Computer Repertories – A Comparison

by Chris Kurz, 1997. No warranty as to the correctness of information is given. Check with the respective company for up-to-date information.

[If you are impatient and want to cut to the chase right away, why not skip ahead to the summary? - Ed.]

So there I was.

One carelessly uttered sentence on the Internet Homeopathy Mailing List, and suddenly I saw the challenge clearly looming ahead. I couldn’t turn back without losing face, and ahead lay only the prospect of many hours of scrutinizing work. There really was only one honorable way out, and that was straight ahead.

Computer programs which help the homeopath find the right remedy have been around for some time. How far have they evolved, and how do they measure up to my stringent standards in terms of user friendliness? Is it time for everyone of us to rush out and plunk down the money for one of these programs? And if so, which one should you buy?

What I set out to do, was to attempt an unbiased, candid and honest comparison of homeopathic repertorization software available today. I limited the scope to three programs: CARA 2.61, RADAR 7.0b, and MacRepertory 5.2, which are, as most professional users will readily acknowledge, the top three in their league. All three versions are the latest ones available as of June 1997. RADAR's 7.0 was tested as a pre-release beta version, but the full release should be available any time now. Because I tested a pre-release version of RADAR, some of the things I objected to will likely be fixed in the released version. According to David Witko, maker of CARA, his program is nearing completion of a major update, for which he promises a significantly expanded set of features.

The electronic soil on which the programs were to be evaluated, was a PC equipped with a Pentium 200, 48 MB RAM, 1.2 GB hard drive, 6x CD ROM, and a HP 682c DeskJet printer (running merrily under Windows 95). If all this means nothing to you, don’t worry. Pretty much any half way decent computer sold today has the oomph to run these programs. Some slower, some faster, but run they will. Of the three, all but CARA are also available in a Macintosh version. Since I tested them on a PC, I cannot make any statements as to their performance on a Macintosh.

My opinion of each program is based on at least two months’ time of exclusive use in practice as well as comparative dry runs on selected test cases. When you base your decision on what you read in this article, bear in mind that what you read has been filtered through my own view of how things ought to be. Although I have tried hard to remain unbiased, there is no better way to find out what is best for you than trying it out for yourself. Whatever bias might have slipped into this report, it is certainly not motivated by any financial benefits to me.

After I installed all three programs, I went back to check how much each was using of my precious hard disk space. A full installation of CARA took 40.6 MB. RADAR, with Synthesis 6.0, gobbled up a whopping 217 MB, the lion share of this being multi media objects like little pictures and sound clips. A minimal installation, which leaves these on the CD-ROM, takes also around 40 MB. MacRepertory fit neatly into a slim 4.1 MB. The big difference comes from the fact that MacRepertory gives you the choice to install only the minimum needed on the hard drive and keep all the repertory and materia medica information on the CD-ROM. Opting for this, will slow down things somewhat, but I never found myself twiddling thumbs for more than a two or three seconds. Installing the Complete repertory and Vermeulen's keynotes to the hard disk in order to improve access time increased the storage space requirements of MacRepertory to 26.8 MB while speeding up access to the repertory considerably. Overall I found the speed with which the programs performed their tasks to be brisk, with no significant difference between them. That statement is based on my subjective observation; I didn't time them with a stopwatch.

Let me tell you right at the beginning who should not rush out and buy one of the programs. If you are a student of homeopathy in the beginning or intermediate phase of your training, this may not be for you. No amount of computing power and software will turn you into a better homeopath. In fact, investing into one of these programs too soon may hamper your education more than it will help it. The situation is like giving word processors to first graders: before you can use the power of word processing to any advantage, you must first learn how to write by hand. For a well trained homeopath, though, these programs are big time savers.

"So what do those programs do for me?", you ask.

They were written to assist the homeopath in analyzing a case and finding the simillimum. This process can be broken down into several steps:

1. Take the case.
2. Select rubrics from the repertory.
3. Analyze the rubric selection in terms of finding likely remedies for the case.
4. Confirm the remedies from the repertorization in the materia medica.
5. Prescribe a remedy and update your records of the case.

