Preventive Medicine 49 (2009) 78–82

Contents lists available at ScienceDirect

Preventive Medicine

j ourna l homepage: www.e lsev ie r.com/ locate /ypmed

Review

Echinacea for prevention of the common cold: An illustrative overview of howinformation from different systematic reviews is summarised on the internet

Anna Hart ⁎, Paola DeySchool of Public Health and Clinical Sciences, University of Central Lancashire, Harrington Building, Adelphi Street, Preston, Lancashire, PR1 2HE, UK

⁎ Corresponding author.E-mail address: AHart@uclan.ac.uk (A. Hart).

0091-7435/$ – see front matter © 2009 Elsevier Inc. Aldoi:10.1016/j.ypmed.2009.04.006

a b s t r a c t

a r t i c l e i n f o

Available online 21 April 2009

Keywords:EchinaceaPreventionSystematic reviewsInternetHealth knowledgeAttitudes and practiceEvidence-based medicineComplementary therapiesUncertaintyCommon cold

Objectives. Three systematic reviews of echinacea for the prevention of colds have somewhat differentconclusions. Our study objectives were to illustrate how selection criteria for trials in each of the reviewscould lead to different conclusions, and to classify the ways in which webpage authors use the reviews toconstruct advice about echinacea.

Methods. A funnel plot was constructed of all treatment comparisons in the primary trials included in thereviews. A World Wide Web search was undertaken using five major search engines in order to locatewebpages that referred to the efficacy of echinacea in cold prevention and also referred to each of thereviews.

Results. Twelve webpages were located. Three webpages presented findings from all three reviews; fivepresented findings from one or two; four used all three reviews as general supporting references. Viewsabout efficacy differed. There were few comments about the sources of heterogeneity between reviews.

Conclusions. Given the residual uncertainty and the gaps between the evidence and the ways that this is

summarised on webpages, it may prove difficult for consumers to assimilate the evidence. As well asundertaking high-quality trials in complementary medicine, we also need to ensure precision in thereporting of uncertainty.

© 2009 Elsevier Inc. All rights reserved.

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Study limitations and strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Introduction

Systematic reviews which summarise evidence from high-qualityrandomised controlled trials are accepted as the best evidence ofefficacy in healthcare. They are used to ameliorate some of theinferential uncertainties associated with the findings of small, under-powered individual trials and single-centre studies (Guyatt et al.,2008). However, in addition to other sources of heterogeneity, theconclusions of systematic reviews depend onwhat decisions aremade

l rights reserved.

about which trials to include and whether or not the results of theseare statistically combined in a meta-analysis. Complementary med-icine is an area in which many small, single-centre randomisedcontrolled trials have been performed and a number of systematicreviews undertaken to summarise the overall evidence of efficacy ofvarious therapies (e.g. Ernst and Canter, 2006; Ernst, 2002; Derryet al., 2006).

Echinacea is a commonly-used, over-the-counter herbal remedy. Itis often used for the prevention of the common cold, but this does notmean that it is effective. There are three systematic reviews of trials ofechinacea for the prevention of the common cold (Schoop et al., 2006;Linde et al., 2006; Shah et al., 2007). The review by Shah et al. (2007)

Table 1Summary of key aspects of the protocols and conclusions of systematic reviews of echinacea for prevention of the common cold.

Review Shah et al. (2007) Schoop et al. (2006) Linde et al. (2006)

Selection criteriafor RCTs

Echinacea products for natural andinduced colds

Echinacea for rhinovirus inoculated colds Echinacea for naturally-acquired colds

Number of trials 9a 3 2a

Analysis Meta-analysis with sensitivity analysis.Data pooled within 3-armed or 4-armed.

Meta-analysis with data pooled fromone 3-arm trial.

Summary of 3 comparisons from 2 trials(meta-analyses were not anticipated).

Treatment Echinacea (1 trial echinacea, propolis,vitamin C)

Echinacea Echinacea

Control groups Placebo (1 magnesium arm in a trial oftriathletes)

Placebo Placebo

Summary intext of review

“An analysis of the current evidence in theliterature suggests that echinacea has abenefit in decreasing the incidence andduration of the common cold; however,large-scale randomised prospective studiescontrolling for variables such as species,quality of preparation, and dose of echinacea,method of cold induction, and objectivity ofstudy endpoints evaluated are needed beforeechinacea for the prevention or treatment ofthe common cold can become standard practice.”

