Bio-genesis.com

Acute treatment of moderate to severe
depression with hypericum extract WS 5570 (St
John's wort): randomised controlled double
blind non-inferiority trial versus paroxetine
A Szegedi, R Kohnen, A Dienel and M Kieser 2005;330;503-; originally published online 11 Feb 2005; BMJdoi:10.1136/bmj.38356.655266.82 Updated information and services can be found at: References
This article cites 22 articles, 5 of which can be accessed free at: 8 online articles that cite this article can be accessed at: Rapid responses
27 rapid responses have been posted to this article, which you can accessfor free at: You can respond to this article at: Email alerting
Receive free email alerts when new articles cite this article - sign up in the box at the top right corner of the article Correction
A correction has been published for this article. The contents of the correctionhave been appended to the original article in this reprint. The correction is available online at: To order reprints of this article go to: Cite this article as: BMJ, doi:10.1136/bmj.38356.655266.82 (published 11 February 2005)
Acute treatment of moderate to severe depression with hypericum
extract WS 5570 (St John’s wort): randomised controlled double
blind non-inferiority trial versus paroxetine
A Szegedi, R Kohnen, A Dienel, M Kieser
Abstract
during six weeks of acute treatment.4 That study, however, wasnot sufficiently powered to demonstrate non-inferiority of the Objective To investigate the efficacy of hypericum extract WS
5570 (St John’s wort) compared with paroxetine in patients with In clinical practice, hypericum extract is better tolerated than moderate to severe major depression.
synthetic antidepressants.7 It may be particularly helpful in Design Randomised double blind, double dummy, reference
severe depression with its high risk of chronicity.8 We compared controlled, multicentre non-inferiority trial.
the efficacy and safety of hypericum extract with paroxetine in Setting 21 psychiatric primary care practices in Germany.
patients with moderate to severe depression.
Participants 251 adult outpatients with acute major depression
Hypericum extract WS 5570 at a dose of 300 mg three times with total score ≥ 22 on the 17 item Hamilton depression scale.
a day has been shown to be more effective than placebo in Interventions 900 mg/day hypericum extract WS 5570 three
patients with mild to moderate major depression treated for six times a day or 20 mg paroxetine once a day for six weeks. In weeks.9 Paroxetine, on the other hand, is a potent selective serot- initial non-responders doses were increased to 1800 mg/day onin reuptake inhibitor with proved efficacy in patients with hypericum or 40 mg/day paroxetine after two weeks.
depression of any severity10 and has a more favourable safety Main outcome measures Change in score on Hamilton
profile than tricyclic antidepressants.11 In major depression, daily depression scale from baseline to day 42 (primary outcome).
doses between 20 mg and 50 mg have been recommended12 and Secondary measures were change in scores on are commonly used in clinical trials and in daily practice.
Montgomery-Åsberg depression rating scale, clinical global In accordance with Kupfer’s model of acute therapy and sub- impressions, and Beck depression inventory.
sequent prophylactic treatment of unipolar depression,13 our Results The Hamilton depression total score decreased by
study included a six week acute phase after which responders mean 14.4 (SD 8.8) points, corresponding to 56.6% (SD 34.3%) undergo four months of prophylactic continuation treatment (to of the baseline value, in the hypericum group and by 11.4 (SD prevent relapse or recurrence, or both).
8.6) points (44.8% (SD 33.5%) of baseline value) in theparoxetine group (intention to treat analysis; similar results were observed in the per protocol analysis). The intention totreat analysis (lower one sided 97.5% confidence limit 1.5 points Protocol, design, and objectives
for the difference hypericum minus paroxetine) and the per This double blind, double dummy, randomised phase III trial protocol analysis (lower confidence limit 0.7 points) showed examined the efficacy of hypericum extract WS 5570 compared non-inferiority of hypericum and statistical superiority over with paroxetine in the acute treatment of moderate to severe paroxetine. The lower limits in both cases exceeded the major depression. After a screening examination participants pre-specified non-inferiority margin of − 2.5 points and the underwent a single blind placebo run-in phase of three to seven superiority margin of 0. The incidence of adverse events was days, during which they received three coated tablets of 0.035 and 0.060 events per day of exposure for hypericum and hypericum placebo per day plus one paroxetine placebo capsule in the morning. After that, we randomised those still meeting the Conclusions In the treatment of moderate to severe major
selection criteria to six weeks of double blind treatment with depression, hypericum extract WS 5570 is at least as effective as hypericum extract or paroxetine. Those who responded to treat- ment (that is, their total score on the 17 item Hamiltondepression scale decreased by ≥ 50%) were invited to participatein a four month double blind maintenance phase (reported else- Introduction
Extract of Hypericum perforatum (St John’s wort) is more effective All patients provided written informed consent. We did not than placebo in the treatment of mild to moderate major use a placebo control group because we considered it unethical depression1 and as effective as several tricyclic antidepressants2–5 to treat severely depressed patients with placebo for six weeks.
