	On November 27, 2013, US President Obama signed the Drug Quality Security Act (DQSA) law into effect. Article II of that law, named the Drug Supply Chain Security Act (DSCSA), requires members of the US drug supply chain to make a series of changes over a 10-year period in order to better secure the US supply chain from counterfeit, diverted, stolen, or mishandled drugs. The changes also seek to address the drug recall process.

	This new federal law preempted all related state laws and was significant in that it created one federal law that took under consideration aspects of four different concerns:

		California's pending Pedigree law (requiring serialization and traceability by 2015)

		Certain established business practices of the supply chain, such as inference (e.g., inferring what is in a sealed case based on the manufacturer's information)

		FDA’s desire to remove bad product and bad actors from the supply chain

		Patient-safety advocates' desire to protect the patient.

	To allow supply-chain partners to budget and prepare for serialization and traceability, the law also sought to spread implementation over a 10-year period. No one party got all of what they wanted, but in the end, everyone got something that could build momentum toward the ultimate goal of removing illegitimate medications and criminals from the supply chain. As with any undertaking of this magnitude, many questions and issues arose and will continue to arise as actual phased implementations take place.

	Most phase dates in the DSCSA are based on a certain period of time measured from when the law was enacted. An exception was the initial implementation phase of sharing transaction statements (TS), transaction information (TI), and transaction history (TH) at the lot/batch level, which was set for implementation on Jan. 1, 2015. Because of delays in Congress, by the time the law was officially enacted on Nov. 27, 2013, the supply-chain participants had little time to think as they switched from preparing for California’s requirements to the first implementation phase of DSCSA. Manufacturers and wholesalers scrambled to cobble together a solution using the systems they had already in place to meet the TS, TI, and TH sharing requirements. The result was that their systems worked with a lot of manual effort and a certain amount of errors-errors that also needed a lot of manual effort to address. It is well recognized by these manufacturers and wholesalers that these manual efforts are unscalable and unsustainable over time as the more demanding requirements of individual item serialization (in 2017) and traceability (in 2023) come into effect. Challenges for implementing serialization are even more magnified considering the growing number of global laws and regulations for serialization, authentication, and traceability.

	This challenge doesn’t mean that the task at hand is impossible. We are, after all, talking about a country that put a man on the moon and an industry that daily discovers and invents new ways of extending life in terms of quality and longevity. It is doable, and this industry is up to the task.

	So what now? Continue to implement as we go? Some hope not. The law provides enough time for the industry to move from the somewhat rushed implementations of 2015 to a well thought-out, globally integrated system for 2023. The problem is that there are still questions to be addressed to achieve new efficiencies that can help fund serialization and provide real benefits, such as more effective and efficient recalls.

	Today, industry trading partners are actively planning and executing traditional, physical pilots, which involve sending trial packages through the system, to ensure that their systems for handling serialization information function in production. To prepare for mass serialization and traceability, however, pilot programs must be conducted on a broader, more industry-wide scale. Issues such as aggregation, inference, scalability, and the core issues related to supply-chain security and patient safety cannot be solved through traditional pilots.

	So, how can these broader issues be addressed? In parallel with the necessary traditional pilots, supply-chain members could undertake a series of “virtual pilots” that would touch on the more difficult, core issues that have yet to be completely resolved. Loosely defined, a virtual pilot is a series of scenarios executed in a simulated environment where alternative courses of action can be experienced, measured, and compared. This concept is easy to understand in terms of extracting TS, TI, TH, or traceability data from one system and devising a method of comparison with other systems. Issues, such as aggregation, however, must consider behavior (resulting in possible changes to standard operating procedures) and measure the seemingly unmeasurable. For example, will this course of action enhance the security of the supply chain and enhance patient safety? To accomplish these tasks, the industry needs to consider the behaviors of its systems and staff; new business rules, laws, and regulations; and the effects of these factors throughout the global supply chain. This complex task requires the cooperation of all participants in the supply chain and an efficient means to track alternative scenarios and their impact on day-to-day events as well as on each other.

