedures shall be
established and followed for “adequacy of mixing
to assure uniformity and homogeneity…” In the
following blend uniformity discussion, we will
examine this rule and demonstrate how guidelines
emerge, how they are interpreted, and their practical
adoption to current industrial methodologies and/or
technologies.

2. 47October 2007 • Volume 12, Number 1
Trudy Yin
mixing and its final blend. Incorrect process parame-
ters, for instance, when discharging the powder with
improper flow rate could create over-flooding or
unsteady flow issues, ultimately affecting the final
blend uniformity. Inadequate mixing time and speed
could also result in poorly blended mixture, clump for-
mation, and segregation. Product characteristics
include particle size, density, morphology, shape, etc.,
of the powder ingredients. Large coarse particles with
higher molecular weight will, for example, settle to the
bottom of the blender vessel; whereas the smaller finer
particles will rise to the top. Particle density affects the
sedimentation or floating ability, for which uniform dis-
tribution of the powder mixture could be warranted.
Shape, stickiness, and electrostatic charge of the parti-
cles may also resist or mark negative impact on the
Category Sub-Category
Physical
Characteristics
of the Ingredients
Actives
• Particle Size
• Density
• Morphology
• Shape
• Amount
Excipients
• Particle Size
• Density
• Morphology
• Shape
• Amount
Chemical
Compositions of
the Ingredients
Actives
• Compatibility with excipients
• Adhesiveness
• Electrostatic charges
• Impurities
• Degradation profile
Excipients
• Compatibility with actives
• Adhesiveness
• Electrostatic charges
Blending
Equipment
Rotating Shell Blenders
• Drum
• Cross-flow
• Double Cone
• Twin-shell
Fixed Shell Blenders
• Ribbon
• Screw
Blending
Methods
Rotating Shell Blenders
• Rotating the blender shell around a
fixed axis.
Fixed Shell Blenders
• Rotating internal blender parts,
such as an impeller or paddle, and
producing a continuous
cutting and shuffling motion
Process
Parameters
• Mixing Time
• Mixing Speed
• Flow Rate
• Volume
Factors Influencing the Quality of the Blend (Source: PAREXEL Consulting)
Factor
Figure 1
Blend Quality Factors

3. 48 Journal of GXP Compliance
Trudy Yin
mixing process. Caution should be taken throughout
product and process development in order to eliminate
or minimize variability. Figure 1 summarizes some
major factors which contribute to the analysis and
determination of the degree of confidence in these
quality attributes, namely, the homogeneity and uni-
form distribution of the final blend mixture.
Scientifically valid procedures, appropriate test
methods, and in-process controls and limits should
be established in the developmental phase to evalu-
ate the adequacy and effectiveness of such blending
process as well as to ensure uniformity and homo-
geneity of the final blend product. The greatest chal-
lenge confronting oral solid dosage form manufactur-
ers today is the difficulty they encounter in applying
sound methods and rationales to the blending opera-
tion and validation, respectively.
FDA GUIDANCE FOR INDUSTRY
FDA announced the availability of a draft guidance for
industry entitled “ANDAs: Blend Uniformity Analysis,”
on August 27, 1999 (64 FR 46917) that was with-
drawn on May17, 2002, after comments from the
public raised scientific issues relating to the scope
and methodology of the blend uniformity draft guide-
line.1,8 This draft guidance was intended to provide
recommendations to abbreviated new drug applica-
tion (ANDAs) filings based on 21 CFR 314, not on
cGMP regulations. In other words, it guided industry
to what information should be provided in an ANDA
in order to demonstrate bioequivalence and to estab-
lish in-process acceptance criteria related to blend
uniformity analysis (BUA). After careful consideration
of the written comments submitted, FDA decided to
withdraw the draft guidance due to (1) adequacy of
current blend sampling techniques and (2) appropri-
ateness of various test methods for assessing blend
uniformity.1 Although this draft document has never
been implemented, and the 21 CFR 211.110 does
not require BUA, an applicant or manufacturer must
still comply with any predicate rules and applicable
regulations to ensure the identity, strength, quality,
purity, and bioavailability of the drug product (21 CFR
314.50(d) (1) (ii) (a)).1 A manufacturer must also
monitor and validate the performance of processes
that could be responsible for variability, including ade-
quacy of mixing to ensure uniformity and homogene-
ity (21 CFR 211.110(a) (3)).1 An evaluation of unifor-
mity and/or equivalent testing may be necessary to
fulfill cGMP requirements.