or fluoxetine.6 In patients with more severe depression, however, Participants
the antidepressant efficacy of hypericum extract is disputed. In a We recruited male and female outpatients in 21 psychiatric pri- comparison of 1800 mg/day hypericum extract (LI 160) and 150 mary care centres in Germany. All participants were 18-70 years mg/day imipramine the effect of both drugs was comparable old and had single or recurrent moderate or severe episodes of BMJ Online First bmj.com
unipolar major depression without psychotic features (Diagnostic Random sequence generation, allocation concealment,
and Statistical Manual of Mental Disorders, fourth edition, implementation
(DSM-IV) 296.22, 296.23, 296.32, 296.33) persisting for two Patients who still met the selection criteria at baseline were ran- weeks to a year. At screening and baseline all participants had to domised at a ratio of 1:1 to hypericum or paroxetine. Randomi- have a total score ≥ 22 points on the 17 item Hamilton depres- sation was performed in blocks stratified by trial centre. A sion scale and ≥ 2 points for the item “depressive mood.” The biometrician otherwise not involved in the trial generated the diagnosis of depression was based on the mini-international code using a validated computer program. The study drugs were neuropsychiatric interview.14 There were no restrictions regard- dispensed to the centres in numbered containers. On inclusion of a patient into randomised treatment the local investigator We excluded anyone with a decrease in total depression score allocated each participant the lowest available number. The of ≥ 25% during the run-in, or with a diagnosis of schizophrenia, block size was withheld from the investigators.
acute anxiety disorder, adjustment disorder, depressive disorderof any type not stated above, bipolar disorder, organic mental Statistical methods, sample size
disorder, acute post-traumatic stress disorder, or substance abuse Non-inferiority is usually established by showing that the true disorder. We also excluded patients with increased risk of suicide treatment difference is likely to be smaller than a prespecified (defined by a score ≥ 4 for item 10 of the Montgomery-Åsberg non-inferiority margin that separates clinically important from depression rating scale), who had previously attempted suicide, clinically negligible (acceptable) differences.15 We considered that or who had not responded to more than one adequate treatment(equivalent to 150 mg/day amitriptyline for ≥ 6 weeks) in the hypericum would not be relevantly inferior to paroxetine if the present episode. Participants were not allowed to take other psy- true decrease in total depression score (primary outcome meas- chotropic medication and psychotherapy during the study (in ure) for hypericum was not more than 2.5 points16 smaller than case of previous antidepressant medication an appropriate wash out period of five half lives had to be observed).
The study was performed with an adaptive interim analysis.
This design includes options for early stopping with rejection of Interventions and blinding
the null hypothesis or for fultility (boundaries We used hypericum extract WS 5570 (Dr Willmar Schwabe = 0.5, respectively) or for re-estimation of sample size in case Pharmaceuticals, Karlsruhe, Germany), a hydroalcoholic extract from herba hyperici (drug to extract ratio 3-7:1) with For the change in total depression score we assessed standardised contents of 3-6% hyperforin and 0.12-0.28% non-inferiority of hypericum by a shifted t test using the hypericin. The coated tablets contained 300 mg or 600 mg of the prespecified non-inferiority margin of 2.5 points and a global extract. Paroxetine was supplied in tablets of 20 mg packed in one sided type I error of = 0.025. We used Fisher’s combination capsules containing one or two tablets. High and low dose tablets test17 in the final analysis, where the null hypothesis can be or capsules were indistinguishable in all aspects of their outward rejected when the product of the P values from both study parts appearance. For each drug an identically matched placebo was falls below c = 0.0038. An analogous approach consists of calcu- available (the success of blinding was evaluated by examining thedrugs before distribution).