“This meta-analysis suggests thatstandardized extracts of echinacea wereeffective in the prevention of symptomsof the common cold after clinical inoculation,compared with placebo. Further prospective,appropriately powered clinical studies aredesired to confirm this finding.”

“Beneficial effects of other echinaceapreparations, in adults or when used forpreventative purposes, might exist buthave not been shown in independentlyreplicated, rigorous randomized trials.”

Summary in abstract “Published evidence supports echinacea'sbenefit in decreasing the incidence andduration of the common cold.”

“This meta-analysis suggests that standardizedextracts of echinacea were effective in theprevention of symptoms of the common coldafter clinical inoculation, compared with placebo.Further prospective, appropriately poweredclinical studies are required to confirm this finding.”

“Echinacea preparations tested in clinicaltrials differ greatly…Beneficial effects ofother echinacea preparations, and echinaceaused for preventative purposes might existbut have not been shown in independentlyreplicated, rigorous RCTs.”

Previous conference abstract (Shah et al. 2006):“The totality of current evidence supportsechinacea's benefit in decreasing the incidenceand duration of the common cold.”

Lay summary: “Two trials investigatedwhether taking echinacea preparationsfor 8 to 12 weeks prevents colds but foundno clear effect.”

a Shah et al. (2007) included two trials of prevention of the development of a full cold after initial symptoms, but these were categorised as treatment trials by Linde et al. (2006).

79A. Hart, P. Dey / Preventive Medicine 49 (2009) 78–82

combines the results from nine trials including the subsets of trialsincluded in the other two reviews. This review was previouslypublished as a conference abstract (Shah et al., 2006). The Cochranereview (Linde et al., 2006) was amended in 2007 including adiscussion of the review by Shah et al. (2007), although the citationwas unchanged. Table 1 summarises the key differences in theprotocols and conclusions of these reviews.

While consumers of echinaceamay not have access to the scientificliterature, many of them will have access to the World Wide Web.Entering the terms ‘echinacea’ and ‘prevention’ and ‘common cold’yields about 47,000 hits on the search engine Google™. Such wealthand accessibility of online health information has generated consider-able concern: there have been numerous empirical studies that haveassessed the quality of information held on the World Wide Web fordifferent topics, and most have concluded that it is a problem.Eysenbach et al. (2002) warned that these types of studies cannotfully quantify the risk posed to the public from inadequate information,as this depends not only on the proportion of inadequate informationon the web but also on how or whether individuals filter theinadequate sites (Eysenbach et al., 2002). They also commented thatit is important to try to understand ‘where andwhy gaps exist betweenevidence-based medicine and health information on the Internet’.

In this paper, we explore gaps between the evidence from the threesystematic reviews of echinacea for the prevention of the commoncold and online reporting of this body of evidence. The first objectiveof this study was to illustrate how the selection criteria for trials ineach of the reviews could lead to different conclusions about theefficacy of echinacea for the common cold. The second objective wasto classify the ways in which webpage authors use the reviews toconstruct advice about the efficacy of echinacea.

Methods

In order to illustrate the heterogeneity related to selection criteria, afunnel plot was produced. In meta-analyses (Schoop et al., 2006; Shah et al.,

2007), data from a three- or four-armed trial were pooled (Berg et al., 1998;Melchart et al., 1998; Turner et al., 2005), but, for the funnel plot, datadescribing each treatment comparison were extracted from the three originalpapers. For the other trials, data were extracted from the reviews. Data wereextracted by one author (AH) and checked against the original papers by theother author (PD). The log of odds ratios and standard errors were computedusing Haldane's method.

We searched for webpages that referred to the efficacy of echinacea in theprevention of the common cold and which also referred to each of the threesystematic reviews. Five major search engines (Google™, Live Search™(MSN), AOL™ UK, Yahoo!™ UK and Ireland and Lycos™) were used. Oneauthor (AH) searched using all four terms: echinacea, Cochrane, Shah andSchoop.We includedwebpages if they referred to the full review by Shah et al.(2007) or to the earlier conference abstract of the same study (Shah et al.,2006). Webpages were excluded if they required a log-in address to access theinformation or were not in English. The other author (PD) checked eachsearch on a separate occasion. The searches were undertaken in December2008. Any discrepancies between reviewers were resolved by discussion.

