Prevalence of Articles With Honorary Authors and Ghost Authors in Peer-Reviewed Medical Journals

Context.— Authorship in biomedical publications establishes accountability, responsibility, and credit. Misappropriation of authorship undermines the integrity of the authorship system, but accurate data on its prevalence are limited. Objectives.— To determine the prevalence of articles with honorary authors (named authors who have not met authorship criteria) and ghost authors (individuals not named as authors but who contributed substantially to the work) in peer-reviewed medical journals and to identify journal characteristics and article types associated with such authorship misappropriation. Design.— Mailed, self-administered, confidential survey. Participants.— A total of 809 corresponding authors (1179 surveyed, 69% response rate) of articles published in 1996 in 3 peer-reviewed, large-circulation general medical journals (Annals of Internal Medicine, JAMA, and The New England Journal of Medicine) and 3 peer-reviewed, smaller-circulation journals that publish supplements (American Journal of Cardiology, American Journal of Medicine, and American Journal of Obstetrics and Gynecology). Main Outcome Measures.— Prevalence of articles with honorary authors and ghost authors, as reported by corresponding authors. Results.— Of the 809 articles, 492 were original research reports, 240 were reviews and articles not reporting original data, and 77 were editorials. A total of 156 articles (19%) had evidence of honorary authors (range, 11%-25% among journals); 93 articles (11%) had evidence of ghost authors (range, 7%-16% among journals); and 13 articles (2%) had evidence of both. The prevalence of articles with honorary authors was greater among review articles than research articles (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.2-2.6) but did not differ significantly between large-circulation and smaller-circulation journals (OR, 1.4; 95% CI, 0.96-2.03). Compared with similar-type articles in large-circulation journals, articles with ghost authors in smaller-circulation journals were more likely to be reviews (OR, 4.2; 95% CI, 1.5-13.5) and less likely to be research articles (OR, 0.49; 95% CI, 0.27-0.88). Conclusion.— A substantial proportion of articles in peer-reviewed medical journals demonstrate evidence of honorary authors or ghost authors.      

Discussion sections in reports of controlled trials published in general medical journals

Context  Reliable interpretation of the results of a controlled trial entails setting its results in the context of similar research. A previous study showed that most reports of controlled trials published in 5 general medical journals in May 1997 were deficient in this respect. We assessed the extent to which reports of controlled trials published in the same 5 journals discussed new results in light of the totality of evidence from other controlled trials. Methods  Assessment of the discussion sections in all 33 reports of randomized trials published during May 2001 in Annals of Internal Medicine, BMJ, JAMA, The Lancet, and The New England Journal of Medicine. Results  Three reports appeared to have been the first published trials to address the questions studied. In none of the remaining 30 reports were the results of the new trial discussed in the context of an updated systematic review of other trials. Although reference was made to relevant systematic reviews in 3 of these 30 reports, there was no integration, quantitative or qualitative, of the results of the new trials in an update of these reviews. In the remaining 27 reports, there was no evidence that any systematic attempt had been made to set the new results in the context of previous trials. Conclusions  Between 1997 and 2001, there was no evidence of progress in the proportion of reports of trials published in general medical journals that discussed the new results within the context of, or with reference to, up-to-date systematic reviews of relevant evidence from other controlled trials.      

Reporting of informed consent and ethics committee approval in clinical trials

Context  To determine whether journals have improved their disclosure of ethical protections in clinical trials. Methods  Comparison of clinical trials published before and after 1997 (July 1995 to December 1996 and January 1998 to June 1999) in Annals of Internal Medicine, BMJ, JAMA, The Lancet, and The New England Journal of Medicine. Sixty articles per journal per period were randomly selected and assessed for rate of reporting on informed consent and on ethics committee approval. Results  Informed consent was not described in 79 articles (26%) published before 1997 vs 53 (18%) published after 1997 (P = .01), and ethics committee approval was not mentioned in 93 (31%) before 1997 vs 54 (18%) after 1997 (P<.001). Neither protection was described in 48 articles (16%) published before 1997 vs 28 (9%) after 1997 (P = .01). In subgroup analyses, those journals with the worst initial rates generally improved the most. BMJ did not describe informed consent in 25 articles (42%) before 1997 vs 15 (25%) after 1997 (P = .05), and JAMA did not describe ethics committee approval in 25 (42%) before 1997 vs 13 (22%) after 1997 (P = .02). BMJ, JAMA, and Annals had the lowest initial rates of reporting on both protections in the same article, with 25 (42%), 32 (53%), and 34 (57%), respectively, but improved markedly to 38 (63%), 43 (72%), and 45 (75%) (P = .02, .04, and .03, respectively). Conclusions  Major medical journals have improved their reporting on informed consent and ethics committee approval; however, 9% of studies still report neither.      