No computer software can really help you with the first step, but right from step 2 through 5, all three programs stand by your side and make your life easier to varying degrees. In fact, they really do more than that. Today's repertorization software also makes studying remedies and differential materia medica much easier.

But still, a computer repertory is nothing but an index to the materia medica, sorted by individual symptoms. There are few things a computer does better than searching databases according to certain criteria, so putting a computer to work on the problem of repertorization is the most natural thing to do. All of the programs perform well in this respect; they find rubrics quickly and obviate the tedious task of writing out remedies in each rubric by hand. The most important question, though, is: How easy is it for the user to get them to do what s/he wants. The consistency and quality of the user interface is probably the most decisive factor in the buying decision.

Repertory books and Materia Medicas

Most programs come with different repertories in electronic form. Depending on the size of your wallet, you start out with one or two and then can add others if you feel the urge. Bear in mind, though, that the list of available repertories for each program is subject to change and will likely grow longer in the near future.

In my experience, I was rarely limited by the choice of repertories available to me. In fact, in more than 90% of all cases I used either the Complete or Synthesis (in RADAR). In a few instances I used Boericke or a special Acute repertory (in MacRepertory). CARA as well as MacRepertory make Murphy's repertory available. Personally, I don't use it, but in testing I found that CARA's implementation was superior to MacRepertory's. In MacRepertory, the chapter headings for this repertory are the first letters of the chapter name. This makes it awkward, for example, to use the "Children" chapter, since two other sections share the same initial "C". CARA gives each chapter its own icon, and therefore improves clarity and accessibility.

As you well know, after narrowing down the remedy selection to a choice of a few, those need to be confirmed in the materia medica. In general, it is a good and comforting idea to have the materia medica within easy reach when repertorizing a case. Each program tries to fill that need and offers you some selection of well known materia medica texts.

The possibility to expand the repertory by adding remedies to existing rubrics or enter entirely new rubrics, exists in every program (except for the bare bones version of RADAR). CARA has you add new repertory additions to a special user area. RADAR and MacRepertory incorporate your changes directly into the main repertory database. In MacRepertory, if your additions become too numerous, the main index on which all searches are based gets out of synch and needs to be rebuilt by running a separate utility program not included on the CD or disks. It is, however, available from KHA, Inc., free for the asking. How quickly this becomes a concern I was not able to evaluate.

There seems to be a big difference in the number of books available for each program. This may be of concern to some people, but to me it made little difference. I never really use more than one repertory, except for the odd case, maybe. As far as repertories go, I used either the Complete or Synthesis. Despite some ideological differences setting those two apart, neither one limited me in my creativity or choice of rubrics. CARA is the only program allowing you to search in all installed repertories at the same time. RADAR and MacRepertory are multi lingual, in that they give you the option to view the repertory in two languages simultaneously (if you purchase the multi lingual version of the book). They also offer single language versions for several European languages.

It never ceases to amaze me, how much information will fit on a single CD ROM disk. Imagine having all your favorite materia medica texts, journal articles of old and recent issues, and then some, at your beckon. Search for any word, combination of words or phrases and find all instances within seconds. You'd never know that the computer was combing through the equivalent of many feet high stacks of books. All programs offer a varying amount of materia medica information as part of the repertorization program. To harness the full power of a dedicated materia medica CD ROM, however, one needs a dedicated materia medica search program. CARA's new version, CARA PRO, integrates Similia into the repertorization program. RADAR allies with Exlibris, and MacRepertory works seamlessly together with ReferenceWorks. A comparison among materia medica programs like these will be subject of another review in a future issue of Homeopathy Online.

Searching for rubrics

Once you decide on a repertory, the next task is to find the appropriate rubrics for your case. Without the help of a computer, you have to know your repertory well. True to history, many repertories are based on J. T. Kent's organization of his repertory and present chapters in the familiar head-to-toe sequence instead of in alphabetic order. There are advantages and disadvantages to each approach, and both sometimes have related symptoms scattered throughout the repertory showing up in places you would have never looked. The computer comes to your aid in that it lets you find rubrics based on words or combinations of words. The downside: if you start relying on this crutch too early, you will never really get to know your repertory; it will always remain a somewhat distant acquaintance to you.

