Defining Your Needs Is a Good Place to Start

A colleague recently told me that he is actively treating more than 250 patients at three separate locations. "Do you think I need a computer?" he asked. That is a question many psychiatrists are asking as they see more patients and do an increasing amount of paperwork to maintain the same income they earned with far fewer patients just five years ago.

To answer my colleague I had to rephrase his question. It should have been, "What type of information do I want to manage in my practice to make me a more effective and efficient physician?" Computers are simply tools for the collection, manipulation and storage of data. In our minds, these tasks should always have priority over an enchantment with the tool used to accomplish them. Once we have identified the relevant information we can legitimately ask how any tool-pen, paper, index card or computer-can help accomplish our task of recording, analyzing and managing that data.

Psychiatrists have managed clinical information with pen and paper for decades. Why should they change now? To answer that question we have to examine the recent history of psychiatric practice. Fifteen or 20 years ago the average psychiatrist saw 30 to 40 patients each week for an hour of psychotherapy in his office. A modest proportion of those patients received medication. Some had insurance for psychiatric care, but many paid for their therapy out-of-pocket. Treatment lasted many months so it was relatively easy for the practitioner to remember important clinical details about each patient. Beyond direct clinical care, the amount of time and effort that the typical psychiatrist devoted to medical records, administrative paperwork, reports, scheduling, communication and billing was relatively small.

The situation is very different today, as my friend's practice suggests. As contemporary psychiatrists we treat a larger number of sicker patients in far fewer sessions than our colleagues of an earlier era. A much higher proportion of these patients require medication. This fact, together with the reimbursement restrictions introduced by managed care, has changed the nature of our practice by dramatically increasing the proportion of event-based medication reviews to traditional time-based psychotherapy. As a result, two or three patients are often scheduled in the time that was previously reserved for one. As this shift has taken place the growing number of specific psychiatric drugs, often used in combination with one another, has made medication treatment increasingly complicated. When we add to these changes an enormous increase in administrative paperwork, our practice begins to resemble the pace and style of data-intensive clinical specialties like internal medicine more than that of our psychotherapeutic predecessors.All these changes have had a significant impact on the type and amount of information that we need to manage in practice. This information can be loosely divided into four overlapping categories.

Patient data. Initial patient evaluation, progress notes, medications, laboratory results and the like. Practice administration. Billing, appointment scheduling, clinical reports, treatment plans. Communications. Messages and correspondence, patient education, informed consent for treatment retrieval of old medical records.

Continuing education. New diagnostic methods and treatments, complications and side effects of current treatments.

None of this information has to be managed by a computer. Many psychiatrists prefer to hand-write descriptive notes in medical charts rather than use a word processor, keep their schedules in a daily planning book instead of on a computer, hire a secretary to answer the telephone rather than using an answering machine, and search for the information they need in printed journals and books instead of using an on-line computer service. They remind me of a friend who is an artist. I once asked her why she didn't work with computer graphics. "I like the feel of the media, the pencil on the paper and the brush on the canvas. I don't get that feel with computer graphics even if I can mimic the same effects," she answered.

Similarly, another friend, who is a psychiatrist, collects pens. He enjoys handwriting his patient notes and is convinced that he can think better while holding a pen in his hand than while typing on a keyboard. This is obviously a matter of personal preference.

There are, however, growing problems with this traditional approach to information management as the number of patients we treat increases, the amount of time we spend with each patient decreases and medication treatment becomes far more frequent and complicated. Since this problem is most evident in the management of patient data, the remainder of this article will focus on that area.

Let's assume that you are actively treating 150 psychiatric patients at two separate locations. Some you see once a week, others once a month, and still others once every six months. Consider two hypothetical scenarios. In the first, a patient calls to say that he lost his prescription. His medical records are in your other office and you can't remember what you prescribed for him. In the second, an internist you don't know calls you at home early one evening and says that she has one of your patients in her office. The patient has been taking verapamil (Calan), a calcium channel blocker, for a week and is now flushed, edematous and complaining of headaches. When the internist originally prescribed the verapamil, your patient denied taking other medications. Today she told the doctor that you gave her a "drug for her nerves," but she can't remember its name. The internist asks what you prescribed.