Prevalence of honorary and ghost authorship in Cochrane reviews

Context  To determine the prevalence of honorary and ghost authorship in Cochrane reviews, how authorship is assigned, and the ways in which authors and Cochrane editorial teams contribute. Methods  Using a Web-based, self-administered survey, corresponding authors for 577 reviews published in issues 1 and 2 from 1999 of The Cochrane Library were invited to report on the prevalence of honorary and ghost authors, contributions by authors listed in the byline and members of Cochrane editorial teams, and identification of methods of assigning authorship. Responses were received for 362 reviews (63% response rate), which contained 913 authors. Results  One hundred forty-one reviews (39%) had evidence of honorary authors, 32 (9%) had evidence of ghost authors (most commonly a member of the Cochrane editorial team), and 9 (2%) had evidence of both honorary and ghost authors. The editorial teams contributed in a wide variety of ways to 301 reviews (83%). Authorship was decided by the group of authors (31%) or lead author (25%) in most reviews. Authorship order was assigned according to contribution in most reviews (76%). The 3 functions contributed to most by those listed in the byline were assessing the quality of included studies (83%), interpreting data (82%), and abstracting data from included studies (77%). Conclusions  A substantial proportion of reviews had evidence of honorary and ghost authorship. The Cochrane editorial teams contributed to most Cochrane reviews.      

What Are the Factors Determining Authorship and the Order of the Authors' Names?: A Study Among Authors of the Nederlands Tijdschrift voor Geneeskunde (Dutch Journal of Medicine)

Context.— Although criteria justifying authorship of scientific medical articles have been formulated, it is not well known how authorship is established in practice. Objectives.— To assess the criteria for authorship used by authors of original articles in Nederlands Tijdschrift voor Geneeskunde (NTVG, the Dutch Journal of Medicine), and to determine whether the criteria for authorship of the International Committee of Medical Journal Editors (ICMJE) are known and applied. Design.— Survey questionnaire. Setting.— Editorial office of the NTVG. Participants.— All 450 authors of 115 original articles published in 1995. Main Outcome Measures.— Author's contribution to study design, material, collection of data, statistics, and writing. Results.— Of 362 forms returned, 352 could be analyzed (78.2% response rate). The 5 questions most frequently answered affirmatively were ICMJE criteria: critical reading (86.1% of the authors), approval of the final version (84.7%), study design (74.7%), study conception (64.2%), and revision (63.4%). Authors rated their contribution 2 points higher than did their coauthors. Interestingly, 64% of the respondents met the ICMJE criteria, although 60% of the respondents did not know them. Conclusion.— Authorship was mostly in accordance with ICMJE criteria although many authors were not familiar with them.      

Extrapolation of correlation between 2 variables in 4 general medical journals

Context  An estimated correlation between 2 variables is valid only within the range of observed data. Extrapolation is risky and should be performed with caution. Methods  To assess the prevalence of problems with data extrapolation in the medical literature, all articles published from January through June 2000 in BMJ, JAMA, The Lancet, and The New England Journal of Medicine (NEJM) were reviewed manually. Articles containing at least 1 scatterplot with raw data and a corresponding fitted regression line were included in the analysis. Articles were considered to involve extrapolation if they contained at least 1 fitted line beyond the observed data in any scatter plot. Results  A total of 178 articles presenting at least 1 scatterplot were identified. Among them, 37 articles (21%) (5 from BMJ, 7 from JAMA, 23 from The Lancet, and 2 from NEJM) were included. Twenty-two articles (59% [95% confidence interval, 42%-75%]) from all 4 journals involved extrapolation. None changed the line type to indicate extrapolation. Four articles (11%) contained a plot in which the fitted line reached unreasonable or meaningless values. Three articles (8%) stated explicit conclusions about values outside the range of the observed data. Conclusions  A high proportion of the articles analyzed from all 4 weekly general medical journals involved extrapolation without indication. Researchers, reviewers, and editors should be aware of this problem and work to eliminate it.      

