See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/51954486

Multi criteria decision making to select the suitable method for the
preparation of nanoparticles using an analytical hierarchy process

Article in Pharmazie · November 2011

DOI: 10.1691/ph.2011.1034 · Source: PubMed

CITATIONS READS

36 22,074

3 authors, including:

Ramaiyan Velmurugan Selvamuthukumar Subramanian

Annamalai University Annamalai University
19 PUBLICATIONS 400 CITATIONS 39 PUBLICATIONS 1,211 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Selvamuthukumar Subramanian on 25 March 2015.

The user has requested enhancement of the downloaded file.

 ORIGINAL ARTICLES

Department of Pharmacy, Faculty of Engineering and Technology, Annamalai University, Annamalai Nagar, Tamil Nadu,
India

Multi criteria decision making to select the suitable method for the
preparation of nanoparticles using an analytical hierarchy process

R. Velmurugan, S. Selvamuthukumar, R. Manavalan

Received March 14, 2011, accepted April 8, 2011

Dr. S. Selvamuthukumar, Department of Pharmacy, Faculty of Engineering and Technology, Annamalai University,
Annamalai Nagar, Tamil Nadu, India 608002
[email protected]
Pharmazie 66: 836–842 (2011) doi: 10.1691/ph.2011.1034

Selecting the right method for the preparation of nanoparticles is a crucial decision. A wrong decision can
result in the product having to be formulated and developed again. One tool that can be useful in determining
the most appropriate method is the Analytical Hierarchy Process (AHP). AHP has been employed in almost
all areas related to decision-making problems. In this paper, the results of a case study illustrate that the AHP
concept can assist designers in the effective evaluation of various methods available for the preparation of
nanoparticles. This paper presents the methodology of selecting the most suitable method for preparing
nanoparticles using the analytical hierarchy process.

1. Introduction theme of selection and evaluation especially in the area of engi-

neering, pharmaceuticals, and personal and social areas (Vaidya
Novel drug delivery systems (NDDS) are among the areas at the and Kumar 2006). Generally, implementing AHP is based on the
cutting edge of research in the pharmaceutical field today. Phar- experience and knowledge of the experts or users to determine
maceutical formulations with NDDS have been introduced with the factors affecting the decision process (Ho 2008; Dweiri and
the aim of optimizing bioavailability by modulating the time Al-Oqla 2006). According to Hajeeh and Al-Othman (2005),
course of the drug concentration in blood (Abu-Izza et al. 1996a, AHP is an intuitive method for formulating and analyzing deci-
b). All sustained and controlled release products have the com- sions whereas Cheng and Li (2001) describe the AHP approach
mon goal of improving drug therapy over that achieved with their as a subjective methodology. AHP is not only used as a stand-
non-sustained and non-controlled release counterparts.(Jain You alone tool but also can be integrated with other techniques.
et al. 2006; Babazadeh 2006). The absorption rate of a drug can AHP can be combined with techniques such as quality func-
be decreased by reducing its rate of release from the dosage form. tion deployment (QFD), and data envelopment analysis (DEA),
The product so formulated is designed as a sustained action, sus- and thus integrated it can be applied to a wide variety of fields
tained release, prolonged action, depot, retarded release, delayed especially in the logistic and manufacturing areas (Ho 2008).
action or timed release medication (Banker and Rhodes 2002). In order to achieve efficiency in selecting the optimum method,
This recent interest has been due to various factors, viz: the appropriate evaluation and decision tools need to be considered.
prohibitive cost of developing new drug entities, expiry of inter- Since AHP applications are related to evaluating and selecting
national patents, discovery of new polymeric materials suitable different alternatives or options, it can also be implemented in
for prolonging drug release, and improvements in therapeutic the product development process, especially to select the most
efficacy and safety achieved by these delivery systems (Gohel appropriate method. At this stage, designers have to consider
and Amin 1998; Jain et al. 2006). Various approaches are avail- a number of factors in order to determine and select the opti-
able for achieving NDDS such as targeted delivery systems, mum decision options, because an inappropriate decision can
nanoparticles, prodrugs, transdermal systems, ocular systems, lead to the eventual product having to be formulated and devel-
intravaginal and intrauterine systems, injections and implants, oped again. The advantages of using AHP include achieving
microencapsulation, matrix devices, and reservoir devices. One higher product quality and shortening the product development
of the most effective approaches comprised nanoparticles. process. AHP helps to capture both subjective and objective eval-
One of the useful tools in method selection is the Analytical uation measures, providing a useful mechanism for checking
Hierarchy Process (AHP) which was developed at the Wharton the consistency of the evaluation measures and alternatives sug-
School of Business by Saaty (Saaty 1980). This is a power- gested by the team, and thus reducing bias in decision-making.
ful and flexible weighted scoring decision making process to AHP allows organizations to avoid common pitfalls of decision-
help people set priorities and make the best decision. AHP has making processes, such as lack of focus, planning, participation
been widely used to solve problems of multi-criteria decision or ownership, which ultimately are costly distractions that can
making in both academic research and in industrial practice. prevent teams from making the right choice (Anonymous 2007).
AHP has been implemented in almost all applications related Some applications of AHP in the pharmaceutical field are in
to decision-making and is currently predominantly used in the selecting alternatives, such as for tablet formulation machines,