lating the one sided repeated 97.5% confidence limit for the During the six weeks of randomised treatment patients allo- treatment difference adjusted for the interim analysis.18 If this cated to hypericum always took three coated tablets of confidence limit is completely above the non-inferiority margin hypericum/day plus one paroxetine placebo capsule in the = − 2.5, hypericum would be judged to be not inferior to morning whereas those in the paroxetine group took one capsule of paroxetine in the morning and three coated tablets of According to applicable guidance19 we reserved the option of hypericum placebo/day. Initially this corresponded to three testing for superiority after establishing non-inferiority of doses of 300 mg/day hypericum or one dose of 20 mg/day par- hypericum. If the lower one sided 97.5% confidence limit lies oxetine. For patients whose total depression score had not above 0, hypericum can be considered superior to paroxetine.
decreased by at least 20% after two weeks of treatment compared We replaced missing values by carrying the last observation for- with baseline we increased the treatment to three doses of 600 ward. The primary analysis was based on the intention to treat mg/day hypericum or one dose of 40 mg/day paroxetine. The analysis to mirror clinical practice. We also performed a per pro- doses for paroxetine were based on published recommenda- tocol analysis to demonstrate robustness of the trial result to the choice of the analysis set.19 All secondary efficacy and safetymeasures were analysed descriptively. For the Hamilton total Outcomes
score, we defined response as a decrease in total score of ≥ 50% We assessed efficacy and safety at screening, baseline, and at the from baseline and remission as a score ≤ 10 points at week six.
end of the first, second, fourth, and sixth weeks. The primaryoutcome measure was the absolute decrease of the Hamilton We calculated the sample size for the first stage of the study total depression score between baseline and week six. Secondary until the interim analysis by assuming equal changes in depres- outcome measures included the Montgomery-Åsberg depres- sion score in each group with a common SD of 6 points. We sion rating scale, the clinical global impressions, and the Beck needed 2×50 patients to attain 90% power for a one sided P depression inventory. We based assessments of safety and value of P ≤ 0.20 in the interim analysis (trend towards tolerability on spontaneous reports of adverse events, a non-inferiority of hypericum). The interim analysis resulted in a semistructured interview exploring known side effects of the one sided P = 0.084 for the primary outcome measure so that investigational treatments, physical examinations, and routine the local type I error level for the second part of the trial was determined as c /P = 0.045. Assuming a common SD of 6 points To assure uniform diagnostic and rating standards, all assess- and equal means in both groups, we needed 2×75 patients to ments were performed by psychiatrists and psychologists who attain a power of 80% for the second stage of the trial, resulting had participated in training before patients were included.
in a total sample size of 2×125 patients.
BMJ Online First bmj.com
Not randomised (n=50) Not meeting selection criteria (n=33) Adverse event (n=1) Enrolment
Informed consent revoked (n=6) Lost to follow up (n=9) Administrative reasons (n=1) Adverse event (n=8) Violation of exclusion criteria (n=1) Lost to follow up (n=8) Other (n=3) Double blind acute treatment
Data sets for analysis
Fig 1 Flow of patients and datasets for analysis
Figure 2 shows the total Hamilton depression scores over time. Between baseline and day 42 scores decreased by an aver- Participants
age of 14.4 (SD 8.8) points (corresponding to 57% (SD 34%) of Between May 2000 and July 2003, we assessed 301 white patients the baseline value) for hypericum and by 11.4 (SD 8.6) points and randomised and treated 251 (125 to hypericum and 126 to (45% (SD 34%)) for paroxetine (lower one sided repeated 97.5% paroxetine). Figure 1 shows reasons for non-randomisation, pre-mature termination, or exclusion. We did not exclude anypatients because we thought they were at increased risk of Table 1 Demographic and clinical characteristics at baseline (intention to
suicide. Among the patients who were not randomised, two were treat analysis; figures are means (SD); medians unless stated otherwise) withdrawn because they responded to placebo during the run-inperiod. All decisions regarding patient eligibility were made Hypericum (n=122)
Paroxetine (n=122)
Baseline demographic and clinical measures were compara- ble in both groups (table 1). Mean age and average duration of the current episode, however, were higher in the hypericum group. The baseline total depression scores ranged from 22 (minimum required) to 34 in both groups. In each group more than half of the patients had a total score ≥ 25 and were thus Investigational treatment
After two weeks of randomised treatment, 69/122 patients in the hypericum group (57%) and 58/122 in the paroxetine group HAMD=Hamilton depression scale; MADRS=Montgomery-Åsberg depression rating scale.
