Content area
Purpose
This work analyzes a pharmaceutical supply chain (PSC) in terms of supply chain visibility (SCV). The current good distribution practice (GDP) guideline demands increased visibility from firms. The purpose of this paper is to propose a solution for SCV enhancements based on automatic identification (Auto-ID) technologies.
Design/methodology/approachThe authors qualitatively analyze data from ten case studies of actors in a PSC. A review of Auto-ID technologies supports the derivation of solutions to enhance SCV.
FindingsThis work shows that the functionalities of Auto-ID technologies offered by current practical monitoring solutions and challenges created by the GDP guideline necessitate further SCV enhancements. To enhance SCV, the authors propose three solutions: securPharm with passive radio frequency identification tags, transport containers with sensor nodes, and an SCV dashboard.
Research limitations/implicationsThis study is limited to a PSC in Germany and is therefore not intended to be exhaustive. Thus, the results serve as a foundation for further analyses.
Practical implicationsThis study provides an overview of the functionality of Auto-ID technologies. In juxtaposition with the influence of the GDP guideline, the use of our Auto-ID-based solutions can help to enhance SCV.
Originality/valueThis work analyzes a PSC in Germany, with consideration given to the influence of current legislation. Based on a multiple-case-study design, the authors derive three Auto-ID-based solutions for enhancing SCV.
Introduction
In the high-selling pharma market (Institute for Healthcare Informatics, 2012), a combination of product counterfeiting, errors in the production and distribution of temperature-sensitive drugs, the theft of expensive products, and faulty drugs resulting from incorrect ingredients lead to multi-billion-dollar revenue losses throughout the world and constitute a serious threat to public health. Against this background, various guidelines support the actors of pharmaceutical supply chains (PSCs) to ensure correct drug handling (EU, 2013; WHO, 2014). The latest European Union guideline for the good distribution practice (GDP) of medical products for human use has been effective since September 2013 and challenges the actors to ensure and prove correct drug handling throughout the entire PSC (Spiggelkötter, 2013). Consequently, actors in PSCs are forced to enhance supply chain visibility (SCV).
SCV can be defined as “the extent to which actors within a supply chain have access to or share information which they consider as key or useful to their operations and which they consider will be of mutual benefit” (Barratt and Oke, 2007, p. 1218). Such visibility helps actors to achieve an enhanced overview of material flows within complex supply chains (Christopher and Lee, 2004; Christopher and Towill, 2001). As Barratt and Oke (2007) mentioned, SCV results from information being shared between supply chain actors (see also Christopher and Lee, 2004) and can have a positive effect on supply chain management (Delen et al., 2007; Jonsson and Mattsson, 2013; Lumsden and Mirzabeiki, 2008). However, SCV and information sharing require the adoption of monitoring solutions based on automatic identification (Auto-ID) technologies (Angeles, 2005; Kärkkäinen and Holmström, 2002; Lee and Özer, 2007), which enable information gathering, sharing (Attaran, 2007; Spekman and Sweeney, 2006) and an overview of material flows (Holmström et al., 2010; McFarlane and Sheffi, 2003).
Against this background, we argue that the functional capabilities of current technologies for practical monitoring solutions, such as barcodes, data matrix codes and data loggers, are insufficient to enhance the pharmaceutical SCV and to ensure and prove that drug handling satisfies the GDP guideline. To support this assertion, we cite other studies on Auto-ID technologies with advanced functional capabilities that are promising in relation to enhancing SCV (Prockl and Pflaum, 2012; Sanchez Lopez et al., 2011; Zheng and Jamalipour, 2009). However, to provide Auto-ID-based solutions for enhancing SCV, it is necessary to conduct a detailed examination of the functional requirements of PSC process activities under the influence of the GDP guideline, and of the functional capabilities of Auto-ID technologies. This study aims to provide such an examination, leading to the following research question:
How can SCV be enhanced using Auto-ID technologies within a PSC?
To answer this research question, we begin by reviewing the SCV literature and its relationship with Auto-ID technology. Next, we present the design of the empirical research scheme based on ten case studies, from which we derive Auto-ID-based solutions for enhancing SCV. To derive the solutions, this study takes a supply chain perspective by considering the actors and their physical process activities (see Hinkka et al., 2012, 2015; Spekman and Sweeney, 2006). This perspective is used because we will only obtain real enhancements in SCV by considering a supply chain perspective (Christopher, 2011; Spekman et al., 1998). Our solutions provide both insights into the implementation of drug-centric monitoring with accurate and secure information handling for SCV, and information that will help to prove that drug handling satisfies the GDP guideline. Finally, this paper emphasizes the importance of Auto-ID technologies for information gathering and sharing in supply chains (Christopher and Holweg, 2011), and is, to the best of our knowledge, the first study that analyzes the influence of the current GDP guideline (legislation) on SCV in a PSC context.
Review of SCV and Auto-ID technology
This section reviews the research on SCV and Auto-ID-based monitoring solutions. The goal is to clarify the essence of SCV and its relationship with Auto-ID technologies. Following Denyer and Tranfield (2009), a keyword search was conducted in the journal databases EBSCO and ScienceDirect to harvest the maximum possible relevant research published to date. SCV, Auto-ID and intelligent products were used as keywords because SCV is central to our study and is realized by products equipped with Auto-ID technology (i.e. intelligent products). Moreover, we also availed ourselves of the references within these first-tier papers to exploit further relevant publications. Publications not sufficiently related to our focus were excluded based on their title, abstract, and, if necessary, the full publication.
SCV
Various definitions of SCV are available within academia. Kaipia and Hartiala (2006, p. 377) define SCV as “[…] the sharing of all relevant information between supply chain partners, also over echelons in the chain,” which emphasizes that SCV only results from the sharing of meaningful information between all relevant supply chain actors. Barratt and Oke (2007) define SCV as the extent to which supply chain actors have access to useful and mutually beneficial information. Thus, SCV emerges from shared information, which offers value and advantages for supply chain management. This information leads to visibility, which in turn improves operational activities, such as inventory management, operational effectiveness, and efficiency. A further definition of SCV is given by Francis (2008), who emphasizes the types of information involved in the entities moving along the supply chain, and the timely delivery of event information. The types of information include identity, location, and status quo. Thus, this definition details which information should be shared in order to achieve SCV. Whereas these definitions focus on the value of information sharing between supply chain actors and the types of information shared, Klueber and O’Keefe (2013) introduce the concept of requisite SCV. They define requisite SCV as the provision of information from different supply chain stakeholders and access to this information for mutual business benefit. Furthermore, this concept stresses the need to assess visibility in a supply chain and to identify aspects that influence SCV. With this definition, Klueber and O’Keefe (2013) extend the perspective of information sharing from supply chain actors to supply chain stakeholders and thereby the scope of such sharing. However, the authors found no single, consistent definition of SCV within the literature (see also Caridi et al., 2014). Table I shows the aforementioned definitions.