Accuracy of data in abstracts of published research articles

Context  The section of a research article most likely to be read is the abstract, and therefore it is particularly important that the abstract reflect the article faithfully. Objective  To assess abstracts accompanying research articles published in 6 medical journals with respect to whether data in the abstract could be verified in the article itself. Design  Analysis of simple random samples of 44 articles and their accompanying abstracts published during 1 year (July 1, 1996-June 30, 1997) in each of 5 major general medical journals (Annals of Internal Medicine, BMJ, JAMA, Lancet, and New England Journal of Medicine) and a consecutive sample of 44 articles published during 15 months (July 1, 1996-August 15, 1997) in the CMAJ. Main Outcome Measure  Abstracts were considered deficient if they contained data that were either inconsistent with corresponding data in the article's body (including tables and figures) or not found in the body at all. Results  The proportion of deficient abstracts varied widely (18%-68%) and to a statistically significant degree (P<.001) among the 6 journals studied. Conclusions  Data in the abstract that are inconsistent with or absent from the article's body are common, even in large-circulation general medical journals.      

Reporting number needed to treat and absolute risk reduction in randomized controlled trials

Context  Ongoing efforts to improve the quality of reporting for randomized controlled trials (RCTs) include the Consolidated Standards of Reporting Trials (CONSORT) statement. We examined the frequency of explicit reporting of the number needed to treat (NNT) and the absolute risk reduction (ARR) in RCTs. Methods  Five frequently cited journals were investigated: Annals of Internal Medicine, BMJ, JAMA, The Lancet, and the New England Journal of Medicine. For each journal, 4 years were evaluated: 1989, 1992, 1995, and 1998. All issues of each journal for each year were reviewed manually. Eligible articles were those in which an RCT was conducted on the use of a medication showing a significant treatment effect. Elements abstracted from each eligible article were the condition investigated, event being treated or prevented, intervention, study results, and reporting methods (relative risk reduction, NNT, and ARR). Results  Of 359 eligible articles, NNT was reported in 8 articles. Six of the 8 studies were from 1998. Absolute risk reduction was reported in 18 articles, 10 of which were from 1998. Conclusions  Despite CONSORT recommendations, few authors expressed their findings in terms of NNT or ARR. Consideration should be given to including these values in reports of RCTs.      

Discussion Sections in Reports of Controlled Trials Published in General Medical Journals: Islands in Search of Continents?

Context.— Several journals have adopted the Consolidated Standards of Reporting Trials (CONSORT) recommendations to make assessment of the quality of randomized controlled trials (RCTs) easier. One of these recommendations is that the trial's results be discussed in light of the totality of the available evidence. Objective.— To assess the extent to which reports of RCTs published in 5 general medical journals have discussed new results in light of all available evidence. Design.— Assessment of the discussion sections in all 26 reports of RCTs published during May 1997 in Annals of Internal Medicine, BMJ, JAMA, The Lancet, and The New England Journal of Medicine. Main Outcome Measure.— The inclusion or mention of a systematic review in the discussion section of each article. Results.— In only 2 articles were the RCT's results discussed in the context of an updated systematic review of earlier trials. In a further 4 articles, references were made to relevant systematic reviews, but no attempts were made to integrate the results of the new trials in updated versions of these reviews. One article was probably the first published trial to address the question studied. The remaining 19 articles included no evidence that any systematic attempt had been made to set the reported trial's results in the context of previous trials. Conclusion.— There is little evidence that journals have adequately implemented the CONSORT recommendation that results of an RCT be discussed in light of the totality of the available evidence.      

Masking Author Identity in Peer Review: What Factors Influence Masking Success?