	An early example of using simulation software and techniques is found in the international commercial banking industry. In 1992, a commercial bank was developing an international funds transfer system to move and reconcile cash between international banks. The bank had over 50 designers/developers, fell six months behind schedule, and was $15MM over budget. The bank followed the latest systems and process technique of the time, creating hundreds of data-flow diagrams, data-model diagrams, and structure charts, but the bank wasn't confident that implementation of the transfer system would actually work if put into practice between the various banks around the world. The bank finally decided to use simulation software, which was typically used by manufacturing companies to study and measure internal processes and simulate hundreds of diagrams. The software revealed that the funds-transfer system didn’t work. One-hundred six major functional flaws were found in the design, and the bank was able to make corrections and implement the entire system within weeks.

	Like the bank pilot program, the pharmaceutical industry could also conduct a simulation for the proposed changes to the global supply chain, adding to the simulation the various worldwide requirements, business processes (local and global), and predicted solutions for the functional flaws that could potentially arise in implementation design. Then, having witnessed how scale and alternative methods impact the safe flow of medication (positively and negatively), key issues such as aggregation, inference, global commerce, local needs, scalability, necessary standards, guidelines, and supply-chain vulnerabilities could be resolved prior to mass implementation-resulting in measurable changes that would positively affect patient safety.

	Consider how supply chain industry players might tackle a “big issue” using a virtual pilot. Take the case of determining the impact of “lot/batch aggregation” (i.e., associating each serialized item to only the lot/batch) versus “packaging aggregation” (i.e., associating each serialized item to each case, pallet, or tote they’ve been in) on identifying a counterfeit. It is assumed that packaging aggregation allows for full supply-chain traceability of each item, but lot/batch aggregation alone could possibly miss a transaction where the outer layer (case) remains intact through a number of trading partner movements. A virtual pilot can be used to compare the two types of aggregation.

	Using simulation software, rules of behavior could be expressed for a series of simulated trading partners, and a counterfeiter could be added to the simulation. Then, two versions of the simulation-one where serialized items are aggregated at the lot/batch level and one where they are also aggregated at the case level and at all subsequent outer containers- could be developed.

	Each simulation would be analyzed to determine whether any party in the legitimate simulated supply chain had the potential to detect a counterfeit product. A virtual pilot could determine, for example:

		Whether there is a threshold of counterfeit products that needs to be reached for the proper detection of falsified medications in each scenario

		Whether there is a time difference in detection and alerts between the scenarios.

	Virtual pilots allow deeper analysis of key questions more cost effectively than physical pilots, which could be expensive and could lead the industry to choose an implementation path that may not necessarily lead to the best outcome. Virtual pilots can make good use of the industry’s time between now and 2023 by ferreting out the best and most efficient ways to secure the supply chain while producing a quality blueprint for physical pilots and implementations.

	Robert Celeste is a founding partner of 

 healthcare consultancy and a former senior director at GS1 US, 

Articles

Packaging

Virtual pilot programs that simulate scenarios can help the pharmaceutical industry address core issues in the implementation of serialization systems that comply with the US Drug Supply Chain Security Act.

Global Serialization: Could Virtual Pilots Be in Our Future?

	Replacing stacks of paper records with easily searched and shared electronic records has clear benefits for both the near and long term, but making the switch requires careful planning and risk management. 

 spoke with Ken Kovacs, quality business systems manager at Fujirebio Diagnostics, about the advantages of using electronic production records and the keys to successfully implementing a system.

: What are electronic production records (EPRs), and how do they relate to electronic batch records (EBRs)?

: Production record requirements are identified in various FDA regulations, such as 21 

	Part 211 for Finished Pharmaceuticals (1) and 21 CFR Part 820 for Medical Devices (2). Production records can be considered a collection of all the records required to manufacture an FDA-regulated product as well as all the records produced during the manufacturing of those products to meet specific regulatory requirements. EPRs provide an electronic equivalent to paper production records as required by agency regulations.