With regard to our research purpose and methodology, it is necessary to provide a conceptual framework for SCV to support our work. Therefore, the authors synthesized the aforementioned definitions of SCV to devise dimensions that operationalize SCV and represent cornerstones of a conceptual framework. Basically, the authors follow the definition of Barratt and Oke (2007) and thus consider the aspect that object information should be permanently available to all supply chain actors (which is beneficial to the entire supply chain and can, for example, help to achieve proof of GDP-consistent drug handling). As this aspect is also represented in the definitions of Kaipia and Hartiala (2006) and Klueber and O’Keefe (2013), the authors derived availability of information as one dimension of a conceptual framework. As mentioned above, this study aims to enhance SCV, which should help to provide proof of GDP-consistent drug handling. Therefore, a conceptual framework for SCV must also consider relevant types of information. In this sense, the authors incorporated the definition given by Francis (2008) and added the aspects of identity, position, and status quo as further dimensions for SCV. The authors did not give further consideration to the requisite supply chain visibility concept given by Klueber and O’Keefe (2013) because it is generic, strategic, and qualitative in nature. It therefore does not support the derivation of detailed supply chain related solutions to enhance SCV, especially from a supply chain perspective. Table II shows our dimensions of SCV, with explanations.
Furthermore, the aforementioned insights into SCV definitions emphasize that information sharing is an important prerequisite for SCV (see also Barratt and Oliveira, 2001; Bartlett et al., 2007; Christopher and Lee, 2004; Holcomb et al., 2011; Lamming et al., 2001; Steinfield et al., 2011). Such information sharing, in turn, requires the use of monitoring solutions with suitable Auto-ID technology (Kärkkäinen and Holmström, 2002) and related information technology (IT) systems to monitor material flows (Hinkka et al., 2012, 2015) and to share valuable information among supply chain actors (Attaran, 2007; Delen et al., 2007; Holmström et al., 2010; Lee and Özer, 2007). Therefore, SCV may be seen as resulting from information gathering and sharing and from applying an appropriate Auto-ID technology. Moreover, to be of value, information must relate to a certain purpose in supply chain management.
Auto-ID technologies and their functions as an enabler for SCV
Auto-ID technologies were originally developed to create a global network of physical objects with a digital identity in information systems (Internet of Things). This should provide detailed information about the flow of materials and improve supply chain processes (McFarlane and Sheffi, 2003). To vertically integrate information and material flows, physical objects are equipped with Auto-ID technologies. Examples of such objects are intelligent containers, which can record temperature and communicate their position (Lang et al., 2011).
Auto-ID technologies are defined as technologies that enable automated, accurate, and detailed identification of objects (Kärkkäinen and Holmström, 2002; McFarlane and Sheffi, 2003), with the possibility of additional acquisition of information for enhanced data quality and vertical integration between information systems and material flows (Holmström et al., 2010; Meyer et al., 2009). In this sense, such technologies help to improve efficient material handling and customization (Kärkkäinen and Holmström, 2002). Auto-ID technologies include radio frequency identification (RFID), RFID with integrated sensors and wireless sensor networks with sensor nodes (WSNs), and focus on Auto-ID and data acquisition (Meyer et al., 2009; Miles, 2008). Sensor nodes are small computer tags that can communicate with one another and build a sensor network with redundant communication paths (Zheng and Jamalipour, 2009). However, based on their technological characteristics, these technologies may contain the following functions (Lempert and Pflaum, 2011; Prockl and Pflaum, 2012):
identification (unique identification of object);
locating (accurate information on object position);
sensors (current status of object);
communication (accessibility of object);
data storage (retention of object history); and
logic (recognition of critical events of object).
Regarding our SCV dimensions and the aforementioned functions, Auto-ID technology is theoretically capable of realizing the dimensions. As our SCV dimension of availability demands information sharing between supply chain actors, Auto-ID technology with a communication function can provide object information for exchange. Additionally, availability can be supported by a data storage function that considers storage volume for object information. Without such storage volume, information can be neither stored nor shared. Moreover, an identification function can support the dimension of identity by providing identity information about an object. Similarly, a locating function can support the dimension of position by providing accurate information about the object’s position. Considering the possibility of a sensors function, the dimension of status quo would be supported because sensors can provide object- or environmental-related information. By recognizing critical events of an object, a logic function can further provide useful information about the dimension status quo. For example, if a logic function is able to recognize a temperature transgression of a temperature-sensitive object, then the status of the object’s quality can be indicated accordingly. Consequently, functions of Auto-ID technology help to realize our SCV dimensions and therewith to create SCV.
Furthermore, the presented Auto-ID functions can be specified with quantitative degrees of performance (assessed on a scale from 0.25 to 1; Prockl and Pflaum, 2012). Table III shows the SCV dimensions, the related Auto-ID functions, and degrees of performance. In this sense, these degrees can be used to operationalize Auto-ID functions and the realization of SCV dimensions. In other words, if an Auto-ID technology allows the provision of object identity at product level by identification, then the degree of performance for the identification function of this Auto-ID technology is unity. Consequently, the SCV dimension of identity would be completely realized. If a function of an Auto-ID technology does not satisfy any related degrees of performance, then the degree of performance of this function is zero. Consequently, the related SCV dimension is not realized. Moreover, if the degree of performance of each function of an Auto-ID technology is unity, then the given technology allows the realization of all SCV dimensions as far as possible, and thereby full SCV in terms of our conceptualization. Finally, a net diagram can visualize the degrees of performance for each function of an Auto-ID technology. The resulting net area then shows how the related SCV dimensions are realized. As Auto-ID technologies and their functions enable SCV, the following sections provide an overview of Auto-ID-based monitoring solutions. This overview results in a functional evaluation of the introduced Auto-ID technologies and a presentation of the theoretical deficiencies and benefits of Auto-ID technologies for realizing the SCV dimensions.