Context.— In a previous study, we found that masking success was higher at a journal that masked reviewers to author identity. We hypothesized that masking policy or other factors could be associated with masking success. Objectives.— To evaluate differences in success of masking reviewers to author identity at 7 biomedical journals and to identify factors associated with these differences. Design.— Written questionnaire. Participants.— Reviewers at 3 journals with a long-standing policy of masking author identity (Annals of Emergency Medicine, Epidemiology, and Journal of the American Geriatrics Society) and 4 journals without a policy of masking author identity (Annals of Internal Medicine, JAMA, Obstetrics & Gynecology, and Ophthalmology). Main Outcome Measures.— Masking success (percentage of reviewers successfully masked) and reviewer characteristics associated with masking. Results.— There was no significant difference in masking success between journals with a policy of masking (60%) and those without (58%) (P=.92). We found no association between masking success and a policy of masking when adjusted for the reviewer characteristics of age, sex, years of reviewing experience, number of articles published, number of articles reviewed, percentage of time spent in research, editorial experience, or academic rank (odds ratio [OR], 1.3; 95% confidence interval [CI], 0.64-2.8; P=.43). In multivariable analysis of reviewer characteristics, reviewers spending a greater percentage of time in research, the only significant predictor of masking success, were less likely to be successfully masked (OR, 1.01; 95% CI, 1.00-1.02) (P=.04). Conclusions.— Masking success appears unrelated to a journal policy of masking, but is associated with reviewers' research experience and could be affected by other characteristics. Using reviewers with less research and reviewing experience might increase masking success, but the effect on review quality is unknown.      

Press Releases of Science Journal Articles and Subsequent Newspaper Stories on the Same Topic

Context.— Scientific journals issue press releases to disseminate scientific news about articles they publish. Objective.— To assess whether press releases about journal articles were associated with publication of subsequent newspaper stories. Design.— Retrospective content analysis of newspaper stories, journal press releases, and journal tables of contents. From December 1, 1996, to February 28, 1997, press releases and tables of contents were collected from BMJ, Nature, Science, and The Lancet, along with newspaper stories on scientific research published in The New York Times (United States), Le Figaro and Le Monde (France), El País and La Vanguardia (Spain), La Repubblica (Italy), and the International Herald Tribune. Main Outcome Measurements.— Number of newspaper stories that contained reference to articles appearing in the 4 scientific journals, number of newspaper stories that referred to journal articles described in press releases, and the order in which journal articles were mentioned in press releases. Results.— Of the 1060 newspaper stories analyzed, 142 referred to journal articles; of these, 119 (84%) referred to articles mentioned in press releases and 23 (16%) referred to journal articles not mentioned in press releases (comparison of proportions, P =.03). Articles described first or second were referenced in more newspapers than articles described later in the press release (P=.01 by ² analysis). Conclusions.— Journal articles described in press releases, in particular those described first or second in the press release, are associated with the subsequent publication of newspaper stories on the same topic.      

Is There a Sex Bias in Choosing Editors?: Epidemiology Journals as an Example

Context.— Editors, authors, and reviewers are influential in shaping science. The careers of women in public health have received less scrutiny than those of women in medicine and other branches of science. The performance of women as editors, authors, and reviewers in epidemiology has not been previously studied. Objective.— To examine changes over time in the representation of women at the editorial level in US epidemiology journals compared with the proportion of women authors and reviewers. Design and Setting.— Cross-sectional study of 4 US epidemiology journals, American Journal of Epidemiology, Annals of Epidemiology, Epidemiology, and the Journal of Clinical Epidemiology (formerly the Journal of Chronic Diseases), for 1982, 1987, 1992, and 1994. Subjects.— Editors, authors, and reviewers for the selected years. Main Outcome Measures.— Sex of editors, authors, and reviewers. Results.— We identified 2415 reports associated with 8005 authors. One of 7 editors in chief was a woman, a position she shared with a man. For all journals, the proportion of editors who were women ranged from 5 (6.5%) of 77 in 1982 to 42 (16.3%) of 258 in 1994. Over all journals and all years, women comprised a higher proportion of authors (28.7% [2225/7743] ) compared with reviewers (26.7% [796/2982]) or editors (12.8% [89/696]). Conclusions.— Fewer women in public health hold editorial positions than are authors and reviewers. The reasons for this important discrepancy, including the possibility of a selection bias favoring men, should be further investigated.      

