Data Integrity Pharma
UNDERSTANDING COMPLEX REGULATIONS AND DESIGNING AN EFFECTIVE IMPLEMENTATION STRATEGY WITHIN A SHORT TIME (PART 2)
In the second part of the article we will delve deeper into key concepts discussed in the first part, we will introduce the topic of Data Governance, we will analyze the main types of data and their sources, and we will see how it is appropriate to treat each of them from the perspective of Data Integrity.
Finally we will try to define the principles on which to base the design of a good strategy to achieve Data Integrity in a pharmaceutical company, identifying the main phases, challenges, opportunities and risks.
The question that every drug producer must answer concretely is: ok, now that I understand the requirements defined by the regulations, how do I achieve the goal and guarantee the integrity of the data in my company?
The number and complexity of regulations and various guides such as GLP (Good Laboratory Practice), GMP (Good Manufacturing Practice) and lately also GAMP (Good Automated Manufacturing Practice) suggest that there is no simple and immediate answer. The road to Data Integrity is long and winding, and must be traveled by all GxP impact departments. Projects of this size require a plan, often a multi-year one, a real strategy.
DAILY CHALLENGES OF DATA INTEGRITY
The laboratory analyst and the production operator in their daily activities follow standard operating procedures (SOPs), analysis methods or batch records. They do this by documenting the entire process and recording the results. To facilitate these processes many companies have adopted digital systems such as LIMS and ELN (Laboratory Information Management Systems and Electronic Laboratory Notebooks) in their laboratories and EBR (Electronic Batch Record) in production departments. These systems are undoubtedly of enormous help in automating operations. However, this conversion is often partial, for two reasons:
These solutions are designed to mainly cover the display part of instruction sheets or analysis methods and to aggregate the results of analytical tests and department operations, entered into the system manually by the operator.
These solutions are not prepared natively for interfacing with electronic devices for the purpose of collecting process and analysis data.
Moreover, the data managed in production and in the laboratory cannot be treated only as a simple set of parameters, as they have a complex and not at all standardized structure:
risultati raccolti dagli strumenti analitici (bilance, titolatori, pH-metri)
production line sensor readings
lo stato dei dispositivi
users and their permissions
i metodi applicati e i metadati
La sola acquisizione elettronica di una quantità limitata di dati di misura senza i relativi metadati (ID dispositivo, utente, tara, dati qualitativi dei campioni, cronologia di taratura/pulizia/utilizzo, SOP adottata, versione del metodo, ecc.) lascia la misura priva del contesto.
Soon it turns out that the process cannot be managed effectively if even every single interaction with the device itself is not described in the work flow during the design of the device and, subsequently, is not tracked during its execution. These types of interactions are as difficult as they are to describe in a paper SOP or Batch Record. It is already so for simple electronic systems, we just can imagine how the challenge becomes unsustainable as the complexity of digital systems increases.
Data transcription is susceptible to errors
Lack of data audit trail
Reduced accuracy/completeness of recordings.
Non-complete compliance with ER-ES (Electronic Record - Electronic Signature) requirements, although the systems are validated.
Documents, and data in general, are not stored in a single location but are distributed in the various departments of the company and saved on different types of media (electronic and paper).
Process managers do not have the necessary knowledge and command of their own processes:
paper procedures are subject to interpretation
it is easy to underestimate the priority of requirements (what data to store, how long, audit trail, ...)
risks are not always adequately managed (impact of failures, durability of magnetic media, ...)
Briefly: failure to comply with GLP, GMP and GAMP regulatory requirements
A STANDARD APPROACH TO DATA INTEGRITY
When launching a Data Integrity project it is necessary to take into account some organizational risks. One of these can be represented by the question "Who should take responsibility, the leadership of the project?". It is not easy to answer to this question: since the DI regulation falls within the scope of quality assurance, it could be said that the person in charge is the Director of QA. 10-15 years ago this might have been the right answer, since at that time a large part of the procedures and GxP data were handled on paper. Today, if the company's QA manager has solid experience in digitization and has already followed a Data Integrity project, he can take on that responsibility. Otherwise, it is better to entrust the guidance to the IT systems manager or a specially selected person. Later on we will see why.
Since 2015, when the topic has become a priority, many companies have already moved with Data Integrity projects, so it is possible to collect testimonials and make an evaluation.