Overview of Auto-ID-based monitoring solutions for SCV
The characteristics of Auto-ID technologies have led to the development of manifold solutions for monitoring supply chains, in which RFID-based solutions are predominant (Hinkka et al., 2015). Kärkkäinen et al. (2003) proposed a system to monitor and control products in investment projects. By endowing products with RFID tags, these products obtain a unique identity and can communicate their information to tracking services (availability). A peer-to-peer information system with distributed software techniques supports the exchange of information between supply chain actors. Kärkkäinen et al. (2004) present a forwarder-independent approach to monitoring objects in short-term supply networks to overcome the difficulties associated with manual information queries and system integration. This solution is similar to that of Kärkkäinen et al. (2003) but applies distributed software components and peer-to-peer information sharing (availability). For the grocery supply chain, Martínez-Sala et al. (2009) propose using intelligent returnable packaging and transport units in combination with an information system to track transport units. While the transport unit is equipped with an RFID tag, the system gathers all the information on the units using RFID technology (identity). This solution enables transport-unit or package-level monitoring and traceability, and the possibility to save additional information at transport-unit level (availability). For the book supply chain, Hinkka et al. (2012) offer recommendations for retailers about where and when to tag books with RFID technology for implementation throughout the supply chain. Their study shows that speculative book tagging as early as possible throughout the entire chain increases supply chain agility by improved sharing of detailed information at object level (availability, identity). Mason et al. (2012) propose an inventory management system based on RFID technology to monitor returnable gas cylinders. Gas cylinders are equipped with an RFID sensor tag capable of peer-to-peer communication (availability) to transmit not only information about the cylinder (identity) and status quo but also a logical mapping information (position) to an RFID reader. Inventory software manages all the information collected and facilitates cylinder monitoring. Hafliðason et al. (2012) present a study on how WSNs can support temperature monitoring in food supply chains to assist decision-support systems. To record both food and ambient temperature, sensors are added to the shipments (status quo). This recorded information enables the identification of failures at shipments within the food supply chain (identity, position), temperature alerts, and the real-time monitoring of shipped products (availability).
Overview of current practical monitoring solutions for SCV
In PSCs, data loggers are used for temperature monitoring (DHL, 2015). Data loggers are small objects that measure temperature and store temperature profiles (status quo). An integrated sensor captures temperature data at predefined intervals and saves this data in the data logger’s memory (availability). The data loggers are then put into the transportation accessories of the drugs to monitor temperature profiles. A data logger records a temperature profile but does not protect the drugs against excessively high or low temperatures. In other words, a logic or regulating function for temperature control is missing.
To monitor deliveries along entire PSCs, barcodes are used to identify and control drugs (Hansen and Gillert, 2008). However, in view of its limited information density, the barcode cannot prevent counterfeit products from entering the supply chain. As a result of the simplicity and benefits of the data matrix code (securPharm, 2014), the pharmaceutical industry demands its use, which allows for identification at product level (identity) and therefore provides greater visibility. A solution associated with the data matrix code is securPharm, which is defined as a drug verification system whereby unique identifiers are stored in a repository and are accessible (availability) by national competent authorities (securPharm, 2014). The development of this system is being expedited by the EU, which demands a secure system against product counterfeiting for all EU member states as of 2017. Medicine labeling consists of applying a data matrix code (standardized ISO coding, ISO/IEC 16022) to ensure global and unique readability. The code consists of a product number, charge number, expiration date, and unique serial number. To allow the verification process, the data matrix code is printed on the product packaging. In parallel, the code is stored in a central database (A) of all participating product manufacturers. Furthermore, the code is copied onto a second database (B) that only pharmacies and wholesalers can access. Additionally, the system entails separate storage and anonymized retrieval sections, which facilitate the protection of sensitive business information and personal data. This separation also ensures that no stakeholder can access the data of other stakeholders. For example, pharmacies can verify the authenticity of product packaging by scanning the product. Thereafter, the information from the code is compared with information from database (B). For an original medical product, scan information and database (B) information are equivalent.
Functionality profiles of reviewed Auto-ID technologies for realizing the SCV dimensions
Table IV presents an overview of the functionalities of the Auto-ID technologies evaluated by using the degrees of performance of functions given in Table III. The preceding literature review provided the information for this overview. In addition, Auto-ID technology information from Finkenzeller (2010), Prockl and Pflaum (2012), Lempert and Pflaum (2011), and Zheng and Jamalipour (2009) supported this evaluation. We thus evaluated the degree of performance of each Auto-ID function shown in Table III for each Auto-ID technology presented in the preceding overview.
Deficiencies to realize the SCV dimensions
The net diagrams in Table IV show the reviewed Auto-ID technologies with their functions and degrees of performance. These net diagrams demonstrate that, in particular, the current practical monitoring solutions exhibit functional deficiencies to realize the SCV dimensions. Specifically, the data matrix code of securPharm and the barcode do not support a sensors and logic function. These missing functions therefore prevent the realization of the SCV dimension of status quo. In addition, barcode and data matrix code do not offer rewritable data storage, which impedes the availability dimension, as no further information can be added for information sharing, especially at object level. The data logger, in turn, does not offer an identification function, which impedes the realization of the identity dimension. Additionally, the missing logic function of a data logger negatively influences the realization of the status quo dimension, as no events of objects can be recognized. Consequently, the review shows that the functions of Auto-ID technologies from the current practical monitoring solutions are not sufficient to realize all dimensions of SCV.
A detailed consideration of the net diagrams of RFID, RFID with sensor, and WSNs shows that the identification, locating, communication, and data storage functions are at least as powerful as the analogous functions of a data logger, a barcode, or a data matrix code. Although RFID technology lacks a logic and sensors function, which prevents it from realizing the SCV dimension of status quo, these Auto-ID technologies are theoretically better suited to the realization of the SCV dimensions. Moreover, RFID with sensor differentiates itself from data logger, barcode, or data matrix code by virtue of its identification and sensor functions. As RFID with sensor lacks only a logic function, realization of the SCV dimension of status quo would only be diminished by the missing recognition of critical object events. Moreover, WSNs are basically capable of all required Auto-ID functions and thereby allow the realization of all SCV dimensions.
In this context, the authors emphasize that Auto-ID technologies from the current practical monitoring solutions exhibit clear deficiencies that prevent them from realizing all SCV dimensions. Therefore, we argue for an examination of the PSC under consideration of, in particular, the Auto-ID technologies RFID, RFID with sensor, and WSN to enhance SCV. However, as these technologies do not completely enable all six Auto-ID functions (not all functions have a unity degree of performance), such an examination must include a detailed analysis of PSC process activities, SCV challenges, and their requirements on the SCV dimensions to derive a feasible Auto-ID-based solution.
