The Role of Speech in Healthcare

According to the American Medical Association (AMA), there is a nationwide shortage of more than 168,000 nurses and other health care professionals. With 60 percent of emergency rooms working at or above capacity, it is essential that hospitals utilize their staff and time carefully, which is becoming more difficult to accomplish as more and more people are turning to the emergency room for care. CDC reports from 2002 estimate that approximately 110.2 million visits were made to emergency rooms in the United States - a 23 percent increase from the 90 million visits in 1992.

Rex Gives Patients and Healthcare Providers Something to Talk About

It is estimated that over 100,000 people die per year from improper use of medicine, which accounts for $93 billion in insurance claims. With more than nine million visually impaired people in the United States and over 90 million adults who have limited ability to read and understand prescriptions or medicine bottles, it is easy to see where a mistake can be made when it comes to reading medicine warnings and instructions. Keeping these statistics in mind, it was the combined work of one blind man, one pharmacist and one speech recognition company that led to the development of a talking pill bottle.

Randy Allnat, VIST coordinator at the Department of Veterans Affairs, was a counselor for visually and cognitively impaired veterans. He also happens to be blind and has been since birth. It was a late night incident that put Allnat in the hospital and inspired him to look for better methods of access to medicines for the visually impaired. Allnat had used a method common among the visually impaired for labeling his medications, rubber bands. He placed two rubberbands on his heart medication, three rubberbands on another medication, etc. Unaware that a band had busted on one of the bottles with three bands, Allnat got up in the middle of the night and took the wrong medication. He wound up in the emergency room where he was treated for an increased heart rate.

This medical mishap was just the incentive that Allnat  needed to begin the search for a better labeling alternative. He and a friend, pharmacist Anthony Mariano, teamed with MedivoxRx to develop a talking pill bottle. MedivoxRx was later acquisitioned by Wizzard.

The ensuing research and testing resulted in "Rex," a disposable talking pill bottle. The 30 dram bottle is approximately four inches tall and two inches in diameter. There is a raised ridge that runs down to the base where the button is located to make Rex talk. The bottle is refrigerator safe and does not damage easily when exposed to moisture. It even has three small feet on the bottom of the bottle to help prevent undue exposure to wet surfaces.

As Gene Franz, general manager of the solutions and channels group at Wizzard, explains, Rex uses text-to-speech technology that recognizes up to 30 languages. The text-tospeech feature works on software installed on the pharmacist's computer and is intercepted, converted to speech and loaded into the bottle's recorder when the pharmacist keys in the prescription label information.

Home recording is also available with the purchase of the Starter Kit, which includes the recorder, microphone and power cords. Recordings can be made in any language by the patient or a caregiver that explains instructions and warnings for either prescription or over-the-counter medications. The bottle can store a message that is a maximum of 32 seconds long. The message can be played a minimum of 300 times. The shorter the message playback, the more uses it can sustain. The button must be held down to play the message; it stops playing when the button is released.

The United States Army recently deployed Rex in Afghanistan to assist Coalition Forces with humanitarian aid efforts for villagers. Recordings are made in the local language, Pashto, and then given to the village elder to administer after the forces have left the area.

In a recent study conducted by the Center for Disease Control (CDC), 72 percent of Americans visit office-based settings for ambulatory care. In 2000, the number of Americans visiting office-based settings averaged 756.7 million. Medications were provided in three quarters of all visits with an average of 2.3 drugs per prescribed case. Even with these impressive numbers, hospital emergency departments have decreased by 15 percent throughout the U.S. The ever increasing call for more efficient methods of servicing and assisting patients has been answered, in part, by speech technologies.

Speech technology can aid institutions in a number of ways, including the use of virtual operators, operating servers, voice activation services, dictation/transcription services, data entry, clinical reports, prescription fulfillment, interpretation services, and assistive technology.

Basic functions necessary to the daily functions of a healthcare institution are made readily available to patients through the use of virtual operators. The virtual operator assists hospital receptionists and operators with the influx of daily calls by answering, transferring and even making calls around the clock. The virtual operatoris a voice-activated recognition system that can also be programmed to place internal and outbound calls by simply speaking the name of an individual, department, hospital or supplier. The database stores the programmed numbers, effectively eliminating the need for internal phone directories. Most of the routine call answering is directed away from the switchboard staff, freeing them up to work with more complicated situations and services. Similarly, the operating server is assisting healthcare providers with everyday tools that they and their clients access.

The operating server is available 24/7 and offers more specific tools for healthcare clients and institutions. The speaker verification function of the operating server allows secure access via voice biometrics for clients to obtain general hospital information, get lab results, fill prescriptions, access personal information, and hear product overviews/information on available health programs. The operating server gives healthcare institutions the ability to retrieve patient medical histories, generate appointment reminders for patients, access work schedules, retrieve personal information and view product overviews. The operating server works less as a directory/switchboard and more as a resource tool for clients and institutions alike, which in turn reduces the number of daily interruptions for nurses and doctors, and speeds the turnaround of information to the patient.

As healthcare providers move away from the traditional means of recording and tracking patient records, they are also becoming more mobile. The stationary desktop systems are being replaced with wireless computers and handheld devices that allow physicians to access and manage patient records remotely. Voice activation of these wireless services permits the physicians to access the hospital's information system. The mobility and hands-free approaches to medical tracking stand out in efforts to streamline dictation, clinical reports and data entry.