(48%) were switched to the higher doses. We assessed *Theoretical range 0–52.
†Theoretical range 0–60.
compliance with treatment by counting tablets; it was 96% (SD ‡Theoretical range 0–63; 119 in hypericum group, 120 in paroxetine group.
7%) for hypericum and 98% (SD 10%) for paroxetine.
§According to clinical global impressions score.
BMJ Online First bmj.com
Table 2 Secondary measures (intention to treat analysis; figures are
numbers (percentages) unless stated otherwise) Fig 2 Total Hamilton depression scores over time (intention to treat analysis,
confidence limit adjusted for the interim analysis18 for the differ- ence hypericum–paroxetine was 1.5 points). In the per protocol MADRS=Montgomery-Åsberg depression rating scale; BDI=Beck depression inventory.
*t test for difference (calculated for pooled data from both study stages).
analysis the decreases in scores during treatment were 14.6 (SD †119 in hypericum group, 120 in paroxetine group.
9.0) points for hypericum and 12.0 (SD 8.5) points for paroxetine ‡ 2 test for difference (calculated for pooled data from both study stages).
(lower confidence limit 0.7 points). Hence, the lower confidencelimits not only exceeded the non-inferiority margin of − 2.5 in 29 patients and 61 events in 43 patients, respectively). Table 3 points but also the value 0, showing that hypericum is statistically shows adverse events that occurred in at least 10 patients in one superior to paroxetine at the one sided 2.5% level.
group. Two serious adverse events occurred in the hypericum According to mean change in depression score from group (psychic decompensation attributable to social problems; baseline, hypericum was descriptively superior to paroxetine in hypertensive crisis); both were thought to be unrelated to 11 of those 13 centres that had two or more patients in each hypericum—that is, a cause other than the administration of group. At the end of the acute treatment phase 86/122 patients (71%) in the hypericum group and 73/122 (60%) in the paroxet-ine group responded to treatment (P = 0.08; 2 test), and 61/122 Discussion
(50%) and 43/122 patients (35%) showed remission (P = 0.02).
A subgroup analysis showed that patients who were switched Principal findings
to 1800 mg/day hypericum or 40 mg/day paroxetine because of We have shown that hypericum extract WS 5570 is at least as lack of efficacy during the first two weeks of randomised effective as paroxetine over six weeks of acute treatment in out- treatment showed marked decreases in total depression score patients with moderate or severe unipolar major depression.
during weeks three to six. By the end of the double blind This finding was stable across several validated investigator and treatment period (day 42) we observed a substantial amelioration self rating scales and across the participating centres as well as in of symptoms compared with baseline in patients with or without different analysis datasets (including or excluding patients with an increase in drug dose in both treatment groups (mean (SD) major protocol violations). The average advantage of 3 points for decrease in total score from baseline to day 42: hypericum 900 the decrease in total Hamilton depression score from baseline mg/day 16.6 (7.5) points, hypericum 1800 mg/day 12.6 (9.3) underlines the clinical relevance of the observed effect,16 as do points, paroxetine 20 mg/day 11.0 (8.9) points, paroxetine 40 the responder rates of 70% v 60% and the remission rates of 50% v 35% for hypericum and paroxetine, respectively. The results Table 2 shows the main results for selected secondary meas- thus indicate that in a group of patients in whom the appropri- ures. For all standardised psychiatric scales we found differences ateness of hypericum extract was previously disputed, the antide- between treatment groups in favour of hypericum, confirming pressant efficacy of the herbal drug is at least comparable with the effect of one of the leading synthetic antidepressants. In Safety and tolerability
During the acute treatment phase 69/125 patients randomised
Table 3 Adverse events that occurred in at least 10 patients in one group
to hypericum (55%) reported 172 adverse events and 96/126 (safety analysis set; figures are numbers (percentages) of patients treated with paroxetine (76%) reported 269 adverse events. The incidences were 0.035 adverse events per day of exposure (0.029 at 900 mg/day and 0.039 at 1800 mg/day) for hypericum and 0.060 (0.062 at 20 mg/day and 0.059 at 40 mg/day) for paroxet- ine. Based on the rate ratio, the incidence of adverse events in the paroxetine group was 1.72 (95% confidence interval21 1.42 to 2.10) of the rate observed for hypericum. The highest incidence was found for gastrointestinal disorders (59 events in 42 patients in the hypericum group and 106 events in 67 patients in the par- oxetine group), followed by nervous system disorders (35 events BMJ Online First bmj.com
Contributors: AS and RK conceived the study. AD conceived the study, and What is already known on this topic
participated in its design and coordination. MK participated in the design ofthe study and was responsible for the analysis. All authors read and Hypericum extract is effective in the acute treatment of approved the final manuscript. AD and MK are guarantors.
patients with mild to moderate depression Funding: Dr Willmar Schwabe Pharmaceuticals, manufacturer of WS 5570.