Issued on October 2003, new draft guidance,
“Powder Blends and Finished Dosage Units- Strati-
fied In-Process Dosage Unit Sampling and Assess-
ment,” was made available and distributed to solicit
public comments and suggestions. “Stratified Sam-
pling” is defined in this new guidance and the sam-
pling method, data evaluation, as well as analysis are
further discussed in this guidance.2,3 However, the
guidance states that the methods described are not
intended to be the only methods for meeting agency
requirements for the demonstration of the adequacy
of powder mix. Traditional powder blend
sampling/testing, finished product content uniformity
(CU) testing or alternate approaches may also be
used to satisfy cGMP requirements. The finalized
guidance represents the agency's current thinking on
“Powder Blends and Finished Dosage Units-Stratified
In-Process Dosage Unit Sampling and Assessment.”
Manufacturers are not encouraged to blindly follow,
but evaluate the methods, criteria, and recommenda-
tions, presented in this new guidance for appropriate-
ness before implementation.
PDA TECHNICAL REPORT NO. 25
In 1997, the International Association for Pharmaceu-
tical Science and Technology (PDA) issued Technical
Report No. 25, “Blend Uniformity Analysis: Validation
and In-Process Testing,” to provide scientifically
sound methods for the appropriate evaluation of
blending processes. The intent of this technical report
was to establish statistically derived methods (e.g.,
Bergum's method) and criteria to overcome the
impact of sampling errors. The report represented the
views of authors, industry experts, and PDA's Solid
Dosage Process Validation Committee. The FDA's
review of the draft of this report resulted in a letter
response from FDA's Center for Drug Evaluation and
Research (CDER) office dated August 29, 1997, in

4. 49October 2007 • Volume 12, Number 1
Trudy Yin
which FDA commented and questioned the rationale
behind the statistical approach advocated in the
report.4 PDA documented its response to FDA in a
letter dated October 9, 1997.3,4 Later, in regard to the
agency's Docket No. 99D-2635, “Draft Guidance for
Industry, ANDA's Blend Uniformity Analysis” of 1999,
PDA was given the opportunity to comment on the
proposed draft guidance, and suggestions were doc-
umented in a letter dated October 26, 1999. A subse-
quent workshop and meeting with the working group
(BUWG) were also conducted in 2000 and a draft
recommendation of the use of stratified sampling of
blend and dosage units to demonstrate adequacy of
mix for powder blends was issued for peer review
and then published in the PDA Journal of Pharma-
ceutical Science and Technology.3 This brief history
of guideline development demonstrates how consen-
sus is reached between FDA and the public. Informa-
tion on Technical Report No. 25 has also been cross-
referenced and incorporated into the FDA 2003 draft
guidance: “Powder Blends and Finished Dosage
Units- Stratified In-Process Dosage Unit Sampling
and Assessment.”
SAMPLING AND ASSESSMENT
In order to adequately measure between and within-
location variability, the most appropriate blend sam-
pling techniques and procedures should be devel-
oped and chosen. Blend sample analysis should be
conducted on both the blend and intermediate bulk
containers (IBCs) to identify the range of blending
times, dead spots in blenders, segregation in IBCs,
and the possibility of sampling errors, etc. Stratified
sampling is defined in the FDA draft guidance, “Pow-
der Blends and Finished Dosage Units- Stratified In-
Process Dosage Unit Sampling and Assessment” as:
“the process of collecting a representative
sample by selecting units deliberately
from various identified locations within a
lot or batch, or from various phases or
periods of a process to obtain a sample
dosage unit that specifically targets loca-
tions throughout the compression/filling
operations that have a higher risk of pro-
ducing failing results in the finished prod-
uct uniformity of content.” 3
A stratified sampling technique is used to collect in-
process dosage units throughout the compression/fill-
ing process. It can be used, therefore, as an alterna-
tive to routine blend sample analysis. Establishing a
sample size is important for the analysis of blend
samples. The FDA Draft Guidance document stated
that the blend sample size should be 1 to 3 dosage
units. The Blend Uniformity Working Group (BUWG)
from Product Quality Research Institute (PQRI) pro-
posed that sample quantities larger than 3X be
allowed if they could be scientifically justified. The
demonstration of content uniformity of the final pow-
der blends cannot be absolutely absent of sampling
bias and errors. Physical design of the thief, physical
and chemical properties of the formulation, and
sampling technique are major factors that can affect
sampling errors.4 Since sampling errors make it a big
challenge to validate the blending operation, blend
sample data should be used in conjunction with in-
process dosage unit data.