There is more to the search capabilities of each program than I described here, and some of it will be explained later, in a different context. All three programs permit you to browse through the repertory much like you would by hand in your trusted old book. Their search capabilities, however, are where the power of the computer shines through. I found myself using the computer search about 50% of the time, the remaining times I just picked the rubric straight from the repertory. Being able to search for rubrics by specifying certain criteria also gave me the comfort of knowing that I really got all rubrics connected with a particular theme. In summary: CARA has the most limited search capabilities (which are, however, adequate most of the time), but it can look through all repertories and chapters simultaneously. RADAR allows you to do pretty much the same things as MacRepertory, but in a way that I found less intuitive to use. It was also particularly annoying to me that the synonym search was not functioning.

As you find the rubrics for your case, you collect them into clipboards. Each clipboard holds a set of rubrics which can be analyzed separately later on. This allows you to take a look at what the mental picture suggests, separately from the physical symptoms, for example. Simply put all the mental rubrics into one clipboard and all the physical ones into another. The uses of these handy clipboards are really manifold, and all three programs have them. CARA offers three clipboards, RADAR has ten, and MacRepertory gives you six. In my own experience, anything less than three is too limiting, but three work fine.

The search capabilities are a feature I tended to use frequently. Therefore, ease of use in this area was a main concern. The hands down winner in this respect, MacRepertory, makes browsing and searching the repertory an easy and quick task. Although search capabilities between RADAR and MacRepertory don't differ much in terms of raw capability, I found MacRepertory's implementation more intuitive to use. CARA offers an adequate set of search features, but is the most limited of the three programs. Its list of synonyms, however, made the most sense. And since it is in the form of a simple text file, it can be expanded and changed to suit your personal needs. RADAR's synonyms were non-existent (although that may be fixed by the time you read this), and MacRepertory routinely provides outlandish suggestions, which make me wonder if it really thought of all the mundane possibilities. Clipboards to collect rubrics for later analysis are a feature common to all programs.

Concepts in RADAR

There is, however, one area of RADAR, which deserves special mention. It doesn’t have a direct counterpart in the other programs, that’s why I am discussing it separately. RADAR allows the user to search for rubrics by specifying or searching for a concept. Two different collections of concepts are offered, one is the "RADAR concepts", and the other one is based on Fonseca and Hyltons’s Semiological Guide.

The RADAR concepts are essentially just the familiar allopathic diagnostic disease labels, grouped by organ system. You can browse them in a format very similar to the familiar repertory book layout. Double clicking on the icon following a listing will let you see a brief textual description or definition of the term. However, I still recommend getting a good medical dictionary, since the information presented here is very terse indeed. The icon gives you instant access to a list of specific rubrics describing the general complaint. Most of the time there are also a few remedy suggestions. So you will find under "Anxiety Neuroses-Plan of Action-Remedies": acon, ARS, calc, phos, puls, sulph. How and why these remedies have been singled out from among hundreds of other – equally likely – candidates, eludes me. As useful as the general information presented may be, the remedy suggestions are, at best, misleading. Whatever happened to individualization and prescribing on the basis of the totality of the patient?

Saving the best for last, let me tell you about Fonseca and Hyltons’s Semiological Guide. And a true guide it is. It leads you to a set of specific rubrics, starting with very general concepts becoming ever more specific. In essence, this is a powerful tool, which helps you translate the patient’s language and your own observations into repertory language. To illustrate, take the symptom of "fear of other people". Which rubrics might we consider here? The semiological guide leads us down a path of choices: Analogic group à in relation to others à Fear of people à intimidation, timidity à intimidated by strangers. At this point, the guide offers you a list of rubrics which describe this state. Double clicking on the icon zaps you to the corresponding rubric in the repertory. Of all additions to RADAR, I found this to be the most useful one, appealing to the student as well as the more experienced prescriber. The rubrics suggested by the guide include, of course, the more obvious choices but make also use of lateral thinking and creative interpretation of the repertory.