	The concept of EPRs has not been used as extensively in industry as EBRs. An International Society for Pharmaceutical Engineering (ISPE) special interest group (SIG), the Good Automated Manufacturing Practice (GAMP) Manufacturing Execution Systems (MES) SIG, used EPR as a more universal term to include all associated production records for both pharmaceutical and medical device industries, which is important because of the increasing overlap in these industries (e.g., combination drug-device products).  Both batch records and device history records can be considered subsets or components of production records. ISPE’s 

GAMP Manufacturing Execution System Good Practice Guide

 (4) provide appendices describing electronic production record systems and associated functionalities, such as ‘review by exception’ and electronic signatures.

	Implementing EPRs creates the ability to generate and distribute electronic records within the enterprise in a timely manner, dynamically control processes to decrease product variability, and assure product quality during production. CGMP-compliance issue investigations and their associated paper-based records may also be simplified through information derived from EPR reports. Over time, EPR systems can evolve to provide real-time product disposition and provide a database of scientific knowledge needed for process understanding and future product development.

: What have you found to be the biggest advantages of EPRs?

: Fujirebio Diagnostics’ current EPR system has saved several thousand hours of people’s time when compared to the previous, manual data-logging process and review of associated paper-based GMP reports. Paper-based logs and reports have been eliminated. Parameter data are automatically logged, analyzed in real time, and retained in a database by the computerized system. Authorized personnel are immediately notified via email in the event an adverse warning or excursion occurs with any monitored parameter, and system functionality allows us to capture personnel comments and electronic signatures by department for each event.

	Monthly departmental summary reports (SRs) are generated for ‘review by exception’ and routed via workflow software to each department for review and electronic signature approval.  The SRs are then routed to quality assurance (QA) for final review and approval. Once approved by QA, the system generates a monthly electronic SR for each department and appropriately files it for retention and future use. All data and electronic records are routinely backed up through information-technology processes and procedures.

	With this system, we now have the ability to more easily analyze equipment and process operations for a better understanding of critical process parameters and product quality attributes.

: How did Fujirebio Diagnostics go about implementing EPR?

: Our executive management team had already identified an electronic initiative as a concept for implementing electronic records. The EPR project team then identified a strong business need for this specific EPR system. A project justification and return-on-investment calculation based on implementing an infrastructure conducive to installing future applied solutions was developed for management review. Based on this strong need and justification, the project was quickly approved. Before developing an infrastructure and architecture, the project team agreed on using International Society of Automation (ISA) 

 (7) standards and models as guidance for system design.

	Personnel from the entire company were engaged to obtain their feedback and expectations for this new system. Twenty-seven individuals from all departments participated in the development of the user requirements specification (URS). This project was truly company-driven by users and management working together.

: What were the keys to successful implementation?

: Direction from executive management provided the support for launching this project and was crucial to getting this system approved. Project implementation was guided by our formal system development life cycle (SDLC) and a step-wise project management methodology that includes a risk-based approach to design and validation. The use of global standards and models greatly reduced the time needed to define a site-wide architecture and system functionality.  Once the URS was completed, a request for proposal package was developed, and leading system integrators (SI) were identified and invited to participate in our selection process. A SI providing what we identified as the best applied solution to address our user requirements was then selected. Deliverables required for each phase of our SDLC were completed as the project progressed.

	Before going live with this system, a comprehensive documentation package was completed, robust qualification testing was executed for validation, and more than 60 users were trained. Completion of system validation allowed us to move one step closer to system turnover. After formalizing a system support team and process, the system was then operated in parallel with existing, manual standard operating procedures (SOPs) for three months. Because this was the first implementation of a record system of this type, we wanted to ensure that everyone was properly trained and that everyone understood the process of generating and approving the e-records. In addition to system-level training, we included training on our e-signature and e-record SOP, which is based on 21 

 Part 11 requirements. The comparison of e-records to paper was integrated in our approach to system qualification for validation. At the end of three months, our user community was comfortable with the new system and was also eager to make the transition to e-records, after seeing how much easier it is to use compared to the time previously spent in manually logging data and reviewing stacks of paper records. Migration to the electronic record system was then made official, with manual activities relegated to a back-up role. Existing SOPs previously used for generating manual logs and paper records will now be used only if there is an operational interruption to the automated system.

	During the entire project, timely meetings were conducted with the SI and contractors, team contributors were acknowledged and thanked, and project update bulletins were frequently posted to keep all company personnel apprised of progress being made. This communication was important because the new system was a significant change that affected people in 18 different departments.