Data on the organisational affiliation of the website holding theinformation and on the type of information were collected. For each webpage,we asked the following questions:

1. What is the overall advice about the efficacy of echinacea for theprevention of common colds?

2. How is the information from each of the systematic reviews used?3. How are the differences between the types of trials included in the

systematic reviews discussed?

One of the authors (PD) extracted data directly from the webpages ontoa spreadsheet and then classified the ways in which the webpages usedthe information from the reviews. The other author (AH) checked thisanalysis against the original webpages. Any differences were resolved bydiscussion.

Results

The funnel plot of treatment comparisons is shown in Fig. 1. Effectestimates (log of odds ratio) are on the horizontal axis, with all values

Table 2URLs of webpages which comment on the efficacy of echinacea in prevention ofcommon cold and which refer to the three systematic reviews.

http://cms.herbalgram.org/press/2007/usatodayechinacea.htmlhttp://www.drugs.com/npp/echinacea.htmlhttp://www.library.nhs.uk/rss/newsAndRssArticle.aspx?uri=http://www.library.nhs.uk/resources/?id=262678http://www.drugdigest.org/wps/PA_1_DHFHCNL218J630276FL8L614D2/pages/common/ddPrintPage.jsphttp://nutrition.about.com/od/dietsformedicaldisorders/f/flunutrients.htmhttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2571112http://nursing.advanceweb.com/Editorial/Content/Editorial.aspx?CC=107731(references at http://nursing.advanceweb.com/Article/Cold-Flu-Season.aspx)http://content.herbalgram.org/new-chapter/herbalgram/articleview.asp?a=3107&p=Yhttp://board.crossfit.com/archive/index.php/t-25124.htmlhttp://pointinstitute.org/resources/standard_v_8.2+immunity.pdfhttp://www.emea.europa.eu/pdfs/human/hmpc/echinaceae_purpureae_herba/47546307en.pdfhttp://www.naturalproductsmarketplace.com/articles/781notes.html

Google™, Live Search™ (MSN), AOL™ UK, Yahoo!™ UK and Ireland and Lycos™ weresearched in December 2008.

80 A. Hart, P. Dey / Preventive Medicine 49 (2009) 78–82

favouring echinacea, and the precision of this estimate is on thevertical axis. The different symbols distinguish between the differenttypes of trials. For some primary trials when there were more thantwo groups and results were pooled in the meta-analyses, the originalcomparisons are shown on the plot using an alphanumeric index, e.g.3a, 3b and 3c define 3 treatment comparisons against the same control(placebo) group.

There are both clinical and statistical heterogeneities betweentrials included in the Shah review (Shah et al., 2007). The two trialsincluded in the Cochrane investigation of prevention tend towardsgreater precision and smaller effect size thanmost of the other subsetsof trials; these trials examined the effect of echinacea to preventnaturally acquired colds (Linde et al., 2006). The Berg study, with alarge effect size but low precision, examines the efficacy of echinaceain triathletes who may be at increased risk of colds because of theimmunosuppressive effects of strenuous exercise (Berg et al., 1998);the three studies included in the review by Schoop et al. (2006)investigate the efficacy of echinacea following the inoculation ofvolunteers by a single rhinovirus; the trial with a large effect size andhigh precision is the only trial investigating the efficacy of echinacea incombination with other remedies, namely vitamin C and propolis(Cohen et al., 2004). The remaining two trials examine the efficacy ofechinacea when taken after initial cold symptoms have occurred:these are classified as treatment trials in the Cochrane review (Lindeet al., 2006).