Comparison of review articles published in peer-reviewed and throwaway journals

Context  To compare the quality, presentation, readability, and clinical relevance of review articles published in peer-reviewed and "throwaway" journals. Methods  We reviewed articles that focused on the diagnosis or treatment of a medical condition published between January 1 and December 31, 1998, in the 5 leading peer-reviewed general medical journals and high-circulation throwaway journals. Reviewers independently assessed the methodologic and reporting quality, and evaluated each article's presentation and readability. Clinical relevance was evaluated independently by 6 physicians. Results  Of the 394 articles in our sample, 16 (4.1%) were peer-reviewed systematic reviews, 135 (34.3%) were peer-reviewed nonsystematic reviews, and 243 (61.7%) were nonsystematic reviews published in throwaway journals. The mean (SD) quality scores were highest for peer-reviewed articles (0.94 [0.09] for systematic reviews and 0.30 [0.19] for nonsystematic reviews) compared with throwaway journal articles (0.23 [0.03], F_2,391 = 280.8, P<.001). Throwaway journal articles used more tables (P = .02), figures (P = .01), photographs (P<.001), color (P<.001), and larger font sizes (P<.001) compared with peer-reviewed articles. Readability scores were more often in the college or higher range for peer-reviewed journals compared with the throwaway journal articles (104 [77.0%] vs 156 [64.2%]; P = .01). Peer-reviewed article titles were judged less relevant to clinical practice than throwaway journal article titles (P<.001). Conclusions  Although lower in methodologic and reporting quality, review articles published in throwaway journals have characteristics that appeal to physician readers.      

Problems with indexing and citation of articles with group authorship

Context  It is not known whether articles with group authorship (ie, with a research group name listed as the author) are difficult to identify or whether use of group authorship may lead to problems with citation. Methods  To examine ways in which reports of controlled trials with group authorship are indexed and citations counted in bibliographic databases, we conducted a cross-sectional study in January 2000. We identified 47 controlled trials funded by the National Eye Institute and 285 associated articles. Between January and August 2000, we searched PubMed and Science Citation Index (SCI) and recorded the citation practices for these articles. Our main outcome measures were ways in which trial reports were listed in PubMed and SCI and number of citations to each report by type of authorship. Results  Of the 285 published reports identified, 126 (44%) had group authorship, 109 (38%) had modified group authorship (listing individual names plus the name of the research group), and 50 (18%) had named authors only. In PubMed, no group authors were listed in the author field (per MEDLINE rules); in SCI, group-authored reports generally were incorrectly attributed (first name on investigator list [35.3%], first name on writing committee [25.5%], contact name [16.7%], anonymous [16.7%], and other [5.9%]). Using the SCI general search, we identified citations to 16.7% of group-authored reports, compared with citations to 96.9% of reports with modified group authorship and 93.9% of citations to reports with named authors only. Other systematic search methods found that more than 98% of group-authored reports actually had been cited and that group-authored reports were cited more than other reports. Conclusions  Indexing systems are not optimally adapted to group authorship. We recommend that indexing services change their practices to include group authors in the author field to help correct the problem.      

