article review
eryrcch
6_articles.pdf
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rjrr20
Download by: [Embry-Riddle Aeronautical University] Date: 27 March 2016, At: 13:53
Journal of Risk Research
ISSN: 1366-9877 (Print) 1466-4461 (Online) Journal homepage: http://www.tandfonline.com/loi/rjrr20
A combined approach for supply chain risk management: description and application to a real hospital pharmaceutical case study
Hatem Elleuch, Wafik Hachicha & Habib Chabchoub
To cite this article: Hatem Elleuch, Wafik Hachicha & Habib Chabchoub (2014) A combined approach for supply chain risk management: description and application to a real hospital pharmaceutical case study, Journal of Risk Research, 17:5, 641-663, DOI: 10.1080/13669877.2013.815653
To link to this article: http://dx.doi.org/10.1080/13669877.2013.815653
Published online: 29 Jul 2013.
Submit your article to this journal
Article views: 420
View related articles
View Crossmark data
Citing articles: 3 View citing articles
A combined approach for supply chain risk management: description and application to a real hospital pharmaceutical case study
Hatem Elleucha, Wafik Hachichab* and Habib Chabchouba
aUnit of Logistic, Industrial and Quality Management (LOGIQ), Higher Institute of Industrial Management of Sfax, University of Sfax, Sfax, Tunisia; bUnit of Mechanic, Modelling and Production (U2MP), Engineering School of Sfax, University of Sfax, Sfax, Tunisia
(Received 24 April 2013; final version received 21 May 2013)
Managing risks in supply chains has emerged as an important issue in supply chain management. This research area has become familiar as supply chain risk management (SCRM). There are numerous approaches and techniques that are proposed in SCRM literature, but little in concrete and systematic approach for SCRM. In this paper, a SCRM framework comprising of different techniques and specialized procedures is proposed that can assist supply chain decision makers to risk identification, assessment and manage- ment. The combined approach consists of including the following. (1) Failure mode, effects, and criticality analysis to identify risk. (2) design of experiment to design risks mitigation and action scenarios. (3) Discrete event simulation to assess risks mitigation action scenario. (4) analytic hierarchy process to evaluate risk management scenarios. (5) desirability function approach to minimize the risk. The proposed approach is illustrated through a real hospital pharmaceutical supply chain case study.
Keywords: supply chain risks management; failure mode effects and criticality analysis; discrete event simulation; design of experiment; analytic hierarchy process; desirabily function approach; hospital pharmacy
1. Introduction
Successful supply chain (SC) management is a significant element of a firm’s ability to fill consumer demand, in any industry case. It is clear that SC performance may be decreased by disruptive events occurring in the supply chain system. SC disrup- tions are ‘unplanned events that may occur in the supply chain which might affect the normal or expected flow of materials and components’ (Svensson 2000). All the parts of supply chain could be impacted by a great variety of risks and the supply chain risks can have critical impact on both firm’s short-term and long-term performance. Managing the risk of these events occurring in the SC has become well-known as supply chain risk management (SCRM) and may be defined as ‘the management of supply chain risks through coordination or collaboration among the SC partners so as to ensure profitability and continuity’ (Tang 2006).
*Corresponding author. Email: [email protected]
Journal of Risk Research, 2014 Vol. 17, No. 5, 641–663, http://dx.doi.org/10.1080/13669877.2013.815653
� 2013 Taylor & Francis
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
In general, a SCRM process consists of four components: (1) Risk identification; (2) Risk assessment; (3) Risk management decisions and implementation; and (4) Risk monitoring (Wu and Blackhurst 2009). During the risk identification step, risks facing the firm’s supply chain are identified. For instance, researches that are applied this step may be found in Chopra and Sodhi (2004) where general risks in the supply chain are categorized and discussed. Some examples of previous risk assessment step include Zsidisin et al. (2004) and Hallikas et al. (2004). Assessing risk is a complicated step and can help a firm to prioritize which risks will affect the vulnerability of a SC. One important constituent of SCRM is the prioritization of risks. Prioritization helps a company to focus the decision making and risk man- agement effort on the most important risks (Hallikas et al. 2004). Prioritization requires comparisons concerning the relative importance of each of the risk vari- ables. However, risk management decision making requires supply chain managers to decide which mitigation strategies should be employed and where scarce resources may be allocated. Certainly, these are by no means easy decisions with many aspects and factors affecting these decisions. Finally, risk monitoring includes monitoring risks over time.
There has been recent research interest from academics and practitioners regard- ing SC disruptions and related issues, certainly because SC risks can potentially be harmful and costly for the whole SC (Craighead et al. 2007). There are numerous frameworks, approaches and techniques that are proposed in the SCRM literature (Pfohl, Kohler, and Thomas 2010; Cagliano et al. 2012), but little in complete and systematic approach for SCRM.
This paper describes a supply network risk approach, to assist supply chain decision makers to risk identification, assessment and management. The proposed approach is based on combining many techniques and methods including the following. (1) FMECA to identify risk and its current location, (2) Design of experiment (DOE) to design risks mitigation and action scenarios, (3) Discrete event simulation (DES) to assess risks mitigation action scenario, (4) AHP method to evaluate risk management scenarios, and (5) Desirability optimization to perform the best risk scenario.
The remaining sections of this paper are organized as follows. In Section 2, a literature review of SCRM is provided. Section 3 presents the proposed approach and a brief theoretical background of the study. In Section 4, the application of the proposed approach is fully illustrated through a real hospital pharmaceutical supply chain. Finally, Section 5 concludes the paper with a summary of the work presented and suggests potential extensions for further research.
2. Literature overview
2.1. Risks related to supply chains
In literature, there are a variety of definitions of risk. Mainly, risk takes into account two aspects: the uncertainty about and the severity of the consequences of an activity having a value for human beings (Aven and Renn 2009). In the context of supply chain, risk has been usually defined by taking a negative perspective. In other words, risk in a supply chain comprises anything that affects the material or information flow between the original supplier and the end customer (Norrman and Lindroth 2002). According to Tang and Nurmaya Musa (2011), any material, financial or information risk can disrupt the normal operations. The issue of supply
642 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
chain risk is attracting increasingly the attention of both researchers and practitio- ners. However, there are diverse perceptions of supply chain risk according to the perspective pursing among researchers from different fields (Sodhi, Son, and Tang 2012).
Various trends that enhance exposure to risks, such as the increased use of out- sourcing, globalization, reduction of the supplier base; reduced buffers, increased demand for on-time deliveries or shorter product life cycles (Norrman and Jansson 2004) are ratcheting up the importance of SCRM. Since the early 1990s, numerous companies have implemented various initiatives in the supply chain to increase revenues and to reduce costs (Sodhi, Son, and Tang 2012). This increases the complexity of operations in supply chains. The complexity has made the chains more vulnerable to various risks from inside and outside (Minahan 2005; Craighead et al. 2007).
Juttner, Peck, and Christopher (2003) identify three risk groups: (1) internal risks arising from the organization; (2) supply chain risks that are external to the organization but within the supply chain; and (3) external risks that are external to the supply chain and arise from the partners or the environment. Risks are signifi- cant if their occurrence would disturb the free flow of materials or information in the supply chain. According to Jüttner (2005), risks in a supply chain, can be classi- fied under four sub-chains: physical, financial, informational, and relational. Physi- cal sub-chain, represent traditionally viewed logistics, in the form of transportation, warehousing, handling, processing, manufacturing, and other forms of utility activi- ties. Financial sub-chain working in parallel deals with the supply chain’s flow of money, while informational sub-chains parallel the physical and financial chains through the processes and electronic systems used for creating events and triggered product movements and service mobilization. Relational sub-chains relate to the chosen linkages between buyers, sellers, and logistics parties in between them.