Case analysis

Here is where having a computer really saves you time. However, the speed with which the computer suggests a remedy to you bears an element of danger. It is deceptive, because one tends to forget that no amount of whiz bang will give you the right remedy from the wrong selection of rubrics. The principle of GIGO (garbage-in-garbage-out) applies here more then anywhere else. Before describing the analysis capabilities of each program, let me explain briefly what analysis strategies are.

Probably the most common way to analyze a case is to take into account the grade of the remedy and the number of rubrics it appears in to compute a final "score". One can modify this, by taking into account the "underlining" (i.e. importance of each rubric to the case) as well. Invariably, however, small remedies will not fare well with either of these strategies. They are underrepresented in our repertories and therefore are statistically less likely to show up in a repertorization, even though they may be the simillimum. To compensate for this, one could try to level the playing field by attaching more significance to a small remedy when it shows up in a rubric than to a polychrest. Among the factors to consider when repertorizing is also the size of each rubric. A large rubric containing 100 or more remedies has comparatively little differentiating power, whereas a carefully chosen small size rubric gives you the ability to narrow the field of remedies considerably. In many cases it may therefore be advantageous to weight smaller rubrics more than larger, more common ones. So you see that rubric size, remedy grade, symptom underlining, and the frequency of the remedy in the repertory all factor into the repertorization and final ranking of remedy choices. There are many more factors one could take into account to fine tune the process of case analysis even further. Consider, for example, remedy families. If in a case you find Phosphorus and Kalium salts well represented, then Kali-phos should be given special consideration. These and many more strategies can be applied automatically by the computer.

An important part of the case analysis is the ability to restrict your attention to only a certain class of remedies, e.g. the minerals, snake remedies, remedies belonging to the sycotic miasm, etc. Sometimes I found myself in a situation, where all I saw was polychrests. In this case, it is a great help to be able to simply tell the computer not to show any polychrests and exclude them from the analysis. The programs differ greatly in their ability to specify exactly on which remedies the analysis is to be performed.

Again, I found that CARA's feature set of analysis strategies and options is sufficient in many circumstances. What RADAR adds on top of it, is its Vithoulkas Expert System (VES) and much expanded capabilities in terms of selecting remedies by family. MacRepertory is, however, the easiest and most straightforward system to use in this respect. In terms of fine tuning and selecting different case analysis strategies, it offers more than the other two programs. Selecting remedies by family and restricting repertorization to certain families is a breeze. Even though RADAR gives me a similar level of control, it feels a bit rough around the edges. Counting the number of mouse clicks necessary to perform a certain task in RADAR, it seems always to be that much more difficult and less intuitive than in MacRepertory.

As to the expert systems. The one in CARA is simply a fixed analysis strategy, which has been honed on a large number of cases. The term "expert system" is, in my opinion, inappropriate. On the other hand, CARA performed well and suggested reasonable remedy choices using the "expert" analysis. Most of the time, I would look at a case using this setting as well as a simple "flat" repertorization. RADAR's VES is touted to be akin to having George Vithoulkas looking over your shoulder. I can't attest to this, since George has never looked over my shoulder, but I can see how it may be attractive to some users. Personally, I found that I preferred most of the intelligence to remain under my control and didn't use the VES frequently. This may to a big part be due to my not using underlining in the manner necessary for the VES to show off its best side.

MacRepertory undoubtedly gives you the most control over the repertorization strategy. Everything is customizable and adaptable to your needs. The sheer number of possibilities can be confusing to the user, but fortunately all but the most frequently used controls are well hidden from plain view. This, I felt, gave me the control I wanted when I needed it, without being right up in my face with every little feature.