Heavy metal content of ayurvedic herbal medicine products

Context  Lead, mercury, and arsenic intoxication have been associated with the use of Ayurvedic herbal medicine product (HMPs). Objectives  To determine the prevalence and concentration of heavy metals in Ayurvedic HMPs manufactured in South Asia and sold in Boston-area stores and to compare estimated daily metal ingestion with regulatory standards. Design and Setting  Systematic search strategy to identify all stores 20 miles or less from Boston City Hall that sold Ayurvedic HMPs from South Asia by searching online Yellow Pages using the categories markets, supermarkets, and convenience stores, and business names containing the word India, Indian cities, and Indian words. An online national directory of Indian grocery stores, a South Asian community business directory, and a newspaper were also searched. We visited each store and purchased all unique Ayurvedic HMPs between April 25 and October 24, 2003. Main Outcome Measures  Concentrations (µg/g) of lead, mercury, and arsenic in each HMP as measured by x-ray fluorescence spectroscopy. Estimates of daily metal ingestion for adults and children estimated using manufacturers’ dosage recommendations with comparisons to US Pharmacopeia and US Environmental Protection Agency regulatory standards. Results  A total of 14 (20%) of 70 HMPs (95% confidence interval, 11%-31%) contained heavy metals: lead (n = 13; median concentration, 40 µg/g; range, 5-37 000), mercury (n = 6; median concentration, 20 225 µg/g; range, 28-104 000), and/or arsenic (n = 6; median concentration, 430 µg/g; range, 37-8130). If taken as recommended by the manufacturers, each of these 14 could result in heavy metal intakes above published regulatory standards. Conclusions  One of 5 Ayurvedic HMPs produced in South Asia and available in Boston South Asian grocery stores contains potentially harmful levels of lead, mercury, and/or arsenic. Users of Ayurvedic medicine may be at risk for heavy metal toxicity, and testing of Ayurvedic HMPs for toxic heavy metals should be mandatory.      

Efficacy of maintenance treatment approaches for childhood overweight: a randomized controlled trial

Context  No trials for childhood overweight have examined maintenance interventions to augment the effects of initial weight loss programs. Objectives  To determine the short-term and long-term efficacy of 2 distinct weight maintenance approaches vs no continued treatment control following standard family-based behavioral weight loss treatment for childhood overweight, and to examine children's social functioning as a moderator of outcome. Design, Setting, and Participants  A parallel-group, randomized controlled trial conducted between October 1999 and July 2004 in a university-based weight control clinic. Participants were 204 healthy 7- to 12-year-olds, 20% to 100% above median body mass index (BMI) for age and sex, with at least 1 overweight parent. Children enrolled in 5 months of weight loss treatment and 150 were randomized to 1 of 3 maintenance conditions. Follow-up assessments occurred immediately following maintenance treatments and 1 and 2 years following randomization. Interventions  Maintenance conditions included the control group or 4 months of behavioral skills maintenance (BSM) or social facilitation maintenance (SFM) treatment. Main Outcome Measures  BMI z score and percentage overweight. Results  Children receiving either BSM or SFM maintained relative weight significantly better than children assigned to the control group from randomization to postweight maintenance (P.01 for all; effect sizes d = 0.72-0.96; mean changes in BMI z scores = –0.04, –0.04, –0.05, and 0.05 for BSM alone, SFM alone, BSM and SFM together, and the control group, respectively). Active maintenance treatment efficacy relative to the control group declined during follow-up, but the effects of SFM alone (P = .03; d = 0.45; mean change in BMI z score = –0.24) and when analyzed together with BSM (P = .04; d = 0.38; mean change in BMI z score = –0.22) were significantly better than the control group (mean change in BMI z score = –0.06) when examining BMI z score outcomes from baseline to 2-year follow-up. Baseline child social problem scores moderated child relative weight change from baseline to 2-year follow-up, with low social problem children in SFM vs the control group having the best outcomes. Conclusions  The addition of maintenance-targeted treatment improves short-term efficacy of weight loss treatment for children relative to no maintenance treatment. However, the waning of effects over follow-up, although moderated by child initial social problems, suggests the need for the bolstering of future maintenance treatments to sustain effects. Trial Registration  clinicaltrials.gov Identifier: NCT00301197      

Respiratory symptoms, pulmonary function, and markers of inflammation among bar workers before and after a legislative ban on smoking in public places