836 Pharmazie 66 (2011)

 ORIGINAL ARTICLES

Table 1: Scale for pair-wise comparisons Saaty TL (1980)

Relative intensity Definition Explanation

1 Equal value Two requirements are of equal value

3 Slightly more value Experience slightly favors one requirement over another
5 Essential or strong value Experience strongly favors one requirement over another
7 Very strong value A requirement is strongly favored and its dominance is
demonstrated in practice
9 Extreme value The evidence favoring one over another is of the highest possible
order of affirmation
2, 4, 6, 8 Intermediate values between two adjacent When compromise is needed
judgements
Reciprocals Reciprocals for inverse comparison

characterization techniques such as PK studies, release behavior, Table 3: Synthesized matrix for criteria
drug content, microbial versus instrumental assays for deter-
Goal PI OS FE SU TI Total row Priority vector
mining the potency of antibodies, blenders for mixing powders,
liquids or semisolids, site selection for pharmaceutical plants,
and very many more. This paper discusses AHP implementation PI 0.484 0.529 0.405 0.529 0.333 2.281 0.456
in the area of the method selection process. Here, using AHP can OS 0.161 0.176 0.243 0.176 0.200 0.957 0.191
FE 0.097 0.059 0.081 0.059 0.200 0.496 0.099
make the job of selecting a method shorter, reduce cost and pro-
SU 0.161 0.176 0.243 0.176 0.200 0.957 0.191
duce higher product quality. Thus, the paper illustrates the use
TI 0.097 0.059 0.027 0.059 0.067 0.308 0.062
of AHP in evaluating and determining the most suitable method
1.0000
for nanoparticle preparation from the methods available.

