Introduction

About 80% of the world’s population in developing countries relies on herbal medicine for some aspect of its health needs.¹ Russian herbal traditions are old and fairly well documented in classic textbooks.² In modern Russia, herbs are regarded as drugs and are used by patients under a physician’s supervision. Physicians themselves usually do not have a preference between herbal and Western conventional pharmaceutical medicines. They also tend to base their recommendations on previous experience as well as the economic status of patients (herbal remedies are often less expensive than pharmaceutical ones that require a prescription). In Russia the practice of herbal medicine (phytotherapy) requires a medical degree from an accredited medical school.

Goal of this Study

General interest in new and traditional uses of herbal medicine in Russia has led to a need for quality assessment of clinical trials published in the Russian language and not yet translated to English. Some published trials include English abstracts and some do not. Not every abstract reflects the study and its results completely and correctly.

This review of human clinical studies recently published in the Russian language was motivated by the general interest of the international traditional medicine community in the quality of clinical studies conducted in Russia and in other countries where research is published in Russian. (The author of this article completed her medical training in Russia and can read and speak Russian fluently.)

Identification and Selection of Studies

A search was conducted on MEDLINE by setting the following parameters: “complementary medicine, clinical trials, human, Russian language, published after 2000.” The decision to select only those studies published after the year 2000 was made in order to limit the number of studies to a manageable size and to analyze only those that are most recent. The search produced 151 articles. Out of these, 49 were studies on herbal preparations. They included 3 review articles, 4 animal studies, 5 homeopathic preparation trials, and 1 nonclinical study, all of which were excluded for the sample. Thus, the final sample of publications satisfying the above criteria contained 36 articles (see Appendix I). Full text copies of all articles were obtained.

Inclusion Criteria

Every journal publication available on MEDLINE and falling under the following criteria was included: Russian-language, original (non-review), clinical studies on herbal remedies in humans.

Extraction and Analysis of Data

For every article, the Medline ID number, name of the journal, date of publication, and city where the study was performed were recorded. Also recorded were specific herb and preparation form, dose, regimen, number of study participants, inclusion and exclusion criteria, length of the study, and parameters of study design. Data were analyzed with Microsoft Excel 2002.

Translation of the names of Russian herbal preparations was conducted using Latin as a common botanical language. When not provided by the authors of the publication, names were cross-referenced through the multilingual Dictionary of Medical Plants.³

Results

The 36 analyzed studies were published in 18 different Russian medical and scientific journals. All of the journals are known to the author of this paper as reputable. The studies were performed in medical centers in 17 different cities. Sixty-four percent of the studies were conducted on patented herbal phytomedicines, with only 36% on herbs and essential oils.

In most of the analyzed studies, study duration and inclusion and exclusion criteria were defined and described (Tables 1 and 2). Institutional review board (IRB) approval or its equivalent was obtained in only 1 of 36 studies (Table 3). None of the studies included the power calculation (p-value) (Table 4). Informed consent was obtained in only 2 studies (Table 5). Placebo control was implemented in 11% of the studies while 28% of the studies were uncontrolled (Table 6). Standard treatment was used for comparison in the rest of the studies (61%). Only 1 trial was double-blind, with almost all studies lacking a blinding procedure entirely (Table 7). Eighty percent of the studies had no randomization performed, while none of the remaining 6 studies described the randomization processes in any detail (Table 8). Most studies had quantitative outcome measures (Table 9). None of the studies described any adverse event reporting system, and only 1 mentioned adverse events experienced or their absence (Table 10). In the cases where study participants stopped their participation before the end of the study, no reasons were provided in any of the studies analyzed. Statistical analysis was done in 75% of the studies (Table 11). Some basic analysis in these trials includes mean and standard deviation percentages, but no p-value for statistical significance. Very few trials list their funding sources and none disclosed authors’ conflict of interests. There is a correlation between geographical location and study quality: all of the best quality studies were conducted in Moscow (Table 12).

It is evident from these tables that many human clinical trials involving herbs and phytomedicines are being conducted in multiple clinical and research centers in Russia and other former republics of the Soviet Union. This reflects the growing interest of the global medical and scientific communities in traditional and new applications of phytotherapy. The treatments examined in these studies reflect the range of conditions that cause significant morbidity in today’s society, not only in Russia but around the world. Cardiovascular diseases and liver problems constitute one third of the health problems studied (Table 13). In total over 100 different herbs and medicinal plant ingredients were included in the 36 trials (Table 14). Some of them—e.g., broomcorn millet (Panicum miliaceum, Poaceae), European white birch (Betula pendula, Betulaceae), and Siberian larch (Larix sibirica, Pinaceae)—
are not very well known or studied in the West.

As noted above, almost all studies (based on the sample that this paper reviews) have deficiencies in the design, quality, and reporting of necessary data in most clinical trials published in Russian medical journals. Various aspects of what is generally considered the gold standard of clinical research—the randomized controlled clinical trial—are frequently lacking in the published studies reviewed in this paper. For example, the concept of IRB approval or some other equivalent seems to still be missing from Russian medical science (Table 3). It is possible that IRB approval has been implemented but is not mentioned in the publication. Even in the studies where masking and randomization were implemented, their method and description was not detailed (Tables 7 & 8). Studies that mentioned and described adverse events never specified the procedure of adverse event reporting (Table 10).

Herbal versus Non-Herbal Studies

All of these design imperfections led to the hypothesis that it may be only herbal medicine studies that are conducted without great attention to study design. This led to the comparative study of publications on clinical trials on non-herbal (pharmaceutical) preparations and other treatment techniques. To accomplish this, a search similar to the original search for herbal studies was performed. Medline was searched with the following parameters: “published after year 2000, in Russian language, human studies, clinical trials, not complementary medicine.” This search produced 1493 publications. Nineteen studies were selected using a table of random numbers (see Appendix II). All the same data as for herbal studies were extracted. Analysis of the non-herbal studies in comparison with herbal studies showed similarities in the samples of both. Almost all non-herbal studies were published in the same academic journals as the herbal studies (Table 15). Their distribution by the year of publication was also similar (Table 16). The hypothesis about differences in study design between the 2 groups found mixed support in this comparison. There were only 2 significant differences:

(1)

greater attention paid to adverse events (over 30% of non-herbal studies registered them, as opposed to only 3% of herbal studies) (Table 10) and

(2)

higher rate for implementation of randomization procedure (26% in non-herbal versus 17% in herbal studies) (Table 8). There was a slight increase in IRB implementation in non-herbal studies as compared to herbal (11% versus 3%) (Table 3). The proportion of studies where informed consent was obtained from participants was also slightly higher in non-herbal studies than in herbal studies (16% versus 6%) (Table 5).

Definitions of inclusion and exclusion criteria were provided in all non-herbal studies and the majority (94%) of herbal studies (Table 2). Power calculation was not provided in any of the herbal or non-herbal studies (Table 4). Use of control in general, and of placebo control in particular, was similar across both groups (Table 6). The same was true of blinding (masking) (Table 7) and randomization procedures (Table 8). In both groups, quantitative outcome measures were the most common (Table 9). Statistical analysis was performed in the majority of both groups (Table 11). Reasons for participants leaving the study before it was complete (dropouts) were not provided in either group.