Both RADAR and MacRepertory offer specialized analysis features which draw upon the experience and techniques of several well known homeopaths. I already mentioned RADAR's Vithoulkas Expert System, but there is also the Vakil and Herscu modules available. See further down for a brief description. MacRepertory extends its suite of analysis strategies by categorizing remedies into groups such as the five elements, or astrological planets, etc. More about this later.

Presentation of the repertorization

Most of the time I used the simple graph type common to all three programs. In terms of comparing different analysis strategies and finding the best approach to interpreting the repertorization, MacRepertory's full suite of graph styles is hard to beat. Much more important than that, however, is the ability to play with the case and see how that affects the repertorization. MacRepertory's graphs are linked "live" to the rubric clipboards. Any change you make there (e.g. underlining, or selecting certain rubrics, etc.) will be reflected immediately in the graph. To me, this immediacy promotes an intimate interaction with the case, which more than once showed me the right way to the simillimum. Although RADAR has, in principle, similar capabilities, the intimacy is lacking. You have to switch between different screens, which are several mouse clicks apart from each other to see the result of any changes in rubrics reflected in the analysis. CARA offers basic graphing features together with simple statistical tools. As bland as this may sound, in practice, few cases required much more than that.

Printing and exporting information

Many practitioners have the need to document a case for presentation at a study group, a journal article, or a seminar. If you are like me, you will want to write the article on your favorite word processor and "stick in" the graph from your analysis. A very crude way of doing this is taking a snap shot of the computer screen with the information you care about displayed. You have to trim the picture to size and cut way unwanted regions. This is a crude approach. Unfortunately that's your only option in RADAR. CARA and MacRepertory allow you simply to cut and paste between the program and your word processor or spread sheet. I was able to write up a case including rubric information and analysis graph quickly and hassle free in this manner.

Materia medica studies

Let's try a more complicated demonstration of how you could use the search capabilities to study remedies. Say, for example, you want to hunt for the "essence" of Magnesium. Is there something common to all magnesium salts? If so, what is it? There are about 10 magnesium salts in our materia medica. Any rubric which contains three or more should be considered as a possible element common to all magnesium remedies. To throw out large, nondescript symptoms, let's narrow our search down to include only rubrics of 50 remedies or less. Clicking on the button labeled "Remedies" in MacRepertory's search dialog allows you to select the Magnesias, whereas the pop up menu shown gives you a handle on the rubric size. You can even restrict your search to certain authors, if you like, or rubrics containing particular words. The sky is the limit. RADAR has the same capabilities, but you can't specify the author(s) as a search condition. Also, in the case of the Magnesias, I would have to enter each one of them separately, which makes this kind of family extraction unwieldy. In addition, I was not able to tell RADAR to find rubrics containing any 3 out of all 10 Magnesias. To perform a family analysis, therefore, is more difficult in RADAR. CARA does not support searches by families.

Incidentally, there are 96 rubrics in the Complete which contain at least 3 of the magnesium salts and have less than 50 remedies in total, 16 of them in the Mind chapter.

Patient charts and case management

In principle, you can type your case notes right into the computer in front of the patient. I know several people who do just that, although I don't do it personally (for one, I am not a good enough typist). You would thereby eliminate the need for a sheet of paper completely, and come a step closer to the paperless office. RADAR has just started to offer, as separate modules, an extensive patient information database system. I saw a brief demo at the IFH conference this May and was assured that it would fit seamlessly into the repertorization and analysis part of RADAR. Its capabilities include patient management, information tracking, follow up sessions, as well as charting trends of certain parameters. If it lives up to the promise, these modules would constitute the most detailed patient management system of all three programs – at an additional price, of course.

Special features in RADAR

With version 7.0, I could see a definite commitment of RADAR to multi media techniques. This shows in the 20 or so images of remedies which are part of their keynote description, the one lonesome video clip of nystagmus, and Vakil's cough module. The information as it stands now is still terribly incomplete, and was therefore of no real value to me. However, I realize that RADAR has just begun to move in the direction of multi media, and future versions will hopefully expand on the beginnings made in this version of the program. It is certainly a feature which has the potential to be of great use and one which distinguishes RADAR from both of its contestants.

