Serialising it right

Anand Modi, VP Sales, India, TraceLink, outlines the five key issues that Indian pharma manufacturers need to consider while building serialisation programmes to make them more effective and efficacious

Serialisation and track and trace regulations are now a strategic requirement for the pharmaceutical industry. By 2019, over 75 per cent of the global drug supply will be covered by requirements for drug serialisation, supply chain traceability, and government reporting across more than 40 countries ranging from India, and the EU member states to the US, South Korea and more. Whether a company owns drug product authorisations for a market, manufacture and package finished dose medicines, distribute medicines across the supply chain or dispense medicines to patients, every partner within the pharma supply chain will be faced with myriad diverse and often conflicting regulations.

As the 2017 US DSCSA deadline approaches, Indian pharma companies are preparing to face unparalleled complexity, cost and risk in implementing strategies to build a serialisation infrastructure that encompasses their entire manufacturing facilities.

When medicines are produced, packaged and readied for shipment, the life sciences cloud manages all serialisation requirements and serialised product inventory data exchanges internally with packaging facilities and warehouses and externally with CMO/CPOs and 3PLs. As medicines are shipped to the target markets, life sciences cloud configures and shares the required compliance data and reports with downstream trade partners and governmental organisations.

DSCSA regulations and timeline

The Drug Supply Chain Security Act’s (DSCSA) ten year timetable outlines critical steps to build an electronic, interoperable system to identify and trace prescription drugs from manufacturer to dispenser across the supply chain serving the USmarket. Implementation of the forty pages of complex DSCSA regulations can be broken down into three general phases:

• 2015: Lot-level traceability and verification of products and transactions
• 2017-2020: Serialisation of drug products and enhanced verification of serialised product identity
• 2023: Unit-level traceability

On the surface, preparing for DSCSA serialisation deadlines doesn’t seem that challenging – at least in comparison to other global serialisation regulations. Pharma companies and their CMO/CPO partners must generate serial numbers for each saleable unit and sealed homogeneous case of drug product produced. The serial numbers with associated National Drug Codes (NDCs), lot numbers, and expiration dates need to be encoded into 2D data matrix barcodes (for units) and either 2D data matrix or linear barcodes (for cases) following generally recognised industry standards.

There are five key issues that Indian pharma companies need to consider as they build their serialisation programmes:

Take a top-down, network approach – Beginning with level 4 and level 5

When determining a strategy for serialisation, the tendency is to start at the package level and focus first on placing the barcode label. However, serialisation goes far beyond simply printing bar codes on a package. Instead, pharma manufacturers should think from the top-down, determining all of the external compliance requirements that will need to be met, the systems that will have to be integrated and the data flows that will need to be managed.

For the serialisation system as a whole, there are multiple interconnected levels of systems which must synchronise together, including the Network, Enterprise, Site-level, Line/ Warehouse, and Device-level systems. Leveraging interconnectivity and interoperability at every level of the stack (Levels 1 – 5) is important.  When specifically implementing serialisation, Level 4 (enterprise system) and Level 5 (network system) are essential systems that enable all regulatory data and business processes throughout the enterprise and beyond, to partners, customers, and regulatory authorities. Below is a breakdown of the different levels within a packaging line and the roles in serialisation:

• Levels 1 -3 (line management system): These systems are provided by line management system (LMS) vendors.

1. u Level 1 includes the physical packaging lines with devices such as printers, cameras, scanners, lights, integrated into machines that produce cartons, cases and pallets.
   u Level 2 and Level 3 are the line controller and site server, providing capabilities at the packaging line and packaging site level, which only assists with the application and verification of serialised packaging.

• LMS vendors providing levels 1 – 3 do not maintain repositories to retain previously commissioned serial numbers for the required regulatory data retention period (for example, 12 years in the US). Some will purge all records associated with the previous batch upon line changeover for the next batch, while others will purge upon shipment or shortly thereafter. Since the LMS isn’t storing this data, there is no ability to check for duplicates against previously commissioned serial numbers from prior batches.

1. Level 4 (enterprise system) and Level 5 (network system): Prior to serialisation mandates, the focus was only on the product. Realising that each serialised product that is shipping now also requires an accompanying ‘shipment of data’ has expanded attention to the Level 4 and 5 systems that need to be in place.
2. Level 4 allows for the management of serialised inventory after it leaves the line, storing all previously commissioned serial numbers in an active available state allowing for immediate verification upon commission. This layer of visibility ensures that no duplicate serial numbers exist from previous batches, offering protection against significant compliance risk.

A Level 5 system provides a supply chain track and trace serialisation data event repository, enabling a secure and accurate data exchange at the correct event trigger across multiple sites.

Serialisation forces supply and trade partner networks to evolve

Indian pharma companies will need to understand all potential network connections (internal, supply, trade, governmental) over which serialisation information needs to flow, not just internal packaging site or CMO/ CPO links, in order to understand the regulatory and business data flows that an infrastructure needs to support. The serialisation infrastructure will likely have to support a surprising diversity of data types, connection methods, business preferences and regulatory interpretations across various network relationships.

