But that's a nearly impossible approach when it comes to digital records. Quite simply, there isn't a market leader. "There is no dominant player yet," according to Tom Handler, a radiologist who is working as a healthcare market analyst for the Gartner Group, a consulting firm headquartered in Stamford, Conn.

That's partly because so few physicians actually have already invested in the systems. Handler estimates that fewer than 3 percent of physicians have an EMR. Even a leading vendor like A4 has only 250 practices using its system, says Bond.

Making matters more confusing, the healthcare technology market is consolidating. According to Vincent Hudson, president of market research firm Jewson Enterprises, there were 1,443 practice management and EMR vendors in 1999. Now, there are only 869. Today's leader may be tomorrow's acquisition or bankruptcy.

Instead of trying to pick a popular winner at this early stage of the contest, define – clearly – what your practice needs from an EMR, then look for the product that best meets those needs.

Check speed of entry

If you still have trouble narrowing the field, look for the fastest charting program. "Every system out there takes longer to create a chart electronically than to dictate it," warns John Cadwell, MD, director of engineering for Cadwell Industries, creators of the QuickMed EMR. Handler agrees, "The systems all take longer to use than pen and paper."

That's a problem because, for physicians, time is money. An extra 15 minutes a day spent charting is one less patient and one less charge a day. "If it takes you more than a minute longer to dictate a note, you are going to lose money," Voelker warns, adding that vendor reps will "tell you they know a doc who can [use their product to] do 30 patients a day, no problem. Well, my response is that there are also people who can paint with their feet." The only way to know for sure is to test it.

Because some programs take longer to learn than others, it's difficult to check the charting speed by timing yourself. You may get faster as you become more familiar with the product. But you can count clicks, Cadwell advises. "Count how many times you need to click or use the keyboard to create one particular note. Forty clicks will take less time than 80."

Ask the right questions

Another helpful technique is to ask colleagues about the systems they have chosen. Most vendors will happily provide the names of clients for you to chat with. However you get your contacts, here are some cagey questions to ask:

• Is the person you talk with using the same version and the same modules of the program that you plan to use – especially those that drew you to the product? If the client has the same modules, are they actually using them regularly and with success? Similarly, make sure the product a vendor shows you in demo is the same one they would actually install. Some are showing product versions that are not yet being produced for sale, Bond cautions.

• How many physicians in your contact's practice are using the software? Of those, how many use it at the point of care instead of batching chart work for the end of the day? Both questions will elucidate how user-friendly the program really is.

• Ask potential vendors how many licenses they have sold in the past six months and, of those, how many are live? That indicates whether a company can deliver on time. Implementation for some companies takes as long as 200 days, if more than one or two physicians in a practice are using their product, Bond adds.

• Is the vendor financially stable? The medical software industry is shifting quickly. Make sure anyone you contract with today will be around tomorrow. "A company's finances are often under-considered," says Jerry Shultz, vice president of sales and marketing for Tustin, Calif.-based NextGen. Physicians shouldn't draw conclusions about stability based on appearances. Lots of companies go public and spend their cash on big trade show booths and fancy brochures, leaving very little in the bank for longevity or product development.

Handler adds that practices should avoid buying software from local, entrepreneurial physicians who have created their own systems. Such software is unlikely to integrate well with practice management software or to be around for very long.