Competing interests: AS has received consultancy fees from Dr Willmar The only randomised controlled trial to date in patients Schwabe Pharmaceuticals. RK is head of a contract research organisation (IMEREM), which is engaged in several clinical trials of hypericum extractfor different pharmaceutical companies. AD and MK are employees of Dr What this study adds
Willmar Schwabe Pharmaceuticals.
Ethical approval: The protocol was approved by the participating centres’ This double blind randomised clinical trial showed that appropriate independent ethics committees.
hypericum extract WS 5570 is at least as effective asparoxetine in ameliorating the symptoms of moderately or Linde K, Mulrow CD. St John’s wort for depression. Cochrane Database Syst Rev Harrer G, Hübner WD, Podzuweit H. Effectiveness and tolerance of the hypericumextract LI 160 compared to maprotiline: a multicenter double-blind study. J Geriatric patients with insufficient response to the initial (lower) dose an Psychiatry Neurol 1994;7(suppl 1):S24-8.
Philipp M, Kohnen R, Hiller KO. Hypericum extract versus imipramine or placebo in increase in dose after two weeks was beneficial.
patients with moderate depression: randomised multicentre study of treatment for It is important to note that for both drugs the higher dose eight weeks. BMJ 1999;319:1534-8.
Vorbach EU, Hübner WD, Arnoldt KH. Effectiveness and tolerance of the hypericum was not associated with a relevant increase in adverse events. In extract LI 160 in comparison with imipramine: randomized double-blind study with particular, none of the patients exposed to hypericum 1800 135 outpatients. J Geriatric Psychiatry Neurol 1994;7(suppl 1):S19-23.
Wheatley D. LI 160, an extract of St. John’s wort, versus amitriptyline in mildly to mod- mg/day for four weeks reported any photosensitivity reactions erately depressed outpatients—a controlled 6-week clinical trial. Pharmacopsychiatry Harrer G, Schmidt U, Kuhn U, Biller A. Äquivalenzvergleich Johanniskrautextrakt Strengths and weaknesses
LoHyp-57 versus Fluoxetin. Arzneimittel-Forschung 1998;49:3-10.
These results contribute to the assessment of the antidepressant Izzo AA. Drug interactions with St. John’s Wort (Hypericum perforatum): a review ofthe clinical evidence. Int J Clin Pharmacol Ther 2004;42:139-48.
effect of hypericum extract in moderately and severely depressed Winkler D, Tauscher J, Kasper S. Maintenance treatment in depression. The role of patients in whom only limited evidence exists. Non-inferiority pharmacological and psychological treatment. Curr Opin Psychiatry 2002;15:63-8.
Lecrubier Y, Clerc G, Didi R, Kieser M. Efficacy of St. John’s wort extract WS 5570 in trials of hypericum extract against synthetic antidepressants have major depression: a double-blind, placebo-controlled trial. Am J Psychiatry been criticised for using doses mostly in the lower therapeutic 10 Bourin M, Chue P, Guillon Y. Paroxetine: a review. CNS Drug Rev 2001;7:25-47.
range of the active comparators.24 This criticism does not apply 11 Preskorn SH. Comparison of the tolerability of bupropion, fluoxetine, imipramine, to our trial, which included a mandatory dose increase in nefazodone, paroxetine, sertraline, and venlafaxine. J Clin Psychiatry 1995;56(suppl6):12-21.
patients with insufficient response after two weeks of treatment.
