Treatment of the Common Cold with Echinacea: A Structured Review
- 1Stanford University School of Medicine, Stanford, CA
- 2Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville
- Reprints or correspondence: Dr. Jack M. Gwaltney, Jr., Dept. of Internal Medicine, Box 800473, Charlottesville, VA 22908 (jmg3u{at}virginia.edu).
Abstract
Background. Echinacea is a herbal preparation that is frequently used to treat the common cold. Spending on echinacea in the United States has risen to >$300 million annually.
Methods. A total of 322 articles related to echinacea and colds, including 9 placebo-controlled clinical trials, were identified using the Medline and PubMed databases. Eleven features of experimental design that affect the accuracy of the measurement of features of interest, the probability of a chance relationship, bias, and blinding were used to evaluate the 9 placebo-controlled studies. The criteria were validated case definition, quantifiable hypothesis, sample-size calculation, randomized assignment, double blinding, proof of blinding, measurement of compliance, measurement of drop-out rate, analysis by intention to treat, description of the methods of analysis, and measurement of probability. Equal weight was given to each criterion, since failure to meet any one of them could potentially invalidate the findings of a clinical trial.
Results. Of the 9 studies, 2 met all 11 criteria. The results of both studies were judged to be negative by the people who performed the studies. Of the remaining 7 studies, 6 were judged to have positive results, and 1 was judged to have negative results. The criterion most commonly not met was proof of blinding.
Conclusions. This structured review suggests that the possible therapeutic effectiveness of echinacea in the treatment of colds has not been established.
Herbal medicine has found its way onto the shelves of drug stores and supermarkets everywhere. The rising costs of prescription drugs and the increased attention to their side effects have contributed to the popularity that herbal treatments have gained in recent years. Also, there is currently a cultural attraction to the “natural” approach to medical care. However, finding medicine in nature is not new. Many well-known drugs, such as aspirin, digitalis, and quinine, originated from plants. These drugs have, over time, been evaluated for efficacy and safety. Many of the herbal medicines found in stores today are untested products, although some have been proven to have no medicinal properties and others have been proven to have adverse effects or to cause drug interactions.
American Indians were the first to use echinacea, a plant found in central and southwestern America, for many different ailments, including cough, sore throat, and tonsillitis. Nine different species of echinacea exist. Three—Echinacea purpurea, Echinacea pallida, and Echinacea angustifolia—are regularly used to treat the common cold [1]. The roots, seeds, flowers, and leaves of the plant are all used in different preparations. Echinacea use in North America peaked in the early 1900s, but its use sharply declined with the advent of antibiotics and after it was denounced in a well-respected medical journal [2]. However, with the recent surge in the use of herbal remedies, spending on echinacea in the United States has risen to >$300 million a year [3].
In this structured review, the use of echinacea to treat the common cold will be examined. Barrett et al. [4], Melchart et al. [5, 6], Immich [7], and Schneider [8] have published reviews on the use of echinacea for colds and influenza-like illness. This report differs from those, in that it focuses only on treatment of the common cold and uses expanded criteria to define the essential features of experimental design. It also uses a dichotomous scale for judging whether criteria were met. In the studies examined, the active ingredient was solely echinacea, and the placebo treatment contained inactive ingredients only. This allowed the focus to remain entirely on echinacea, with no confounding from the effects of other ingredients in the treatment.
Search Methods
The Medline database (which covers the years 1966–2003) and the PubMed database were used to search for the keywords “echinacea,” “common cold treatment,” and “URI treatment.” Searches were narrowed by using other key words: “echinacea common cold,” “echina* common cold,” “echina* URI,” and “echina* treatment cold and URI.” A total of 322 articles were identified. Also, previous reviews of echinacea and colds were examined for studies that may have been overlooked. Studies were excluded if the echinacea formula used for treatment included other active ingredients, such as baptista [9], when the echinacea treatment was compared not with placebo but rather with a vitamin treatment [10], or when the study dealt with prevention of the common cold. A total of 9 placebo-controlled studies were identified that examined the effect of echinacea treatment on decreasing the symptoms, severity, and duration of the common cold.
Criteria Applied to Experimental Design
Eleven criteria were selected to evaluate the adequacy of experimental design (table 1). In some previous reviews [4–8], criteria were given different weights, depending on their presumed importance in affecting the validity of the studies. In this structured review, all criteria were weighted equally on the basis of the assumption that the failure to meet just 1 criterion could potentially invalidate the entire study.
Table 1
Criteria used to evaluate the experimental design of clinical trials of echinacea treatment for the common cold.
The rationale for the selection of the 11 criteria, perhaps obvious in most cases, is as follows (table 1). The validated case definition ensures high “signal quality” and reduces “noise” by ensuring that the cases have the disease of interest on the basis of validated criteria for diagnosis. “Signal quality” refers to the accurate measurement of events of interest (the signal)—in this case, cold symptoms—without dilution of the signal by similar events, such as respiratory symptoms due to other causes such as allergy or respiratory tract irritants that would represent “noise.” A quantifiable hypothesis provides an end point, which allows for the calculation of sample size relative to the defined statistical power, thus addressing the problem that an effect might be due to chance. The randomization of assignment of case and control subjects and the blinding of subjects and investigators control for biases that may affect the validity of the results. Also, data should be supplied to show that blinding was effective. This is especially important for studies of liquid echinacea preparations, which characteristically have a “medicinal” taste. A separate, controlled experiment to actually test the adequacy of blinding should be performed before the clinical trial is conducted. Signal quality is also dependent on the degree of compliance with therapy, which should be monitored to evaluate the amount of bias resulting from subjects who failed to take the drug as directed. Measurement of the drop-out rate should be recorded to evaluate whether statistical power was maintained. Results should be analyzed by the intention-to-treat principle to maintain randomization and statistical power. Although intention to treat ordinarily applies to the number of subjects enrolled in the experiment, it also can apply to the number of symptoms being evaluated. In one study included in this review, an analysis was reported for all subjects enrolled but for only 7 of the 13 symptoms that were recorded. A description of the methods of statistical analysis should be provided, and information should be supplied on the precision of the findings.