Speech-to-text recognition creates a hands-free and eyes-free environment for physicians, automating the transcription process and reducing the time it would take to dictate, listen and type the information to a two-step process of dictation and editing only. Clinical reports, such as discharge and radiological reports, are completed by the physician in a timely manner without stopping to type or play with buttons. This can be especially helpful for hospital employees with learning disabilities and physically debilitating problems, for example: dyslexia - a person can speak their ideas clearly, but has difficulty converting those concepts to print; and arthritis or carpal tunnel syndrome - problems that limit the use of hands and/or the upper body. Speech recognition permits the user to minimize use of the keyboard by speaking the text into the software.

Even more impressively, speech can assist with both the placement and error avoidance of a drug order. Using the spoken order, pharmacists are not required to decipher messy or rushed handwriting. This can be very important when filling a patient's prescription as mistakes in this step can lead to serious or fatal reactions. The physician can also confirm his/her order with the computer before sending it to the pharmacy by having the computer repeat the drug order back in a synthesized voice. Keeping the diagnosis and prescriptions straight can be further complicated when language barriers and physical communication impairments are thrown in the mix.

Many hospitals, particularly in metropolitan areas, run into language barriers when working with patients from various ethnic and geographical backgrounds. Calling in interpreters to assist with translation can be difficult, time consuming and costly. It is often difficult to predict how long it will take to reach a translator and get them into the hospital area. Once the translator arrives, many other difficulties may arise - errors in interpretation can be a serious factor in providing proper healthcare to patients. Software for language translation can offer a readily-available solution with an array of languages and dialects programmed into the system. Even the methods for selecting the required language can be speech-enabled menus that allow the healthcare provider to speak the name of the language rather than manually searching or scrolling through each language for the appropriate one.

Communication barriers can also develop in situations involving patients who are deaf, hard of hearing, blind, or have low vision. The CDC recorded 30.8 million non-institutionalized adults in America who suffer from hearing loss. Efforts have been made to open the communication gap between healthcare givers and patients who are hearing impaired or deaf. Using software that allows the healthcare provider to speak into a microphone system which enables speech to be converted into a number of formats including sign language, patients can view the translated communication on a computer screen.

Another format involves converting the care giver's speech into text that can be read by the patient on the screen. This saves valuable time in emergency situations where a sign language interpreter would usually be required before the patient could receive treatment.

Outside of the hospital setting, communicating dosage instructions to those who are hearing impaired is rarely as problematic as to the visually impaired community. According to the National Eye Institute's Archives of Ophthalmology, blindness and low vision affect 3.3 million Americans aged 40 and over - the equivalent of one in 28 people. This number is projected to reach 5.5 million by the year 2020. Progress has been made in speech recognition technology to prevent misuse of prescriptions by visually-impaired patients who can not see well enough to read the labels on pill bottles. A pill bottle with text-tospeech technology converts the pharmacy instructions to a synthesized voice that is read to the patient at the touch of a button, preventing any confusion of medications or doses.

Other patients with disabilities can benefit from the uses of speech technology. Among them are quadriplegics, individuals with autism and with Parkinson's disease. Researchers, developers, healthcare providers and educators are all seeking better ways to make life a little smoother and learning a little more successful for people with disabilities. Some of the more recent developments and concepts have involved voice-operated wheelchairs with digital signal processing technology and beds with environmental controls for quadriplegics. Even household devices that are controlled by voice, such as the lights, television, computers and telephone, can be highly effectual in giving people with these types of limitations more freedom and individuality in their everyday lives.

As for those suffering from chronic illnesses such as Idiopathic Parkinson disease (IPD), aphasia and autism, there are therapies available to these people without necessitating a constant one-on-one interaction with another human. Computer programs and animations containing speech recognition software recognize and interact with the patients to provide the intense and repetitive instruction and practice needed to help them advance. The Center for Spoken Language Research explains that children with autism have cognitive, memory and attention constraints, which often necessitate over-learning information for it to be automatic and permanent. While many teachers are restrained by time andresources, the speech recognition software will repeat the information as many times as needed to help the patient respond correctly. There are no time restraints on the trainer; and it can be customized to use examples and tools familiar to the individual's environment. Victims of Parkinson's disease have similar options with programs that interact through speech and its correspondent facial expressions to help patients improve and control their own speech. Aphasia patients look at a picture on a computer monitor and receive feedback from the computer about the sentences that they have used to describe the picture. To help those suffering from Parkinson's disease, the software records the person's voice as they repeat sounds queued by the computer and then responds and interacts with the patient to help them train their voices to maintain strength and loudness. It is possible to fight the effects of diseases such as these with the help of assistive technologies backed by speech.

It is for all of the above reasons, as well as the reduction of emergency departments, low staffing situations and rise in hospital visits across the United States, that make it imperative that the basic functions and communications of healthcare providers be as simple and straightforward as possible. Speech technology developers have recognized the potential that speech has to offer the healthcare industry and have realized this potential through the development of software that enables healthcare institutions to maximize valuable staff time and skills. Although transcriptionists and care providers must still spend time making corrections and approving the output, the time and money saved using the speech recognition system versus manual input is significant.

Stephanie Owens is the associate editor for Speech Technology Magazine. She can be reached at stephanie@amcommpublications.com.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues