AL 721 Survey Results: Preliminary Report

Last month this writer mailed a questionnaire on AL 721 and other treatments to all 898 subscribers to AIDS Treatment News. We asked people who have used any form of AL 721 to report on their personal experiences with it, as well as with any other AIDS/ARC treatments which were important to them. We received 147 completed questionnaires, with information not only on AL 721 but also on AZT, vitamin C, homeopathic treatments, Chinese herbs, lysine, DNCB, zinc, diet, spiritual practices, and many other treatment approaches. We are now analyzing the information; this article is the first published report.

We believe that this study will add important new knowledge about the effectiveness of AL 721--as well as some
other experimental and alternative treatments, many of which have never received formal clinical investigation.


Goals and Design of the Study

We started with several questions. First, how many of the people who are using AL 721, etc. are getting good results?
Are some of the versions proving better than others? What other treatments are people combining with it? Which patients are most likely to benefit? What improvements do they report-- improvements others could watch for to see if the treatment is working for them?

And aside from AL 721, this study asked which treatments of any kind are proving most valuable.

We chose to use open-ended questions, to let people answer in their own words instead of giving multiple-choice responses. Open-ended questions take longer to analyze, but they let people tell us what is happening even it doesn't match any of our preconceptions. We asked respondents to ignore our questions if necessary; several did so and wrote long letters instead.

How does one analyze such open-ended replies? Our basic approach was first to use certain questions or other
information provided by the respondents to divide the set of questionnaires into subsets. One breakdown, for example, would divide the questionnaires into those from people who believed AL 721 helped them, those who found no benefit, and those who were uncertain. Other breakdowns could be those who used egg- derived vs. soy-derived products, those who used AZT vs those who didn't, those who did or did not have KS, etc. After each division or subdivision, we cover each subset journalistically, reporting what those people have to say and how their experience differs from that of the others.

By analyzing the questionnaires this way we can use all the information people provided, whether or not they followed
the formats we suggested. (This preliminary report gives mainly tabulations, however, due to time constraints.)

This survey could not rely on any standardized treatments, protocols, laboratory tests, or data collection. So we
designed it as a study of peoples' beliefs about their treatments--admittedly subjective--rather than attempting to make it a scientific study of the treatments themselves. We asked people to tell us what they believed and to back up their beliefs with any evidence available, such as laboratory tests, symptoms, or how they feel overall.

Beliefs are not the same a proof; placebo effects and medical treatment fads illustrate errors the human mind does
make. Those who consider patients' beliefs irrelevant will find nothing of interest in this study. But we think that people's beliefs about what works for their health are worthy of respect, and are valuable guidelines in the absence of perfect knowledge. It is better to face the danger of error by educating ourselves about how these errors happen, than by refusing to listen in order to protect ourselves from being misled.

How do we answer the criticism that any good results about AL 721 might be false "artifacts", because people who think the treatment is helpful might be more eager to tell us about it, biasing the statistics?

First, we tried to minimize such biases. We carefully worded the questions in a neutral way, and emphasized that we
wanted to hear from everyone who used any form of AL 721, whether it worked or not. To increase the response rate, so
that we wouldn't hear only from the most committed, we included a self-addressed stamped envelope with the questionnaire, set an early deadline for its return, and later sent a reminder with a short extension of the deadline. Almost all of the responses arrived on the forms and in the envelopes we provided--a protection against any possibility of a malicious attempt to "pack" the study, and also against well-meaning efforts of enthusiasts to get other enthusiasts involved. We found no indication of any such problems.

Second, this study has generated many results, often unexpected and unpredictable, which do fit together into
coherent pictures or patterns. Attributing the whole outcome to errors and artifacts would require belabored, contorted
explanations.

Ultimately, however, this study is descriptive; it does not claim to show conclusive or scientific proof that any
treatment works. This survey cannot substitute for clinical trials. But it provides the clearest picture we could obtain
of what peoples' experience has been.


Early Results

Exhibit II below summarizes the results so far.

* How much success are people having with AL 721?
Of the 147 persons who returned the questionnaire, 110 had used AL 721 or a substitute for at least three weeks, a cut-off point which we chose before starting this analysis. Half of the 110 (50 percent) found it helpful, 15 percent found it not helpful, and 35 percent were uncertain. No one found the treatment harmful.

The actual picture is considerably brighter than these figures suggest. See other results and discussion below.

* Does the version matter?
Few Americans can obtain the "real" official AL 721; only three of the 110 reported above had done so. But various
substitutes have come into use. This study confirmed that only two of the substitutes had been widely available before August of 1987: the "home formula" made with PC-55, a commercially available soy lecithin concentrate, and the egg-lecithin lipids from the PWA Health Group, a non-profit buyers' club in New York. Sixty nine of the 110 (63%) had used the home formula, 33 (30%) used the Health Group version, and 8 (7%) used other versions, including the three who obtained the official one.