A total of 12 webpages on 11 websites were identified and arelisted in Table 2. The affiliation of the websites comprised nationalhealth and regulatory agencies; publishing houses; nutraceutical andfitness companies; not-for-profit and non-commercial organisationsand an independent trust. Of the 12webpages, four could be describedas providing consumer information, two as providing information forprofessionals and one as providing information for retailers; twocontained journal papers (one of which had an impact factor); onewas a consultation on product licensing; one was a press statementand the remaining source was a comment in a chat room. Evidencefrom the systematic reviews was summarised in different ways on the

Fig. 1. Funnel plot showing all treatment comparisons from the original trials.● trials ofnaturally-acquired colds, ○ trials of full cold development after initial symptoms, □trials of rhinovirus induced colds, + trial of echinacea with vitamin C and propolis, Xtrial on triathletes. All 9 trials were included in the review by Shah et al. (2007), witharms within 3 trials pooled (1a and 1b placebo and magnesium control groups, 2a and2b different varieties of echinacea, 3a 3b and 3c different extracts of echinacea: theresults for 3b and 3c were identical). Schoop et al. (2006) only included the 3inoculation trials with arms 3a, 3b and 3c pooled. Linde et al. (2006) included 2 trials ofnatural colds, and 2 trials of development of full colds after initial symptoms butcategorised these as treatment trials. Only Shah et al. (2007) included the trial ofechinacea with vitamin C and propolis and the trial on triathletes (with pooled controlgroups).

webpages. These included presenting or summarising the findings ofall three reviews (3 webpages); presenting or summarising thefindings from one or two of the reviews and either making a generalstatement about, or just mentioning the existence of, the remainingreview(s) (5 webpages); using all three reviews as general supportingreferences (4 webpages).

The overall advice on webpages about the efficacy of echinacea inpreventing common colds varied: views expressed included that it didprevent common colds, that it had only modest impact and that it didnot prevent colds. There were no views suggesting echinacea was lessbeneficial than placebo. There were few comments relating to theheterogeneity between reviews as illustrated in the figure. Remarkswere made on the specific objectives and inclusion criteria in thereview by Schoop et al. (2006), some exclusion criteria in the reviewby Linde et al. (2006), and the fact that trials excluded by Linde et al.(2006) were included in the other two reviews. Implications for theinterpretation of the evidence were not discussed. Some webpagesnoted reviewers' concerns about the variety of preparations ofechinacea used in trials.

Discussion

Consumers and health professionals look to systematic reviews toprovide definitive answers about the efficacy of an intervention to aiddecisions about treatment or prevention. But systematic reviews willfall short of this expectation. They can only summarise evidence fromthe available studies pertinent to the question posed by the authors.Often, uncertainty remains because there are too few trials or theselected trials have limited relevance to usual care or situations. Meta-analysis cannot overcome this problem: in the presence of publicationbias (Dwan et al., 2008), p-values and confidence intervals are stillunreliable (Copas and Malley 2008), and clinical heterogeneitybetween trials can greatly affect the results and subsequentinterpretation of a meta-analysis.

The latter is illustrated in our figure for all the trials included inthe three reviews of echinacea for common cold prevention. Wehave suggested some reasons for the clinical heterogeneity observedin our figure; others have also been described, such as the differentpreparations (Shah et al., 2006; Linde et al., 2006; Woelkart et al.,2008). This clinical heterogeneity may explain why authors of thethree reviews vary as to whether they consider that the preventiveefficacy of echinacea has been established. Schoop et al. (2006)restricted their study to trials investigating the efficacy of standar-dised extracts of echinacea used in ‘the prevention of symptoms ofthe common cold after clinical inoculation’ of single strains of

81A. Hart, P. Dey / Preventive Medicine 49 (2009) 78–82

rhinovirus. The findings of this review may have less relevance tothe prevention of colds acquired in the community. The Cochranereviewers took a strict approach to answer a precise question aboutechinacea itself (i.e. not combination products) in naturally-acquiredcolds (Linde et al., 2006). This was likely to give an uncertain resultas ‘the more rigorous the review, the less evidence there will be thatthe intervention is effective’ (Pettigrew 2003), and they did notexpect to be able to undertake meta-analyses. Shah et al. (2007)took an inclusive approach to the selection of trials which appears toreduce the uncertainty in the overall estimate of effect from themeta-analysis, but increases the uncertainty about what thatestimate means.