2.2. Supply chain risk management
SCRM is a field of escalating importance and is aimed at developing approaches to the identification, assessment, analysis and treatment of areas of vulnerability and risk in SCs (Neiger, Rotaru, and Churilov 2009). It has been gaining considerable attention in the last decade as an autonomous subject in the field of supply chain management (Macgillivray et al. 2007; Verbano and Venturini 2011). SCRM is most often a formal process that involves identifying potential losses, understanding the likelihood of potential losses, and assigning significance to these losses (Giunipero and Eltantawy 2004). According to Brindley (2004), SCRM is the management of supply chain risk through coordination or collaboration among supply chain part- ners so as to ensure profitability and continuity.
According to Hallikas et al. (2004), SCRM comprises four main elements: (1) risk identification, (2) risk assessment, (3) risk management, and (4) risk monitor- ing. SCRM starts from the identification and evaluation of potential risks and their impacts on operations in supply chain processes. Identifying the risks is a key activ- ity on which all other aspects of the process are based (Waters 2007). Numerous techniques have been used in the risk identification and assessment phases. Neiger, Rotaru, and Churilov (2009) have proposed a novel value-focused process engineer- ing methodology for process-based supply chain risk identification with the aim to increase value to supply chain members and supply chain as a whole. Trkman and
Journal of Risk Research 643
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
McCormack (2009) have presented preliminary research concepts regarding a new approach to the identification and prediction of supply risk. The findings of their approach are explained within the contingency theory. Mohd (2009) has used a multi criteria approach to assign relative importance to various risks in a supply chain and develop plans accordingly to mitigate them. Tuncel and Alpan (2010) have investigated the disruption factors of the supply chain network by a failure mode, effects and criticality analysis (FMECA) technique and have integrated there- fore the most critical failure modes into the supply chain network under study. Recently, Diabat, Govindan, and Panicker (2012) and Elmsalmi and Hachicha (2013) have applied a structural analysis based-model which prioritizes the various risks involved in a food supply chain with the help of interpretive structural model- ing ISM and MICMAC analysis.
Waters (2007) defines the activity of SCRM following risk analysis as ‘design- ing an appropriate response’, otherwise, determining the best strategies of dealing with the risks. Waters (2007) suggest the following range of responses to risk: ignore or accept it, reduce the probability, reduce or limit the consequences, trans- fer, share or deflect the risk, make contingency plans, adopt it, oppose it, or move to another environment. Tummala and Schoenherr (2011) present a list of ‘risk trig- gers,’ and like many other authors divide the risks according to their ‘consequence severity level’ and ‘risk probability,’ which finally determines the severity.
Numerous studies have been published on SCRM and in related fields in the last decade. Chopra and Sodhi (2004) highlight mitigation strategies for different types of risks, which manufacturing organizations apply to deal with uncertainty. They identify drivers for a wide variety of different risks and pinpoint alternative proac- tive mitigation strategies for each corresponding risk. Goh, Lim, and Meng (2007) have provided a stochastic model using the Moreau-Yosida regulation and design an algorithm for treating the multi-stage global supply chain network problem with profit maximization and risk minimization objectives. Sodhi and Tang (2009) pro- posed a time-based SCRM in order to motivate research on the practice of system- atic and preplanned response to rare events that can cause disruptions to the supply chain. Goh and Meng (2009) proposed a stochastic model for SCRM using condi- tional value at risk. They used the sample average approximation method for solv- ing the underlying stochastic model. Datta et al. (2009) used GARCH proof of concept as a forecasting and risk analysis tool in supply chain management for decision support in supply chain scenarios and provides preliminary simulation results from their impact on demand amplification.
Yang (2011) used a bow tie diagram to investigate appropriate risk management strategies to deal with maritime security risks. Wagner and Neshat (2010) developed an approach based on graph theory to quantify and hence mitigate supply chain vulnerability.
Cagliano et al. (2012) proposed a framework to integrate both risk identification and analysis in extensively applied supply chain management practices, like process mapping and performance measurement. Their framework is based on data currently recorded by companies for purposes other than risk investigation. In particular, for each supply chain process of the SCOR Model, risk sources are identified and con- nected to elementary activities through a standard framework. After that, the effects of risky events due to the defined sources are assessed by means of data taken from the performance measurement system of an organization.
644 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
The simulation has been used in SCRM as a useful modeling and analysis tool. Cigolini and Rossi (2006) have used as a simulated-based decision support system model for collaboration level selection. DES have been used to provide an innova- tive formalized methodology in mitigating disruptions in SCs (Melnyk, Rodrigues, and Ragatz 2009); Vilko and Hallikas (2012) proposed a Monte-Carlo-based simu- lation for analyzing the risk impacts in terms of delays in the maritime supply chain. Petri net-based simulation has been used to integrate the risk management procedures into design, planning, and performance evaluation process of supply chain networks (Tuncel and Alpan 2010).
Other studies have proposed different tools and frameworks for SCRM. Harland, Brenchley, and Walker (2003) proposed a conceptual framework as a tool to help in the identification, assessment and management of risks in supply networks. The tool consists of a set of successive stages. Beginning by the map supply network, fol- lowed by, identification of risks and its current location, risks assessment, risks man- agement, form collaborative supply networks risk strategy and finally, implement supply network risk strategy. Jüttner, Peck, and Christopher (2003) proposed a framework and foundation for systematically exploring the concept of risk manage- ment in supply chains by delineating the domain of risk management in supply chains in order to identify an agenda for future research. Christopher and Lee (2004) proposed a contention suggesting that end to end visibility is the one key element in any supply chain risks mitigation strategy by restoring supply chain confidence throughout the chain. Huang and Tseng (2009) proposed a semi-structured knowl- edge model to represent SCRM knowledge and uses Resource Description frame- work and Resource Description framework system as metadata languages to annotate semantic metadata. Tang and Tomlin (2008) presented a unified framework and five stylized models to illustrate that firms can obtain significant strategic value by implementing a risk reduction program that calls for a relatively low level of flex- ibility. Dong, Wang, and Wu (2009) proposed a generalized simulation framework for responsive supply network management. They used a General Business Simula- tion Environment that provides useful insight network’s real operations and allow assessing the responsiveness of a supply network. Qiang, Naurney, and Dong (2009) proposed an integrated modeling and robustness analysis framework through a new supply chain network model with multiple decision-makers associated at different tiers and with multiple transportation modes for shipment of the good between tiers. Recently, Giannakis and Louis (2011) proposed a framework for the design of a multi-agent-based decision support system for the management disruptions and miti- gation of risks in manufacturing supply chains. Hahn and Kuhn (2012) developed a framework for value-based performance and risk optimization in supply chains.
Some research papers are focused on real case studies. For instance, Thun and Hoeing (2011) have presented an empirical analysis of the SCRM in the case of the German automotive industry based on a survey with 67 manufacturing plants. The authors have examined the vulnerabilities of supply chains in general and investigated instruments for dealing with supply chain risks. They also have analyzed the relationship between the implementation of these instruments and different performance criteria. Yang (2011) proposed a methodology of risk management in the case of the maritime supply chain security by evaluating the container security initiative on the maritime supply chain in Taiwan. Vilko and Hallikas (2012) proposed preliminary research concepts and finding concerning the identification and analysis of risks in the cases of multimodal supply chains.
Journal of Risk Research 645
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
Finally, some authors have studied other aspects of risk management, such as the risk communication strategy. In this context, Lofstedt (2006) and Smillie and Blissett (2010) provided some methodologies enabling all communicators to reliably appraise risk in the context of the current risk environment, allowing the successful design and implementation of an effective communication strategy.