12 Dunner DL, Dunbar GC. Optimal dose regimen for paroxetine. J Clin Psychiatry For paroxetine, 40 mg/day correspond to the established use of 13 Kupfer DJ. Lessons to be learned from long-term treatment of affective disorders: the drug in clinical trials and daily practice.12 The trial’s assay potential utility in panic disorder. J Clin Psychiatry 1991;52(suppl);12-7.
sensitivity is supported by the observed treatment effect for par- 14 Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. The mini- international neuropsychiatric interview (MINI): the development and validation of a oxetine which was in line with previously published data from structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry trials against placebo and synthetic antidepressants.10 Another 15 Jones B, Jarvis P, Lewis JA, Ebbutt AF. Trials to assess equivalence: the importance of indicator of a pharmacological effect is that in both study groups rigorous methods. BMJ 1996;313:36-9.
a (single blind) dose increase in initial non-responders was 16 Montgomery SA. Clinically relevant effect sizes in depression. Eur Neuropsychopharma- followed by a substantial decrease in depression score that was 17 Bauer P, Köhne K. Evaluation of experiments with adaptive interim analyses. Biometrics comparable with the effect observed in those patients who were adequately treated with the initial (lower) dose. A placebo control 18 Brannath W, Posch M, Bauer P. Recursive combination tests. J Am Stat Assoc could not be used in this group of predominantly severely 19 Committee for Proprietary Medicinal Products. Points to consider on switching between depressed patients for ethical reasons, particularly as comedica- superiority and non-inferiority. London: European Agency for the Evaluation of Medici-nal Products, 2000.
tion with benzodiazepines was not permitted. For the same rea- 20 Paykel ES. The classification of depression. Br J Clin Pharmacol 1983;15(suppl 2):155- son we had to refrain from including patients at high risk of 21 Ederer F, Mantel N. Confidence limits on the ratio of two Poisson variables. Am J Epide- suicide. As we did not actually withdraw any patient because of increased risk of suicide, however, this restriction does not 22 Golsch S, Vocks E, Rakoski J, Brockow K, Ring J. Reversible Erhöhung der Photosensi- tivität im UV-B-Bereich durch Johanniskrautextrakt-Präparate. Hautarzt 1996;48:249- adversely affect the external validity of our data.
23 Schulz V. Incidence and clinical relevance of the interactions and side effects of Implications for clinicians
Hypericum preparations. Phytomedicine 2001;8:152-60.
Our results support the use of hypericum extract WS 5570 as an 24 Hypericum Depression Trial Study Group. Effect of Hypericum perforatum (St. John’s wort) in major depressive disorder. JAMA 2002;287:1807-14.
alternative to standard antidepressants in moderate to severedepression, especially as it is well tolerated.7 As in any effective antidepressant, potential interactions with other drugs deserve The convincing results for hypericum extract WS 5570 Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin, Department of observed in this trial deserve independent confirmation by other Psychiatry and Psychotherapy, Eschenallee 3, 14050 Berlin, GermanyA Szegedi managing senior physician research. We are assessing efficacy in long term treatment, for Institute for Medical Research Management and Biometrics GmbH, Scheurlstraße which the drug can be an interesting option because of its favourable ratio between efficacy and tolerability, in the ongoing R Kohnen head of scientific affairs Dr Willmar Schwabe Pharmaceuticals, PO Box 410925, 76209 Karlsruhe,Germany We thank the investigators and patients, St Klement for project A Dienel head of clinical trials department management, T Konstantinowicz for the data analysis, T Utz for project M Kieser head of biometry department assistance, and A Völp for help with the manuscript.
Correspondence to: M Kieser meinhard.kieser@schwabe.de BMJ Online First bmj.com
Pelizzari E. Academic staff use, perception and expectations about open-access University, 2004. http://ciber.soi.city.ac.uk./ciber-pa-report.pdf (accessed archives. A survey of social science sector at Brescia University. http:// Richardson M, Saxby C. Experimenting with open access publishing.
Academic_staff_perception_about_Open_archives.htm (accessed 7 July www.nature.com/nature/focus/accessdebate/12.html JISC/OSI. Journal authors survey report. www.jisc.ac.uk/uploaded_ Cozzarelli NR, Fulton KR, Sullenberger DM. Results of a PNAS author documents/JISCOAreport1.pdf (accessed 7 July 2004).
survey on an open access option for publication. Proc Natl Acad Sci2004;101:1111.
www.pnas.org/cgi/doi/10.1073/pnas.0307315101 Rowlands I, Nicholas D, Huntingdon P. Scholarly communication in the dig- ital environment: what do authors want? Findings of an international survey ofauthor opinion: project report. London: Centre for Information Behaviour and Evaluation of Research, Department of Information Science, City Commentary: Open access publishing: too much oxygen?