: How does the new infrastructure provide a basis for further improvements to manufacturing equipment systems?

: During initial discussions on what an electronic initiative meant to Fujirebio Diagnostics, a ‘future state’ of paperless operations was envisioned and various strategies identified to make it all happen. The infrastructure implemented with our first project was designed to provide an architecture that addresses equipment integration, data acquisition, and information management from the shop floor through manufacturing operations and up to our business systems. System hardware and software products from global suppliers were chosen to provide for easy expansion and compatibility with future applied solutions.

	Since this system was originally installed, we have doubled the number of GMP-related equipment that is being monitored, added more electronic record capabilities, and used system information to support other operations and quality processes. We are now scoping out a new project to further expand data acquisition into other areas in manufacturing with added functionality for manufacturing operations management. By implementing an infrastructure up front with capabilities to meet future needs for data management and EPRs, we now have an established step-wise strategic approach to rolling out applications that support our electronic initiative path forward.  

, Title 21, Food and Drugs (Government Printing Office, Washington, DC) Part 211, Subpart J-Records and Reports (2014).
	2. 

, Title 21, Food and Drugs (Government Printing Office, Washington, DC), Part 820, Subpart M-Records (2014).
	3. ISPE, 

GAMP 5: A Risk-Based Approach to Compliant GxP Computerized Systems

GAMP Good Practice Guide: Manufacturing Execution Systems-A Strategic and Program Management Approach

ANSI/ISA-95.00.01-2010 (IEC 62264-1 Mod) Enterprise-Control System Integration - Part 1: Models and Terminology 

(Research Triangle Park, NC, 2010).
	6. ISA, 

ANSI/ISA-95.00.03-2013 Enterprise-Control System Integration - Part 3: Activity Models of Manufacturing Operations Management

 (Research Triangle Park, NC, 2013).
	7. ISA, 

ANSI/ISA-88.00.01-2010 Batch Control Part 1: Models and Terminology 

 When referring to this article, please cite it as J. Markarian, “Implementing Electronic Production Records,” 

Troubleshooting

Paperless operations improve efficiency and increase assurance of product quality.

Implementing Electronic Production Records

	Manufacturing execution systems (MES) automate production processes and, although often perceived as high-cost investments, can offer significant return on investment by improving efficiency and reducing errors. In addition, new web-based approaches to implementation promise to reduce cost and improve interoperability between different vendors’ systems. 

 interviewed three of the industry experts who presented at CBI's 10th annual conference, MES 2015, (Aug 12–13), about the benefits of implementing MES and their vision for the future of MES.

	Over the past 10–15 years, pharmaceutical and biopharmaceutical companies that have implemented MES have seen ongoing, quantifiable benefits from improved efficiency. One of the most significant benefits is the reduction of human error by, for example, automating data collection, entry, and reporting through use of electronic batch records (EBR) and electronic production records (EPR).

	"People make mistakes-often simple mistakes that are not caught by other people.  Our attention wanders, we lose focus, and we get distracted and forget important steps," notes Dennis Brandl, immediate-past chairman of the Manufacturing Enterprise Solutions Association (MESA) Americas board, president of BR&L Consulting, and keynote speaker at the MES conference. "By reducing the need for human input, providing auto-checking workflows, and letting the computer double-check work, productivity in the pharma industry has been continually improving. Companies that have implemented MES and other workflow automation systems have seen year-to-year productivity increases that don’t go away."

	In addition to automating mundane, manual data-collection tasks and getting more product through an existing manufacturing environment without increasing headcount, MES increase the efficiency of the product-release process, adds Keith Bowen, general manager of LZ Lifescience. Bowen gave a presentation at the MES conference on developing a business case for MES. "Following an automatic executing sequence for making a product, with the quality team working in parallel with the production team, leads to at least a 25% increase in productivity," explains Bowen. 

	A high level of automation is not required to achieve benefits, he says. "In a small pharma solid-dose plant that I visited recently, the per-year measurable benefit was half a million dollars, with an ROI of less than one year," he reports. A key to developing a business case is targeting measurable, objective savings and mapping these to MES product modules for specific products. It is important to use actual manufacturing activity data and to purchase only what is needed, Bowen says, because large MES modules may contain functionalities that aren’t required for a particular company's process.