3. The proposed SCRM approach
The proposed approach aims at determining the most significant supply chain risks mitigation strategy for a supply chain. For these reason, the procedure of the pro- posed approach is based on a set of tools for the risks identification, assessment and management (Figure 1). The approach is divided into five stages including the following: (1) Failure mode, effects, and criticality analysis (FMECA) to identify risk and its current location and assess risks, (2) DOE to design risks mitigation and action scenarios, (3) DES to assess risks mitigation action scenario, (4) Analytic hierarchy process (AHP) method to evaluate risk management scenarios, and (5) Desirability function approach (DFA) to perform the minimize risk and to find the optimal risk mitigation scenario.
This section is composed of two subsections. The aim of the first subsection is to describe all used techniques which are FMECA, DOE, DES, AHP, and DFA, while the second subsection is allowed to present the procedure of the proposed approach by describing the relationship between these techniques.
3.1. Background
3.1.1. Failure mode, effects, and criticality analysis
FMECA was first developed as a formal design methodology in the 1960s by the aerospace industry with their obvious reliability and safety requirements (Bowles and Bonnell 1995). Since then, it has come to be used extensively to help assure the safety and reliability of products in a wide range of industries – particularly the
Supply chain network
(1) FMECA (identifies risk and its current location and assesses risks)
(2) Design of experiment (design of risks mitigation action scenarios)
(3) Discrete event simulation (assess risks mitigation action scenarios)
(4) AHP method (select risk management scenario)
Supply network risk strategy
(5) Desirability optimization of the final score (performs the best risk strategy)
Objectives of the study
Figure 1. Flowchart of the proposed SCRM approach.
646 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
aerospace, automotive, nuclear, and medical technologies industries. FMECA is characterized by a bottom-up approach by which any complex production system is decomposed into its constituent parts, which are successively analyzed to find all the potential failure causes and their effects. The analyst builds a table with all fail- ure causes and performs a criticality assessment to measure the risk level for each fault, in terms of criteria such as the chance of failure or the severity of the fault itself.
In doing the analysis, the system behavior is evaluated for every potential failure mode of every system component. Where unacceptable failure effects occur, design changes must be made to either eliminate the causes of the failures or to mitigate their effects. The criticality part of the analysis prioritizes the failures for corrective action based on the probability of the item failure mode and the severity of its effects. Traditionally, the criticality assessment has been performed by either: developing a Risk Priority Number (RPN) or calculating an item criticality number. The RPN calculation uses linguistic terms to rank the probability of failure (P), the severity of its failure effect (S), and the chance of the failure being undetected (D) on a 1 to 10 numeric scale. Well-known ‘conversion’ tables (e.g. Gilchrist (1993)) report the typical basis for the linguistic judgment scales used to estimate the three crisp (in the sense of ‘net’, ‘precise’) quantities which are used to calculate the RPN value in the following manner:
RPN ¼ P � D � S ð1Þ
Failure modes having a high RPN are assumed to be more important and given a higher priority than those having a lower RPN (Bowles and Peldez 1995).
3.1.2. Design of experiment
The goal of performing a DOE-based method is to obtain information as efficiently as possible. An experiment is a series of planned trials in which factors (indepen- dents variables) that are thought to affect the outcome are varied systematically and the outputs (dependant variables, also called responses) are measured and recorded. An experiment provides insight into how a SC behaves and, perhaps, why it behaves as it does. There are principally two types for designing an experiment: full factorial design and fractional factorial design. The full factorial design has the advantages that all kinds of main effects and interactions can be considered. How- ever, since all combinations are to be tested, the number of experiments increases exponentially (Montgomery 2005).
An objective of DOE is to find an appropriate approximation for the true func- tional relationship between response and the set of independent variables. If the response is well modeled by a linear function of the independent variables, then the approximating function is the first order model. In general, the relationship between the input and output variables (i.e. SC performance measure) is assumed to be a first-order model with two-factor interactions.
3.1.3. Discrete event simulation
Various alternative methods have been proposed for modeling supply chains (Goyal and Giri 2001). According to Beamon (1998), they can be grouped into four
Journal of Risk Research 647
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
categories: deterministic models where all the parameters are known, stochastic models where at least one parameter is unknown but follows a probabilistic distri- bution, economic game-theoretic models and models based on simulation, which evaluate the performance of various SC strategies. Literature in simulation research area can be classified in four simulation types (Kleijnen and Smits 2003): spread- sheet simulation; system dynamics; DES; and business games. The type of simula- tion that must be used will depend on the problem to be solved by the problem in each specific case.
Today, DES is a popular tool for the analysis and/or design of existing or proposed complex systems. This popularity is partly due to its flexibility, and to its ability to model real-world systems in some detail, which, in turn, leads simulation to be used as a decision support tool in supervising and controlling the underlying system. Though, simulation models require fewer restrictive assumptions than mathematical models when representing complex, dynamic systems, these models themselves are usually fairly complex and of relatively high dimensionality. That is, the performance of a simulation model mostly depends on a large number of parameters or factors that act and interact in a complex manner. However, with simulation modeling, the relationships between the design parameters and their resulting performance measures are not explicitly known. Therefore, simulation modeling becomes a trial-and-error process in which a set of input factors is used to predict a set of output performance measures. If the desired performances are achieved, a good system design has been attained; otherwise the process is repeated until a satisfactory set of performance measures is obtained.
3.1.4. Analytic hierarchy process
The AHP developed by Saaty (1980) is a multi-criteria decision-making tool that can handle unstructured or semi-structured decisions with multi-person and multi- criteria inputs. The AHP is designed to handle those decision environments in which subjective judgments are inherent in the decision making process. It also allows users to structure complex problems in the form of a hierarchy or a set of integrated levels. In addition to this, AHP is easier to understand and can effectively handle both qualitative and quantitative data. The AHP has been used widely to solve decision problem and attracted the interest of many researchers for long because of its easy applicability and interesting mathematical properties (Ramanjaneyulu and Sasamal 2008). AHP involves the principles of decomposition, pair wise comparisons, and priority vector generation and synthesis. On the negative side, however, while the AHP allows for checking the consistency of an individual’s judgments, adequate consistency is not always achieved. This is a disadvantage. Another disadvantage is the inability of AHP to handle correlation among criteria.
3.1.5. Desirability function approach
Desirability appears to have been first proposed as a criterion for response optimization by Harrington (1965) and popularized by Derringer and Suich (1980). DFA transforms a predicted response value (e.g. the predicted response value ŷ) into a scale-free value, called desirability, which is a value between 0 and 1. The
648 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
value of di increases as the corresponding response value becomes more desirable. In another words, the first step in defining a desirability function is to assign values to the responses that reflect their desirability. The multi-objective desirability optimization method involves transformation of each predicted response, ŷ, to a dimensionless partial desirability function, di, which includes the researcher’s priorities and desires when building the optimization procedure. One- or two-sided functions are used, depending on whether each of the n responses has to be maximized or minimized, or has an allotted target value. If the response i is to be maximized the quantity di is defined as:
di ¼ ŷ �AB�A � �wi
If A � y^ � B di ¼ 1 If y
^ � B di ¼ 0 If y
^ � A
8>>< >>: ð2Þ
Likewise, di can be defined when the response is to be minimized or if there is a target value for the response. In Equation (2), A and B are, respectively, the lowest and the highest values obtained for the response i, and wi is the weight. di ranges between 0, for a completely undesired response, and 1, for a fully desired response. The partial desirability functions are then combined into a single compos- ite response, the so-called global desirability function D, defined as the geometric mean of the different di-values. This desirability D, another value between 0 and 1, is obtained by aggregating the individual desirability as indicated in Equation (3). Then the objective is to find the input variable setting x⁄which maximizes the value of D.