2. Investigations, results and discussion selection is discussed in this paper. In order to choose the most
2.1. Analytical hierarchy process principles suitable method for the preparation of nanoparticles, the fol-
lowing AHP steps, as listed in Fig. 1, should be considered:
Generally, AHP consists of three main principles, including
hierarchy framework, priority analysis and consistency verifi-
cation (Saaty 1980; Adhikari et al. 2006; Cheng et al. 2007).
Formulating the decision problem in the form of a hierarchy 2.2.1. Step 1: Define the problem
framework is the first step of AHP, with the top level represent- A case study for this research is about selecting the best method
ing the overall objectives or goal, the middle levels representing for the preparation of nanoparticles. After performing several
criteria and sub-criteria, and the decision alternatives being at steps of method selection, there are seven possible methods
the lowest level. Once a hierarchy framework has been con- remain, as listed below. Thus, it is necessary to choose the
structed, users are requested to set up a pairwise comparison most suitable of these methods by using AHP: M1 Polymer
matrix at each hierarchy and compare options by using a scale precipitation, M2 Interfacial polymer deposition, M3 Complex
pairwise comparison as shown in Table 1. Finally, in the pri- coacervation, M4 Cross linking, M5 Emulsion solvent diffusion,
ority stage synthesis, each comparison matrix is solved by an M6 Homogenization, M7 Polymerization technique.
eigenvector method to determine the importance of the crite-
ria and the performance of alternatives (Cheng et al. 2007).
These principles can be elaborated by structuring them in a 2.2.2. Step 2: Develop a hierarchy model
more encompassing nine step process as shown in Fig. 1. In this section, a hierarchy model for method selection using
AHP is introduced. A four-level hierarchy decision process, as
shown in Fig. 2, is described below:
2.2. AHP at the conceptual design stage – case study
Generally, there are six stages in the product development pro-
cess. One of them is conceptual design. It consists of three Table 4: Calculation to obtain new vector
processes, namely concept generation, concept evaluation and
1 3 5
concept development. Be this as it may, concept evaluation or
1/3 1 3
Table 2: Pairwise comparison of criteria with respect to overall 0.456 1/5 1/3 1
goal 1/3 1 3
1/5 + 0.191 1/3 + 0.099 1/3
Goal PI OS FE SU TI
New Vector
Process Information (PI) 1 3 A=5 3 5
Operation Skill (OS) 1/3 1 3 1 3 + 0.191 3 5 2.409
Feasibility (FE) 1/5 1/3 1 1/3 3 1 3 1.017
Supplier (SU) 1/3 1 3 1 3 1/3 + 0.062 3 = 0.503
Technical Information (TI) 1/5 1/3 1/3 1/3 1 1 3 1.017
Total Column 2.067 5.667 10.333 6.333 15.0 1/3 1 0.314

Pharmazie 66 (2011) 837

 ORIGINAL ARTICLES

Table 5: Random index of analytic hierarchy process

Size of matrix (n) 1 2 3 4 5 6 7 8 9 10 11 12

Random index (RI) 0 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.58

2.2.2.1. Level I. First, the objective, or the overall goal of the for the preparation of nanoparticles. The main criteria can be
decision, is presented at the top level of the hierarchy. Specif- classified into five aspects:
ically, the overall goal of this application is to ‘select the most Process Information (PI), Operational Skill (OS), Feasibility
suitable method for the preparation of nanoparticles’. (FE), Supplier (SU) and Technical Information (TI).

2.2.2.2. Level II. The second level represents the main criteria 2.2.2.3. Level III. The sub-criteria are represented at the third
that help to reach the goal i.e., selecting the most suitable method level of the hierarchy. There are five sub-criteria that refer to pro-

Table 6: Consistency test for criteria

1 Define problem
Goal PI OS FE SU TI Priority vector (PV) New vector (NV) NV/PV

PI 1 3 5 3 5 0.456 2.409 5.279

OS 1/3 1 3 1 3 0.191 1.017 5.312
FE 1/5 1/3 1 1/3 3 0.099 0.503 5.075
Develop hierarchical SU 1/3 1 3 1 3 0.191 1.017 5.312
2 TI 1/5 1/3 1/3 1/3 1 0.062 0.314 5.089
framework Total 26.067
Maximum eigenvalue 35.21

Consistency index CI = 0.053 Consistency ratio CR = CI/RI = 0.05

3 Construct pairwise comparison

matrix Table 7: Consistency test for sub-criteria

G/PI TD PC GR CO TR Priority vector (PV) New vector (NV) NV/PV

TD 1 3 5 3 3 0.415 2.283 5.501

Perform judgement for pairwise PC 0.333 1 3 3 3 0.251 1.392 5.551
4 GR 0.200 0.333 1 0.333 0.200 0.056 0.295 5.259
comparisons CO 0.333 0.333 3 1 1 0.127 0.669 5.251
TR 0.333 0.333 5 1 1 0.151 0.781 5.176
Total 26.739
Maximum eigenvalue 5.348
Synthesize pairwise Consistency index (CI) = 0.087 Consistency ratio (CR) = 0.08
5 comparisons Note: As the value of CR is less than 0.1, the judgements are acceptable because CR< 0.1 and Table 11
represent the consistency test for the sub-criteria and alternatives. As the value of CR is less than 0.1
for all sub-criteria and alternatives, the judgements are acceptable.

