Voice-recognition technology developers are introducing a slew of products aimed at getting doctors to document patient information more immediately and thoroughly. The technology has advanced far enough, its proponents say, that it can now do more than just passively receive doctors' input.

Nuance, the maker of Dragon voice-recognition software, is marketing a new product that is embedded into a hospital's electronic medical record system. As doctors input a patient's information via voice, it can highlight and validate medical facts, spot inconsistencies and ask follow-up questions, says Nick Van Terheyden of Nuance. They can edit to ensure accuracy.

M-Modal, a Nuance competitor, is working on a similar product that allows physicians to record information in real time on mobile devices as they consult with patients. "Physicians prefer to narrate and dictate. They don't want to point and click," says Juergen Fritsch, co-founder of M-Modal.

The companies' push into health care, where accuracy is of utmost importance, is enabled by constant improvement in voice technology. Apple's introduction of Siri on iPhone has also raised the public's awareness of its potential. The federal government's call for hospitals to install electronic medical record systems has also stirred the industry's curiosity about the technology, says Mary Griskewicz of the Healthcare Information and Management Systems Society.

Doctors have used dictation for years as they compile patient care history and post-care summaries, orders and prescriptions. Many speak into recording devices and pay transcribers to type the notes. In recent years, they've moved on to software that transcribes directly as they dictate.

But Nuance's technology -- in testing at the University of Pittsburgh Medical Center and a few other hospitals -- takes the process a step further, employing a mixture of data mining and more "intelligence" software behind the content. For example, a doctor who is writing a note about a heart failure, but fails to register the level of acuity, would be prompted to give more details on whether it was diastolic or systolic, says Rasu Shrestha of UPMC. "A physician may know what to say but not always how to say it," he says. "The idea is to get actionable data while you're interacting with the patient."

The University of California-Irvine Medical Center is testing M-Modal's software that enables doctors to use voice to locate and dictate information to files in its electronic record system. The hospital will use a desktop version when it launches in October, but plans to deploy it on iPads in the next generation, says Jim Murry, the hospital's CIO.