D ¼ Yn i¼1
d pi i
!1 n
ð3Þ
A value of D different from zero implies that all responses are in a desirable range simultaneously and, consequently, for a value of D close to 1, the combina- tion of the different criteria is globally optimal, so the response values are near the target values. In Equation (3), pi is the relative importance assigned to the response i. The relative importance pi is a comparative scale for weighting each of the result- ing di in the overall desirability product and it varies from the least important (pi = 1) to the most important (pi = 5). It is noteworthy that the outcome of the overall desirability D depends on the pi value that offers users flexibility in the definition of desirability functions (Hachicha et al. 2010).
3.2. Research procedure
The framework of the proposed approach is constituted by a mix of methodologies for risk assessment and management. First, the hospital pharmaceutical supply chain was described and modeled. Then, FMECA method was applied for the identifica- tion and assessment of the major risks in supply chain, corrective actions was also provided in FMECA. Given the major risks, all possible mitigation scenarios were conceptualized through DOE method based on risks exposure levels. The
Journal of Risk Research 649
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
performances of scenarios under study were evaluated through simulation based on criteria that were set by the managers. Subsequently a multi criteria decision aid method was resorted to with the aim to give the best scenario according to five criteria. Finally, a desirability optimization approach has the role of performing the optimized scenario based on the parameters found in the last stage.
4. Case of the hospital pharmaceutical supply chain
As a validation step, the proposed approach is illustrated with computational study involving a real hospital pharmaceutical supply chain case. The hospital pharmacy service is responsible for the dispensing of prescriptions but also the purchase, gale- nic preparation and quality testing of all medicines used in the hospital. This section will follow the five-stage proposed approach as indicated in Figure 1.
4.1. Supply chain description
The supply chain of the hospital pharmacy can be modeled as a network of three subsystems (see Figure 2). These subsystems are identified as the supplier of the hos- pital, the inbound/outbound logistics at the drugstore department and the operations at the Care departments. The establishment under study is placed as the drugstore of the hospital which includes the central and the internal drugstores. The supply chain network considered for modeling and analysis is limited to the public drugstore as a supplier, the drugstore of the hospital which comprises the inbound logistics which provides the drugs from suppliers, outbound logistics which is responsible for distri- bution of drugs to the care departments, and 16 care departments which dictate the demand structure.
The drugs are provided from the public drugstore by their means of transport and stored in the central drugstore of the hospital. When the drugs are received at the drugstore site, they are controlled to check if there is any quality problem. If a drug has a quality problem, it is returned to the supplier (public drugstore). Then, drugs are delivered to the internal drugstore as part of the resupply. Finally, they are distributed to the care departments in detail according to the need of each one and following the prescriptions. It should be noted that the pharmacy of the hospital made some galenic preparation according to the prescriptions from the different care departments. The emphasis in this study is focused on the physical and informa- tional flows and only the drug flows are considered.
(Supplier) (Drugstore department) Inbound and outbound logistics
Care Department 16
Care Department 1
Care Department 2
Internal drugstore Central drugstore
Public drugstore
Private drug manufacturers
Figure 2. Structure of the hospital pharmaceutical supply chain.
650 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
To achieve FMECA and AHP steps, the Risk Analysis Group was created (our acronym RAG). The RAG was made up of six members including the hospital managers (three administrators and the head of the drugstore department), and two university professors.
4.2. FMECA tool for use in identifying and assessing potential risks
In order to identify the major risks to the hospital pharmaceutical supply chain, a brainstorming process has been conducted with the hospital managers. Based on the collected data, a FMECA is build. In this FMECA table, each subsystem is consid- ered as an entity of the supply chain network which may be exposed to different risks. For each of these subsystems, the failure mode (column 2 of Table 1) is defined. Once the potential failures are identified, the potential effects (column 3) are listed. The effect of each probable fault on the overall system performance is examined and each failure mode is associated with a severity index (S) (column 4). The severity index is used to classify the relative importance of the effects due to a failure mode. Historical records stored in databases were used to determine the severity index. Then, the potential causes of the failure modes (column 5) are listed and evaluated the probability of their occurrences (column 6) using the information obtained by past statistical data sources, monthly reports and daily operator perfor- mance evaluation results. Occurrence rate is a numerical subjective estimate of the likelihood that the cause, if it occurs, will produce the failure mode and its particu- lar effect. Finally, the failure detection and possible correction actions to prevent the cause and occurrence for the potential fault scenarios (columns 7 and 8) are determined following the discussion within the RAG. In column 8, the possible control and detection process is assessed to determine how well it is expected to detect or control the failure modes or the probability that the proposed process con- trols will detect a potential cause of failure. As seen in Table 1, 11 common supply chain disruptions have been determined and classified into three main groups. The criticality analysis is performed to enable a priority ranking among the identified risks. The ranking is done using the severity S, the occurrence O and the detection difficulty D values and referred to as the risk priority number (RPN) as mentioned in Equation (1).
The information provided in the FMECA table helps doubly: first of all, the risk priority number (in column 9) reveals the most critical failure modes of all the processes studied. The strategy adopted in this study consists of: (1) taking the risks which IPR are lower at 100 and (2) managing the risk which IPR are superior to 100.
In the following sections, the failure modes that will be dealt with in the supply chain network under study are:
(1) Compliance problem related to the supplier subsystem (public drugstore), has the highest RPN (RPN = 168).
(2) Lack of personnel related to the inbound and outbound logistics in the hospi- tal pharmacy subsystem (central and internal drugstore), has the highest RPN (RPN = 280).
(3) Time limit related to the customer subsystem (care departments), has the highest RPN (RPN = 336).
Journal of Risk Research 651
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
T ab le
1 .
A p p li ca ti o n o f F M E C A
to th e h o sp it al
p h ar m ac eu ti ca l su p p ly
ch ai n .
E n ti ty /p ro ce ss
st ep
P o te n ti al
fa il u re /
er ro r m o d es
P o te n ti al
ef fe ct (s ) o f
fa il u re s
S ev er it y
(S )
P o te n ti al
ca u se
(s ) o f fa il u re s
O cc u rr en ce
(O )
D et ec ti o n /c u rr en t
co n tr o ls
D et ec ti o n
sc o re
(D )
R is k
P ri o ri ty
N u m b er
(R P N )
P u b li c d ru g st o re
(S u p p li er )
P o o r q u al it y in
th e p u rc h as ed
d ru g s fr o m
su p p li er
H ea lt h ri sk s, q u al it y
p ro b le m s o f th e
p h ar m ac eu ti ca l
p re p ar at io n s
7 L o w
te ch n ic al
re li ab il it y
2 S ta ti st ic al
q u al it y
co n tr o l, in sp ec ti o n ,
q u al it y co n tr o l
5 7 0
S h o rt ag e o f
d ru g s (w
it h o u t
su b st it u te )
L au n ch
o f o rd er ,
d el ay ed
ar ri v al
o f
d ru g s
6 L u ck
o f d ru g s
in th e ce n tr al
d ru g st o re
7 F o re ca st in g , E R P,
E D I,
co m m u n ic at io n an d
in fo rm
at io n sh ar in g
w it h su p p li er s
2 8 4
C o m p li an ce
p ro b le m
(t im
e li m it , b re ak ag e,
em p ty
o r
m is si n g b o x es
… )
R et u rn
o f in ad eq u at e
d ru g s, d el ay
in d el iv er y,
fi n an ci al
lo ss
d u e to
su b se q u en t
o rd er s
4 H u m an
er ro r,
b ad
in te n ti o n ,
fa il u re
in th e
p ro ce ss es
6 V ig il an ce
in th e
re ce p ti o n o f d ru g s in
th e d ru g st o re ,
in sp ec ti o n , tr ac ea b il it y
an d in fo rm
at io n sy st em
7 1 6 8
C en tr al
an d in te rn al
d ru g st o re
o f th e
h o sp it al
p h ar m ac y
(I n b o u n d an d
o u tb o u n d
lo g is ti cs )
L ac k o f
p er so n n el
D el ay
in m an u fa ct u ri n g , d el ay
in d el iv er y
7 A b se n te ei sm
, st ri k e,
d is sa ti sf ac ti o n
w it h w o rk
8 M o ti v at io n , re la ti o n
w it h la b o r u n io n ,
re w ar d sy st em
, ap p ro p ri at e
ap p o in tm
en t, ca re er
m an ag em
en t
5 2 8 0
T ec h n ic al
p ro b le m s o f th e
m ac h in er y (i n
g al en ic
p re p ar at io n )
D el ay
in p re p ar at io n ,
d el ay
in d el iv er y an d
in ad eq u at e d ru g s fo r
re q u is it e q u al it ie s
5 N eg li g en tl y
m ai n te n an ce ,
L o w
te ch n ic al
re li ab il it y
4 P er io d ic
m ai n te n an ce ,
st at is ti ca l p ro ce ss
co n tr o l
4 8 0
(C o n ti n u ed )
652 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
T ab le
1 . (C o n ti n u ed
).