Design of empirical research and methodology
Design of empirical research
In our study, we elaborate the requirements on our SCV dimensions and a feasible Auto-ID-based solution through detailed analysis of a PSC. Thus, we chose a case-study research design (Eisenhardt, 1989; Yin, 2014) because it allows for the gathering of detailed information (Aastrup and Halldórsson, 2008; Harding, 2013; see e.g. Hammervoll et al., 2014), and serves our exploratory objective. Particularly because the PSC network is complex (Rossetti et al., 2011) and because very little knowledge exists regarding how to enhance SCV in this network, we considered a qualitative research approach to be appropriate. To ensure the quality of our case-study research, we focussed on the criteria recommended by da Mota Pedrosa et al. (2012).
The first important aspect of conducting case-study research is the design of the study (Yin, 2014). In line with our research objective, the PSC network forms the coherent context of this study. Figure 1 shows the actors within the PSC network in Germany, with transportation service providers (TSPs) represented by black arrows (Abramovici et al., 2010). The difference between a TSP and a logistics service provider (LSP) here is that an LSP realizes value-added services, such as the labeling or creation of drug-promotion packages. A TSP is only responsible for drug transportation without realizing any value-added services. As a result of the complexity of the PSC network, the study focusses only on the dashed line in Figure 1, which represents the most common chain in drug supply. Accordingly, this focus constitutes our research premise. We chose an embedded multiple-case study, in which the actors along this PSC represent the various cases. As the cases represent the actors, our design enables a detailed end-to-end analysis of process activities and drug flow along our focussed PSC. By this design, we are able to completely describe the most common chain of drug supply and identify process related requirements on SCV.
Next, we acquired different companies from the pharmaceutical sector that match our focus in Figure 1. Table V gives an overview of the case companies. According to Yin (2014), the sampling of a multiple-case design follows replication logic that also supports external validity. To assure external validity, we selected at least one case company for each actor in our focussed PSC. To simplify the study, we did not include a packaging service provider. However, this does not affect the results of our study because the packaging service is integrated into the upstream value-creation stage (pharmaceutical manufacturer).
We began our study with a drug producer and acquired two other drug producers with packaging activities to replicate insights into process activities and SCV challenges. Similarly, we acquired different LSPs and a TSP to obtain insights into process activities and SCV challenges. In so doing, we also tried to acquire firms of varying sizes and revenues to increase the external validity of our analysis (Yin, 2014). Next, we included two pharmacy cases to gather replicated insights into this stage of the focussed PSC. Finally, we included one wholesaler company to analyze the related process and SCV challenges. The study includes only one wholesale partner because only a few wholesalers are active in the German market. This wholesaler is active throughout Germany and a member of a nationwide union of pharmaceutical wholesalers, making it a representative case company for this stage in the supply chain.
Data collection
The case studies were conducted between the winter of 2013 and the spring of 2014. During the studies, data collection involved expert interviews, firm visits, discussions, and documents. These multiple sources support data triangulation and construct validity (Yin, 2014). The goal was to understand the viewpoint of the companies (key informants) regarding process activities and SCV challenges. We conducted one management-level expert interview with each case company, with ten expert interviews in total. Table V shows the position of the experts involved in the interviews. Furthermore, each case company was visited to understand and analyze process activities for drug handling, information gathering and sharing, and the challenges of SCV. In so doing, we obtained ten more interviews with employees of the case companies who were responsible for the processes on a physical supply chain level. These interviews involved further process discussions.
We applied a structured interview guideline (see the Appendix) to allow a complete and comparable information inquiry and to ensure research reliability (Yin, 2014). We developed the guideline following Bryman and Bell (2011). Based on our central research question and our conceptual framework, we derived an initial structure with topics and open questions. This guideline was pre-tested in terms of comprehensiveness, complexity, and interview duration. After revision, the interview guideline comprised four topics, beginning with an explanation of our research purpose. After gathering facesheet information, the third topic focussed on processes and SCV. The experts were asked to describe their process activities and SCV challenges, their current use of Auto-ID technologies, information sharing, and the requirements needed to enhance SCV. The interview guideline finished with the possibility for interviewees to make further remarks. All 20 interviews were audio-recorded and transcripts were created to support data analysis.
Moreover, we considered the notes of our accompanying discussions in the analysis. The study partners were allowed to check our transcripts, notes, and interpretations to ensure their veracity. In addition, firm visits allowed us to observe the processes in real time. Nevertheless, interviews can be biased and reflexive (Yin, 2014); it is difficult to achieve broad coverage of all processes during firm visits, and related activities might proceed differently (Yin, 2014). For these reasons, the case companies provided documentation on the process activities and the pharmaceutical sector to support the study. Although such documents provide specific insights and are unobtrusive, they also exhibit the weaknesses of selectivity and reporting bias (Yin, 2014).
Data analysis
Data analysis basically consisted of a time-series analysis, which implies the examination of how events influence activities and allows the development of theoretical solutions (Yin, 2014). In this sense, we examined the process activities[1] and determined the challenges that are needed to overcome to enhance SCV and enable proof of adherence to the GDP guideline for drug handling. Based on this examination, we identified requirements for our SCV dimensions to overcome the challenges. We subsequently discussed these requirements with the reviewed Auto-ID technologies to determine which Auto-ID technology is capable of meeting the requirements of overcoming the challenges, respectively. We then derived solutions of Auto-ID technologies from the findings of the analysis and from the insights provided by the case companies. Our derivation was further supported by discussions with the case companies and by comparisons with published studies to ensure construct- and internal-validity and to provide feasible solutions for our focussed PSC. At the end of the study, we presented the final solutions to our case companies. The next section describes the process activities in our studied PSC. This description is based on our empirical research and should facilitate an understanding of the process activities, challenges, and the identification of requirements on SCV dimensions.
Case description and identification of requirements for SCV
Against the background of past pharmaceutical drug disasters (e.g. Lipobay in 2001), and because of the importance of drugs to public health, controlling institutions such as the European Medicines Agency (EMA) in Europe and the Food and Drug Administration (FDA) in the USA were established and now issue strict regulations. The FDA is concerned with the safety of human and veterinary drugs to protect public health, especially in the USA (FDA, 2015). This agency also provides consumers with detailed and scientific information on medicines. For pharmaceuticals, both prescription and non-prescription drugs fall under the FDA’s regulatory jurisdiction. The European counterpart of the FDA is the EMA, which is responsible for the scientific evaluation of new medicines and the delegation of marketing authority in Europe (EMA, 2015). Moreover, to protect public health, the EMA’s responsibility includes monitoring the safety status of medicines. Through the new GDP guideline, actors in the PSCs are even more obliged to prove correct drug handling (Spiggelkötter, 2013).