Context  Scotland prohibited smoking in confined public places on March 26, 2006. Objective  To investigate the association of smoke-free legislation with symptoms, pulmonary function, and markers of inflammation of bar workers. Design, Setting, and Participants  This prospective observational study was conducted in Tayside, Scotland from February-June 2006. One hundred five nonasthmatic and asthmatic nonsmoking bar workers were initially enrolled, of whom 77 completed the study per protocol. Main Outcome Measures  Respiratory and sensory symptoms, spirometry measurements, serum cotinine levels, peripheral inflammatory cell count, asthma quality-of-life scores, and exhaled nitric oxide levels were evaluated before and after introduction of the smoking ban. Results  For the per-protocol analysis, the percentage of bar workers with respiratory and sensory symptoms decreased from 79.2% (n = 61) before the smoke-free policy to 53.2% (n = 41) (total change, –26%; 95% confidence interval [CI], –13.8% to –38.1%; P<.001) and 46.8% (n = 38) (–32.5%; 95% CI, –19.8% to –45.2%; P<.001) 1 and 2 months afterward. Forced expiratory volume in the first second increased from 96.6% predicted to 104.8% (change, 8.2%; 95% CI, 3.9% to 12.4%; P<.001) and then 101.7% (change, 5.1%; 95% CI, 2.1% to 8.0%; P = .002), and serum cotinine levels decreased from 5.15 ng/mL to 3.22 ng/mL (change, –1.93 ng/mL; 95% CI, –2.83 to –1.03 ng/mL; P<.001) and then 2.93 ng/mL (–2.22 ng/mL; 95% CI, –3.10 to –1.34 ng/mL; P<.001). The total white blood cell and neutrophil count was reduced from 7610 to 6980 cells/µL at 2 months (–630 cells/µL; 95% CI, –1010 to –260 cells/µL; P = .002) and from 4440 to 4030 cells/µL (–410 cells/µL; 95% CI, –740 to –90 cells/µL; P = .03), respectively. Asthmatic bar workers also had less airway inflammation, with a reduction in exhaled nitric oxide from 34.3 parts per billion (ppb) to 27.4 ppb 1 month after the ban (0.8-fold change; 95% CI, 0.67 to 0.96 ppb; P = .04), and Juniper quality-of-life scores increased from 80.2 to 87.5 points (7.3 points; 95% CI, 0.1 to 14.6 points; P = .049). Conclusions  Smoke-free legislation was associated with significant early improvements in symptoms, spirometry measurements, and systemic inflammation of bar workers. Asthmatic bar workers also had reduced airway inflammation and improved quality of life.      

Qualitative and quantitative measures of indexed health sciences electronic journals

Context  Little is known about qualitative and quantitative characteristics of indexed health sciences electronic journals (e-journals). Methods  To determine peer-review practices and qualitative and quantitative characteristics of different types of indexed health sciences e-journals, 3 types of e-journals indexed in MEDLINE were compared (type 1, completely electronic with no print counterpart; type 2, print and electronic versions with the same title but each publishing some unique content; and type 3, print and electronic versions containing equal content). Results  There were 13 type 1 journals, 16 type 2 journals, and 16 type 3 journals. Most journals in each category (85%-94%) imply or state the use of peer review. Significant differences (P<.05, analysis of variance) exist among the e-journals for the inclusion of complex types of publications (clinical trials, randomized controlled trials, meta-analyses, and practice guidelines) (15%-100%), editorials (0%-75%), letters to the editor (10%-88%), and case reports (17%-94%); the average number of items indexed in MEDLINE (22.5-544.5); and the number of complex publication types, editorials, letters, and case reports. Conclusions  Type 1 e-journals do not have the qualitative or quantitative complexity of traditional print journals. Although editors' statements on editorial peer review are similar, there are differences in number and type of materials included in the 3 different types of e-journals.      

Stereotactic radiosurgery plus whole-brain radiation therapy vs stereotactic radiosurgery alone for treatment of brain metastases: a randomized controlled trial