Table 8: Consistency test for sub-criteria

6 Check for consistency G/OS TE KN Priority vector (PV) New vector (NV) NV/PV

TE 1 3 0.750 1.500 2.000

KN 0.333 1 0.250 0.500 2.000
Total 4.000
7 Perform steps (3-6) for all Maximum eigenvalue 2.000
levels in the hierarchy
Consistency index (CI) = 0.000 Consistency ratio (CR) = 0.000 Note: As the value of CR is less than
0.1, the judgements are acceptable because CR< 0.1

Table 9: Consistency test for sub-criteria

8 Develop overall priority ranking
G/FE VE CO Priority vector (PV) New vector (NV) NV/PV

VE 1 3 0.750 1.500 2.000

CO 0.333 1 0.250 0.500 2.000
Total 4.000
9
Select best alternative Maximum eigenvalue 2.000

Consistency index (CI) = 0.000 Consistency ratio (CR) = 0.000 Note: As the value of CR is less than
Fig. 1: Steps of the analytical hierarchy process (AHP) 0.1, the judgements are acceptable because CR< 0.1

838 Pharmazie 66 (2011)

 ORIGINAL ARTICLES

Table 10: Consistency test for sub-criteria for this calculation are shown in Table 3. In mathematical form,
the vector of priorities can be calculated as
G/TI LT MN Priority vector (PV) New vector (NV) NV/PV

1  aij
n

LT 1 3 0.750 1.500 2.000 wi = n , i, j = 1, 2, · · · n (1)

MN 0.333 1 0.250 0.500 2.000
n i aij
j=1
Total 4.000
Maximum eigenvalue 2.000 For instance, the calculation for the first priority vector is as
follows 
n
Consistency index (CI) = 0.000 Consistency ratio (CR) = 0.000 Note: As the value of CR is less than Firstly, i aij hence, 1 + 1/3 + 1/5 + 1/3 + 1/5 = 2.067.
Secondly ijn hence, 1/2.067 = 0.484.
0.1, the judgements are acceptable because CR< 0.1
a
aij
n j

Thirdly, j=1
aijn hence, 0.484 + 0.529 + 0.405 + 0.529 +
aij
cess information: Type of drugs used (TD) Processing conditions i
0.333 = 2.281 and finally, divide this sum by the number of
(PI), Growth in the field (GR) Cost (CO) and Hands-on train-
elements (n = 5) hence, 2.281/5 = 0.456.
ing (TR). Techniques of the method (TE) and the Knowledge
about it (KN) add value for Operational Skill. Versatility (VE)
and Complexity (CO), and Literature (LT) and Manuals (MN) 2.2.6. Step 6: Perform consistency verification
are the subcriteria that add values to Feasibility and Technical Since the comparisons are carried out through personal or sub-
Information respectively. jective judgments, some degree of inconsistency may occur. To
ensure the judgments are consistent, a final operation called
2.2.2.4. Level IV. Finally, at the lowest level of the hierarchy, consistency verification, which is regarded as one of the most
the alternative methods (M) for nanoparticle preparation are advantageous features of the AHP, is incorporated in order to
identified, which are the decision options: measure the degree of consistency among the pairwise compar-
M1 Polymer precipitation, M2 Interfacial polymer deposition, isons by computing the consistency ratio (10). The consistency
M3 Complex coacervation, M4 Cross-linking, M5 Emulsion is determined by the consistency ratio (CR). Consistency ratio
solvent diffusion, M6 Homogenization, M7 Polymerization (CR) is the ratio of consistency index (CI) to random index (RI)
technique. for the same order matrices. To calculate the consistency ratio
(CR), there are three steps to be implemented as follows:

2.2.3. Step 3: Construct a pairwise comparison matrix 2.2.6.1. First calculate the Eigenvalue (␭max ). To calculate the
One of the major strengths of AHP is the use of pairwise eigenvalue (␭max), multiply the right of judgement matrix by
comparisons to derive accurate ratio scale priorities. Pairwise the priority vector or eigenvector, obtaining a new vector. The
comparisons are fundamental to the AHP methodology. Thus, calculation to give a new vector is shown in Table 4.
a pairwise comparison matrix (size n × n) is constructed for the For instance, the calculation for the first row in the matrix is
lower levels with one matrix in the level immediately above. The 0.456(1) + 0.191(3) + 0.099(5) + 0.191(3) + 0.062(5) = 2.409
pairwise comparisons generate a matrix of relative rankings for Then, dividing all the elements of the weighted sum matrices or
each level of the hierarchy. The number of matrices depends on new vector by their respective priority vector element, hence
the number of elements at each level. The order of the matrix at 2.409/0.456 = 5.279; 1.017/0.191 = 5.312; 0.503/0.099 = 5.075;
each level depends on the number of elements at the lower level 1.017/0.191 = 5.312; 0.314/0.062 = 5.089
that it links to. Then calculate the average of these values to obtain
␭max = (5.279 + 5.312 + 5.075 + 5.312 + 5.089)/5 = 5.213

2.2.4. Step 4: Perform judgement for pairwise comparison 2.2.6.2. Second: Calculate the consistency index (CI).
Pairwise comparison begins with comparing the relative impor- CI = (λ max −n)/(n − 1) (2)
tance of two selected items. There are n × (n−1) judgments
required to develop the set of matrices in step 3. The deci- Where n is the matrix size.
sion makers have to compare or judge each element by using CI = (5.213−5)/(5−1) = 0.053
the relative scale pairwise comparison as shown in Table 1.
The judgements are decided on the basis of the decision mak- 2.2.6.3. Finally calculate consistency ratio (CR). The CR can
ers’ or users’ experience and knowledge. The scale used for be calculated using the formula
comparisons in AHP enables the decision maker to incorpo-
rate experience and knowledge intuitively. For example, when CR = CI/RI (3)
making pairwise comparisons as shown in Table 2, if Pro-
cess Information (PI) is strongly more important or essential Selecting the appropriate value of random index (RI), for the
than Feasibility (FE), then a = 5. Reciprocals are automatically matrix size of five using Table 5, RI = 1.12. Then calculate the
assigned to each pairwise comparison. consistency ratio (CR), CR = CI/RI = 0.053/1.12 = 0.05. As the
value of CR is less than 0.1, the judgements are acceptable. If
CR > 0.1, the judgement matrix is inconsistent. To obtain a con-
2.2.5. Synthesizing the pairwise comparison sistent matrix, judgements should be reviewed and improved.
The summary results for this calculation are shown in Table 6
To calculate the vectors of priorities, the average of normalized
column (ANC) method is used (18). In ANC the elements of
each column are divided by the sum of the column and then the 2.2.7. Step 7: Steps 3–6 are performed for all levels in the
elements in each resulting row are added and this sum is divided hierarchy model
by the number of elements in the row (n). This is a process of
averaging over the normalized columns. The summary results The elements in Tables 7–10.
Pharmazie 66 (2011) 839
 ORIGINAL ARTICLES