At the pharmaceutical manufacturing stage, outgoing consignments demand a temperature check, currently conducted using an infrared thermometer. As the GDP guideline stipulates reliable temperature surveillance, the interviews revealed that, because of reflective materials in drug packages and different temperatures both within and outside pallets, automated temperature recording and temperature checks are a challenge to increasing check reliability. Thus, temperature sensing and a record of temperature data were identified as requirements. The temperature record should also allow the temperature profile to be created upon delivery, as requested by the GDP guideline. Furthermore, the process of loading incoming trucks is not reliable. For customized consignments, incorrect loading sequences resulting from precarious identification by barcodes lead to incorrect deliveries, which subsequently reduce the lifetime of the drugs. At this point, approximately 40-50 trucks must be loaded (full truckload) per day. Consequently, this process constitutes a challenge to the implementation of accurate drug identification. Therefore, to avoid loading failures, the authors identified the requirement of an accurate identification of drugs.
In the following stage, which falls under the responsibility of the TSP, the consignments must pass various scan points. In the event of cross-docking, the consignments must pass approximately 20 scan points. Otherwise, they must pass approximately 40 scan points due to interim storage. Despite the use of barcodes, consignments are not guaranteed to reach the correct transport vehicle for further transport, even if they have been scanned. Human errors and scan activities before the loading process hinder these activities. Therefore, the challenge lies in accurately scanning the drugs throughout this process. Such scanning activities and the communicated drug identity would support employees in correctly assigning the consignments to the transport vehicles. To further increase the reliability of assignments, the authors recognized a demand for assignment verification. Consequently, we identified the requirements of an accurate identification of drugs and a kind of logic that enables automated assignment verification (i.e. to ensure that the drugs are consigned to the correct vehicle).
At the stage of LSPs, we identified various processes that represent challenges to enhancing SCV with the introduction of the GDP guideline. Currently, LSPs manually measure the temperature of consignments using infrared pistols after truck docking. If the temperature is acceptable, the drugs are unloaded and scanned. If not, the drug quality is requested from the pharmaceutical manufacturer. If drug quality is still acceptable, the drugs are unloaded and scanned; otherwise the consignment is refused. The recognized challenge here is to provide transport-temperature information in advance, which would accelerate quality requests and thereby help to maintain the expected distribution rate. Currently, vehicles are equipped with telematic modules, which transmit information about temperature and position to the IT system of the LSP. Providing such information in advance therefore requires a central database in which information is stored and accessible to actors.
A further process activity for LSPs deals with checking the batches (approximately 20-30 pallets) and the size of consignments. The size of consignments differs between, for example, two pallets of a highly specialized drug and a full truckload of a standard drug. These checks are currently carried out manually and randomly. Moreover, the following activities in which drugs are selected and shipped, involve weighing containers to ensure that orders are complete. This method involves a certain degree of inaccuracy because different drugs may have similar weights. Of greater concern, however, is the fact that the expiration date of the drugs (printed on the package) selected is not verified. To increase the rate of verification while maintaining the expected distribution rate, the case companies must accurately check the consignment, order completeness, and expiration date. Therefore, we identified an accurate identification of drugs in the consignments and containers through a scanning activity as a necessary requirement. To verify the expiration date, we also incorporated a kind of logic into the requirements, which would allow reliable verification by automatically comparing the current date with the expiration date.
Furthermore, LSPs pack individual drug packages together on a pallet in the consolidation area of the shipping department. In the following process of interim storage, pallets are stored before being handed over to the transport service provider. In both steps, we recognized the challenge of correct classification; for example, drug packages may be placed on the wrong pallet, or pallets may be parked in the wrong temporary storage space. In turn, these classifications influence the rate of correct deliveries, the distribution rate, and, finally, the drug’s lifetime. Consequently, the authors identified a logic requirement for verifying classification (i.e. verifying drugs, pallets, and storage space). Additionally, the process examination showed that, in the case of misclassification, the challenge is to rapidly find the right packages or pallets. Such time-consuming searches decelerate the expected distribution rate. To facilitate such searches in the issuing department, the authors also considered the provision of spatial position information about missing objects to be a requirement.
At the pharmaceutical wholesaler stage, both the content announcement of combined incoming consignments and the scan to ensure the order completeness of drugs are performed manually. The size of consignments encompasses approximately two to three five-tonne trucks (full truckload). As a result of unreliability, these activities lead to errors in the subsequent order-selection process of drugs and in consolidation for delivery tours. Accordingly, the challenge is to ensure accurate content announcements and order-completeness scans. Therefore, the authors identified accurate identification as a requirement to support these manual tasks. Moreover, the order-selection process by which drugs are put into containers requires a further scan process and a final check for order completeness. According to the study partner, this selection activity should also be optimized, which also requires accurate identification. Finally, the consolidation of containers to delivery tours reveals the challenge of correct container classification. Approximately 1,500-2,000 containers must be classified for shipment per day. Due to human errors, this classification requires an accurate identification of drugs and verification of the container classification through the comparison of destination information (i.e. comparison between container and drug destination).
The final process activities include the transport of containers from the pharmaceutical wholesaler to the pharmacy (final distribution) and the process activities within the pharmacy itself. These activities revealed only one challenge: automated information logging. Currently, the wholesaler’s TSP records the delivery of drugs only in exceptional cases, such as anesthetics. Therefore, no information tracking is possible between the time that the drugs are handed to the TSP and the end of their arrival at the pharmacy (a black box). This circumstance also breaks the chain of continuous temperature measurements, preventing the provision of a temperature profile. In this regard, the case study revealed the requirement for an accurate identification of the drugs at the onset of transport and an automatic logging of transport and temperature information at drug level and in a database. Consequently, a record of temperature information and an information exchange between database and drug level are required to allow complete information tracking.
Finally, with respect to drug transport, the case companies also mentioned the challenge posed by theft and counterfeiting. According to the study partners, secure and accurate identification of the drugs at scan points would facilitate the recognition of stolen and counterfeit drugs. Consequently, to increase protection against theft and counterfeiting, we included a requirement for the recognition of unauthorized drug removal with associated date and position.
Analysis of requirements for SCV and derivation of Auto-ID-based solutions
This section first presents the relationship between the identified requirements and the SCV dimensions and an analysis of the necessary functional degrees of performance. In a second step, these functional degrees are used to derive a feasible solution for SCV enhancements.
Analysis of requirements on the SCV dimensions
Availability
As the dimension availability focusses on sharing and accessibility of information, we assigned all related requirements to this dimension. Accordingly, the requirement automatic logging of information at both drug and database level was assigned to this dimension because drug information should be available at both levels. Such logging, however, requires a bi-directional communication function of Auto-ID technologies so that information can be automatically exchanged between drugs and databases at scan points. Considering Table III, a performance degree of 0.5 was derived for the necessary communication function. In this context, a helpful Auto-ID technology requires an appropriate data storage function to allow for information logging at drug level. Based on the case studies, kilobyte-range data size was deemed sufficient to store temperature and additional (transport, position) information. Under consideration of Table III, we evaluated the necessary data storage function with 0.5. Finally, a central database must be considered in order to store information and provide actors with access to this information.
Identity
This dimension focusses on the provision of identity information. In this sense, we distinguished accurate and secure identification at drug level as a requirement of this dimension. Accurate identification logically demands an identification function, specifically, identification at product level. Giving consideration to Table III, this identification function has to exhibit a performance degree of one for the fulfillment of this requirement. Additionally, identification at product level must include security. This means that we have to consider a protection mechanism in the derivation of a feasible solution.
Position
This SCV dimension includes the provision of position information about objects. In this sense, we assigned the identified requirement provision of spatial position information to this dimension. To realize the requirement, a locating function is necessary. As the case study results revealed spatial position accuracy, we selected a required performance degree of 0.5 for the locating function because this degree corresponds to room-level precise localization (Table III).
Status quo
This dimension focusses on the provision of information about the status quo of an object. The information can entail object- and environmental-related aspects. Thus, we assigned all the identified requirements to this dimension that deal with information about drugs and their environment. In this sense, temperature sensing and a record of this information were assigned to this dimension. To realize temperature sensing and recording, a sensors function is essential. As we only want to measure and record one parameter (temperature), a single sensor for this function is sufficient. According to Table III, the resulting degree of sensors performance must be 0.25.
Furthermore, we assigned the identified verification and comparison requirements to the status quo dimension because they deal with drug related information (i.e. verifying drugs, a comparison of drug and container destination). In this regard, we also dedicated the requirement recognition of unauthorized drug removal to the status quo dimension because any information about removal relates to the drug itself. However, the realization of verification, comparison and recognition demands a logic function. As verification, comparison and recognition constitute three different events, the logic function must be able to recognize more than one event resulting in a 0.5 degree of performance (Table III).
Summary of requirements for SCV
To summarize our first analysis, Table VI shows the relationships between the identified requirements and the SCV dimensions.
Moreover, we illustrated the required degrees of performance for each Auto-ID function in a net diagram (Figure 2). Accordingly, Figure 2 shows to what extent the functions have to be met in order to realize the requirements and thereby SCV.
Derivation of Auto-ID-based solutions
By comparing the profile of required functionality (Figure 2) with the functionality profile of Auto-ID technologies (Table IV), only WSNs are able to meet all the required degrees of performance without any adverse deviations for the discharge of the requirements. As WSNs also allow secure identification (Zheng and Jamalipour, 2009), this technology was derived for the proposition of a first solution to enhance SCV. However, this finding was discussed with the case companies to ensure the derivation of a feasible solution. Although the selected WSN technology fulfills all the required functionalities, it involves the implementation of completely new technology within the focussed PSC. In this context, the case companies indicated the need for a more practical solution, as they are currently working with different existing practical monitoring solutions (barcode, securPharm with data matrix code, data logger) and separate IT systems. In this regard, legislation (controlling institutions with legislative influence and regulatory requirements) and drug (object) characteristics have already promoted the implementation of current practical monitoring solutions within PSCs. Thus, the case companies were also open minded about a solution that is easy to adopt, or one overall solution, no isolated solution, to enhance SCV. Nevertheless, the capacity to provide accurate and secure information should be ensured.
Transport containers with sensor nodes
Based on the above discussion, the authors propose the implementation of WSN technology in the opaque final distribution stage from pharmaceutical wholesalers to pharmacies (a complete black box). Here, transport containers are used in closed-loop chains and are properly constructed to be equipped with sensor nodes, in parallel with the use of a central database. The nodes accompany the containers throughout this stage, and record and store all required information (Zheng and Jamalipour, 2009). In addition to providing automated information logging, closed-loop implementation is economical because the containers are reused (see Mason et al., 2012). However, it is important to link the containers with a pharmacy order and the corresponding delivery with all the available information on both the drug and the transport. This allows the unique assignment of drugs to each individual transport container (identification at drug level). Verifications are supported because the communication capacity of sensor nodes (Zheng and Jamalipour, 2009) enables them to trigger an alarm if the transport container is classified to the wrong delivery. Moreover, the nodes can recognize when transport containers are removed from a vehicle without authorization (logic; Zheng and Jamalipour, 2009) and can forward this event, providing both date and position, to the database.
The use of sensor nodes in the stages before final distribution is theoretically possible but requires the equipment of each pallet used with a node. However, equipment with sensor nodes is not economical as a result of the use of pallets in open loops. Additionally, if a pallet is equipped with a sensor node then temperature is only measured at the bottom of this pallet. During our process observations, we also validated the fact that equipment of each drug package with a sensor node is neither feasible nor economical. Against this background, we propose the use of sensor nodes in the black box final distribution, in addition to the use of existing monitoring solutions in the upstream stages of the PSC.
However, as the case companies prefer a solution that synthesizes current practical monitoring solutions that have already been implemented, we derived further solutions to improve SCV. Nevertheless, we gave consideration to accurate and secure identification and Figure 2.
securPharm with RFID tags
As a result of the benefits of securPharm (2014) and its database, the authors included this approach to provide a further, synthesized and more refined solution. A comparison of Figure 2 with our Auto-ID review shows that RFID is only lacking in a sensors and logic function. This means that the use of RFID would only impede the realization of the SCV dimension status quo. However, the comparison also shows that RFID with sensor exhibits inadequate degrees of performance in terms of locating, communication, and logic function impairing the SCV dimensions position, status quo, and availability. Based on this comparison, we propose the use of rewritable RFID tags (Finkenzeller, 2010) inside drug packaging in combination with securPharm. Regarding its functionality profile in Table IV, RFID can fulfill the required functionalities of identification, locating, communication, and data storage. Accordingly, the requirements 1.1, 1.3, 2.1 and 3.1 from Table VI can be met. In this context, further information on temperature and transport would be stored at drug level along the PSC but would require manual scanning. Therefore, only manual information logging is possible (Req. 1.2). Unfortunately, as the required sensors and logic functionalities cannot be realized by RFID tags, requirements 4.1, 4.2, and 4.3 from Table VI cannot be met. Consequently, related activities, such as temperature recording with infrared pistols, must be carried out manually using existing techniques or they will require further improvements. However, as information is also stored on the central securPharm database, the provision of temperature data in advance is ensured.