The most common line followed by companies is, of course, to ensure coverage of ALCOA principles for data generated during the production, analysis and movement of drugs.
However, in the context of the high complexity of data and the large number and variety of systems, this approach can prove to be far from easy.
Digital system and device manufacturers, however, leveraging the trend and sniffing out new business opportunities, propose to interface their devices to a PC or, in the worst cases, to provide devices with the ability to export data manually. Even if the first method (interfacing with a PC) was adopted, this could be a not definitive or optimal solution to the problem.
WHAT DO COMPANIES DO BASICALLY?
So, with or without a well-defined strategy, here is a range of actions normally taken by companies?
If possible, adaptation (configuration) of existing systems
Purchase of new complete systems (hardware-software) to replace existing systems
Purchase of new software to upgrade existing systems
Training and awareness of the staff to the D.I. concepts.
Back to the MHRA recommendations, let's focus on the word relevance. There is a strong temptation, in fact, to order the list of systems identified for increasing complexity and to plan the work according to this order, believing that what counts is the number of "regularized" devices.
However, remembering that the subject is the data and not the system, we can say that with 20% of the carefully selected systems it is possible to reach a Data Integrity coverage level of 80%.
Solo dopo aver stilato questa TOP 10 (o TOP 20, o TOP50, dipende dalla dimensione e complessità dell’azienda) si procede all’individuazione dei sistemi coinvolti nella loro gestione. Per attribuire una priorità ai sistemi individuati sarà necessario ordinarli, questa volta in base alla linea temporale della vita del dato (generazione, trasferimento, trasformazione e conservazione del tato). Ma questo non è altro che mappare, ad alto livello, i processi attraversati dai dati critici. Per cui, volendo dare un ordine di importanza, la sequenza della nostra indagine avrà questo ordine: Dati - Processi - Sistemi. Questi ultimi, a loro volta, si trascinano dietro l'analisi dell'infrastruttura sulla quale dovranno funzionare.
All this will help to understand the real data stream, its complexity and possible inconsistencies and redundancies. Yes, because one of the actions could also be the elimination of a system, if this introduces redundancies and unjustified complexity.
Esiste un modo consistente e valido per arrivare in breve tempo ad un piano operativo (detto anche "rimediation plan")? Sì.
L'obiettivo di questo articolo è proprio la definizione di questo metodo. Ma essendo, appunto, un metodo, e non una soluzione pronta all'uso, è necessario comprendere i suoi principi di base.
TOP-DOWN, BOTTOM-UP... INSIDE-OUT!
Often we hear of a "top-down and bottom-up approach to ... achieve data integrity" and "Data Integrity by design". But what does it mean exactly?
One way to achieve DI could be to identify all gaps in data management, monitor all errors, sporadic or systematic, and then plan corrective actions. But this would mean chasing problems, a path where all actions are of equal importance and never leads to a conclusion, simply because it has not been previously defined.
Moreover, because of the constant "fear" of inspections and receiving notifications, a habit has been established in companies to hide errors and inconsistencies. This definitively eliminates any hope of success, in case the method described above was adopted.
Non possiamo di certo pretendere che FDA non rilevi le mancanze durante le ispezioni. È il loro mestiere. Ci vuole quindi tanto coraggio e determinazione ai piani alti delle aziende farmaceutiche per adottare un approccio diametralmente opposto.
Innanzitutto, è importante comprendere che il Data Integrity non è un problema di un solo reparto, come Assicurazione Qualità o IT. La qualità e la disponibilità del farmaco, che sono gli obiettivi primari di qualsiasi azienda farmaceutica nel suo complesso, dipendono sempre di più dai dati, e non solo quelli GxP critici. Di conseguenza è importante avviare, a livello aziendale, un ben definito programma culturale di data governance.
Everyone in the company needs to be conscious and feel the problem of DI. But above all, employees should be encouraged to report problems instead of "burying them", while management must take responsibility for collecting, analyzing and implementing a structured plan to resolve them.
Tra le prime soluzioni, quindi, è necessario implementare (o ottimizzare se esistono già) strumenti e processi che facilitano la sorveglianza dei sistemi. E se il monitoraggio non è automatico, com'è tuttora nella stragrande maggioranza delle aziende, una volta fornito lo strumento bisogna predisporre le procedure di segnalazione manuale, con meccanismi di stimoli per le loro adozioni.