Both of these problems can be solved using traditional methods of patient data management. It would be easy to retrieve the two prescriptions if you routinely carried an index card for each of your patients listing their medications and other important clinical information. Of course, this would require you to compulsively record the data in two places, the patient's medical record and the index card. Let's assume that you don't have index cards. You decide to answer the internist's question by driving to your office and retrieving the information from the patient's chart. As you look through the medical record you begin to wonder whether some of your other patients are also taking calcium channel blockers. Unfortunately, the only way to find out is to manually search through all of your patient charts. This is the traditional method of information management in psychiatry.

There is, however, another way that may better meet your clinical needs. Both problems can be solved with a clinical computer data base.

What is a computer data base? It's an organized collection of information stored on a computer's hard disk. The information is collected and managed using computer software that is generically called a data base system. There are many different data base systems on the market. Some of these run on personal laptop computers, or even hand-held computers, while others require a larger, institutional computer.

All data base systems have four basic components: a user interface or front end, a search engine, a report module and data tables. The database front end is the user's visual window into the system. It stipulates how information is displayed on the screen and entered into the database. The search engine is the component that retrieves information according to the user's specific criteria. The report module formats data to be printed on paper. The tables contain the actual information and are stored as files on the computer's hard disk. Each table holds an unlimited number of records. Each record contains one or more specific data slots called fields. There must be one field for each element of data you wish to collect: name, address, telephone number and so on.

A clinical computer data base is an organized collection of clinical information stored in a computer system. Its fields hold specific clinical information such as the patient's diagnosis, medication, progress notes and response to treatment, in addition to demographic data such as name, age, address, etc. Each patient is assigned one record. If you treat 200 patients there will be 200 records in your clinical data base. Each record is like a small medical chart. You can think of the collection of 200 records as electronic charts filed side by side in your computer just as the paper charts sit in your filing cabinet.

So, what's the advantage of a clinical computer data base versus a traditional medical record? There are several advantages. First, a computer data base is compact. Much of the information contained in a filing cabinet of paper medical charts can be stored on the hard drive of a small, portable, laptop computer. Second, you have complete control over how the information for each patient is entered into the data base, displayed on the screen, or printed on paper. Third, patient data can be displayed in graphs or other visual formats that may make it easier to discover important underlying patterns that can't easily be seen by looking at the numbers. Fourth, and perhaps most important, you can retrieve information from the data base using multiple search criteria. You might, for example, ask the system to find all your patients, older than 25 years, who are taking more than 20 mg of haloperidol (Haldol) per day. The search function becomes increasingly important the more patients you see.

How do you determine if you need a clinical computer data base? The key issue to consider in making this decision is the break-even point in effectiveness between traditional methods of clinical information management and a data base system. Since every new tool exacts a price in time and resources, you must decide whether a data base offers a significant enough bonus in efficiency and efficacy to justify the additional work it takes to learn the system and maintain it on a day-to-day basis.

We have all heard stories about hospitals that lost millions of dollars when their new sophisticated and expensive computer billing systems failed to work as promised. The potential problems are well-illustrated by an article in the April 21 issue of Time magazine that described the experience of a large software company after it installed a new e-mail system. The system completely changed the corporate culture. People stopped talking and began bombarding each other with hundreds of trivial e-mail messages a day. The result was a significant decrease in work efficiency, a good example of technology run amok. The problem was eventually solved when the company banned all e-mail messages during the five busiest working hours of the day. The message is clear; keep technology simple and keep it in its place or it will cause far more trouble than it's worth.

So where is the break-even point? How can you determine whether you need a clinical computer data base? Here are few guidelines that may help you decide. A computer data base can be helpful if you have more than 100 patients in active treatment, if the majority of your patients are taking medications, if you practice from multiple locations, if you see patients who also have nonphysician cotherapists, and if you see patients for several different managed care companies.Suppose your practice meets those criteria or you're just interested in trying a clinical data base. How do you begin? I would recommend the same guidelines that are used in prescribing psychiatric medication to the elderly. Start low and go slow. Buy an inexpensive laptop computer and a basic data base system like Alpha Five (Alpha Software), FileMaker (Claris Software) or Microsoft Works (Microsoft), a multi-feature program that is included with many new computers. If you're adventurous try one of the more powerful programs like Microsoft Access, Borland Paradox or Lotus Approach. If you have trouble using the system take a brief introductory class at your local computer store.

Start by creating a simple clinical data base. Don't use it in your practice until you are sure that it works smoothly. Although the experience may be frustrating at first, the eventual rewards can be substantial. Not only will you have a working clinical data base, but you will also gain a growing insight into a technology that will continue to have a profound influence on 21st century medicine.