	In addition to perceptions of high cost, another barrier to MES adoption in pharma has been lack of cross-vendor compatibility, says Brandl, "Workflows developed in one system cannot easily interact with workflows in other systems; data collected in one system is not easily shared with other systems. This means you have to pick one vendor and stick with them, even if they are only adequate, or even inadequate, in some areas."

	Brandl suggests that MES suppliers embrace the “app” model that uses independent internet application programs to share data from multiple vendors in one human-machine interface environment. A virtualized system on a smart phone or tablet would create a lower entry cost and enable more companies to use MES more fully. "Even with measurable economic advantages, companies are still reluctant to spend hundreds of thousands of dollars to implement MES," says Brandl. "With a lower entry cost...benefits can be applied to all areas of production, not just those with the biggest problems."

	Web-based MES products are already available, Bowen says. "Web-based MES products reduce the cost of MES solutions by hundreds of thousands of dollars and even more for large sites, and that is just on the hardware costs. The software sequences can be configured more quickly using a library of modules and operations which provide the building blocks for new-product master-batch records, resulting in a faster time to benefit from an MES solution."

	A long-term vision for MES in the pharmaceutical industry is to have "streamlined, efficient, out-of-the-box business process library templates for manufacturing facilities (of each type)," says Bowen. MES will allow the facility of the future to combine the best workflow practices of the past 20 years with new, more efficient, corporately controlled business processes, he notes.

	Analyst and advisory firm Axendia presented findings from a 

survey of the MES life-science community 

at the conference. The objective of the survey was to identify current issues and benefits, as well as requirements for next-generation MES.

	The survey found that many in industry view MES primarily as a tool to solve the problem of eliminating human error and ensuring that records are complete (i.e., regulatory compliance), notes Daniel R. Matlis, president of Axendia. "FDA wants industry to focus on quality rather than compliance, but two-thirds of brand owners surveyed identified ‘improved regulatory compliance’ as the main benefit when implementing MES, while only 44% see improving product quality as the top benefit. We would hope to see that product quality is the driver, and that compliance is the baseline and a direct side effect." Software vendors who answered the survey had yet another perspective, with 60% selecting a reduction in data entry errors and reducing manufacturing cycle time as the most important benefits of MES.

	The survey went on to show an even bigger disconnect between vendors and brand-owners regarding the top future requirements of MES. Most vendors (67%) chose "Internet of Things" as a top requirement, but brand-owners selected "global visibility," "predictive maintenance," and "manufacturing optimization."

	"This dichotomy is a concern because vendors may be building software to support an internet of things, but this may not be a fit for what industry says it needs," notes Matlis. "The ability of MES to support global visibility is key because industry is globalizing. Most MES today are designed for one plant, resulting in the need to aggregate data for multiple sites. To optimize capacity and improve product quality, however, decision makers need to be able to compare parameters at multiple sites to identify critical to quality attributes and prepare accurate forecasting."

	In addition to global visibility, next-generation MES should provide visibility across the entire supply chain. "Much of the metrics data that FDA is looking for in the new quality metrics initiative could come from MES. Visibility throughout the supply network is important, however, because the brand owner may need to report data from outsourced processes, not just the in-house manufacturing pieces," explains Matlis.

	Improved predictive maintenance is also part of optimizing manufacturing processes. "Software should provide tools that enable industry to become proactive and ultimately predictive of product quality, which ties in to operations, maintenance, and training," says Matlis.  

	These capabilities either don't currently exist in MES or are not user friendly, concludes Matlis. "It is important for vendors to support the needs of brand-owners and for both to look forward to ensure that the systems being implemented address the future needs of industry rather than merely fixing issues of the past." 

Production Trends

MES has already rewarded early adopters by reducing errors; next-generation technology will enable broader data analysis.

Equipment and Processing Report-08-19-2015

Global Serialization: Could Virtual Pilots Be in Our Future?

Implementing Electronic Production Records

Proactive Optimization: the Future of Manufacturing Execution Systems