Home

Cover Story

Table of Contents

E-Newsletter

Article Archive

Editorial Calendar

Datebook

Writers' Guidelines

Orgs/Links

Opinion Polls

Reprints

Search

For other articles and previous issues click here.

June 3, 2002

GOING DIGITAL WITHOUT GOING BROKE
By Melinda Decker

Contemplating making the switch to digital sound files? Concerned about the cost of such a venture? There may be a simple solution. Among medical transcriptionists, it’s fairly well agreed upon that making the switch from tapes to digital is a good move. Digital sound files can be handled and transported like most other computer files via disk or modem with ease of storage and duplication. Digital files can be sent to remote transcriptionists across the nation or, in some cases, offshore within seconds to remote transcriptionists. Unlike tapes that can wear out and become scratchy, digital files generally contain better quality voice recordings. Best of all, no driving is involved to pick up a digital recording. Only a few clicks of the mouse are necessary.

The problem, though, is in making the actual switch. There are many systems that have great features, including dictation capture, automated transfer to a designated medical transcriptionist (or pool of transcriptionists) then on to a quality checker, and subsequent automated delivery of the document back to the originating medical facility. With such a system, a physician might telephone the dictation server located at the transcription company site to make his or her recording, following the prompts given over the phone. Also, he or she might dictate into a system onsite at the medical facility, and the residing codec (digital to analog converter) within the system will convert the analog signal of his or her voice into a digital file format. The digital file would go into a queue along with other voice files, and the system would automatically send it to the transcription service via file transfer protocol (FTP) or other file transfer methods.

At the medical transcription service site, varying degrees of automation can be utilized to send the digital voice file to a particular transcriptionist. If not available, the medical transcription service organization (MTSO) might opt to send the voice file to an alternative transcriptionist (or even to a pool of transcriptionists), who accesses the files as they become available and downloads them into his or her own computer.

At the transcriptionist’s end, a player will convert the digital signal back to an analog one so that the transcriptionist can hear the recording from his or her computer, operating the process with a foot pedal. The transcriptionist types his or her work into a word processor, and the file is sent back to the MTSO or to a proofreader for a quality check. When a completed document is received back at the MTSO location, it is moved into a queue for automated delivery to its intended location.

Such a system is wonderful, but also rather expensive. For the small MTSO or the independent medical transcriptionist, it is prohibitively expensive, and another solution must be sought. Have you ever wondered, for example, how your teenager can zap those songs back and forth to friends over the Internet while you sit befuddled? Have you pondered how to send voice or text files in a manner that’s HIPAA-compliant? If your computer can play a Bach concerto, it ought to be able to play a doctor’s voice. The technology exists, and, in many cases, it’s right in your own PC or inexpensively available on the Web.

A quick review of the basics of digital voice technology will demystify the world of voice transmission. Just as word processing programs create different file formats (eg, .doc and .wpd), digital sound technology has different file formats as well. Some of the formats created are .wav, .vox, .dsp, and .dss. What’s the difference? It’s file size, or compression ratio. The Windows audio file format is .wav. In digital systems, voice, which is an analog signal, is converted by a codec to a digital signal. Depending on the codec, that signal is compressed at varying rates into a smaller file size. The difference in the end is the compression factor (ie, size of the file and quality of the voice).

Higher compression results in lower quality. DSP TrueSpeech is one such codec, and ADPCM is another. The difference is in compression.

TrueSpeech files are approximately 1/15 to 1/8 the size of other .wav codecs and approximately 1/4 the size of the smallest ADPCMs. Sound quality is essentially unchanged for dictation and transcription purposes. This is done so that files can be stored and transferred more efficiently because .wav files are very large. The resultant voice file must be converted back to the analog signal to be heard by the transcriptionist.

By far, the simplest means of recording a voice file and having it available for transcriptionists is for a physician to dictate directly into a dictation server via telephone. If you have an extra PC lying around—and who doesn’t these days—you may be able to turn it into a dictation server. Voice files are generally large and require a great deal of storage, so a large amount of hard drive space (approximately 20 gigabytes) is warranted. A faster and newer system will help things to work even better. It can be done using inexpensive proprietary software that turns your PC into a dictation server.

One module allows you to specify and use any audio compression format (codec) that already exists on your computer. Windows comes with a wide variety of codecs, including DSP TrueSpeech, that offer greatly reduced file sizes. You may need to experiment with different compression formats to find the one that offers the right balance between file size and sound quality. Recording, replay, editing, and file management are controlled by pressing numbers on the telephone keypad.

Once you have chosen the compression format to use, the recorded voice file is saved in a proprietary audio format that requires the use of free downloadable player software in order to be played back.