E n ti ty /p ro ce ss
st ep
P o te n ti al
fa il u re /
er ro r m o d es
P o te n ti al
ef fe ct (s ) o f
fa il u re s
S ev er it y
(S )
P o te n ti al
ca u se
(s ) o f fa il u re s
O cc u rr en ce
(O )
D et ec ti o n /c u rr en t
co n tr o ls
D et ec ti o n
sc o re
(D )
R is k
P ri o ri ty
N u m b er
(R P N )
H u m an
er ro r (i n
g al en ic
p re p ar at io n )
N o n co n fo rm
it y w it h
th e p re sc ri b ed
d o se s,
h ig h ra te
o f re w o rk
n ee d .
1 0
L o ss
o f
m o ti v at io n ,
L ac k o f
tr ai n in g ,
u n fa v o ra b le
w o rk
co n d it io n s
1 T ra in in g , am
el io ra te
th e w o rk in g
er g o n o m ic , re w ar d
sy st em
re v is io n
5 5 0
H u m an
er ro r (i n
h an d li n g an d in
st o ri n g th e
d ru g s)
D ef ec ti v e
p h ar m ac eu ti ca l
p ro d u ct s (t h e m o st
se n si ti v e)
4 U n tr ai n ed
st af f,
L ac k o f
h an d li n g
m at er ia l
3 P er so n n el
tr ai n in g ,
in v es tm
en t in
h an d li n g
m at er ia ls
6 7 2
T h ef t in
th e
st o re s an d in
th e
d el iv er y se ct o rs
L ac k o f d ru g s,
F in an ci al
lo ss
6 D is ta n t
d ru g st o re s,
L ac k o f
in te rn al
au d it
an d tr ac ea b il it y
sy st em
5 F it ti n g o u t th e
d ru g st o re s, in te rn al
au d it sy st em
.
3 9 0
C ar e d ep ar tm
en ts
(C u st o m er )
F lu ct u at io n in
cu st o m er
d em
an d s
B u ll w h ip
ef fe ct s,
u rg en t la u n ch
o f
o rd er s
6 F o re ca st
p ro b le m s,
u n fo re se ea b le
si tu at io n s.
4 C o ll ab o ra ti o n w it h th e
m an ag er
o f th e ca re
u n it s, fo re ca st in g
4 9 6
T im
e li m it o f
d ru g s in
th e
m ed ic in e ca b in et
D ep ri v at io n o f o th er
d ep ar tm
en ts /p at ie n ts
o f d ru g s, li m it at io n o f
th e av ai la b il it y,
F in an ci al
lo ss
6 L ac k o f
tr ac ea b il it y
7 T ra ce ab il it y sy st em
, in te rn al
au d it
8 3 3 6
T ra n sp o si ti o n
er ro r o f th e
p re sc ri p ti o n to
th e n o te b o o k o f
o rd er s
R et u rn
o f d ru g s,
ex ch an g e o f d ru g s,
re tu rn
o rg an iz at io n
p ro b le m s
2 M an u el
in fo rm
at io n
sy st em
, lo n g
w o rk in g ti m e,
st re ss
4 E R P,
in v o lv em
en t o f
d o ct o rs
an d n u rs es
o f
th e ca re
u n it s.
3 2 4
Journal of Risk Research 653
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
Secondly, the potential effects of risks are expressed either in terms of delay or in terms of cost. In the simulation section, these two indicators will be used as a criterion for evaluating the supply chain risk strategies.
4.3. DOE for designing risk management scenarios
A qualitative description or analysis of the simulation results does not provide a deep understanding of the supply chain behavior and could lead to erroneous con- clusions in the decision making process. It is generally implicit that experiments are natural part of the engineering and scientific process because they help us in under- standing how systems and processes work. The validity of the decisions taken after an experiment strongly depends on how the experiment was conducted and how the results were analyzed. For these reasons, the proposed approach suggests to use the DOE technique jointly with simulation, because it permits experiments planning and understanding how factors (input parameters) affect the supply chain behavior.
Before using DOE notation to divide each failure mode into several levels, it is necessary to assign the variables for the operational parameters. In this study, two levels are chosen for each failure mode as mentioned in Table 2 (the high and the low levels). For instance, level 1 of compliance problem (CP) consists of 0.24 as a high-risk exposure level and 0.12 as a low-risk exposure level. FMECA is used to establish the risk probabilities for the high scenario: risk probability rates determined by multiplying the severity score and the occurrence rate of the related failure modes obtained in Table 1. Then, the risk probabilities are calculated for the low scenario probabilities by multiplying the severity score and the new occurrence rate (e.g., for the compliance problem risk, the low-risk exposure level = 4 � 3%). Likewise, the new occurrence rate in the low level for the lack of personnel and the time limit risks is also set to 3.
Checking all possible factors levels combinations is based on the factorial exper- imental design, which requires 8 (23) possible scenarios as explained in Table 3. This scenario reflects the case where the system is exposed to risks but no action is taken. This scenario is excluded temporarily from the study.
4.4. DES for assessing risks mitigation scenarios
In order to formalize the hospital pharmaceutical SC, then to study the impact of different risk factors and mitigation actions on the system performance, ARENA software is chosen as a modeling and analysis tool. It is used to observe the effects of risk factors on performance measures for various scenarios.
The conceptual model of the hospital pharmaceutical supply chain (see Figure 3) consist in supplying the drugstore by consistent drugs from the public drugstore as
Table 2. Risks exposure levels.
Failure mode
Risks exposure level
High (H) Low (L)
Compliance problem (CP) 0.24 0.12 Lack of personnel (LP) 0.56 0.21 Time limit (TL) 0.42 0.18
654 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
the first activity then the next activity regroup the drugs stocking in the drugstore, the resupplying of the distribution sectors (internal and external drugstores) and the removal of drugs to the care units. The final activity consists of drugs unloading and cross-checking when they have arrived to the care units. Periodic checking of drugs is a separate activity that consists in checking the medicine cabinets by the medical staff in each care units. Based on the preference of the hospital managers, the corrective actions that could be implemented are estimated as follows:
Compliance problem = purchase a software package at a cost of 50 thousand dinars, ‘Td.’ Lack of personnel: recruitment of two nurses at a cost of 25 Td; Time limit = recruitment of an internal auditor at a cost of 12 Td.
Figure 3 illustrates the conceptual model of the pharmaceutical supply chain, the risk factors are already included in the conceptual model, the squares in red represent the risks and the squares in green represent the backup or the control of the risk.
The ‘Cilastatine Imipenen drug’ has been chosen in the implementation of simu- lation. As parameters of simulation, it should be noted that the arrival rate of the medicament (Cilastatine Imipenen) is a batch of 70 small bottle per day, while the operations at the drugstore department (drugs stocking, resupplying and removal) take four hours, and the operations at the care units (drugs unloading and cross- checking) take one hour.
