David J. Balderson, vice president, Global Safety Operations at Sciformix - a Covance company - examines how automation is helping to reduce negative drug outcomes.
Automation
Pharmacovigilance (PV), the process of identifying, tracking, evaluating and preventing negative outcomes from drug therapies, is a sector that has seen huge growth in recent years. According to a Global Market Insights report, the PV market is predicted to exceed US$8 billion by 2024, with its growth attributed to factors such as: a rising number of adverse drug reactions (ADRs), an ageing global population and rising chronic disease burden; an increasing number of pharmaceutical companies and the emergence of personalised medicine. While PV starts during clinical development, it is not limited to clinical trials alone as post-market surveillance is crucial to monitoring a drug’s safety after it has been approved.
There are a range of challenges the pharmaceutical industry faces when establishing and maintaining increasingly complex PV systems. The evolving regulatory environment in a progressively global industry places demand on pharmaceutical companies to manage PV activities more efficiently than before. With the ongoing pressures of optimising costs, traditional PV strategies must be revised and revitalised with smarter spending in mind. The focus is now shifting from primarily safety operations to proactive risk management, personalised medicine, and completely transparent data between pharma, patients, healthcare providers and regulatory agencies.
Technological advances are playing a major role in pharmaceutical PV strategy updates. For example, more companies are looking towards cloud-based solutions, mobile applications, robotic automation, artificial intelligence (AI) and big data analytics as a vital part of clinical, safety and regulatory operations in the pharmaceutical industry. Applying innovative technology automation tools and processes to PV strategies is now a critical requirement for managing the safety of pharmaceutical products.
Optimising efficiency
As one of the fastest growing life science disciplines, PV strategies must be optimised for peak efficiency. A well-established principal information technology (IT) framework provides organisations with high performance and scalability, together with system validation and information security, for effective design and distribution of automation initiatives. However, many of the problems surrounding PV systems are not IT issues, but are actually down to the processes or individuals managing the system holding onto personal preferences rather than absolute requirements. Addressing process improvements and organisational requirements in parallel with IT solution improvements will enable operational efficiency to go further and drive a proactive PV strategy.
AI has the potential to fill the gaps that traditional PV services currently leave, such as the ability to assimilate large volumes of cloud-based data and map patterns, in order to effectively predict ADRs. Genetic information and real-world patient data can also feed into this more streamlined approach to make PV more of a predictive science. Integrated IT solutions that combine scientific and technological expertise are capable of delivering high operational efficiency, quality and regulatory compliance.
The four stages of automation
Many current IT systems and applications are capable of automating case processing and reporting activities, but the overall process still requires significant manual effort, particularly regarding case intake and data entry. There are multiple levels of automation that can serve to make end-to-end safety processes more streamlined and strip down redundant, non-value added steps in existing processes, while increasing human labor efficiency.
The first stage is basic process automation, which involves tracking and monitoring tasks and enables the collection of continuous metrics. Basic automation provides reporting and dashboards, and automates a workflow that involves multiple roles, but still requires manual entry, processing and analysing of safety data into a database or system. Robotic process Automation (RPA) is the next level, and helps to reduce or eliminate these manual tasks. RPA is often combined with the subsequent level, cognitive automation, which leverages Natural Language Processing (NLP) to assist human decision-making. The system engages in human interaction, whereas the final level, AI, requires little or no human interaction and self-learns through experience, to make predictions based on patterns observed in large volumes of data with the help of machine learning (ML).
Regulatory responses to industry change
Ever increasing volumes of drug data places an urgent need on the development and implementation of technology capable of providing a secure, integrated big data repository. For example, all AEs, regardless of their degree of severity and source, should be stored in a single drug safety database.
Cloud-based capture and reporting is a key trend in the PV space, and is now being used to bring a fully-integrated database to all stakeholders. Integrating cloud technology can further optimise data intake, storage and analysis, and even provide regional and temporal insights into ADR patterns. Healthcare providers, physicians, users and research institutions can store and access drug safety information, whether reported during clinical trials or post-market experience. Cloud-based systems can also reduce latency between reporting and analysis, leading to more timely and qualitatively improved regulatory decision-making surrounding crucial public health issues.