The software can handle from one to 32 phone lines. In regard to the telephone ports, you will need separate incoming lines. However, oftentimes all lines are assigned to a single number. This is called a “hunt group” because, if someone calls in on the number and one or more lines are already in use, the phone network will “hunt” for another of the available lines. That way, you would only need one main number.

However, there’s certainly nothing to prevent each line from having its own number, if that is preferable.

Adding additional ports becomes a matter of adding hardware—in this case a modem or multiport telephony card. Most modern PCs can run more than one standard modem simultaneously. Again, RAM and processor speed will play an important role in how many different applications you can run simultaneously. As long as you have a free expansion slot in your computer, you can simply add another modem and the software will reconfigure accordingly. A port is the digital dictation term for a connection for an incoming phone line. A computer with one modem has one port; two modems would make two ports, and so on. Only one incoming call can be taken on a given port (modem) at once; in order to have more than one dictation coming in at the same time, you would need two or more ports.

For more than two ports, a multiport telephony card is recommended. This is basically a modem that can handle four or more phone lines at once. These are not cheap—starting at around $500 to $600—but significantly less expensive than buying an off-the-shelf digital dictation system.

Alternatively, dictation can be captured at the physician site by using handheld digital recorders or by dictating directly into the office PC.

Various recorders will create varying file formats, all of which will need to be transferred to the MTSO. If choosing e-mail for delivery, one must keep in mind that the large file size can slow down the process and some Internet Service Providers (ISPs) don’t appreciate the bottleneck on the e-mail server. Even the use of WinZip, preinstalled on many PCs, will not prevent a slow delivery. All transfers must also use at least 128-bit encryption (for now, until HIPAA guidelines suggest otherwise).

Some handheld digital recorders, especially the newer ones, offer additional file management functionality.

An alternative is available on the Internet on a per-send basis. Rather than sending files to an e-mail server, the software relies on the secure data transfer protocols used by Web browsers (HTTPS and SSL). Unlike e-mail, which comes to rest on multiple mail servers as it makes its way through the Internet, every delivery is sent as a secure, encrypted stream of data directly to the servers. Copies of files never exist on mail servers, nor do they remain on the software’s servers after delivery. HyperTerminal dials another computer’s modem for direct modem-to-modem connectivity and is preinstalled on some PCs.

Did you know that within Windows resides an FTP delivery function? Using a relatively simple programmer’s script, you can check any folder or group of folders on a PC, generate batches of files, and use FTP to send them to a specified Internet address. Using Windows scheduler, a “task” can be scheduled to regularly run the script, thereby enabling one to automatically screen folders and send files to a given location. FTP, as with all file transfers, should be encrypted for security.

If the idea of programming causes consternation, there is an inexpensive software that enables dictation directly into a PC via a microphone or cordless headset. The software allows files to then be automatically transmitted to the medical transcriptionist or MTSO via FTP, e-mail, or local computer network with no further intervention required on the client’s part.

At the MTSO location, where voice files are received, a system of delegating files to individual transcriptionists must be created. Again, e-mail can be used, but, as mentioned, it has inherent problems. It is possible to set up an ordinary PC for use as an FTP server using additional, inexpensive software that provides support for SSL connections between the server and a secure client. However, the software alone will not enable you to operate as an FTP server yet.

Here’s why: When you log on to your ISP, you are operating from what’s known as a dynamic IP address. Your personal IP address, therefore, is different every time you log on. Can you see why it would be difficult for anyone to send you files with a dynamically changing IP address? You need to obtain what’s known as a static IP address. An Internet search of “static IP” reveals a number of providers and solutions.

Running customized scripts again on an FTP server will automatically put files where they belong for transcription pickup. They can, however, be moved manually. Using an inexpensive client program, the transcriptionist can log on to the server, access the folder to which he or she is assigned, and download the voice files in order to begin work.

Alternatively, a file manager program will send the recordings automatically or manually to remote transcriptionists. Users send their recordings to specified directories on the MTSO’s server. The software then will regularly scan these directories and automatically (or manually) forward the files to the correct transcriptionist.

The dictation capture segment of the transcription cycle is one with several options, many of which already reside within one’s computer or are accessible via inexpensive download. For the faint at heart, when it comes to programming, there is bundled software that accomplishes the same thing and comes with a fair amount of technical support. The result is an alternative method of going digital without going broke, and a way for the local independent transcriptionist or small MTSO to meet the growing digital and electronic needs of the physicians they serve.

— Melinda Decker is CEO of MedicalTranscription.com, a Web site that brings medical transcriptionists together with the physician groups that hire them.

Subscribe to For the Record Magazine!