Effective communication is a core part of any encounter between a patient and the healthcare industry. Yet it’s often hard to achieve good communication, particularly when there’s a language barrier between patient and provider.
As technological solutions to resolve these language gaps become ever more sophisticated, there’s real potential to improve health outcomes thanks to better translation technology.
In a clinical setting, it’s not always easy to get the language support necessary to ensure the dialogue runs smoothly. That’s particularly true when patients seek urgent care and there’s no time to arrange for a human translator.
A shortage of translators working in the medical field doesn’t help the situation. There’s also the question of cost – although America’s Affordable Care Act specified that clinics should provide translation services, often these are provided at prohibitive cost via a pay-per-minute phone line.
There are multiple studies showing how patients presenting with limited language abilities often receive compromised care. It’s possible to improve outcomes by offering trained professional interpreters and bilingual health care providers but this isn’t always feasible.
Technology is starting to offer solutions that are increasingly accurate and, even more importantly, both practical and cost-effective to use in a clinical setting.
Mobile technology is at the core of this revolution in translation in a medical setting. Paramedics have used their own initiative to download translation apps such as SpeechTrans Ultimate in order to better serve patients. But new translation technology isn’t just confined to the smartphone.
Translation of written information
When you’re translating written medical information such as prescription labels and medication documentation, accuracy is everything.
When translations are completed in a formal environment, the translation team will follow strict processes to ensure the quality and reliability of their work. But what happens when materials are translated in a more ad hoc manner?
A study in an area of New York where Spanish is widely spoken as a primary language found that many pharmacies were providing translation services that were alarmingly flawed.
Although not all pharmacies provided translation services with those that did so used machine translation to give patients their prescriptions in a more familiar language. Around half of these translations were found to be seriously flawed. A third was incomplete, meaning that information was missing.
More worryingly, some of the pharmaceutical information was often mistranslated where a word on the English prescription had a Spanish language homonym. One example would be when a prescription advised that a drug needed to be taken once a day.
‘Once’ in Spanish means eleven, a translation error that led to a patient taking far more medication than was intended. The problem with machine translation is that it really needs a human translator to check the result – but it’s usually used when human translators aren’t available, such as in small pharmacies.
Can more intelligent machine translation services improve translation outcomes, and help avoid risks to patients? The rate of improvement in translation technology just in the 7 years since the New York study suggests that there’s every reason to believe the technology already exists to improve these on-the-spot translation problems.
Of course, this presupposes that businesses such as pharmacies will have access to advanced translation technology.
For small businesses, it’s important that access is affordable. Whilst pharmacies in cosmopolitan areas may see the value in providing translated materials to customers, those in more mono-lingual areas may feel less need to invest in this kind of technology to suit a smaller minority of customers.
There are alternatives to giving pharmacies access to machine translation. One would be for standard terms and phrases in a variety of languages to be collected into an accessible language bank where pharmacists can pull together the important information, such as ‘take twice a day with meals’ for customers.
If the resultant translations are pre-approved by a human translator, they are likely to be more robust compared to machine-translated prescriptions that have not be checked by a human before they are given to the patient.
Formal translation processes, with expert human medical translators and formal editing and review processes, are extremely important to the creation of reliable medical information. But informal and ad hoc translation is an important part of making sure medication is used effectively.
Improved performance and growing availability of translation solutions at the point of access to the healthcare mean more patients can get the advice they need in the language in which they need it.
That’s particularly important as language speakers become more dispersed around the world and language diversity increases at local level.
It’s thought around 25% of New Yorkers speak Spanish at home so it’s not an unrealistic prospect to be able to access healthcare in Spanish, either because it would make sense for a translator to be on hand in busy practices or because the healthcare worker is a Spanish speaker themselves.
But New York’s also home to many hundreds more languages, 800 if you believe The New Yorker.
For many of the more obscure languages, it may be more realistic to use machine translation to support patients who need help in their native language. Being able to access vital health information in a familiar language is highly important to successful health transactions between patient and provider.
Trying to achieve this in a way that’s cost-effective is an important part of the solution. Perhaps the most important contribution machine translation can make to healthcare is in providing a cost-effective solution that helps ensure patient safety.