Next to Vithoulkas and Vakil, Paul Herscu has lent his wisdom to RADAR. The Herscu module provides an interface to Herscu's case analysis approach. It would be too much to describe its features here, but I can only recommend to read Herscu's new book. Based on my personal style of case analysis, I found the Herscu module enlightening and easy to use. It has the potential of providing a new view on an old case, which hitherto proved resistant to any other strategy.

Special features in MacRepertory

One are where MacRepertory really shines is the analysis of a case based on natural families. With only two mouse clicks you see your case projected onto the periodical table of elements and can easily locate the center of gravity and likely remedy relationships there. Or, maybe, you suspect a plant remedy and want to see which botanical families the repertorization favors. Nothing easier than that. You can view the botanical families in hierarchical order, while the darker shades of green show you where the repertorization falls. I really enjoyed playing with this part of the program, not the least because the pictures are well done and fun to look at. Besides chemical and botanical relationships, there are several others you can explore in a similar fashion (e.g., zoological or miasmatic).

CARA PRO on the horizon

CARA PRO comes with a full fledged materia medica search engine already built in. This search engine, a program named Similia, was previously available only separately. Now it has become an integral part of CARA PRO and offers many classical as well as contemporary materia medica texts. As the second player in the multi media field, after RADAR, CARA PRO supports the display of color remedy images and playing of audio and video clips. David Witko claims that full color photographs of the remedies are included, but how many and of which quality remains to be seen. There will also be spoken text on some remedies by a few of the world's leading homeopaths.

Apart from that, CARA PRO will also offer a new repertory search method, which is by "theme". This sounds reminiscent of RADAR's semiological concepts. The appearance of rubric presentations has been improved, so that CARA PRO on screen looks like your familiar repertory book. In addition, it appears that several previous limitations of the program have been removed and enhanced.

CARA PRO will be available on CD ROM as well as floppy disks and be fully supportive of Windows 95. Later, a version for the Macintosh is also on the schedule. If you can't wait, you can wet your mouth with screen shots of CARA PRO on Miccant's web site: http://www.miccant.co.uk/preview.htm.

The gist of it

Pricing information is subject to change, and student or group discounts are sometimes available. Prices may also differ in other countries. Therefore, I strongly urge you to contact the companies for a local sales representative and current pricing. Be careful when comparing the prices from the table above, since they are really apples and oranges. What they tell you is how little (or how much) you can get the program for. Also, be sure you know what you are getting. The base versions may give you less than you need in practice. I, personally, would want an extensive repertory (Kent, in my opinion, is not enough for a professional practice today) and access to modern materia medica and keynote information. What you need may not be the fully loaded version, but actually something in between the bare bones and the deluxe. Don't be shy, call the company and ask. In most cases you can get what you want without having to buy into something you don't want.

All three programs are also offered in a "Lite" version, which, essentially, is a stripped version of the program with only Kent's repertory. Prices run in the neighborhood of $500, and trade-up offers to the full versions exist for a limited time after purchase. However, think twice before going this way. If you really need a program, chances are you won't be satisfied with the feather weight. If you can't afford to shell out the money for the full version, don't go and buy the light one; save some more and go the whole nine yards.

The software field, like no other, is developing at such a rapid pace, that features only dreamed of yesterday are already being implemented today. Areas where you should expect some major advances in the near future are multi-media and repertory searches by themes.

Homeopathy has seen significant new development in the recent past. For example, we are now starting to build a bridge between the remedy source and our remedy images. Jan Scholten, with his latest book Homeoathy and the Elements has been a leader in this area, but others have worked along similar lines in the area of botany and zoology. It is easy to give a computer the capability to incorporate this into the repertorization and case analysis. The harder part is up to us, though. As homeopaths we have to educate ourselves and become familiar with these new tools. We have to learn how to wield the electronic power properly. Therefore I see the biggest potential of computers in that their raw computing power will stimulate growth in the community and eventually make us all better homeopahts.