Reply to Professor Ernst

I thank Prof Ernst for his valuable comments. This paper was originally published in 2005, it was reprinted as part of the Centenary issue of Homeopathy. This is clearly stated: ‘This article is a reprint of a previously published article. For citation purposes, please use the original publication details Homp 2005; 94: 75–80. DOI of original item: doi:10.1016/j.homp.2005.01.002. Received 3 August 2004; revised 11 January 2005; accepted 26 January 2005’. I was not involved in selecting the paper for reprint. I agree that in PubMed it appears as if it was a new paper, we have requested that this be clarified.

Taking Prof Ernst’s points as numbered in his letter:

1. If real, the findings reported are of ground-breaking importance and could save many lives. Did Frass et al. not feel a responsibility to conduct a more definitive, appropriately powered study during the 6 years since the first publication of their trial? Is such a study in progress? If yes, why re-publish this preliminary report?

With regard to responsibility to conduct a more definitive, appropriately powered study, I would have been willing to do so. However, I changed my position at the Vienna Medical University to the “Outpatient Unit Homeopathy in Malignant Diseases” 7 years ago and I am no longer working at the ICU. I want to take this opportunity to discuss the many problems a homeopath is facing when performing research: (1) High quality studies can only be performed at universities or university – affiliated institutions. To my knowledge, there are very few homeopaths working at such institutions fulltime. (2) In my experience, it is extremely difficult to obtain financial resources for studies of classical homeopathy. The reasons are diverse: the size of the homeopathic pharmaceutical industry is much smaller than conventional pharmaceutical industry. The homeopathic pharmaceutical industry is interested in the evaluation of so-called “complex” remedies, which contain a number of substances, which is not in the interest of classical homeopaths. Studies performed according to classical homeopathy using the extremely cheap and non-patentable remedies are not in the focus of the industry. I have applied for other homeopathic studies (e.g. for instance in Crohn’s Disease), which received Ethical approval; however, I was unable to obtain financial support despite applying to several different institutions. (3) There are few facilities designated for research in homeopathic at academic institutions. (4) A good climate facilitates performance of studies: however research in homeopathy often faces inappropriate, unscientific, negative and even hostile reactions which inhibit acceptance of research at academic institutions.

2. Can the authors confirm that the two study endpoints were the only ones evaluated according to the protocol approved by their ethical committee?

The two study endpoints were the only ones to be evaluated according to the protocol.

3. If there were other endpoints (which their phrase “the evaluated endpoint was death…” seems to imply), do the authors agree that the p-values should have been corrected and are therefore misleading?

There were no other endpoints. I agree that it should read “the evaluated endpoint was survival …” But since death and survival are directly, inversely related, this makes no practical difference.

4. Do the authors agree that the endpoint at 180 days is likely to be entirely unrelated to sepsis?

Yes, the endpoint is entirely unrelated to sepsis; however, it is often used in intensive care studies as endpoint and is a meaningful endpoint. It may also be pointed out that homeopathy does not cure a disease, but the patient. Therefore, it is not the disease sepsis alone but the constitution of the patient which is improved and/or healed by homeopathy.

5. Could they provide the sites of sepsis by treatment group, as this is an important determinant of prognosis?

The primary sites of sepsis were (homeopathy vs. placebo group):

Pulmonary infection 27 vs. 28.

Abdominal infection 4 vs. 5.

Urinary tract infection 2 vs. 1.

6. Could the authors provide causes of deaths?

Infections and cardiac failure accounted for almost all of the causes of deaths.

7. Could they also provide the details of the concomitant therapies administered to both patient groups?

In critical care medicine, concomitant therapies comprise of at least 15, often more than 25 different pharmaceutical agents plus mechanical therapies. Therefore, the details would have exceeded the size of the article. All patients received antibiotics, and randomization produced very similar groups. In addition, as pointed out above, the sites of sepsis were very similar. So it is reasonable to assume there were no important differences in concomitant therapies.

8. Could the authors explain the figure of 67.7% as it seems to imply that 21 of 31 patients survived at day 30? If that is correct, what happened to the other three patients in the control group?

Recalculation of the reported number of 23 divided by 34 multiplied by 100 (%) shows the same result except there is a small error at the first decimal place: it should read 67.6% instead of 67.7%. There were no other three patients in the control group.

9. One of the most puzzling features of this study is the fact that 70 patients “were assessed eligibility, all were included in the study”. How do the authors explain that not the usual percentage of patients declined to participate?

The Austrian population has a very positive attitude towards complementary medicine especially homeopathy. Since classical homeopathy does not cause harm if applied professionally, and these patients had little to lose, they, or their authorized representatives, had no objection to the use of homeopathy.

10. As this small study was conducted in Vienna, what were the roles and contributions of the coauthors many of which did not work in Vienna? What were their conflicts of interest? Who financed this study?

The following list describes the roles of the authors. At the time of the study, Dr Banyai was working in Vienna. The coauthors located in Graz cooperated by email and telephone. There were no conflicts of interest, there was no financing of the study, and the authors participated without a fee.

M Frass: principal investigator, decision about remedies, developing protocol, preparation of the manuscript.

M Linkesch: developing protocol, preparation of the manuscript.

S Banyai: collecting data.

G Resch: participating in choosing remedies.

C Dielacher: administration of remedies.

T Löbl: administration of remedies.

C Endler: developing protocol, preparation of the manuscript.

M Haidvogl: developing protocol, preparation of the manuscript.

I Muchitsch: pharmaceutical advice.

E Schuster: statistical analysis.

Again, we thank Prof Ernst for his interest in our study. Since we have other studies in mind, I ask whether he and his Department at Peninsula Medical School would be willing to support our studies theoretically and financially.