For DSCSA, there will be multiple serial number requests and responses with the different packaging line systems, commissioning events will need to be captured and potential aggregation data exchanges managed as product travels from packaging line to packaging site warehouse, through internal warehouses and into 3PL facilities. Shortages, damaged product and other exceptions need to be dealt with in new ways where serialised product is involved.

Downstream, pharma companies may have to send aggregation data to direct trade partners and capture and respond to SNI verification inquiries from any entity that has the product in their possession. Each of these partners may have its own preferences, system capabilities or interpretations which can impact how the serialisation system is designed and how the packaging lines are deployed.

Serialisation creates unprecedented scalability challenges

Today, the pharma industry is accustomed to work on batches, where shipping a batch of 10,000 units requires just four simple transactions. The batch is created, the pallet is picked and shipped and an advanced shipping notice (ASN) is sent.

However, serialisation requires the management of a unique identity for every single product. Now, this shipment of a 10,000 unit batch will require at least 60,000 database transactions to be processed in order to update the data corresponding to every single unit. Provisioning serial numbers will require 10,000 transactions and commissioning the batch will be 10,000 more. And new tasks, such as aggregation or shipment notifications, introduce thousands of new transactions. The bottom line is that the four transactions completed today to ship 10,000 units will be – at minimum – 60,000 transactions in a serialised world. And as the transaction volumes expand, so will the information processing requirements: they will be 15,000 times more and the size of the database will be 5,500 times larger, for the same volume of products shipped today.

Most solution providers use an EPCIS scheme with a relational database management system (RDBMS) for serialisation. Maintaining performance in a relational database focuses on keeping the number of rows at a reasonable level, sized according to the hardware that is available. Typically, all the processing of a relational database goes through a single machine. To do this, database administrators are forced to invest time and effort managing the number of rows in their database to maintain performance. If they don’t, they run the risk that performance degrades so far that the database stops working altogether. The only other option in these cases is to scale up, which means they must acquire bigger hardware and more powerful machines.

Even if your company manages low volumes, a RDBMS will begin to fail when handling the exponential rate in which serialisation data is created and replicated across tables, along with the transaction processing speeds required to perform at operational scale.

In the case of serialisation, pharma manufacturers cannot risk performance, otherwise product will cease to ship.

Knowing that RDBMS won’t scale, a majority of the pharma companies facing serialisation requirements have quickly realised that NoSQL is the preferred approach. Used by Facebook, Twitter and Google, NoSQL is a non-relational database widely recognised for its high scalability and performance that remains near-constant, regardless of the data store.

When evaluating serialisation providers, pharma manufacturers should ask for a demonstration on serial number provisioning, commissioning, aggregating, shipping, and read/write speeds on a database containing several years of accumulated data. A demonstration on a clean, empty database – or a pilot that doesn’t reflect anticipated daily volumes – will not provide any insight into the solution’s operational scalability and performance and may likely raise a red flag.

Serialisation fundamentally changes how a company conducts business

It’s crucial to reach out across the organisation, from quality and artwork, to supply planning, trade relations and commercial operations, to understand how corporate functions are impacted by serialisation. Continuous education on serialisation regulations and their implementation rules, data standards and industry implementation trends across the organisation is important to gain informed feedback on organisational needs, preferences and requirements to optimise serialisation planning. For example, good distribution practices in the warehouses may conflict with the needs to maintain aggregation relationships across the organisation.

Most companies will undergo a transition from lot-level identified to serialised product over time. So the systems, processes and connections to supply and trade partners will need to be flexible in managing both serialised and lot-level product throughout an organisation and network.

Serialisation preparation timelines are always longer than they appear

Enabling a packaging site for serialisation is a long and complicated project. In fact, many companies have started planning for serialisation 18 months ahead of the November 2017 deadline. In order to leave enough time to test new systems and processes, as well as make any necessary adjustments. The preparatory work is multifaceted and not accounting for all factors could lead to a flawed implementation.

Depending on product stock levels and velocity through the supply chain, it may take months to bleed out existing lot level product from internal warehouses and external distribution sites. Existing supply plans need to be incorporated to determine how often production runs are executed, how long they take and when existing production lines can be idled for retrofit.

Rarely can all serialisation lines be upgraded in parallel due to cost or resource constraints. So, projects must be staged in phases. As each line and site is being serialisation-enabled, it must be integrated into the internal serialisation architecture linking the enterprise systems, warehouse management systems, edge devices and trading partner systems across which serialisation data and serialisation events must flow. This places a requirement on the business to understand how this architecture should be developed well in advance of line deployment. With only one year left before the DSCSA serialisation deadline for pharma manufacturers, there are still many issues to address. The requirements far exceed just putting numbers on bottles, as those that have started serialisation projects can attest. Indian pharma companies that need to comply with DSCSA regulations should be starting now and building a strategic plan to meet both DSCSA and global regulatory requirements.