Although Google Translate was less than 100 percent accurate in translating English emergency discharge instructions for patients who speak Spanish and Chinese, researchers are cautiously supporting its use, a study published in JAMA Internal Medicine found.
The researchers, from the University of California San Francisco, analyzed 100 sets of emergency discharge instructions translated by Google's new machine learning algorithm, released in 2017. They found the algorithm was 92 percent accurate for Spanish and 81 percent accurate for Chinese.
Only a small percentage of the inaccurate translations — 2 percent in Spanish and 8 percent in Chinese — had the potential to cause clinically significant harm, the study found.
The translation errors were mostly grammatical or typographical in the original written English instructions, and patients who could read English would have been able to understand correctly, the study concluded.
"Google Translate is more accurate than a lot of clinicians believe, and I think it's definitely more useful than not providing anything at all," said first study author Elaine Khoong, MD. "We cautiously support its use."
The algorithm faced difficulty when physicians used colloquial terms such as "skip a meal," a phrase that Google translated into Chinese as "jump over" a meal.
A troubling mistranslation happened when a physician told a patient to "hold the kidney medicine," meaning to stop taking it. The algorithm translated that into Spanish as, "keep the medication" and into Chinese as "keep taking" the medication. The researchers called these mistakes "life-threatening."
Machine translation errors also were associated with medical jargon and long, complicated sentences.
The researchers said Google Translate is best used in combination with human interpreters who can interpret the physician's verbal instructions for the patient, even over the phone or via video, while the physician shows the patient the written translation.