Assessment and validation (i.e., exhibit and
process validation batches) should optimally be per-
formed separately on the uniformity of the powder
blend, the in-process dosage units, and on the fin-
ished product. The FDA has proposed the following
acceptance criteria3:
1. Assay one sample (n) per location, n≥10
2. RSD of all individual results ≤5.0%
3. All individual results within 10.0% of the mean
Only after an initial assessment/validation is com-
pleted, routine manufacturing controls and criteria
shall follow. All these cGMP process controls and
product testing are necessary to ensure batch homo-
geneity and batch-to-batch reproducibility. Procedures
will be written for material handling, machine opera-
tions, cleaning, maintenance, as well as sample col-
lection, and analysis as minimal. In analysis of batch
normality, the FDA has proposed results to be classi-
fied as readily or marginally pass. Readily pass
homogeneity shall meet the following criteria3:

5. 50 Journal of GXP Compliance
Trudy Yin
1. RSD ≤ 4.0%, for each batch n≥ 60
2. Each location mean within 90.0%-110.0% of
target strength
3. All individual results within 75.0%-125.0% of
target strength
Marginally pass homogeneity shall meet the following
criteria3:
1. RSD ≤ 6.0%, for one batch n≥ 140
2. Each location mean within 90.0%-110.0% of
target strength
3. All individual results within 75.0%-125.0% of
target strength
Passing the readily pass criteria for all validation and
exhibit batches qualifies a product to use standard test-
ing during routine manufacture. Otherwise, tightened
testing must be used for products that marginally pass
during validation. If test results fail readily pass, remain-
ing dosage units shall be tested and compared with
marginally pass criteria. If tested samples again fail
marginally pass criteria, the company shall investigate,
make corrections, remix and reassess the blend and
the product accordingly. Standard and tightened tests
for routine verification of manufacturing batches are
described as per FDA draft guidance: “Powder Blends
and Finished Dosage Units- Stratified In-Process
Dosage Unit Sampling and Assessment.” Evaluation
can be done in the two proposed methods: Standard
Criteria Method (SCM) and the Marginal Criteria
Method (MCM); commensurate with the compendial
procedure described in the USP. Likewise, the PDA
Technical Report No. 25 introduces the conception of a
holistic approach whereby the mean and variances of
the blend are compared with those in the finished prod-
uct. In the report, it simply states that “the overall mea-
surement of the effectiveness of the blending and the
extent of post blending particle segregation are
obtained by a thorough evaluation of the blend and the
product.” To evaluate product uniformity, detailed statisti-
cal methods, such as standard deviation prediction,
Bergum's method, variance component analysis, etc.,
are recommended for validation purposes.
MODERN TECHNOLOGIES ON
ANALYTICAL METHOD
Near-InfraRed Spectroscopy (NIRS) and other imag-
ing techniques have drawn much attention in the
chemical industry over the past decade. NIRS is a
noninvasive on-line monitoring system with measure-
ments based on the standard deviation of the spectra
obtained.5-8 Because the conventional method of in-
process thief sample collection, laboratory testing,
and verification is considered labor intensive for many
manufacturers and results are irreproducible often
times, this type of process analytical technology
(PAT) offers a solution while providing “real-time” in-
process control without having to collect samples.6
Information provided by NIRS chemical imaging from
different time points throughout the blending cycle,
can be analyzed statistically to permit quantitative
evaluations of the blend content uniformity.