Table 11: Consistency test for alternatives

Priority vector/eigenvector
Goal
PI OS FE SU TI
TD PC GR CO TR TE KN VE CO LT MN

M1 0.175 0.215 0.093 0.170 0.112 0.144 0.189 0.126 0.124 0.120 0.262 0.195
M2 0.104 0.081 0.074 0.290 0.061 0.055 0.058 0.229 0.227 0.066 0.191 0.316
M3 0.140 0.145 0.066 0.140 0.118 0.092 0.174 0.126 0.124 0.120 0.144 0.125
M4 0.126 0.044 0.070 0.127 0.061 0.062 0.055 0.229 0.227 0.050 0.093 0.101
M5 0.323 0.201 0.420 0.050 0.311 0.269 0.174 0.051 0.047 0.322 0.045 0.034
M6 0.080 0.162 0.074 0.167 0.084 0.147 0.174 0.169 0.182 0.100 0.198 0.165
M7 0.051 0.152 0.204 0.056 0.253 0.230 0.174 0.072 0.070 0.223 0.068 0.065
Consistency test
␭max 7.5000 7.638 7.172 7.315 7.307 7.617 7.027 7.436 7.437 7.388 7.654 7.490
CI 0.083 0.106 0.029 0.052 0.051 0.103 0.005 0.073 0.073 0.065 0.109 0.082
RI 1.320
CR 0.063 0.081 0.022 0.040 0.039 0.078 0.003 0.055 0.055 0.049 0.083 0.062

2.2.8. Develop overall priority ranking The elements in Table 14 show the overall priority vector of
the alternatives with respect to the criteria. The overall priority
After the consistency calculation for all levels has been com-
vector can be obtained by multiplying the priority vector for
pleted, further calculation of the overall priority vector to
the Method alternatives by the priority vector of the criteria. An
select the best preparation method must be performed. The ele-
example of the overall priority calculation is as follows:
ments/points in Table 12 represent priority vectors for criteria,
0.170 (0.456) + 0.156(0.191) + 0.125(0.099) + 0.120(0.191) +
sub-criteria and alternatives.
0.245(0.062) = 0.158
The elements in Table 13 represent the overall priority vector for
seven alternative methods with respect to the sub-criteria. The
overall priority vector can be obtained by multiplying the priority 2.2.9. Selection of most suitable method
vector for the alternative methods by the vector of priority of the Table 15 shows that the emulsion solvent diffusion technique
sub-criteria. An example of the overall priority calculation is as (M-5) has the highest value (0.236 or 23.6%) of the alternative
follows: methods that are applicable to the preparation of the nanoparti-
0.175(0.415) + 0.215(0.251) + 0.093(0.056) + 0.170(0.127) + cles. The second highest is (M-1) with a value of 0.158 (15.8%),
0.112(0.151) = 0.170 and the lowest value or last choice is method 4 (M-4) with a value
Goal
Selection of the best method for the preparation of Nanoparticle (Level 1)

Criteria
Process Information Operation Skill Feasibility Supplier Technical Information
(Level 2)

Type of Drugs Technique Versatility Literature

Processing Knowledge Complexity Manual

Condition
Sub Criteria
(Level 3)
Growth in the
field

Cost

Training

Alternatives
M1 M2 M3 M4 M5 M6 M7 (Level 4)
Fig. 2: Hierarchy model for selection of method for preparation of nanoparticles

840 Pharmazie 66 (2011)

 ORIGINAL ARTICLES

Table 12: All priority vectors for criteria, sub-criteria and alternatives

Priority vector
Goal
Criteria PI OS FE SU TI
0.456 0.191 0.099 0.191 0.062
Sub-criteria TD PC GR CO TR TE KN VE CO LT MN

0.415 0.251 0.056 0.127 0.151 0.750 0.250 0.750 0.250 0.750 0.250
Alternatives
M1 0.175 0.215 0.093 0.170 0.112 0.144 0.189 0.126 0.124 0.120 0.262 0.195
M2 0.104 0.081 0.074 0.290 0.061 0.055 0.058 0.229 0.227 0.066 0.191 0.316
M3 0.140 0.145 0.066 0.140 0.118 0.092 0.174 0.126 0.124 0.120 0.144 0.125
M4 0.126 0.044 0.070 0.127 0.061 0.062 0.055 0.229 0.227 0.050 0.093 0.101
M5 0.323 0.201 0.420 0.050 0.311 0.269 0.174 0.051 0.047 0.322 0.045 0.034
M6 0.080 0.162 0.074 0.167 0.084 0.147 0.174 0.169 0.182 0.100 0.198 0.165
M7 0.051 0.152 0.204 0.056 0.253 0.230 0.174 0.072 0.070 0.223 0.068 0.065