It is also important to have a clear understanding of the complete data stream, how it is shared between the various actors, and the systems that make up the company's ecosystem.
The main steps for the design of the strategy can be summarized in:
Understanding of the processes and data involved within them
Identification of risks and integrity gaps
Analysis and design of the remediation plan
However, before we begin, we need to establish what the real goal is. Alignment with ALCOA principles? That is, "I do as much as I need to". Or do I take all possible advantages from this opportunity?
For CEOs and IT Managers data integrity can be an opportunity to build the digital environment of their dreams. Taking advantage of the historical period, characterized by pressure from regulatory agencies, they could finally justify and get the budget they have been waiting for a long time. Benefiting, of course, is the entire company. It's just that it's harder for non-experts to imagine these benefits before the whole strategy is fully achieved.
Data is king
DATA INTEGRITY OR DATA GOVERNANCE?
We have already mentioned Data Governance before and now it is time to structure the ideas. The two concepts are not an alternative to each other. Data Integrity is a key component of Data Governance. Moreover, as you can imagine, they do not belong exclusively to the Pharma world, and since data has been crucial for any modern company for years now, Data Governance is vitally important at a corporate level.
The range of measures put in place to ensure that data, regardless of the format in which it is generated, is recorded, processed, stored and used in a way that ensures its integrity (i.e. completeness, consistency and accuracy) throughout its life cycle.
The main Data Governance issues include availability, usability, consistency, integrity and security. Data Governance involves the establishment of processes that ensure effective data management throughout the company and the allocation of precise responsibilities for emergency management, data quality and availability within the organization.
Since this article is dedicated to pharmaceutical companies we will focus on the most critical aspects of Data Governance for the Pharma world.
The Data Governance model in the pharma industry must necessarily be determined by the criticality of the data, rather than the complexity of the system that deals with it.
Making a complete mapping of all data managed in a company from scratch could be costly.
The good news is that, for the classic management systems in use in pharmaceutical companies, this analysis should already have been done. Simply ask your colleagues in the Quality Assurance department about the outcome of the risk assessment for ERP and LIMS made at the time of their validation. This is likely to contain a criticality ranking for the key data as well as the processes implemented. With a little luck you can even acquire valuable information about the data managed in the laboratory.
The "bad news" is that, thanks to technological progress and new digital systems, the data is increasingly detailed, complex and pervasive. The miniaturization of the sensors and their diffusion have made it possible to collect a multiplicity of data that was previously collected manually and only if strictly necessary.
Not only that. Applications for the management of workflows are no longer exclusive to accounting processes or Master Data (e.g. multi-role approvals of master data). Thanks to the spread of HMIs, it has been possible to define digital workflows also for those departments that until not so long ago were managed only through paper SOPs and Batch Records. And once the execution of processes becomes digital, numerous opportunities arise for data collection, not only environmental (from sensors) but also operational, such as the execution of process methods and instructions.
THE IMPORTANCE OF RISK ASSESTMENT AND THE CHALLENGE OF THE PROPER DATA CLASSIFICATION
Parlando in un linguaggio tecnico, bisogna considerare parametri critici di processo (CPP) e degli attributi critici di qualità (CQA) dei prodotti.
By correlating the characteristics thus identified with the work instructions (manufacturing process) and by identifying the ways in which the data make the transition from one phase to another, we obtain a scheme that represents the journey of critical data through business processes. It is not required to analyze the processes in great detail, which would lead to an unjustified waste of time. The real focus is always the data, so it is enough, starting from the identified input attributes and parameters, to track all their steps, enhancements and transformations.
By doing so we obtain the "data - processes" link. The next step is to cross-reference the results obtained with systems (applications), devices and infrastructures.
Another way to speed up the assessment phase is to design the strategy is to follow known patterns and try to drop one on the company using the documentation already available together with the knowledge and experience of the staff.
We can take as reference the best known standard in the automation field. ISA-95 is an international standard from the International Society of Automation for automated interfacing between company systems and control systems. In practice it has proved to be a very useful tool for understanding and structuring the panorama of business systems.
The standard does not give priority to any particular type of data but rather focuses on the exchange of data between different levels of the architecture. To show a practical example we analyze the flow of master data. The master data flow could be synchronized between systems in a one-way manner. The meaning of the standard is that it could be advantageous to manage each type of master data (product codes, recipes, methods, MBR, etc.) in a single repository, a DB for each master data or for more than one (even if some companies opt for a single system dedicated to the management of all company master data).