The European Medicines Agency (EMA) is in the process of implementing the standards developed by the International Organisation for Standardisation (ISO) for the identification of medicinal products (IDMP), in a phased program. The overall aim is to simplify the exchange of information between stakeholders and improve international system interoperability, and is expected to benefit PV processes. AE reports will be based on a harmonised set of medicinal product definitions, improving the quality of data used for safety signal management and speeding up communication, decision-making and regulatory actions. The EMA also monitors ADR data using the EudraVigilance database that facilitates data collection, management and analysis. It can determine whether there are new or changed risks associated with a drug and if these risks impact the benefit-risk balance.
In the US., the Food and Drug Administration (FDA) launched its Sentinel Initiative in May 2008, which is an integrated electronic system that enhances the FDA’s ability to proactively monitor and assess post-marketed medical product safety. Sentinel complements the existing FDA Adverse Event Reporting System (FAERS) – a database that contains AE reports, medication error reports and product quality complaints resulting in AEs that were submitted to the FDA. The FDA sees how RWD and RWE can be especially useful for post-market monitoring of the safety of products during their use in real world settings. For example, its use of RWD and RWE, derived from its Sentinel system, eliminated the need for post-marketing studies on nine potential safety issues involving five products, making the FDA’s post-market evaluation of safety timelier and more effective. By using RWD and RWE, we may be able to provide patients and providers with important answers sooner - identifying a broader range of safety signals and following up on them more efficiently.
Automation strategy implementation
Higher levels of automation, such as RPA, cognitive automation, or AI, enable organisations to identify patterns in unstructured data, and can automate the whole process, from case receipt to reporting. Implementing an automation strategy can not only reduce costs, but also eliminate the chance of human error, and therefore, improve the quality and accuracy of safety data processing. This improved quality and accuracy, in addition to faster turnaround times, can result in almost 100% regulatory compliance. There are three main areas within the realm of safety that can be transformed through appropriate and effective use of technology:
Standardisation and automation of PV processes and safety data management
Integration of safety data by applying appropriate data and system interoperability standards, implementing best practices and technological models, ensures transparency and accessibility of safety information. Work flow management technology is one key example, which also allows targeted automation for enhanced end-to-end PV processes along the safety and risk management (SRM) continuum, ultimately leading to efficiency gains while maintaining quality and compliance. Some technology solutions are already available, including a cloud-based call centre and medical information solutions which, when coupled with a range of automation tools for downstream PV processing, can result in efficiencies along the SRM continuum.
Proactive PV and risk minimisation
Implementing data mining techniques is an important way to identify and predict emerging safety signals. Detecting the majority of safety signals is challenging and can require analysis across a number of data sets, which is time-consuming and labor intensive. Data mining technology enables the systematic eradication of data noise to determine which signals require further investigation.
Open and transparent data sharing
Companies are required to share their data with regulators, prescribers and patients, to protect patients and to build public trust and confidence. Work flow management technology can be applied to PV to identify and distribute information to stakeholders in line with a pre-defined set of rules. AE data can be automatically extracted and processed with advanced cognitive solutions, with real-time views of safety issues enabled by AI. This increases transparency with stakeholders, drives more trust and collaboration and, ultimately, results in fewer ADRs. For example, the FDA Sentinel system and the EMA EudraVigilance database and WEB-RADR are paving the way for this level of transparency. EudraVigilance governs the degree of access stakeholder groups have to ADR reports, and one of the deliverables from the WEB-RADR initiative is a mobile application which allows patients to directly report potential side effects.
The future of automated PV processes
The changing PV landscape is seeing regulatory authorities using more sophisticated tools to collect, characterise and evaluate data on AEs, so pharmaceutical organisations can implement successful PV programs and more efficiently manage the safety of drugs. An increasing population, more unique and highly specialised therapies for unmet medical needs, and rising number of pharmaceutical organisations are drivers for the technology revolution the industry is experiencing. Automation is vital if the costs and complexities of clinical trials are to be minimised, and the collaboration between stakeholders for real-time decision-making is to be improved.
Robotic automation, cognitive computing and AI technologies will allow organisations to reduce the manual effort and cost of safety case processing, meaning resources can be redirected to the proactive identification, evaluation and minimisation of risks. Cloud-based solutions and big data analytics technology are helping companies achieve end-to-end automation across the SRM continuum, while adhering to regulatory guidelines.