Prediction of product performance is more reliable
using the pattern recognition tools on the NIR spec-
trums to the identification and characterization of the
materials being tested. The disadvantage of tradi-
tional equipment such as the error prone manual
sample thief can be eliminated; and therefore,
reduces sampling bias and errors. With the continu-
ous efforts of new development and emerging tech-
nologies, a more thorough (i.e., API plus excipients)
and higher analytical accuracy with improved produc-
tion efficiency is just around the corner.6-8
REFERENCES
1. http://www.fda.gov/OHRMS/DOCKETS/98fr/051702a.htm,
ANDAs: Blend Uniformity Analysis; Withdrawal of Draft
Guidance, as of 10/22/06.
2. http://a257.g.akamaitech.net/7/257/2422/14mar20010800/
edocket.access.gpo.gov/2003/03-28045.htm, Draft Guid-
ance for Industry on Powder Blends and Finished Dosage
Units-Stratified In-Process Dosage Unit Sampling and
Assessment; Availability, as of 10/22/06.
3. www.fda.gov/cder/guidance/5831dft.pdf, Guidance for
Industry Powder Blends and Finished Dosage Units-
Stratified In-Process Dosage Unit Sampling and Assess-
ment, as of 10/22/06.

6. 51October 2007 • Volume 12, Number 1
Trudy Yin
4. Berman, J, Ph.D., Elinski, D.E., Gonzales, C.R., Hofer, J.D.,
Jimenez, P.J. Ph.D., Planchard, J.A., Ph.D., Tlachac, R.J.,
Vogel, P.F., 1997, “Blend Uniformity Analysis: Validation and
In-Process Testing,” Technical Report No. 25. PDA 51:S1-
S99.
5. http://www.fda.gov/ohrms/dockets/ac/01/slides/
3763s1_14_hussain/tsld001.htm, Optimal Applications of
“In-Line” or “At-Line” Manufacturing Controls in Pharmaceuti-
cal Production, as of 11/24/06.
6. Sullivan, M., 2006, “The Use of NIR as a PAT Tool for
Measuring Blend Uniformity,” SPECTROSCOPY SUPPLE-
MENT: Process Analytical Technologies, AN ADVANSTAR
PUBLICATION, U.S.A.
7. El-Hagrasy, A.S., Morris, H.R., D'Amico, F., Lodder, R.A.,
Drennen III, J.K., 2001, “Near-Infrared Spectroscopy and
Imaging for the Monitoring of Powder Blend Homogeneity,”
Journal of Pharmaceutical Sciences, Vol. 90, No. 9: 1298-
1307.
8. Popo, M., Romera-Torres, S., Conde, C., and Romanach,
R.J., 2002, “Blend Uniformity Analysis Using Stream Sam-
pling and Near Infrared Spectroscopy,” AAPS Pharm-
SciTech, article 24.
ABOUT THE AUTHOR
Trudy Yin is a consultant with PAREXEL Consulting, a
business unit of PAREXEL International and a leading
global life sciences consultancy serving the bio/phar-
maceutical and medical device industries. Ms. Yin pro-
vides strategic compliance and operational perfor-
mance excellence services to clients. Her expertise
includes design, qualification, and implementation of
guidelines that are regulatory compliant. Ms. Yin holds
a B.S. degree in Biological Sciences from University
of California, Irvine. She is a Quality Systems regula-
tion Certified Compliance Professional from the Cen-
ter for Professional Innovation & Education (CfPIE)
and a senior member of the International Society for
Pharmaceutical Engineering (ISPE) and the American
Society for Quality (ASQ). You can contact Ms. Yin at
Trudy.Yin@PAREXEL.com.
Article Acronym Listing
ANDA Amended New Drug Application
API Active Pharmaceutical Ingredient
BUA Blend Uniformity Analysis
BUWG Blend Uniformity Working Group
CDER Center for Drug Evaluation and
Research
CFR Code of Federal Regulations
cGMP Current Good Manufacturing Practice
CU Content Uniformity
FDA Food and Drug Administration
IBC Intermediate Bulk Container
MCM Marginal Criteria Method
NIRS Near Infrared Spectroscopy
OSD Oral Solid Dosage
PAT Process Analytical Technology
PQRI Product Quality Research Institute
SCM Standard Criteria Method