This facilitates the application of ALCOA and Data Governance principles as it allows to intercept the birth of data from a single source and its distribution along all defined paths. A similar reasoning can also be applied to transactional data such as batch, process order, department scheduling, planned and confirmed quantities, etc.
Following this approach, in a single analysis session it is possible to:
classificare tutti i sistemi esistenti (in base ai livelli ISA-95 e alla tipologia digitale/cartaceo)
stratify the data itself on the basis of these layers
Finally, track the "routes" that every critical data travels from its birth to archiving.
Here one must not be misled by indicators. Theoretically each level, each system can cover at least the first 2 phases of the data life cycle.
processing and reporting
storage and recovery
Of course, the higher the level of membership of the system, the more data managed is structured, aggregated and includes data from lower levels.
TYPES OF EQUIPMENTS
In chemical industries, the production department is usually less automated compared to purely pharmaceutical companies. As a result, production processes are less complex and efforts related to Data Integrity are more focused in laboratories, being the bulwark of raw material quality. The most complex systems in the laboratory are, of course, HPLC.
Companies whose core business is drug manufacturing have a hybrid situation. Both the production and the laboratory produce and process a lot of critical data.
Un caso a parte sono le aziende CMO (terzisti) che coprono solo i processi di packaging per le aziende che hanno deciso di esternalizzare il confezionamento. Nel loro caso la complessità maggiore è concentrata nell’automazione delle linee di packaging, non dovendo eseguire le analisi di laboratorio approfondite, già effettuate dai loro clienti.
IT IS EASY TO UNDERESTIMATE THE COMPLEXITY OF THE SYSTEMS
In the first version of the "GMP Data Integrity Definitions and Guidance for Industry" the MHRA suggested a classification of systems according to the complexity of their configuration, specifying that the more complex the system, the greater the effort required for its validation. The logic is simple: where more configuration is required, the greater the risk of data manipulation (it is therefore Data Integrity).Purtroppo questa figura viene spesso mal interpretata e alcune aziende confondono la semplicità di convalida con la facilità di raggiungere la conformità di Data Integrity, sottovalutando così il rischio.
Paradossalmente la tipologia di sistemi più difficili da mettere in conformità sono gli strumenti semplici, come bilance, pH-metri, conduttivimetri, ecc..
Stessa MHRA mette in guardia su questo rischio:
However, it is common for companies to overlook systems of apparent lower complexity. Within these systems it may be possible to manipulate data or repeat testing to achieve a desired outcome with limited opportunity of detection (e.g. stand-alone systems with a user configurable output such as FT-IR, UV spectrophotometers).
Moreover, the complexity is given by the number of these instruments, their variety and the diversity of data formats that each device provides.
We can also add that as the complexity of the system typically increases, so does the "paperless" level, without the need to generate reports on paper. This too, however, is a rule that now tends to remain in the past, given the spread of digital, integrated and paperless systems also in the left part of the scheme.
This diagram is particularly useful because it represents an overlap of the ISA-95 standard with the types of typically pharma systems/devices, all with a view to Data Integrity. The focus, therefore, is on the relevance of the paper/electronic data for system adaptation and subsequent validation activities.
WE HAVE STRATEGY. NOW WHAT?
What we have seen so far is already enough to make an assessment and outline a Data Integrity strategy. The primary objective is to identify critical data and prioritize them. Having done this we have seen methods that allow us to quickly list and classify company systems and devices. By cross-referencing the results of the two analyses we can prepare a high-level strategic plan that will lead us towards our goal of operating under Data Integrity.
There is still so much to do. In fact, the plan obtained must be articulated in an operational plan, also called Remediation Plan. Since each system, each device, is special, the remedial action to be taken for each can also be different. For one system it will be sufficient to define the roles and permissions, for another it will be necessary to purchase an additional SW. Other systems that handle critical data may be so outdated that they will not offer any remedial action, and therefore they will be intended for replacement.
The transformation of the strategy into a remediation plan, therefore, is another important and challenging step. In the next chapters, we will no longer talk about data. On the understanding that, from what we have seen in the previous chapters, we already know what our data is to be protected. Instead, we will present elements of system and infrastructure assessment that will help us to choose the appropriate remedial action.