Table 13: Overall priority vectors for sub-criteria with respect The priorities of the alternatives would indicate their relative
to criteria strengths as the best method.
The alternative with the highest priority would be the most
Overall priority vector
suitable method and the ratios of the method priorities would
indicate their relative strength (Saaty 1980). Obviously, the pri-
M1 0.170 0.156 0.125 0.120 0.245
ority of the goal (to select the most suitable method) would be
M2 0.114 0.056 0.228 0.066 0.222
1.000.
M3 0.134 0.112 0.125 0.120 0.139
The priorities of the criteria Process Information (PI) Opera-
M4 0.093 0.060 0.228 0.050 0.095
tion Skill (OS), Feasibility (FE), Supplier (SU), and Technical
M5 0.262 0.246 0.050 0.322 0.042
M6 0.112 0.154 0.173 0.100 0.190 Information (TI), are 0.456, 0.191, 0.099, 0.191 and 0.062
M7 0.116 0.216 0.071 0.223 0.067 respectively (Table 12):
The priorities of the subcriteria Type of Drugs used (TD),
Processing Conditions to be carried out throughout the experi-
ment (PC), Growth of the technique in the field (GR), Overall
cost (CO), Hands-on training designed for the technique (TR),
of only 0.092 (9.2%). M-5 is the preferred choice since it has
Knowledge (KN), Versatility (VE), Complexity (CO), Litera-
the highest value among the seven alternatives.
ture (LT) and Manuals (MN) are 0.415, 0.251, 0.056, 0.127,
In the case study, the AHP technique was applied to selecting
0.151, 0.750, 0.250, 0.750, 0.250, 0.750 and 0.250 (Table 12).
amongst alternative preparation methods for nanoparticles and
The overall priorities of the subcriteria with respect to the criteria
thereby opting for the best technique. The composite score is
for each method are given in Table 13.
used for the final ranking of the alternatives. The solution of
The overall priorities for the alternatives with respect to the
the problem involves finding the composite score that reflects
criteria are given in Table 14. Method M5 emulsion solvent dif-
the relative priorities of all the alternatives at the lowest level of
fusion technique scored 0.236, M1 polymer precipitation scored
the hierarchy. The composite score favored the selection of the
0.158, M7 polymerization technique scored 0.148, M6 homog-
emulsion solvent diffusion technique (score = 0.236, Table 14,
enization techniques scored 0.129, M3 complex coacervation
15) over the other alternative methods.
scored 0.126, M2 interfacial polymer deposition scored 0.111
In the present study, a hierarchy was designed containing the
and finally M4 crosslinking technique scored 0.029.
decision as goal, the alternatives for reaching it and the criteria
The alternative with the highest priority would achieve the goal
for evaluating the alternatives and subcriteria for evaluating the
as per Saaty.
criteria. A priority was determined for each node based on its
The alternative in this case, M5 emulsion solvent diffusion tech-
strength, relative to the other nodes. The priorities of the criteria
nique, scoring 0.236 was the one with highest priority among
would indicate their relative importance in reaching the goal.
all the other alternatives.
Hence, having worked out the AHP technique, the emulsion
solvent diffusion technique is judged to be the most suitable
Table 14: Overall priority vector for alternatives with respect method for the preparation of nanoparticles.
to criteria

Priority Vector Table 15: Result of selection

PI OS FE SU TI Overall Priority
No Best selection
0.456 0.191 0.099 0.191 0.062
M1 0.170 0.156 0.125 0.120 0.245 0.158 1 M5 0.236
M2 0.114 0.056 0.228 0.066 0.222 0.111 2 M1 0.158
M3 0.134 0.112 0.125 0.120 0.139 0.126 3 M7 0.148
M4 0.093 0.060 0.228 0.050 0.095 0.092 4 M6 0.129
M5 0.262 0.246 0.050 0.322 0.042 0.236 5 M3 0.126
M6 0.112 0.154 0.173 0.100 0.190 0.129 6 M2 0.111
M7 0.116 0.216 0.071 0.223 0.067 0.148 7 M4 0.029