Nevertheless, RFID can be read out or manipulated (Finkenzeller, 2010), whereby secure identification requires a further protection mechanism. The authors thus propose secure identification by the redundant storage of drug information in the central securPharm database and on RFID tags. Accordingly, counterfeit drugs may be identified if the tag and database information are not identical. However, despite its implementation in open-loop chains, this solution does not realize automatic information logging.
Finally, our solution relating to securPharm with RFID is similar to the reviewed monitoring solution of Kärkkäinen et al. (2003); however, it synthesizes that work by including the securPharm solution. In other words, drugs carry their own identity and additional information, and deliveries can be managed from the inside out. However, our solution differs from that of Kärkkäinen et al. (2003) in that it builds on the securPharm architecture with information-exchange channels, which is easy to adopt. Thus, our solution applies an existing communication code and a reliable hardware infrastructure, which facilitates the adoption of a drug-centric approach, as opposed to the installation of a completely new peer-to-peer communication network.
SCV dashboard
Considering the interest of the case companies in an overall solution to improve SCV, the authors further propose the implementation of an SCV dashboard that merges all the information collected by current practical monitoring solutions applied within the PSC, combining the functions of these solutions. Such a dashboard would allow the continuous provision of collected information (Davenport and Brooks, 2004) and thereby complete drug-centric monitoring without additional Auto-ID technology implementation. The dashboard should be implemented on a central database, providing all PSC actors with access to the information. However, this dashboard only combines information from current practical monitoring solutions applied. This means that a logic function is missing and the purely optical barcode or data matrix code do not ensure a sensor functionality. Moreover, accurate identification function is influenced by the fact that identification by barcode and data matrix code requires line-of-sight scanning. Neither code allows the required bi-directional communication function. Additionally, no information can be added at drug level during the process activities because barcode and data matrix code are read only (data storage function). Thus, by constant scanning to feed information to the database, this solution would only accomplish Req. 1.1 from Table VI. Moreover, the logical assignment of drugs to scan points (locating function) does not allow for the provision of the required spatial position information. In terms of secure identification, we propose redundant storage of information in both the database and on the barcode or data matrix code. However, the shortcoming of such an overall solution is that requirements other than 1.1, 2.1, and 2.2 from Table VI cannot automatically be met in the way identified by the Auto-ID technologies of the current practical monitoring solutions used. Consequently, the SCV dimensions position and status quo are not realized by this solution.
However, the advantage of the SCV dashboard is that it allows the use of the central database of the securPharm approach, even when securPharm is not implemented. The ongoing securPharm initiative in Germany can facilitate implementation efforts of this solution. With these modest efforts, also claimed by the pharmaceutical industry (Al-Kassab and Rumsch, 2008), this solution can lead to an extension of the existing IT infrastructure, instead of building a new solution. Furthermore, the dashboard solution can be implemented in an open-loop system. To implement such a cross-actor system, an SCV dashboard should be operated by a third party. This allows all the supply chain actors to access the dashboard without requiring direct access to the in-house IT structure of one of the supply chain actors involved (availability). Actors within the supply chain act as senders and receivers, with previous authentication to ensure eligible access and data privacy. Moreover, the dashboard can provide a useful possibility to involve the EMA because of its surveillance responsibility.
Discussion and concluding remarks
The current GDP guideline forces PSC actors to ensure and prove correct drug handling across the entire supply chain. This study responds to this obligation by proposing solutions for SCV enhancements based on Auto-ID technologies. To provide feasible solutions, our study first presents an overview of SCV and its relationship with Auto-ID technologies. In this context, the authors provide a conceptual framework with the dimensions availability, identity, position, and status quo for the analysis of SCV. In connection with Auto-ID functions, this framework helps to operationalize SCV and supports the selection of helpful Auto-ID technologies. Based on this conceptual framework, the authors derived three Auto-ID-based solutions: a securPharm approach with passive RFID tags, transport containers with sensor nodes, and an SCV dashboard. To overcome all the identified requirements of the SCV dimensions, our study theoretically recommends implementation of the sensor nodes’ solution. If actors require the continuation of their current practical monitoring solutions and hazard deficiencies in the realization of SCV dimensions, the SCV dashboard provides an overall solution in our studied PSC. However, our solutions demonstrate to researchers and practitioners approaches to implementing drug-centric monitoring with stringent requirements for accurate and secure identification and information handling.
Insights from the case study also show that current practical monitoring solutions (barcode, data matrix code, data logger) that have already been implemented rather impede the implementation of advanced Auto-ID technologies (RFID, RFID with sensor, WSNs). Although advanced Auto-ID technologies help to enhance SCV, the case companies, however, demanded an overall solution that takes into consideration those that have already been implemented. This issue is also recognizable in previous studies (Kärkkäinen et al., 2004; Mirzabeiki et al., 2014) and can therefore be seen as a barrier to SCV enhancements. Against this background, the analysis of our study has led to a dashboard solution that synthesizes monitoring solutions that have already been implemented and combines their functions to improve SCV. Similar solutions can be found in the studies of Houghton et al. (2004), Mirzabeiki et al. (2014) and Sanders (2016) that deal with an analogical background. In this context, we consequently argue the following propositions:
Already implemented current practical monitoring solutions present a barrier to enhancing SCV in a PSC according to legislative influence.
A dashboard as an overall monitoring solution that synthesizes monitoring solutions that have already been implemented will improve SCV in PSCs.
Furthermore, our empirical study shows that legislation with regulatory requirements clearly influences process activities, and thereby creates requirements for the realization of SCV. This insight confirms the finding of regulatory impact on SCV by Klueber and O’Keefe (2013). Our study additionally demonstrates that object characteristics also present requirements for SCV. For example, sensitivity, identity, and weight of drugs need to be regarded when clarifying the status quo of drugs in the supply chain. Therefore, we also support the works of Francis (2008) and Hafliðason et al. (2012), who emphasized the consideration of object-related characteristics as important for the realization of SCV. As our analysis of the study shows, the resulting requirements for SCV of both legislation and object characteristics demand an Auto-ID technology with advanced functionality and exclude the use of current practical monitoring solutions. Our analysis demonstrates that sensor nodes of WSNs based on their accurate and secure identification, automatic information logging, locating, sensors, and logic functions are able to meet regulatory and object related requirements for SCV. Thereby, our work corroborates current studies with similar requirements (Hafliðason et al., 2012; Lang et al., 2011; Wang et al., 2015), which show an increasing use of sensor nodes to enhance monitoring and visibility. Consequently, we argue the following proposition:
Object characteristics and regulatory requirements present requirements for SCV that demand a sensor node based Auto-ID technology (WSNs) in order to achieve enhanced SCV.