IDEAL DATA INTEGRITY ENVIRONMENT
Prima di immaginarci una soluzione di Data Integrity Ideale rinfreschiamo le principali osservazioni provenienti dalle ispezioni FDA. Nella seguente tabella vengono proposti dei suggerimenti utili alla gestione delle osservazioni sopra riportate.
|- Mancato controllo degli accessi
|Assignment of authorizations for:
- use of automation / production / laboratory systems
-- access to factory area
|-- profile setup with different privileges (segregation of duties)
- badge, fingerprint scanner etc. for system access
|- input not contemporary to the activity and backdating
-- use of unofficial sheets and notes
- input the same data in different batch records
|- registrazione dati contestualmente con le attività
- revisione periodica degli audit trail
|- sensibilizzazione su DI e training del personale
- supporto alle attività da parte dei consulenti qualificati
|Mismatch between reported and recorded data
|- acquisizione dei dati tramite sistemi elettronici e tracciatura digitale del flusso dei dati
-revisione periodica di audit trail
|- interfacciamento dei sistemi digitali e convalida di questi
- training al personale
“Your quality system does not adequately ensure the accuracy and integrity of data to support the safety, effectiveness, and quality of the drugs you manufacture. We strongly recommend that you retain a qualified consultant to assist in your remediation.”
ELECTRONIC OR PAPER?
Among the first actions during the definition of a remediation plan, it is necessary to take a choice: continue to manage the identified data on paper, or implement a digital system? In the case of already installed digital systems, it is possible to consider whether or not to replace them. While it is highly unlikely that the choice to switch from a digital system to a paper one.
Dal punto di vista delle normative, entrambi i tipi di supporto sono accettabili, purché rispettino criteri definiti da ALCOA:
an electronically generated raw data can be stored in an acceptable paper format or in pdf, where it is justifiable for a "static" record to preserve the integrity of the original data.
In these cases, however, it must be proven that the data retention process includes verified copies of raw data, metadata, audit trails and verification files, system configuration, specific settings of each production or analytical process execution, including the data processing processes necessary to reconstruct a specific raw data set.
a document certifying that the printouts are an accurate copy of the original data is also required.
The choice between Electronic and Paper depends on the criticality of the data and the complexity/cost of the system to manage the data in electronic format. We summarize the main ALCOA recommendations trying to decline them for both types.
Per gli strumenti "semplici", come bilance e pH-metri, le linee guida [WHO] considerano accettabile che i dati vengano stampati su uno "scontrino". Purtroppo, queste stampe sono carenti e spesso è necessario arricchirli con metadati, come la data e l'ora dell'evento e firma dell'operatore che ha eseguito l'azione, l'ID del campione. Lo scontrino deve poi essere fissato sul quaderno analitico. Tutti questi passaggi richiedono delle procedure chiare e robuste e un'attenzione da parte degli analisti. Attenzione sottratta alle attività principali, e cioè l'esecuzione delle analisi.
DIGITAL SYSTEM LIMITATIONS
To overcome the lack of instruments the alternative is to replace them with models featuring an internal clock (synchronized with the company one) and user authentication tools. Unfortunately, such an expensive solution, even if desirable to renew the toolkit, is not always sustainable.
It is clear that if new purchases are planned for the laboratory, it is possible to choose the instrument already compliant with Data Integrity. And what about the existing stock of instruments?
Although it is possible to adapt existing instruments or to purchase new instruments that already comply with the requirements, as long as these instruments remain "stand-alone" they cannot be considered fully compliant as they can still lead to a loss of data or worse consequences. Just consider the risks of manual back-up via USB stick. This type of manual storage cannot be sufficient as it does not guarantee the completeness of each set of results.
There is another aspect that needs to be considered. Laboratory Information Management Systems (LIMS, ELN, LIS) can be easily integrated with each other to work in synergy, thus covering Data Integrity requirements. While the integration of analytical instruments can be a difficult task as you often operate in an environment with many different instrument brands and many disaggregated SW. Integrating them all directly into the LIMS means giving it responsibility and functionality it was not designed for. One solution to this could be the adoption of middleware.
As mentioned in the first part, regulatory agencies have recognized both the advantages and limitations of electronic data systems and have increased the controls for their use. Although they do not explicitly push for the adoption of digital systems, they much prefer them to paper systems because of the opportunities they offer for compliance.