Pharmazie 66 (2011) 841

 ORIGINAL ARTICLES

In conclusion, this paper presents the methodology for evaluat- Babazadeh M (2006) Synthesis and study of controlled release of ibuprofen
ing and selecting the most appropriate method for preparation from the new acrylic type polymers. Int J Pharm 316: 68–73.
of nanoparticles by implementing the analytical hierarchy pro- Banker GS, Rhodes CT (2002) Modern pharmaceutics, Marcel Dekker, New
cess (AHP). AHP can be used to help designers to evaluate York, 121: 501–527.
Cheng EWL, Li H (2001) Information priority-setting for better resource
and select the best method based on the criteria and sub-criteria allocation using Analytic Hierarchy Process (AHP). Inform Manage
aspects of a decision. The analysis reveals that Method 5, the Comput Security 9(2): 61–70.
emulsion solvent diffusion technique, is the most appropriate Cheng SC, Chen MY, Chang HY and Chou TZ (2007) Semantic-based facial
for the preparation of nanoparticles because it has the highest expression recognition using Analytical Hierarchy Process. J Expert Syst
value (0.236 or 23.6%) of any of the methods. The application Applic 33: 86–95.
of the AHP for selecting the most suitable method for the prepa- Dweiri F, Al-Oqla FM (2006) Material selection using Analytical Hierarchy
ration of nanoparticles can improve the quality of the product Process. Intl J Comput Applic Technol 26 (4): 82–189.
Gohel MC, Amin AF (1998) Formulation optimization of controlled release
and shorten the product development process.
diclofenac sodium microspheres using factorial design. J Control Rel 51:
115–122.
Acknowledgement: The authors wish to thank the Department of Pharmacy, Hajeeh M, Al-Othman A (2005) Application of the Analytical Hierar-
Annamalai University for supporting this research. chy Process in the selection of desalination plants. Desalination 174:
97–108.
References Ho W (2008) Integrated Analytic Hierarchy Process and its applications - a
Abu-izza KA, Garcia-Contreras L, and Lu DR (1996) Preparation and eval- literature review. Eur J Operational Res 186: 211–228.
uation of sustained release AZT-loaded microspheres: Optimization of Jain SK, Agrawal GP, Jain NK (2006) A novel calcium silicate based
the release characteristics using response surface methodology. J Pharm microspheres of repaglinide: in vivo investigations. J Control Rel 113:
Sci 85: 144–149. 111–116.
Abu-izza KA, Garcia-Contreras L, and Lu RD (1996) Preparation and You J, Cui F-D, Han X, Wang Y-S, Yang L, Yu Y-W, and Li Q-P. (2006) Study
evaluation of zidovudine-loaded sustained-release microspheres. 2. Opti- of the preparation of sustained-release microspheres containing zedoary
mization of multiple response variables. J Pharm Sci 85: 572–575. turmeric oil by the emulsion-solvent diffusion method and evaluation
Adhikari I, Kim SY, Lee YD (2006) Selection of Appropriate Schedule of the self-emulsification and bioavailability of the oil. Colloid Surf B:
Delay Analysis. Biointerfaces 48: 35–41.
Anonymous (2007) Analytical Hierarchy Process: http://www.rfp- tem- Saaty TL (1980) The Analytic Hierarchy Process. New York, McGraw Hill.
plates.com/Analytical Hierarchy-Process-(AHP).html. Accessed on 10th Vaidya OS, Kumar S (2006) Analytical Hierarchy Process, an overview of
October 2007. applications. Eur J Operational Res 169: 1–29.

842 Pharmazie 66 (2011)

View publication stats