These are complex issues but they have important implications forconsumers who are seeking advice on whether or not to takeechinacea to prevent colds. However, our review of webpagesdemonstrated that such issues were rarely mentioned. We cannottell whether those summarising the evidence from the reviews did notunderstand the implications of these issues or felt it was unnecessaryor too difficult to convey them to their potential readers. Knowing thismight explainwhywebpages did not always summarise or present thefindings of each of the three reviews and why they had differentconclusions about the overall efficacy of echinacea, despite referringto the same body of evidence. However, somewebpage authors mightalso summarise or present evidence in different ways and come todifferent conclusions because of their own beliefs as to the preventiveproperties of echinacea.

There is limited research about whether consumers are influencedby how evidence is summarised. Research suggests that consumersmay ignore information from sites that they perceive as less credible,but definitions of credibility vary (Dutta-Bergman 2003). Tools toguide consumers' appraisal of online information exist (Charnock andShepperd 2004); one highlights the discussion of uncertainty as amarker of high-quality information (http://www.discern.org.uk/discern_instrument.php). Consumers may also hold prior beliefsabout the health properties of echinacea. These beliefs can be basedon a variety of sources such as personal or anecdotal experience andmaterial from other media. There is some evidence that those withstrong beliefs are most likely to search out online information thatconfirms their viewpoint and that prior beliefs can affect consumerinterpretation of online information (Keselman et al., 2008; Lau andCoiera 2007).

Study limitations and strengths

There is a lack of empirical studies about online informationabout complementary therapies. Our study was exploratory. It waslimited to one topic area and only a small number of webpageswere identified. Therefore, we cannot be sure that the classificationof the way webpages summarise evidence is comprehensive orgeneralisable to other areas. We focussed on webpages that referredto all three reviews, and it is likely that some webpage authorsmight handle uncertainty by omitting conflicting or controversialevidence. We can only suggest the reasons why the differentconclusions of the systematic reviews and the clinical heterogeneityof the trials in the reviews were not always discussed in theidentified webpages.

It was not the intention of this review to undertake anycomparative analysis, for example, between the type of website andhow the evidence from the systematic reviews was summarised, or toundertake a formal evaluation of the quality of the information on thewebpages. Without involving webpage authors, it is not possible to besure of the reasons and motivations for variations in how theyconstruct advice from a body of evidence. Without involving internetusers, it is impossible to estimate the risk of misinformation frominadequate summarising or presentation of evidence from thesystematic reviews. Further work is clearly needed, but there are

also more immediate lessons for those who undertake systematicreviews.

All three systematic reviews concluded that more high-qualitytrials are needed in order to answer the question about theprophylactic efficacy of echinacea; this is likely to result in furthersystematic reviews. Similar issues will pervade other areas ofcomplementary medicine. However, elements of uncertainty willremain. Uncertainty is difficult to understand and to convey, but theproblem does not rest solely with the non-scientific community. In2003, Alderson and Chalmers noted that ‘It is never correct toconclude that treatments have no effects’, but they identifiedinappropriate summaries of ‘negative’ results in Cochrane reviewabstracts (Alderson and Chalmers, 2003). While such errors occur inthe scientific literature, one cannot expect lay summaries to beaccurate. This study reinforces the need for precision in the way thatevidence is summarised and presented in systematic reviews,especially resistance of the temptation to sensationalise results or tosummarise them to ‘the extent that the correct meaning is lost’(Alderson, 2004).

Conclusions

The three systematic reviews of echinacea for the prevention ofthe common cold come to somewhat different conclusions. Thisevidence is potentially difficult for consumers to interpret, especiallyas there are gaps between the body of evidence and the diversesummaries on the World Wide Web. More high-quality trials areneeded in complementary medicine, but areas of uncertainty mayprevail. It is important to foster a culture in which it is possible toaccept and discuss uncertainty, otherwise consumers may not be ableto make informed treatment decisions and will rely on their ownbeliefs. Precision is vital in the reporting of uncertainty in scientificfindings.

Conflict of interest statement

The authors declare that there are no conflicts of interest.

References

Alderson, P., 2004. Absence of effect is not evidence of absence. BMJ 328, 476–477.Alderson, P., Chalmers, I., 2003. Survey of claims of no effect in abstracts of Cochrane

reviews. BMJ 326, 475.Berg, A., Northoff, H., Konig, D., et al., 1998. Influence of Echinacin (EC31) treatment on

the exercise-induced immune response in athletes. J. Clin. Res. 1, 367–380.Charnock, D., Shepperd, S., 2004. Learning to DISCERN online: applying an appraisal tool

to health websites in a workshop setting. Health Educ. Res. 19, 440–446.Cohen, H.A., Varsano, I., Kahan, E., Sarrell, E.M., Uziel, Y., 2004. Effectiveness of an herbal

preparation containing echinacea, propolis and vitamin C in preventing respiratorytract infections in children: a randomised controlled trial. Arch. Pediatr. Adolesc.Med. 158, 217–221.