However, official regulations do not extend beyond the coverage of requirements and do not state anything about the optimization of processes that can be achieved with the introduction of these systems, as if this were not relevant to Data Integrity. Let's see how the introduction of digital workflows, in addition to process improvement, can help to achieve excellence in Data Integrity.
DIGITAL. AUTOMATIC. CENTRALIZED.
FREE YOUR HANDS. AND YOUR MIND!
Quali sono le attività principali dei analisti di laboratorio o degli operatori di produzione? Fabbricare i medicinali, eseguire le analisi, o fare ricerca. Attività ad alto valore aggiunto che richiedono concentrazione, approfondimenti, oltre alle competenze ed una forte specializzazione. Invece, una buona parte del loro tempo è dedicato al controllo di processo ed alla produzione della carta.
Unfortunately, the time they have left to perform the main tasks is not an algebraic difference between the total time available and the time spent managing and controlling the process. Being two very different types of tasks, any person, regardless of their experience, needs time to refocus by moving from one to the other and viceversa.
Studies estimate that 20 to 30 minutes are needed to reach the right concentration on a complex activity. This factor is a real burden for any company, even a structured one, and is often underestimated by managers.
That's why the real advantage of being paperless is not the elimination of paper and the automation of data collection. It goes without saying that the elimination of these routine tasks simplifies core operations. These factors free the hands of operators, while the real process boost is achieved by freeing their minds, when they can not think about "what" and "when" they have to do, but only "how".
There are various possibilities and options to automate workflows:
Automatic identification of samples, and materials in general, by bar-code labels.
Planning of activities through a dedicated SW and sending the schedule to the workstations/operators.
Digital management of analysis methods and process instructions, both preparation and execution.
Automatic decisions taken by the systems according to pre-established rules and collected data.
Automatic generation of accurate reports.
Automatic sending of data to interested parties (people, departments, entities, or digital systems).
In general, regarding the future, the skills required from the staff will change, going towards a strong specialization, on the one hand, and towards the ability to bring continuous improvement and automation of processes on the other.
Is there a way to seize the benefits of digital systems and avoid the disadvantages described above?
There is another dimension of automation that is worth examining further. Today many manufacturers of laboratory instruments and production lines claim that their instruments comply with CFR21 and Data Integrity standards. Usually this applies to each device individually, which for the company means managing both the master data and the data generated by this device separately. Each with its own format.
Since this fragmented management is extremely expensive, it becomes the subject of continuous internal discussions between departments on who should be responsible for it. At best, each department takes over this burden. At worst, it all falls to the IT department, for a simple logic "digital means IT", without understanding that digital is now a problem for everyone. And it takes a strategic vision and strong leadership from those who lead the company to turn the problem into an opportunity and make the only right decision: adopt a centralized and automated system on which to "transfer" the responsibilities described above.
Centralization is synonymous with standardization which is synonymous with simplification and cost reduction. A unique, digital and centralized tool helps, but we can also say forces, to adopt uniform templates for master records, methods and batch records.
Centralization also requires unique user management. A single database containing all the accounts automatically means a single digital model, so-called "digital twin", of the entire company organization chart (instead of as many incomplete models as systems and devices) with all the defined roles and permissions. You can imagine how the management model and administrative activities are simplified. But above all, how this approach, being the organization chart one of the pillars of the company, clarifies and simplifies all processes, helping to detect possible conflicts and redundancies (another important source of inefficiencies) in data flows.
The creation of a centralized database makes it possible to set the clear boundary between operational processes and processes of verification and interpretation of results. Those involved in the verification of collected data and batch approval do not have their desk cluttered with folders of documents from different sources. Nor should they be in the vicinity of the data source. This offers the opportunity, for example, to have a single detached department for monitoring and verification of results, which serves multiple laboratories around the world.
Finally, centralization and the resulting standardization allow the systems validation cycle to be optimized. The centralized software only needs to be validated once and each new connected analytical instrument will simply require configuration and creation of the respective master data, as the system itself, data management, audit trail, user management, result flow, etc. are already validated.