Our study has certain limitations. Because of its complexity, it focusses on a special chain within the PSC network. Therefore, care must be taken when extending the solutions to other chains. Next, this study focusses on Germany; however, this is mitigated by the fact that the GDP guideline is valid throughout Europe. In addition, the solutions do not claim to encompass all Auto-ID technologies because we were unable to consider all the technological possibilities.
Further research should investigate in more detail the implementation and costs (business case) of the solutions. Moreover, other chains within the PSC network should be analyzed in terms of SCV. Additionally, further studies can contrast our solutions and other monitoring solutions of different industries to derive industry-specific context factors for monitoring solutions (Al-Kassab and Rumsch, 2008; Holmström et al., 2010). Moreover, the derived propositions from our empirical findings need further empirical investigation in other pharmaceutical and industry supply chains. Regarding the SCV dashboard, possible incentives to stimulate the exchange of information between actors must be investigated, because the exchange of semi-sensitive information is always a critical concern. In this regard, jurisdictional transaction flows should also be considered. Moreover, our SCV dimensions and applied operationalization should be tested in terms of completeness to support SCV analysis. In addition to our focus on enhancing SCV using Auto-ID technologies, further research can examine how to improve this aspect for in-transit services; for example, drugs used in hospitals. For such services, Auto-ID technologies offer promising control and visibility improvements in material flows (Arnäs et al., 2013).
The authors are extremely grateful to the anonymous reviewers. The authors would also like to sincerely thank the Guest Editors, Lise Lillebrygfjeld Halse and Trond Hammervoll, and the Senior Assistant Editor, Patrik Jonsson, for their support during the review process.
Appendix. Interview guideline
Introduction and explanation of research purpose
Facesheet information
Name of the company, year of establishment, employees, revenue
Name and position of interview partner
Process activities and SCV
What are the required process activities of this company for order fulfilment?
How does this company organize the handover of drugs to the following supply chain stage?
How does this company measure and log drug temperature and identify temperature deviation?
How does this company ensure drug authenticity and quality?
What is the average size of shipping unit?
What process related challenges and improvement potential does this company see in terms of SCV, giving consideration to the current GDP guideline?
What technologies does this company currently use to realize SCV?
Does this company use automatic identification technologies in the process activities to realize SCV?
If so, how does this company apply such technologies?
Possibility for further remarks and wrap-up
Note
1.A detailed process model is available upon request.
Figure 1
Actors within the PSC network
[Figure omitted. See PDF]
Figure 2
Profile of required functionality
[Figure omitted. See PDF]
Table IDefinitions of SCV
| Author | Definition |
|---|---|
| Kaipia and Hartiala (2006, p. 377) | “The sharing of all relevant information between supply chain partners, also over echelons in the chain” |
| Barratt and Oke (2007, p. 1218) | “The extent to which actors within a supply chain have access to or share information which they consider as key or useful to their operations and which they consider will be of mutual benefit” |
| Francis (2008, p. 182) | “Supply chain visibility is the identity, location and status of entities transiting the supply chain, captured in timely messages about events, along with the planned and actual dates/times for these events” |
| Klueber and O’Keefe (2013, p. 300) | “Requisite supply chain visibility (RSCV) is the ability to provide and access information elements at a level chosen by the relevant supply chain stakeholders […]” |
Conceptual framework with SCV dimensions and explanation
| SCV dimension | Explanation |
|---|---|
| Availability | Accessibility of information by all eligible supply chain actors Availability demands information sharing |
| Identity | Provision of identity information Level of detail depends on the demands of supply chain actors; and Example: identity at object level or identity of a complete shipment |
| Position | Provision of position information about an object within the supply chain Level of detail depends on the demands of supply chain actors; and Example: GPS information about an object or information about which supply chain actor is currently responsible for which object |
| Status quo | Provision of information about the status quo of an object within the supply chain Information can entail object- and environmental-related aspects; and Example: information about transport, handling, storage, and temperature |
SCV dimensions and related Auto-ID functions with degrees of performance
| Quantified degrees of performance | |||||
|---|---|---|---|---|---|
| SCV dimension | Auto-ID function | 0.25 | 0.5 | 0.75 | 1 |
| Availability | Communication | Line of sight | Availability at base stations or scan points (bi-directional) | Complete cross-linkage | Integration into telecommunication networks |
| Data storage | One identification number | Data volume in kilobyte range | Data volume in megabyte range | Data volume in gigabyte range | |
| Identity | Identification | Identification of charges | Identification of outer packaging | Identification of product category | Identification of products |
| Position | Locating | Logical assignment | Room-level localization | Localization in meter range | Localization in centimeter range |
| Status quo | Sensors | Single sensor | Up to five sensors | More than five sensors | Nervous system |
| Logic | One event | More than one event | Business logic | Artificial intelligence | |
Functionality profiles of Auto-ID technologies from reviewed monitoring solutions
Overview of case companies
| Case | Actor within our focussed PSC | Position of experts in interviews | Employees (2013) | Revenue (2013) |
|---|---|---|---|---|
| 1 | Pharmaceutical manufacturer | Head of supply chain | <250 | >1,000 MEUR |
| 2 | Transportation service provider for wholesalers and pharmacies | CEO | <250 | >1 MEUR |
| 3 | Logistics service provider | Head of quality management | >1,000 | >100 MEUR |
| 4 | Pharmaceutical manufacturer | Head of supply chain | >50,000 | >10,000 MEUR |
| 5 | Pharmaceutical manufacturer | Head of supply chain | >10,000 | >10,000 MEUR |
| 6 | Logistics service provider | CEO | <250 | >1 MEUR |
| 7 | Logistics service provider | Head of supply chain and quality management | >1,000 | >1 MEUR |
| 8 | Pharmacy | CEO (pharmacist) | <250 | >1 MEUR |
| 9 | Pharmacy | CEO (pharmacist) | <250 | >1 MEUR |
| 10 | Pharmaceutical wholesaler | Head of supply chain | >1,000 | >1,000 MEUR |
SCV dimensions and requirements
| SCV dimension | Requirement (Req.) |
|---|---|
| Availability | 1.1 A central database where information is stored and accessible to actors 1.2 Automatic logging of transport and temperature information in a database and at drug level; and 1.3 Information exchange between database and drug level |
| Identity | 2.1 Secure and accurate identification at drug level |
| Position | 3.1 Spatial position information |
| Status quo | 4.1 Temperature sensing and a record of temperature data 4.2 Verification of assignments, comparison of current and expiration date, verification of classifications (i.e. verifying drugs, pallets, and storage space i.e. a comparison of container and drug destination); and 4.3 Recognition of unauthorized drug removal with the associated date and position |
© Emerald Group Publishing Limited 2016