ow far can intelligent automation, including machine learning, go in transforming global content management and compliance for international companies with extensive content needs and tight timelines? The life sciences industry’s specialist requirements are reaching a peak, and neural machine learning now offers a practical solution.

With business now more global than ever, and technology making it possible to bridge distance and time zones to a degree that was once unthinkable, it follows that approaches to international information translation will also need to be transformed so that market expansion plans are not hampered by language barriers.

As with so many digital developments today, consumer experience shines a light on what’s possible. Instant phrase translation and real-time conversations between people with different mother tongues, enabled by tools such as Google Translate, iTranslate and Waygo, have raised expectations of what should be possible in a business context. This is especially the case given how much budget and time is allocated to maintaining international consistency and messaging, and containing the risk of meaning being lost or skewed as content is adapted for different markets.

There are few industries where the need for transformation is more acute than life sciences. In response to intense global competition, with great pressure on speed to market, new drug development is happening at an accelerating pace — 2018 saw a record number of drugs approved by the FDA in the United States. At the same time, the industry is subject to increasingly rigorous and demanding regulatory standards; the scale of documentation now needed to bring products to market is immense. Medical device manufacturers now face similarly strict controls too, as governments act to improve patient safety in the wake of some high-profile public scares.

The risks of decentralized global content management

All aspects of the product life cycle, from development and clinical trials through to post-marketing compliance and safety vigilance, must be tracked and documented in very specific ways in every market, in line with local as well as regional and global marketing authorization and reporting standards. Noncompliance in content can result in delay to market, recalls, fines and most notably risk to patient safety, as well as associated reputational damage.

Traditionally, the translation-related activities in this area have been managed somewhere between local market affiliates and professional translation agencies or language service providers (LSPs), but almost always in a decentralized way, largely out of view and beyond the reach of corporate quality control teams. Processes are labor-intensive, costly and inefficient in terms of speed to market, and there is considerable risk of inconsistency and compliance failure — especially as the international regulatory environment is so volatile. Requirements across every stage in the product life cycle are being enhanced and updated with regularity, introducing new work, new cost and new risk at every juncture if life sciences companies aren’t on top of things.

Yet moving to a centralized content management model is onerous too, akin to trying to turn around an oil tanker, in terms of the likely displacement and resistance involved. Trying to link previously unconnected systems so that they can talk to each other is a significant and expensive undertaking in its own right, and something companies can’t expect to achieve overnight, however valuable the exercise will ultimately prove to be.

Another approach has been to create regional capabilities — teams structured to look after the content needs of groups of countries, which share at least some of the same characteristics or requirements. But these plans place too much emphasis on people to handle all of the work and quality checks, incurring considerable expense and processing time.

Translation technology catches up with more specialist needs

In the meantime, translation technology has moved on at a phenomenal rate and continues to do so, offering companies a lifeline. Specifically, it offers a way to accelerate local content compliance and speed of processing clinical trial documentation, marketing authorizations and pharmacovigilance activities with a reduction in the time, cost and risk compromises that have had to be worked through previously.

While translation automation is nothing new, it is now being transformed at lightning speed by artificial intelligence and in particular neural machine translation (NMT) learning.

Until about two years ago, translation automation offered limited value in specialist fields such as life sciences, because machine vocabularies didn’t cater to industry terminology — the specific wording, phrases and codifications mandated by regulators for describing products, ingredients, processes and outcomes. It hasn’t helped that regulated terminology is continuously evolving and expanding. So, for the most part, life sciences companies and their LSPs have had to persist with heavily human-oriented translation cycles, slowing down market access and swelling their costs as regulatory demands increase.

In 2016/2017 this all began to change with the coming-of-age of NMT and the ability of automated systems to recognize, learn and adapt to new vocabularies at high speed. The result was translation accuracy and quality that can make a meaningful difference to life sciences workloads, as well as associated time and cost metrics. Already, today, NMT technology is at such a point that the overwhelming majority of international life sciences organizations are planning for its usage, and are putting pressure on their translation agencies to adopt it.

In clinical trials, where the timelines for producing localized content are extremely tight, NMT offers a very real solution and is already having an impact today. Notifications of adverse events, for example, can now be translated and understood almost immediately. Compare that to taking days or more depending on the volume to process. The potential for saving time is tremendous and can be a game changer in the clinical trials process where every day saved in the timeline counts.