The version does matter. Of those who used the soy formula, only 43 percent reported it helpful, vs 22 percent not
helpful. Of those who used the egg version (from the PWA Health Group), 58 percent reported it helpful and only 9
percent not helpful. These percentages were computed from numbers given in Exhibit II, below.

These numbers are conservative because the questionnaires made it easy to get into the large "uncertain" category. The 33 percent uncertain about the egg formula (100 minus 58 minus 9) and 35 percent about the soy include several different groups of people. Some checked "uncertain" on the questionnaire, or expressed uncertainty in their own words. Others had improved but were using many treatments and could not know what benefits if any to attribute to AL 721. Others did not answer Question IV at all but left it blank without explanation. Others had been healthy throughout--such as those who were asymptomatic HIV positive--and therefore had no basis of comparison about whether any treatment worked.

If we consider only those who answered the question and had an opinion, then 67 percent of those using the soy version, and 86 percent using the egg, reported benefits. So the real fraction reporting benefit seems to be between 43 and 67 percent for the "home formula" soy version, and 58 to 86 percent for the egg version.


Future Reports

Analysis not yet completed includes what improvements, side effects, or other changes people found while using AL 721, etc., and which patients were most or least likely to benefit.

We will also analyze the dozens of other treatments reported--especially those volunteered in response to Question
III, "What treatments--of any sort--seem to have been most valuable to you?". For each treatment in common use, we can also compute the proportion of its users who listed it as "most valuable", to obtain an index of the value of a treatment independent of its prevalence or popularity. For this analysis we will use all the questionnaires, not only those reporting use of AL 721.

We will also look at particular groups of patients, such as those with KS, to see what treatments they reported as most
valuable.



Exhibit I: The Questionnaire

The following questionnaire was mailed to the 898 subscribers to AIDS Treatment News, on July 8. About two-
thirds of these subscribers have AIDS or ARC.

The text below is identical to the form which was mailed, except that blank spaces for replies have been deleted.

AL 721 Survey

AIDS Treatment News is conducting a survey of AL 721 (and substitute or related treatments such as the "home formula"). The results, to be published here and elsewhere, will provide much-needed information which may help people make better treatment decisions.
If you have used AL 721 or any substitute whether it helped you or not, or if you have cared for anyone who has, you
can participate by returning this survey before July 20.
Answer in your own words. Use the back of this sheet or other paper if necessary to continue. Don't hesitate to change or ignore our questions; just tell us what matters. We will read and consider everything you have to say.
***********************************************************
I. What version(s) of AL 721 (or substitutes) have you used?
What daily doses and schedule? For how many weeks? (You can identify the version by telling us where it came from. If you are using an unusual product, tell us whether it comes from egg or soy lecithin.)
II. What other treatments have you used during this time? How long have you been using each one?
III. What treatments--of any sort--seem to have been most valuable to you?
IV. Do you believe that AL 721 has been very helpful? somewhat helpful? no noticeable change? harmful? Are you very confident of this belief, somewhat confident, or uncertain?
V. What changes in symptoms, overall condition, or laboratory tests have you noticed while using AL 721?
VI. Diagnosis. Have you been diagnosed with AIDS? KS? ARC?
Lymphadenopathy? Low helper T-cells but no symptoms? HIV- positive but no symptoms? No medical diagnosis but suspect you may have been exposed? Or are you using AL 721 or substitute for a condition unrelated to AIDS? How long have you had each diagnosis?
VII. Any other thoughts or comments?
(optional) Include your name, and phone or address, if we may contact you if we have questions. OK to remain anonymous if you prefer.
Return this survey to: John S. James, P.O. Box 411256, San
Francisco, CA 94141.


Exhibit II: Preliminary Tabulations


147 (16%) of the 898 questionnaires were returned by the August 1 deadline.

Of these 147:
110 had used AL 721 or substitute for three weeks or more.
18 had used it, but didn't say how long
7 had used it less than three weeks
1 was sent by a physician reporting on four patients
11 had not used AL 721.

Of the 110 who had used AL 721 or substitute for 3 weeks or more: 55 (50%) found the treatment either somewhat
helpful, helpful, or very helpful; and they did not check "uncertain" or otherwise express uncertainty about this
estimate. 17 (15%) found the treatment not helpful. 38 (35%) either checked "uncertain", said it was too early
to tell or otherwise expressed uncertainty, or left the question blank. No one rated the treatment harmful.

Effect of different versions of AL 721, etc:
Only two versions, the "home formula" using the "PC-55"
soy lecithin concentrate (PC), and the all-egg version from
the PWA Health Group in New York (HG), were reported often
enough to give reliable information. The tabulation below
shows that the egg-lecithin version gave better results.

result PC (soy) HG (egg) other

HELPFUL 30 19 6
UNCERTAIN 24 11 2
NOT HELPFUL 15 3 0

This table shows that 43% of the users of the soy formula rated it helpful, 22% found it not helpful. But 58% of the
users of the egg formula found it helpful, and only 9% not helpful.

Those who started with one version and then switched to another were tabulated as using the later version, unless
it could be determined that they had used it less than three weeks.