Copas, J.B., Malley, P.F., 2008. A robust P-value for treatment effect in meta-analysis withpublication bias. Stat. Med. 27, 4267–4278.

Dwan, K., Altman, D.G., Arnaiz, J.A. et al., 2008. Systematic review of the empiricalevidence of study publication bias and outcome reporting bias. PLoS ONE 3,e3081.doi:10.1371/journal.pone.0003081.

Derry, C.J., Derry, S., McQuay, H.J., Moore, R.A., 2006. Systematic reviews of systematicreviews of acupuncture published 1996–2005. Clin. Med. 6, 381–386.

Dutta-Bergman, M., 2003. Trusted online sources of health information: differences indemographics, health beliefs, and health-information orientation. J. Med. InternetRes. 5, e21 doi:10.2196/jmir.5.3.e21.

Ernst, E., 2002. A systematic review of systematic reviews of homeopathy. Br. J. Clin.Pharmacol. 54, 577–582.

Ernst, E., Canter, P.H., 2006. A systematic review of systematic reviews of spinalmanipulation. J. R. Soc. Med. 99, 192–196.

Eysenbach, G., Powell, J., Kuss, O., Sa, E.R., 2002. Empirical studies assessing the qualityof health information for consumers on the World Wide Web: a systematic review.JAMA 287, 2691–2700.

Guyatt, G.H., Mills, E.J., Elbourne, D., 2008. In the era of systematic reviews, does the sizeof an individual trial still matter? PLOS Medicine 5, 3–5. doi:10.1371/journal.pmed.0050004.

Keselman, A., Browne, A.C., Kaufman, D.R., 2008. Consumer health information seekingas hypothesis testing. J. Am. Med. Inform. Assoc. 15, 484–495.

Lau, A.Y., Coiera, E.W., 2007. Do people experience cognitive biases while searching forinformation? J. Am. Med. Inform. Assoc. 14, 599–608.

82 A. Hart, P. Dey / Preventive Medicine 49 (2009) 78–82

Linde, K., Barrett, B., Bauer, R., Melchart, D., Woelkart, K., 2006. Echinacea for preventingand treating the common cold. Cochrane Database of Systematic Reviews, Issue 1.Art. No.: CD000530.DOI:10.1002/14651858.CD000530.pub2.

Melchart, D., Walther, E., Linde, K., Brandmaier, R., Lersch, C., 1998. Echinacea rootextracts for the prevention of upper respiratory tract infections: a double-blind,placebo-controlled randomized trial. Arch. Fam. Med. 7, 541–545.

Pettigrew, M., 2003. Why certain systematic reviews reach uncertain conclusions. BMJ326, 756–758.

Schoop, R., Klein, P., Suter, A., Johnston, S.L., 2006. Echinacea in the prevention ofinduced rhinovirus colds: a meta-analysis. Clin. Ther. 28, 174–183.

Shah, S.A., Sander, S., White, C.M., Rinaldi, M., Coleman, C.I., 2006. Evaluation ofechinacea for the prevention and treatment of the common cold: a meta-analysis ofrandomized, controlled trials. J. Clin. Pharmacol. 46, 1076.

Shah, S.A., Sander, S., White, C.M., Rinaldi, M., Coleman, C.I., 2007. Evaluation ofechinacea for the prevention and treatment of the common cold: a meta-analysis.Lancet Infect. Dis. 7, 473–480.

Turner, R.B., Bauer, R., Woelkart, K., Hulsey, T.C., Gangemi, J.D., 2005. An evaluation of Echi-nacea angustifolia in experimental rhinovirus infections. N. Eng. J. Med. 353, 341–348.

Woelkart, K., Linde, K., Bauer, R., 2008. Echinacea for preventing and treating thecommon cold. Planta Medica 74, 633–637.