Table 3. The 23 factorial design configurations (risk management scenarios).
Risk modes
Scenario Compliance Problem (CP)
Lack of Personnel (LP) Time Limit (TL)
1 L L L 2 L H H 3 L L H 4 L H L 5 H L L 6 H L H 7 H H L 8 (without any change) H H H
Reception of consistent drugs
Drugs stocking, resupplying and
removal
Drugs unloading and cross-checking
Periodic checking
Compliance problem risks Backup
Lack of personnel risks
Backup Time limit risks
Backup
Figure 3. Pharmaceutical supply chain conceptual model.
Journal of Risk Research 655
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
The objective of the simulation application is to understand the effects of failure mode levels on the performance measures. The criteria that were used for this purpose are as follows:
• Reliability: The compliance of drugs with care unit’s orders (in percentage). • Fluidity of drug circuit: Can be translated by the time delivery of drugs to the care units (in hours).
• Budget consistency: The appropriateness of the risk management scenario expenses with budget.
Global cost is calculated by adding the cost of each corrective action in the sce- nario. The estimation of the hospital managers is to consider that scenarios with a global cost under 65 Td are the most consistent with budget. However, scenarios with a global cost above 65 Td are the least consistent. The outcomes are input– output analytical relations (called the meta-models of the simulation model). Check- ing all possible factors levels combinations (factorial experimental design) requires 8 simulation runs; if each run is replicated three times we have 24 replications. The simulation results (means) are presented in Table 4.
The bold numbers in Table 4 show the best values of performance measures. The hospital managers would enhance the performance at the Reliability and the Fluidity of Drug; however, they aim at reducing the global cost in a way prefered scenarios are the most consistents with budget.
According the above experiments, major conclusions could be drawn as follows:
• Obviously, none of the scenarios under study has simultaneously performed the decision maker objective.
• Scenario 1 has simultaneously the best values of performance measures in the Reliability and the Fluidity of Drug but it’s the costliest one.
• The best values of performance measures in the Reliability are located in Scenario 1 and Scenario 4. However, the best values of performance measures in Fluidity of Drug are located in Scenario 1, Scenario 3, Scenario 5 and Scenario 6.
In terms of optimization, the results issued by the simulation offer an overview on the performance of each criterion individually, but don’t enable the hospital
Table 4. Performance measures of the risks mitigation scenarios.
Scenarios
Risk exposure levels
Reliability (%) Fluidity of drug (h)
Budget consistencyCP LP TL
Scenario 1 (CP: L, LP: L, TL: L) 0.12 0.21 0.18 72.66 5.27 87 (LC) Scenario 2 (CP: L, LP: H, TL: H) 0.12 0.56 0.42 51.77 5.73 50 (MC) Scenario 3 (CP: L, LP: L, TL: H) 0.12 0.21 0.42 51.77 5.27 75 (LC) Scenario 4 (CP: L, LP: H, TL: L) 0.12 0.56 0.18 72.66 5.73 62 (MC) Scenario 5 (CP: H, LP: L, TL: L) 0.24 0.21 0.18 63.26 5.27 37 (MC) Scenario 6 (CP: H, LP: L, TL: H) 0.24 0.21 0.42 44.65 5.27 25 (MC) Scenario 7 (CP: H, LP: H, TL: L) 0.24 0.56 0.18 63.26 5.73 12 (MC)
Note: LC: Least Consistent; MC: Most Consistent.
656 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
managers to choose the best scenario. That’s why a decision aid multicriteria method (AHP method) has been used to assist the hospital managers to choose the compromise scenario. The results obtained in this section will be taken into account by the managers to draw their preferences to the scenarios compared to the criteria.
4.5. AHP method for classifying risk management scenarios
AHP method was carried out thanks to the managers’ knowledge by perception or preference, through an evaluation process. Given the hierarchical structure (Figure 4), the managers were asked to perform a ranking of the scenarios based on Saaty scores; they also had to perform pair wise comparisons with less subjectivity and more consistency.
Analytic hierarchical model (Figure 4) consists of an objective which is the selection of the best risk mitigation scenario. The first job of the hospital managers was to identify the pertinent criteria to be used in the scenarios evaluation process. The three performance measures used in simulation are kept as part of criteria in this analytical section (budget consistency, reliability, and budget consistency). However, non-quantifiable measures (flexibility improvement [FI] and bullwhip effect control) have been added as pertinent criteria in selecting the best scenario.
• The FI. The improvement of the hospital pharmacy service ability to change. • The bullwhip effect control (BEC). The control of the amplifications through the pharmaceutical supply chain caused by the care units orders variability.
To perform AHP application, the hospital managers had to compare all pairs of criteria and scenarios using a ratio scale. The scores admitted in the study are the compromise of the evaluations of the three hospital managers. Expert choice soft- ware has been used to solve the decision problem. The criteria priorities calculated by following the standard AHP are given in Table 5. The consistency ratio for the five criteria, calculated following the standard AHP, resulted in an acceptable value of 0.05. The scenario priority for each criterion is calculated by the expert choice software as given in Table 6.
The overall rank of each scenario is calculated by multiplying a criterion priority (Table 5) by a scenario priority for each criterion (Table 6) and summing the results for all the criteria. The final scenario ranking is given below (Table 7).
In this case, budget consistency criterion was found to have the highest priority, that is, importance, while FI criterion had the lowest. The criterion regarding the
Alternatives actions
Criteria (Objectives)
Overall Goal Selection of the SCRM scenario
BC RE FDC FI BEC
S1 (Cp=L;Lp=L;Tl=L)
S2 (Cp=L;Lp=H;Tl=H)
S3 (Cp=L;Lp=L;Tl=H)
S4 (Cp=L;Lp=H;Tl=L)
S5 (Cp=H;Lp=L;Tl=L)
S6 (Cp=H;Lp=L;Tl=H)
S7 (Cp=H;Lp=H;Tl=L)
Figure 4. Hierarchical model for selection of the SCRM scenario.
Journal of Risk Research 657
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
reliability had a higher priority than the criterion related to fluidity of drug circuit. The fifth scenario was found to have the highest priority based on all criteria. The seventh scenario had a higher priority than the sixth one for all criteria. Overall results indicate that the fifth scenario has a ranking well higher than either the seventh or the sixth scenarios. The seventh scenario has a slight advantage on the sixth one.
4.6. Results analysis and desirability optimization
Based on the output of the AHP method, further analysis is conducted for the fundamental understanding of failure modes into risk management. Due to multiple decision variables involved, main effects plots (Figure 5) and second-order interac- tion effects plots (Figure 6) are obtained using Minitab 14 software to demonstrate the magnitudes of each failure mode.
In this case, it seems that DOE is pointing out a little interaction between the factors. But, in other situations, DOE can give more interesting results. After
Table 5. Performance measures score.
Performance measures (criteria) Priority
Budget consistency 0.510 Reliability 0.226 Fluidity of drug circuit 0.161 Flexibility improvement 0.049 Bullwhip effect control 0.055
Table 6. Scenario relative weight compared to criterion.
S1 S2 S3 S4 S5 S6 S7
Budget consistency 0.026 0.104 0.053 0.052 0.207 0.264 0.294 Reliability 0.362 0.048 0.227 0.076 0.171 0.082 0.035 Fluidity of drug circuit 0.358 0.042 0.148 0.100 0.245 0.066 0.042 Flexibility improvement 0.341 0.050 0.128 0.151 0.194 0.061 0.074 Bullwhip effect control 0.340 0.068 0.184 0.164 0.148 0.048 0.048
Table 7. Scenarios ranking.