Jeffrey K Aronson
“We hold these truths to be self-evident . . .” This asser- access on day one is basically desirable? But we need to tion of the US founding fathers betokened their zeal be completely sure that if we open the tap on the cylin- for human equality and rights. But such an attitude can der of this 100% oxygen the benefit to harm balance betoken intellectual arrogance. It was, for example, self will be favourable, for we will not be able to turn the tap evident to paediatricians in the 1950s that it would be off—there will be no way back to subscription based Jeffrey K Aronsonreader in clinical beneficial to give premature babies 100% oxygen with- journal publishing. As the third author of the above out proper trial. But 100% oxygen caused blindness, paper1 has written elsewhere, “think harm always.”5 and the balance of benefit to harm was unfavourable.
Competing interests: JKA is a fellow of the British Pharmaco- In their survey of the attitudes of a small sample of logical Society and chairman of the editorial board of the Brit- scientists to open access1 Schroter and colleagues don’t ish Journal of Clinical Pharmacology, which is published on the actually trumpet its self evident benefits, but their call society’s behalf by Blackwell Publishing, as a subscription for evidence refers to the author pays model, not open journal with free access after 12 months; the complete archivesof the journal are about to be digitised for free access.
access publishing itself, although open access will notbe possible without an author pays scheme or Schroter S, Tite L, Smith R. Perceptions of open access publishing: inter- something comparable. But scientists’ opinions should views with journal authors. BMJ 2005;330:756-9.
not frame policy without supporting evidence. We Delamothe T, Smith R. Open access publishing takes off. The dream isnow achievable. BMJ 2004;328:1-3.
need to ask whether immediate free access to readers, Katikireddi SV. HINARI: bridging the global information divide. BMJ with whatever method of payment is used, would ben- Merton RK. The unanticipated consequences of purposive social action.
efit science (not the scientists or the grant giving bod- Am Sociol Rev 1936;1:894-904.
ies, who are also zealous about this idea) and hence Smith R. Think harm always [editor’s choice]. BMJ 2004;329. (3 July.) society. To zealots (“the dream is now achievable”2) thebenefits of this 100% oxygen may be self evident. Butwe have little evidence about the balance of benefits A summary of advantages and disadvantages of the authorpays model is on bmj.com. and harms. I believe that the potential advantages arefew and the disadvantages many; I have summarisedthem on bmj.com.
Why should we uncritically adopt this system? We Corrections and clarifications
already have a better one, operated by many journals Acute treatment of moderate to severe depression with currently and in increasing numbers, in which readers hypericum extract WS 5570 (St John’s wort): pay for immediate access and access becomes randomised controlled double blind non-inferiority trial universally free after a delay, for example 12 months, as required by the National Library of Medicine and the An editing error may have caused confusion in theabstract of this paper by A Szegedi and colleagues Wellcome Trust in their current initiative to digitise (BMJ 2005;330:503-6, 5 Mar). The initial daily dose back issues of journals. Schemes such as HINARI of hypericum WS 5570 was 900 mg split into three (Health InterNetwork Access to Research Initiative) doses of 300 mg—that is, 300 mg three times a day.
and AGORA (Access to Global Online Research in NICE proposes to withdraw Alzheimer’s drugs from Agriculture) will maximise opportunities to access In this News article by Zosia Kmietowicz we In any system the burden of cost should be spread mistakenly referred to donepezil, rivastigmine, and across those who are advantaged. A mixed model galantamine as anticholinesterase inhibitors (BMJ might be appropriate, maintaining subscriptions while 2005;330:495, 5 Mar). They are not; they areacetylcholinesterase inhibitors.
allowing authors who want or are forced to pay forimmediate free access to pay for it, and those who do Children may die when left in overheated cars not want it or cannot afford it, not to. Currently, some In this item in the “BMJ family highlights” section journals adopt author pays access, others do not. But by Harvey Marcovitch, we wrongly said: “A fewchildren were deliberately restrained in a safety belt there are many more readers than authors, which any so that adults could sleep, work, use drugs, or gamble” (BMJ 2005;330:564, 12 Mar). In fact, The uncritical application of basic values is a major according to the original study, the children were source of unforeseen undesirable consequences of restrained in a safety seat, not a belt.
social actions.4 Who doesn’t instinctively feel that free BMJ VOLUME 330 2 APRIL 2005

Source: http://www.bio-genesis.com/productpages/eskaloft/BMJ%20hypericum%20versus%20paroxetine%202005.pdf