Documents that must be flawless in their translation accuracy, like patient informed consent forms, will still require human oversight. However, even for these, potential savings range from 30-50%. This is because initial rounds of translation can be automated, producing high enough quality that leaves just the final honing and checks to human editors or compliance and quality control teams.

The benefits become more pronounced as volumes of content rise and where language pairing is favorable (English to Spanish being more common than English to Malay, for example).

Ability to train the neural machine engines on custom vocabularies pushes the potential for higher quality and therefore savings in time and effort on the part of the human input even further.

Vigilance demands a clearer big picture

The scope for combining NMT with regional and eventually more centralized content processes is considerable. For instance, as pharmaceutical and medical device companies look to take a more holistic and efficient approach to monitoring market authorization requirements and compliance, they may opt to translate everything into English, or standardize in another corporate language, for the purpose of corporate visibility and quality control checks.

Such an approach offers those with overall responsibility the chance to verify what the equivalent document says for each market. Alternatively, in the case of pharmacovigilance, central responsible teams are able to collectively view all real-world feedback/adverse event reporting about a product from across geographical boundaries, enabling speedier and more precise decision-making, with a positive impact on risk control.

Although companies can’t (yet) rely entirely on machines alone to pick up everything, neural networks are already having an impact on data mining by quickly learning the signs to look out for. As a result, these systems’ accuracy in flagging meaningful events from the vast data depths and market noise can quickly reach a level superior to anything that could be achieved by people alone. Human translators still have a role to play, though — for instance, in validating whether red flags in translated feedback warrant further exploration and action.

Life sciences companies are already making good headway with intelligent automation of data mining in pharmacovigilance, for instance for combing the published literature and public patient forums for red flags. For example, mentions of potential adverse effects relating to their products in the market. So their desire to extend the same capabilities to cross-market content translation is logical and rapidly becoming established.

As they have already seen, the cost, time and risk-management benefits are potentially very impressive, especially for high-volume translation services for which raw machine translation output is adequate. The processing cost per word through a neural machine engine is nominal compared to translation costs by a human. For the more exact needs of standard publishable content, the scope for savings is still substantial because of the time and labor saved in initial rounds of translation.

Keeping automation options open

When translation needs are highly specialized, companies may prefer to invest in custom-trained engines that will utilize the company’s preferred terminology and will use more precise vocabulary and style for the industry.

How does a life sciences company decide on whether the investment is warranted? An LSP partner can help navigate the fast-changing intelligent translation automation technologies. Machine translation systems and processes are not one-size-fits-all and they are constantly evolving.

Within a system-neutral approach, an LSP can be the gatekeeper for all options and the latest best systems. By keeping their finger on the pulse of the evolving technology, they are best positioned to not only manage the automated translation process for their life sciences clients, but they are uniquely qualified to provide guidance on the most suitable systems and processes to maximize the output quality and ROI.

In fact, in life sciences, as the need for efficient management of international content rises and global compliance demands become an increasingly onerous issue, companies are looking less for just translation suppliers and more for complete solutions/process partners — service providers that can turn projects around quickly, efficiently using the best route to accuracy, quality and ROI for the workload at hand.

As to where this could all end, machine translation technology is improving at such a fast pace that we can already expect translation automation solutions to be a standard part of commercial translation processes — even before 2019 is out. Demands for the capability are soaring already, so being able to offer more intelligent options, and report on associated performance metrics, is now imperative for any LSP.

As things stand, the conditions in life sciences are ripe for positive change. The volume of content that companies need to translate has exploded, timeframes to process these expanding workloads have shrunk and budgets are being squeezed, so technology-enabled efficiencies are a must. Intelligent translation automation solutions which bridge the gap between efficiencies and high quality have become a vital consideration as companies look to address these challenges — whether across clinical trial documentation, patient labeling or post-marketing pharmacovigilance activities.

Other industries are already further along with NMT and it’s true that the life sciences market has been behind the curve with AI until now. But necessity is the mother of innovation as well as invention, and the developments in the global and fast-changing world of pharmaceuticals and medical devices will be interesting for the wider world to watch.