Context  In patients with brain metastases, it is unclear whether adding up-front whole-brain radiation therapy (WBRT) to stereotactic radiosurgery (SRS) has beneficial effects on mortality or neurologic function compared with SRS alone. Objective  To determine if WBRT combined with SRS results in improvements in survival, brain tumor control, functional preservation rate, and frequency of neurologic death. Design, Setting, and Patients  Randomized controlled trial of 132 patients with 1 to 4 brain metastases, each less than 3 cm in diameter, enrolled at 11 hospitals in Japan between October 1999 and December 2003. Interventions  Patients were randomly assigned to receive WBRT plus SRS (65 patients) or SRS alone (67 patients). Main Outcome Measures  The primary end point was overall survival; secondary end points were brain tumor recurrence, salvage brain treatment, functional preservation, toxic effects of radiation, and cause of death. Results  The median survival time and the 1-year actuarial survival rate were 7.5 months and 38.5% (95% confidence interval, 26.7%-50.3%) in the WBRT + SRS group and 8.0 months and 28.4% (95% confidence interval, 17.6%-39.2%) for SRS alone (P = .42). The 12-month brain tumor recurrence rate was 46.8% in the WBRT + SRS group and 76.4% for SRS alone group (P<.001). Salvage brain treatment was less frequently required in the WBRT + SRS group (n = 10) than with SRS alone (n = 29) (P<.001). Death was attributed to neurologic causes in 22.8% of patients in the WBRT + SRS group and in 19.3% of those treated with SRS alone (P = .64). There were no significant differences in systemic and neurologic functional preservation and toxic effects of radiation. Conclusions  Compared with SRS alone, the use of WBRT plus SRS did not improve survival for patients with 1 to 4 brain metastases, but intracranial relapse occurred considerably more frequently in those who did not receive WBRT. Consequently, salvage treatment is frequently required when up-front WBRT is not used. Trial Registration  umin.ac.jp/ctr Identifier: C000000412      

Persistence of contradicted claims in the literature

Athina Tatsioni, MD; Nikolaos G. Bonitsis, MD; John P. A. Ioannidis, MD

JAMA. 2007;298(21):2517-2526.

Context  Some research findings based on observational epidemiology are contradicted by randomized trials, but may nevertheless still be supported in some scientific circles.

Objectives  To evaluate the change over time in the content of citations for 2 highly cited epidemiological studies that proposed major cardiovascular benefits associated with vitamin E in 1993; and to understand how these benefits continued being defended in the literature, despite strong contradicting evidence from large randomized clinical trials (RCTs). To examine the generalizability of these findings, we also examined the extent of persistence of supporting citations for the highly cited and contradicted protective effects of beta-carotene on cancer and of estrogen on Alzheimer disease.

Data Sources  For vitamin E, we sampled articles published in 1997, 2001, and 2005 (before, early, and late after publication of refuting evidence) that referenced the highly cited epidemiological studies and separately sampled articles published in 2005 and referencing the major contradicting RCT (HOPE trial). We also sampled articles published in 2006 that referenced highly cited articles proposing benefits associated with beta-carotene for cancer (published in 1981 and contradicted long ago by RCTs in 1994-1996) and estrogen for Alzheimer disease (published in 1996 and contradicted recently by RCTs in 2004).

Data Extraction  The stance of the citing articles was rated as favorable, equivocal, and unfavorable to the intervention. We also recorded the range of counterarguments raised to defend effectiveness against contradicting evidence.

Results  For the 2 vitamin E epidemiological studies, even in 2005, 50% of citing articles remained favorable. A favorable stance was independently less likely in more recent articles, specifically in articles that also cited the HOPE trial (odds ratio for 2001, 0.05 [95% confidence interval, 0.01-0.19; P < .001] and the odds ratio for 2005, 0.06 [95% confidence interval, 0.02-0.24; P < .001], as compared with 1997), and in general/internal medicine vs specialty journals. Among articles citing the HOPE trial in 2005, 41.4% were unfavorable. In 2006, 62.5% of articles referencing the highly cited article that had proposed beta-carotene and 61.7% of those referencing the highly cited article on estrogen effectiveness were still favorable; 100% and 96%, respectively, of the citations appeared in specialty journals; and citations were significantly less favorable (P = .001 and P = .009, respectively) when the major contradicting trials were also mentioned. Counterarguments defending vitamin E or estrogen included diverse selection and information biases and genuine differences across studies in participants, interventions, cointerventions, and outcomes. Favorable citations to beta-carotene, long after evidence contradicted its effectiveness, did not consider the contradicting evidence.

Conclusion  Claims from highly cited observational studies persist and continue to be supported in the medical literature despite strong contradictory evidence from randomized trials.