Scenarios Ranking
Scenario 1 (CP: L, LP: L, TL: L) 0.172 Scenario 2 (CP: L, LP: H, TL: H) 0.079 Scenario 3 (CP: L, LP: L, TL: H) 0.112 Scenario 4 (CP: L, LP: H, TL: L) 0.074 Scenario 5 (CP: H, LP: L, TL: L) 0.202 Scenario 6 (CP: H, LP: L, TL: H) 0.178 Scenario 7 (CP: H, LP: H, TL: L) 0.183 Scenario 8 (CP: H, LP: H, TL: H) 0.000
Note: Bold values represent the largest ranking score.
658 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
planning the experiments scenarios and identifying the most important failure modes of the model, final score, which is calculated based on AHP method, is used as input data for the desirability optimization. This optimization tool is integrated in many software packages such as Minitab, Statistica, JMP software, etc.
Applying Equation (2) for only one response measure, the individual desirability is equal to 0.948. The response optimization consists in determining how the solu- tion has maximized the final score. Composite desirability has a range of zero to one. One represents the ideal case; zero indicates that response measure is outside their acceptable limits. Composite desirability is the weighted geometric mean of the individual desirability for the responses as presented in Equation (3). The composite desirability in this case (mono-objective optimization) is equal to the
Figure 5. Main effects of failure modes levels on final score.
Figure 6. Interaction effects of failure modes levels on final score.
Journal of Risk Research 659
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
individual desirability. To obtain this desirability, we would set the failure mode levels at the values shown under global solution in the Figure 7. That is, CP would be set at 0.235, LP would be set at 0.27, and TL would be set at 0.18. The final score of the obtained scenario (solution) is 0.2126 instead the maximum of all scenarios (0.202) which correspond to Scenario 5 (see Table 7).
5. Conclusion
This paper presents a new concrete and systematic approach for the SCRM to assist supply chain decision makers to risk identification, assessment and management. The proposed approach is based on combining many techniques and methods include the following. (1) FMECA to identify risk and its current location and assess risks. (2) DOE to design risks mitigation and action scenarios. (3) DES to assess risks mitigation action scenario. (4) AHP method to evaluate risk manage- ment scenarios. (5) Desirability optimization to perform the best risk scenario.
For a validation purpose, a base case has been developed by applying the pro- posed approach to a real hospital pharmaceutical supply chain. The contribution of the article is certainly very important to this base case. But in terms of risk and risk management, it is not employment of other companies in the pharmaceutical indus- try. Each company must carefully follow the five steps of the proposed approach, leading to the process management of these critical own risk.
Our perspective should give more extensive testing on different problems to support the proposed approach application in supply chain management. Details and others real benefits of the proposed approach into other real-life case study will be presented at our subsequent publications. Here we considered only the downstream of the pharmaceutical supply chain. However, as the size of the sup- ply chain networks increase, the dependency between the entities in the supply chain network increases as well. Thus, the concept of risk sharing becomes an
Figure 7. Desirability optimization for risk minimization.
660 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
issue of pharmaceutical SCRM and can be considered as a second perspective for the current paper.
Acknowledgements The authors would like to express their thanks to all the personal of the drugstore department of Habib Bourguiba Hospital, especially the head of the department, for their support during this research and for their help in the data collecting stage. They would also like to thank the managers’ team of the hospital for valuable help to realize this study. Lastly, the authors warmly thank the editor and the anonymous reviewers for their comments and recommendations.
References Aven, T., and O. Renn. 2009. “On Risk Defined as an Event where the Outcome is
Uncertain.” Journal of Risk Research 12 (1): 1–11. Beamon, B. M. 1998. “Supply Chain Design and Analysis: Models and Methods.”
International Journal of Production Economics 55 (3): 281–294. Bowles, J. B., and R. D. Bonnell. 1995. “Failure Mode and Effects Analysis.”
In Ann. Reliab. & Maint. Symp, Tutorial Notes, Washington DC, USA, January 16–19, 1995.
Bowles, J. B., and E. Peldez. 1995. “Fuzzy Logic Prioritization of Failures in a System Failure Mode, Effects and Criticality Analysis.” Reliability Engineering and System Safety 50 (2): 203–213.
Brindley, C. 2004. Supply Chain Risk. Aldershot: Ashgate. Cagliano, A. C., A. De Marco, S. Grimaldi, and C. Rafele. 2012. “An Integrated Approach
to Supply Chain Risk Analysis.” Journal of Risk Research 15 (7): 817–840. Chopra, S., and M. S. Sodhi. 2004. “Managing Risk to Avoid Supply Chain Breakdown.”
MIT Sloan Management Review 46 (1): 53–61. Christopher, M., and H. Lee. 2004. “Mitigation Supply Chain through Improved
Confidence.” International Journal of Physical Distribution and Logistics Management 34 (5): 388–396.
Cigolini, R., and T. Rossi. 2006. “A Note on Supply Risk and Inventory Outsourcing.” Production Planning and Control 17 (4): 424–437.
Craighead, C. W., J. Blackhurst, M. J. Rungtusanatham, and R. B. Handfield. 2007. “The Severity of Supply Chain Disruptions: Design Characteristics and Mitigation Capabili- ties.” Decision Science 38 (1): 131–156.
Datta, S., D. P. Graham, N. Sagar, P. Doody, R. Slone, and O. P. Hilmola. 2009. “Forecast- ing and Risk Analysis in Supply Chain Management: GARCH Proof of Concept.” In Managing Supply Chain Risk and Vulnerability, edited by T. Wu, and J. Blackhurst, 187–203. London: Springer.
Derringer, G. C., and R. Suich. 1980. “Simultaneous Optimization of Several Response Variables.” Journal of Quality Technology 12: 214–219.
Diabat, A., K. Govindan, and V. V. Panicker. 2012. “Supply Chain Risk Management and its Mitigation in a Food Industry.” International Journal of Production Research 50 (11): 3039–3050.
Dong. J., W. Wang, and T. Wu. 2009. “A Generalized Simulation Framework for Responsive Supply Network Management.” In Managing Supply Chain Risk and Vulnerability, edi- ted by T. Wu, and J. Blackhurst, 67–88. London: Springer.
Elmsalmi, M., and W. Hachicha. 2013. “Risks Prioritization in Global Supply Networks Using MICMAC Method: A Real Case Study.” International Conference on Advanced Logistics and Transport (ICALT’2013), IEEE proceeding in Sousse Tunisia, May 29–31, 394–399.
Giannakis, M., and M. Louis. 2011. “A Multi-Agent Based Framework for Supply Chain Risk Management.” Journal of Purchasing & Supply Management 17 (1): 23–31.
Gilchrist, W. 1993. “Modelling Failure Modes and Effect Analysis.” International Journal of Quality and Reliability Management 10 (5): 16–23.
Journal of Risk Research 661
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
Giunipero, L. C., and R. A. Eltantawy. 2004. “Securing the Upstream Supply Chain: A Risk Management Approach.” International Journal of Physical Distribution and Logistics Management 34 (9): 698–713.
Goh, M., J. Y. S. Lim, and F. Meng. 2007. “A Stochastic Model for Risk Management in Global Supply Chain Networks.” European Journal of Operational Research 182 (1): 164–173.
Goh, M., and F. Meng. 2009. “A Stochastic Model for Supply Chain Risk Management Using Conditional Value at Risk.” In Managing Supply Chain Risk and Vulnerability, edited by T. Wu, and J. Blackhurst, 141–157. London: Springer.
Goyal, S. K., and B. C. Giri. 2001. “Recent Trends in Modeling of Deteriorating Inventory.” European Journal of operational research 134 (1): 1–16.
Hachicha, W., A. Ammeri, F. Masmoudi, and H. Chabchoub. 2010. “A Multi-Product Lot Size in Make-To-Order Supply Chain Using Discrete Event Simulation and Response Surface Methodology.” International Journal of Services, Economics and Management 2 (3–4): 246–266.