Centralized software, therefore, can play a key role in a data integrity solution offering:
Reducing the cost of analysis per sample
Better use of analysts' skills, so that they can focus on the development of new methodologies, data analysis and their clients
Miglior qualità e omogeneità dei risultati, per un confronto coerente con i risultati di altri laboratori
Improvements in accuracy and precision
Higher analysis speed
Consequential increase in the demand for laboratory services
All this, in short, leads to a reduction in operating costs and annual manpower, which in the laboratory represent a significant share of total costs, unlike production whose costs are distributed between machinery and materials handling.
Of course this increase in efficiency, and therefore in the value of its operations, are long-term benefits. And as with any long-term return, companies, who recognize its strategic importance, must be prepared to undertake major investments both in technology and in the culture and skills of their employees.
DATA INTEGRITY INFRASTRUCTURE
PCs, servers with databases, service software, hardware servers, operating systems are the infrastructure components that underpin GMP applications. To these, we need to add the different IT tasks such as user management, server management, system management, backups, licensing, etc. It should be noted that the quality of the infrastructure is mandatory.
As mentioned in the previous chapters, since the goal is the integration of everything with everything, the crucial role is played by the component that connects different systems that speak different languages. Therefore, a middleware is not linked to any particular system, but was created to adapt to any communication protocol and method. The introduction of middleware facilitates:
measurement detection directly from the instrument
sample identification by integration with a barcode/datamatrix reader
data e ora vengono presi dalla rete
Possibly integrates electronic signature SW, or other user identification tools: fingerprint, heartbeat, facial recognition (in environments that do not require wearing a mask), etc.
When we talk about the need for a centralized DB we automatically think of the Cloud. Preferably a Data Lake. Nowadays everyone talks about it, even in the pharmaceutical industry, however, the same companies are still wary of delegating the management of their data to external actors.
We can't really know if data confidentiality is guaranteed once they reach the cloud. The only way to protect ourselves here is to have a contract with the supplier and a solid insurance policy that guarantees adequate compensation for the damage suffered in the event of a given breach. However, we cannot deny that, from the point of view of data back-up and availability, no pharmaceutical company can compete with a good cloud service provider. However, guidelines for the qualification of these types of infrastructure are still being defined by regulatory agencies.
Once the cloud is "cleared", all the opportunities it provides are unlocked. Access to external data and services, building complex and integrated applications, Big Data with its advanced analytics, machine learning and predictive algorithms. All this, or at least critical services such as production execution, cannot migrate to the cloud until there is a fast, secure and always available internet. But this is the subject of another article.
So what is the common feature of companies that are successful in Data Integrity?
Since the biggest risk is that DI can be seen as a problem and not an opportunity, we can say that the winning companies are the ones that use Data Integrity as a way to reach an advanced level of their growth and market presence.
Those companies, who see Data Integrity as an enabling factor for digitization, certainly have a far-sighted vision.
What is the point of investing a lot of money in making paper processes and data compliant with DI if they will have to be digitized very soon anyway?
Il budget per DI e per la digitalizzazione, dunque, può essere visto come un'unica voce del budget globale, egli investimenti non devono solo coprire la compliance ma soppratutto mirano al miglioramento dei processi. Perché la compliance è un beneficio secondario rispetto ai vantaggi che si ottengono dal raggiungimento del Data Integrity.
From here we can see another important requirement. In order for the company to be able to achieve the structure outlined above, the IT department needs to be able to make choices that "invade" the territories of other company departments. IT can no longer, therefore, be seen as a service department. As a "pit stop mechanic" that it used to be, it must now necessarily sit by the driver's side and suggest the way. And the driver is the only one who can afford it, apart from the owner of the racing stable, of course.
Ympronta and Data Integrity
OUR services and products
At Ympronta we have been working in the pharmaceutical industry since the early 2000s. We have seen the creation and evolution of CFR21 and GMP. We have observed the definition of Data Integrity and its ALCOA+ principles.
But we haven't just been passive bystanders. Operating mainly in the Lifecience, each of our projects had to comply with regulations so the experience we gained led us to an important decision: to found a company to offer our expertise directly to our customers!
The result has been the creation of a department dedicated to compliance and Data Integrity and the development of digital systems dedicated to the most complex areas:
Digital Data Integrity in the analysis laboratory (Y-Lab)
Pharmaceutical Serialization / aggregation (SERYAL)
Production automation (Y-Mes)
Contact us to learn more and to book a meeting where we can assess your challenges and objectives.