38-39 - AE reporting turnarounds present a multilingual challenge

Pharmacovigilance has never been a bigger undertaking than it is today. During the COVID-19 pandemic, workloads have soared and despite self-isolation and social distancing, those already vulnerable have become more so. Patients who are taking drugs long-term due to pre-existing conditions tend to be at higher risk for COVID-19 meaning that some risk factors have multiplied and clinicians have had to be extra vigilant.

Simultaneously, the global market has been flooded with new or updated products as regulators worldwide have worked alongside pharma companies to speed through medicines and vaccines to combat the pandemic.

While the entire health ecosystem is clear that no corners have been cut, numerous unknowns have had to be tracked meticulously as new virus strains and new pharma formulations have converged around the world. Usually, new drugs or vaccines are rolled out in a carefully controlled manner, region by region: global roll-outs on the scale witnessed over the past 12-18 months are unprecedented.
To keep on top of the risks, regulatory authorities have remained insistent that adverse event (AE) reports must be filed within 24 hours to keep patients safe and prevent products having to be pulled from the market later. But since notifications of adverse effects come in via a whole host of sources, and in the native language of the patient or healthcare provider, this has knock-on effects. Within the single-day reporting window, pharma companies must convert details of all of these incoming AE events into the common business language of English so they can be submitted to the regulators.

It’s not uncommon to process many hundreds of thousands of AE reports annually but few organisations can afford their own bank of dedicated translators, so many are having to lean heavily on external resources. In some cases they are spending a fortune, contracting large clinical research organisations (CROs) to do the work – even though this may not be their core competence – or engaging linguists/translators on an ad hoc basis, thus losing the economies of scale of a retainer arrangement.

Given that translations ideally need to be turned around within a couple of hours to keep companies’ AE reporting on schedule, this is playing with fire.

The best hope that companies have of keeping on top of their global AE reporting is to co-ordinate it centrally. If in-country teams are left to their own devices to engage local support as needed, achieving economies of scale and consistent quality/speed becomes very difficult, as does the ability to aggregate data across what might be 150+ countries. But once companies have central control over everything that’s coming in, they have greater scope for process streamlining, automation and efficient turnaround.

Where pharma companies have sought help up to now, the approach has not tended to be well-co-ordinated. For instance, local drug safety offices might struggle on with their own manual processes, or enlist the help of a contractor or a local language agency to convert incoming reports to business English. This ad hoc approach fails to make best use of external resources and may not pick up on trends in emerging safety data, because these siloed resources do not have access to the bigger picture.

Aggregation and analysis of safety data as a means of tracking emerging trends should be a critical focus for pharma companies – especially now. As the pandemic evolves, the discovery of positive coincidences between existing medicines and COVID severity is hugely important – the case of a routine asthma drug (budesonide) having a beneficial effect on COVID sufferers’ symptoms is just one discovery made in recent months.

So what’s the best approach to centralising and streamlining AE reporting, with a view to staying ahead of regulatory deadlines and maximising the strategic insights from incoming safety signals?

Here are some practical steps that companies can take:

Agree a central capability and process. This can be established with the right support from a specialist managed service provider – one with not just appropriate language skills across all markets, but also the safety/PV experience to improve how the company handles incoming AE information
Get ahead with data quality and consistency. Where companies have relied on cleaning up data after the fact, and on pulling it into data lakes to perform analysis and extract meaningful insights, the ideal should be to achieve cleaner, more centralised and reliable data earlier in the AE data capture process
Take the portal approach for stronger and more consistent connectivity between information silos. One option, for instance, might be to build a portal through which translations of AE reports are submitted and managed by a central service, and which then pushes the content directly into a central drug safety database for ongoing analysis
Optimise economies of scale in translation by centralising resources. This could help with retaining on-demand experts in languages subject to high-volume demand. Having experts on standby can cut AE report translation costs down to as little as an eighth of buying in ad hoc support
Create a foundation for AI-based automation. The greater the consistency in the created record, the greater the scope for repetitive pattern recognition and standardised learning by machines, including trend identification
Think digital. As digital/online/mobile AE reporting increases, including voice-to-text capture, the scope for near real-time AI and machine learning-enabled trend capture is likely to grow. But only if information is captured and processed centrally, enabling a 360-degree view of global input. The more reliable the trend information that’s captured, the greater the potential for enhancing patient safety and making timely improvements to drug formulations
Use AI/ML technology to process content from published global medical safety literature and threads in public online forums. Together these form a large and growing element of pharmacovigilance, existing in all languages and character sets – content which all needs to be interpreted, cross-referenced and included in PV reports.

‘Webcrawler’ software may make light work of ‘reading’ and filtering this content, but key information still needs to be translated, examined and amalgamated: that’s labour-intensive work that can add up over the course of a year – a million words of such content annually is not uncommon for larger companies. The more the processing of this work can be centralised and managed consistently via a single specialist resource, the greater the scope for cost-efficiencies – and for breakthrough insights.