Results
The 9 studies were evaluated on the basis of the 11 criteria presented in table 2. The dosages of the echinacea treatments are also presented. Two studies [12, 13] fulfilled all 11 criteria; however, in both, proof of blinding was assessed only after the experiment had been completed. This deficiency is acceptable, since the findings of these studies were negative. One study met 10 of the criteria [14], 1 met 8 [17], 1 met 7 [11], 2 met 6 [15, 18], 1 met 5 [16], and 1 met 3 [19] (table 3). One study with positive results [11] also examined the effectiveness of blinding after the trial was completed. In such a circumstance, there is no way to determine whether judgments were made on the basis of a therapeutic effect. Of the 9 studies, 7 lacked proof of blinding, making it the criterion most frequently not met [11, 14–19] (table 4). Half of the studies did not include a quantifiable hypothesis, sample-size calculation, measurement of compliance, information on the method of randomization and similarity of groups, or validated case definitions.
Table 2
Summary of placebo-controlled studies of the effect of echinacea on the severity and duration of the common cold.
Table 3
Quality of experimental design and reported results of studies of the effectiveness of echinacea for treatment of the common cold.
Table 4
Experimental-design criteria not met in studies of echinacea treatment for the common cold.
The 9 studies were ranked in accordance with the number of criteria fulfilled, and their results were compared (table 3). The 2 studies [12, 13] that fulfilled all 11 criteria had negative results, as judged by the investigators conducting the work. The remaining 7 studies fulfilled 3-10 criteria. Of these studies, 1 had negative results [18], and 6 were judged as having positive results by the investigators [11, 14–17, 19]. The study with positive results [14] that met 10 of the 11 criteria lacked proof of blinding.
Discussion
This structured review was conducted to determine what information has been provided by controlled clinical trials on the effectiveness of echinacea, a popular “natural” remedy, for treatment of the common cold. Eleven features of experimental design were identified as being essential to providing valid results. Both of the 2 studies that met all 11 criteria were judged as having negative results. Of the 7 studies that failed to meet ⩾1 of the 11 criteria, 6 were judged to have positive results (P =.08, by 2-sided Fisher's exact test). It was not felt that a combined analysis of the results of the 9 studies would have been useful, since the validity of the findings from the 7 studies with deficiencies in experimental design was in question.
The most common deficiency was the lack of proof of blinding, which occurred in 7 of the 9 studies. The importance of the placebo effect in cold treatment has been known since 1933, when Diehl [20] showed that placebo capsules containing lactose had a beneficial effect for persons with the common cold. Approximately 30% more of the people given lactose reported improvement, compared with those receiving no treatment. More recently, the placebo affect was documented in a large study of cold treatment [21] that had three arms: treatment, placebo, and no treatment. Reduced illness severity was observed in the group receiving placebo, compared with the group not given treatment.
With these past findings on record, it is not sufficient to assume that blinding is successful in cold treatment studies, and proof of blinding should be supplied in future studies. Chalmers [22] showed the extent to which curious subjects will go to unblind themselves. He described a study that initially showed a positive therapeutic effect of capsules containing vitamin C on the common cold. However, blinding was not maintained, because many subjects bit through the capsules to taste the contents, which they correctly identified. When data from the unblinded subjects were discarded, “there were no differences in the durations of colds” [22, p. 534].
In the present structured review, the second most-common deficiencies, each occurring in 4 trials, were lack of a quantifiable hypothesis, lack of calculation of sample size, lack of a validated case definition, and lack of information on the method of randomization and similarity of groups. Also, compliance was not measured in 4 trials. The other criteria were not met in ⩽3 trials. No studies failed to describe the methods of statistical analysis or probability testing. Thus, the quality of the experimental design of these studies was generally satisfactory, except with regard to proof of blinding and sample size. Of these 2 criteria, the lack of proof of blinding appears to be the most important and that which may have invalidated the results because of bias associated with a placebo effect.
Only 9 placebo-controlled trials of echinacea treatment for colds were identified, and, of these, only 2 met the predetermined criteria judged acceptable for experimental design. On the basis of this limited information, this review suggests that the possible value of echinacea in treating the common cold is not established. Data from further well-designed studies are needed to address this question. Cold remedies are available that have been well studied, are effective, and have acceptable safety levels [23]. Also, new approaches to the treatment of colds, such as use of antiviral drugs, are being developed that have a sounder theoretical basis than does use of echinacea. These approaches are based on increasing knowledge about the pathogenesis of colds [24], whereas the theoretical basis for the use of echinacea is unclear, other than an ill-defined effect on “immunity.”
Acknowledgments
We thank Jacqueline Grubbs for help in preparing the manuscript.
Potential conflicts of interest. J.M.G. and T.J.C.: no conflicts.
- Received September 20, 2004.
- Accepted October 28, 2004.
- © 2005 by the Infectious Diseases Society of America
References
This Article
- AbstractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Classifications
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 8.186
5-Yr impact factor: 7.898
For The Media
For Authors
- Services for authors
- Self-archiving Policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