Hahn, G. J., and H. Kuhn. 2012. “Value-Based Performance and Risk Management in Supply Chains: A Robust Optimization Approach.” International Journal of Production Economics 139 (1): 135–144.
Hallikas, J., I. Karvonen, U. Pulkkinen, V. M. Virolainen, and M. Tuominen. 2004. “Risk management processes in supplier networks.” International Journal of Production Eco- nomics 90 (1): 47–58.
Harland, C., R. Brenchley, and H. Walker. 2003. “Risk in Supply Networks.” Journal of Purchasing & Supply Management 9 (2): 51–62.
Harrington, J. 1965. “The Desirability Function.” Industrial Quality Control 21: 494–498. Huang, C. C., and T. L. Tseng. 2009. “Supply Chain Risk Management: Annotation of
Knowledge Using a Semi-Structured Knowledge Model.” In Managing Supply Chain Risk and Vulnerability, edited by T. Wu, and J. Blackhurst, 205–232. London: Springer.
Jüttner, U. 2005. “Supply Chain Risk Management: Understanding the Business Require- ments from a Practitioner Perspective.” The International Journal of Logistics Manage- ment 16 (1): 120–141.
Jüttner, U., H. Peck, and M. Christopher. 2003. “Supply Chain Risk Management: Outlining an agenda for Future Research.” International Journal of Logistics: Research and Appli- cations 6 (4): 197–210.
Kleijnen, J. P. C., and M. T. Smits. 2003. “Performance Metrics in Supply Chain Manage- ment.” The Journal of the Operational Research Society 50 (5): 507–514.
Lofstedt, R. E. 2006. “How We can Make Food Risk Communication Better: Where Are We and Where Are We Going?” Journal of Risk Research 9 (8): 869–890.
Macgillivray, B. H., J. V. Sharp, J. E. Strutt, P. D. Hamilton, and S. J. T. Pollard. 2007. “Benchmarking Risk Management within the International Water Utility Sector. Part II: A Survey of Eight Water Utilities.” Journal of Risk Research 10 (1): 105–123.
Melnyk, S. A., A. Rodrigues, and G. L. Ragatz. 2009. “Using Simulation to Investigate Supply Chain Disruptions.” International Series in Operations Research & Management Science 124 (1): 103–122.
Minahan, T. 2005. The Supply Risk Benchmark Report. Boston, MA: Aberdeen Group. Mohd, N. F. 2009. Managing Supply Chain Risk and Vulnerability-Tools and Methods for
Supply Chain Decision Makers. London: Springer. Montgomery, D. C. 2005. Design and Analysis of Experiments. 6th ed. New York, NY: John
Wiley & Sons, 160–335. Neiger, D., K. Rotaru, and L. Churilov. 2009. “Supply Chain Risk Identification with Value
Focused Process Engineering.” Journal of Operations Management 27 (2): 154–168. Norrman, A., and U. Jansson. 2004. “Ericsson’s Proactive Supply Chain Risk Management
Approach after a Serious Sub-Supplier Accident.” International Journal of Physical Dis- tribution and Logistics Management 34 (5): 434–456.
Norrman, A., and B. Lindroth. 2002. “Supply Chain Risk Management: Purchasers’ vs. Planners’ Views on Sharing Capacity Investment Risks in the Telecom Industry.” Paper presented at the 11th International IPSERA conference, Enschede, The Netherlands.
662 H. Elleuch et al.
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
Pfohl, H. C., H. Kohler, and D. Thomas. 2010. “State of the Art in Supply Chain Risk Management Research: Empirical and Conceptual Findings and a Roadmap for the Implementation in Practice.” Logistic Research 2 (1): 33–44.
Qiang, Q., A. Naurney, and J. Dong. 2009. “Modeling of Supply Chain Risk Under Disrup- tions with Performance Measurement and Robustness Analysis.” In Managing Supply Chain Risk and Vulnerability, edited by T. Wu, and J. Blackhurst, 91–111. London: Springer.
Ramanjaneyulu, K., and S. Sasamal. 2008. “Condition Evaluation of Existing Reinforced Concrete Bridges Using Fuzzy Based Analytic Hierarchy Approach.” Expert Systems with Applications 35 (3): 1430–1443.
Saaty, T. L. 1980. The Analytic Hierarchy Process. New York: McGraw Hill. Smillie, L., and A. Blissett. 2010. “A Model for Developing Risk Communication Strategy.”
Journal of Risk Research 13 (1): 115–134. Sodhi, M. S., B. G. Son, and C. S. Tang. 2012. “Perspectives on Supply Chain Risk Man-
agement.” International Journal of Production and Operations Management 21 (1): 1–13.
Sodhi, M. S., and C. S. Tang. 2009. “Managing Supply Chain Disruption via Time-Based Risk Management.” In Managing Supply Chain Risk and Vulnerability, edited by T. Wu, and J. Blackhurst, 29–40. London: Springer.
Svensson, G. 2000. “A Conceptual Framework for the Analysis of Vulnerability in Supply Chains.” International Journal of Physical Distribution & Logistics Management 30 (9): 731–749.
Tang, C. 2006. “Perspectives in Supply Chain Risk Management.” International Journal of Production Economics 103 (2): 451–488.
Tang, O., and S. N. Nurmaya Musa. 2011. “Identifying Risk Issues and Research Advancements in Supply Chain Risk Management.” International Journal of Production Economics 133 (1): 25–34.
Tang, C., and B. Tomlin. 2008. “The Power of Flexibility for Mitigating Supply Chain Risks.” International Journal of Production Economics 116 (1): 12–27.
Thun, J. H., and D. Hoeing. 2011. “An Empirical Analysis of Supply Chain Risk Manage- ment in the German Automotive Industry.” International Journal of Production Econom- ics 131 (1): 242–249.
Trkman, P., and K. McCormack. 2009. “Supply Chain Risk in Turbulent Environnements – A Conceptual Model for Managing Supply Network Risk.” International Journal of Pro- duction Economics 119 (2): 247–258.
Tummala, R., and T. Schoenherr. 2011. “Assessing and Managing Risks Using the Supply Chain Risk Management Process (SCRMP).” Supply Chain Management: An Interna- tional Journal 16 (6): 474–483.
Tuncel, G., and G. Alpan. 2010. “Risk Assessment and Management for Supply Chain Networks: A Case Study.” Computers in Industry 61 (3): 250–259.
Verbano, C., and K. Venturini. 2011. “Development Paths of Risk Management: Approaches, Methods and Fields of Application.” Journal of Risk Research 14 (5): 519–550.
Vilko, J. P. P., and J. M. Hallikas. 2012. “Risk Assessment in Multimodal Supply Chain.” International Journal of Production Economics 140 (2): 586–595.
Wagner, S. M., and N. Neshat. 2010. “Assessing the Vulnerability of Supply Chains Using Graph Theory.” International Journal Production Economics 126 (1): 121–129.
Waters, D. 2007. Supply Chain Risk Management: Vulnerability and Resilience in Logistics. London: Kogan Page Limited.
Wu, T., and J. Blackhurst. 2009. “Book Introduction.” In Managing Supply Chain Risk and Vulnerability, 1–5. London: Springer-Verlag London Limited.
Yang, Y. C. 2011. “Risk Management of Taiwan’s Maritime Supply Chain Security.” Safety Science 49 (3): 382–393.
Zsidisin, G., L. Ellram, J. Carter, and J. Cavinato. 2004. “An Analysis of Supply Chain Assessment Techniques.” International Journal of Physical Distribution & Logistics Management 34 (5): 397–413.
Journal of Risk Research 663
D ow
nl oa
de d
by [
E m
br y-
R id
dl e
A er
on au
ti ca
l U
ni ve
rs it
y] a
t 13
:5 3
27 M
ar ch
2 01
6
