Advertisement

What Happens to a Published Article if a Cited Article Is Corrected?

      Abstract

      When one considers that as many as 2.5 million scientific articles are published each year, it is likely that more than a few contain errors. Probably, most go undetected. In theory, scientific literature is self-correcting, and the truth will eventually be revealed. However, to maintain the integrity of our literature, it is best to correct errors. Fortunately, when it comes to an errant citation, most scientific citations provide background, and errors in background citations should not change the conclusion of a study. However, for systematic reviews that quantitatively synthesize published research findings in a meta-analysis, an error in (or retraction of) an included citation will affect the study results. Such errors require correction, revision of the meta-analysis, and electronic attachment of the notation to the publication.
      What happens to a published article if a cited article is corrected or retracted? Guidance regarding this question is limited, but Google Search directs us to the Retraction Watch website where an article titled, “What Should You Do if a Paper You’ve Cited Is Later Retracted?”
      What should you do if a paper you’ve cited is later retracted. Reaction Watch.
      shares advice from board members of the Retraction Watch parent organization, The Center for Scientific Integrity. We found this reference helpful, and because it is presented as a discussion or “blog,” it allows us to generate our own opinions.

      What Happened at Arthroscopy?

      What happened at Arthroscopy? Here’s the story: Your editors received a Letter to the Editor
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      concerning the meta-analysis, “Operative Versus Nonoperative Treatment of Femoroacetabular Impingement Syndrome” by Dwyer, Whelan, Shah, Ajrawat, Hoit, and Chahal.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis of short-term outcomes.
      This study, which notably won the Arthroscopy Journal 2020 Systematic Review/Meta-analysis Research Excellence Award, showed that “patients with FAI syndrome treated with hip arthroscopy have statistically superior hip-related outcomes in the short term compared with those treated with physical therapy alone.” To be fair, in an Editorial Commentary, Kemp wrote, “Hip arthroscopy should not be a first-line treatment but can be necessary in cases in which high-quality, exercise-based nonsurgical treatment options have been exhausted.”
      • Kemp J.L.
      Editorial commentary: A commentary on a meta-analysis of short-term outcomes.
      We conclude that patients with femoroacetabular impingement should undergo a trial of nonoperative management, and if physical therapy fails, surgical treatment is likely to result in a positive patient outcome and a better outcome than after rehabilitation.

      About That Letter

      Now about that letter…
      Ishøi et al.,
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      a group of attentive Danish research scientists, discovered that one of the studies cited in Dwyer’s meta-analysis contained an error. In that study published in BMJ, “Arthroscopic Hip Surgery Compared With Physiotherapy and Activity Modification for the Treatment of Symptomatic Femoroacetabular Impingement: Multicentre Randomised Controlled Trial” by Palmer et al.,
      • Palmer A.J.R.
      • Ayyar G.V.
      • Fernquest S.
      • et al.
      Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: Multicentre randomised controlled trial.
      it happens that Palmer et al. improperly measured and reported the International Hip Outcome Tool-33 on a visual analog scale using centimeters rather than millimeters.
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      One thing leads to another. Our award-winning researchers, Dwyer et al.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis of short-term outcomes.
      not only cited Palmer et al.
      • Palmer A.J.R.
      • Ayyar G.V.
      • Fernquest S.
      • et al.
      Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: Multicentre randomised controlled trial.
      but quantitatively synthesized the results of the BMJ article in their meta-analysis. Thus, correction of the BMJ article changed the results of Dwyer et al.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis of short-term outcomes.
      To paraphrase our Journal Board of Trustees Chairman, Nick Sgaglione, “You can’t make this stuff up!”

      The Cited Article Was Corrected

      In the end, the cited article was corrected in an Erratum in BMJ.
      Rapid responses to “Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: multicentre randomised controlled trial.”.
      This correction changed the results of Dwyer et al.’s meta-analysis.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Author reply to “can’t see the right forest plot for the wrong trees!”.
      Fortunately, despite the error in the BMJ article, the conclusion of our award-winning meta-analysis did not change. In fact, after: (1) Palmer et al. corrected their International Hip Outcome Tool-33 visual analog scale from millimeters to centimeters
      Rapid responses to “Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: multicentre randomised controlled trial.”.
      ; and then (2) our letter authors pointed out the error
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      ; and then (3) our meta-analysis authors updated their quantitative synthesis using the proper data
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Author reply to “can’t see the right forest plot for the wrong trees!”.
      ; (4) it turns out that the conclusion of Dwyer et al. that the treatment effect favoring operative management over nonoperative management of FAI was even more strongly favored.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Author reply to “can’t see the right forest plot for the wrong trees!”.
      Like Nick said, “You can’t make this stuff up!”

      An Editorial Decision

      After due diligence,
      What should you do if a paper you’ve cited is later retracted. Reaction Watch.
      collaborative discussion, and thoughtful consideration, your editors decided that since the conclusion of the meta-analysis remained unchanged, a formal correction calling attention to the Dwyer et al. author reply
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Author reply to “can’t see the right forest plot for the wrong trees!”.
      to the Ishøi letter
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      was sufficient notice of the update, bearing in mind that the letter,
      • Ishøi L.
      • Nielsen M.F.
      • Hölmich P.
      • Thorborg K.
      Can’t see the right forest plot for the wrong trees!.
      the author reply,
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Author reply to “can’t see the right forest plot for the wrong trees!”.
      Kemp’s reply,
      • Kemp J.
      Regarding “Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis.”.
      and the correction (published in this issue) will be electronically attached to the meta-analysis in perpetuity.

      Good News, Bad News

      It is likely, when one considers that as many as 2.5 million scientific articles are published each year,
      • Boon S.
      21st century science overload. CDN SciencePub.com.
      that more than a few published scientific articles contain errors. Probably, most go undetected. The good news is that in theory, scientific literature is “self-correcting.” In other words, since additional research continues to be conducted, the truth is eventually revealed.
      How is science self correcting? Quora.com.
      The bad news is that some scholars disagree with this theory.
      • Engber D.
      Is science broken? Slate.com.
      As a result, to maintain the integrity of our literature, errors require correction.
      International Committee of Journal Medical Editors. “Recommendations: Corrections, Retractions, Republications and Version Control.”.
      ,
      • Linton J.
      All journals need to correct errors.
      More good news—most scientific reference citations provide background or context. Errors in background citations should not change the conclusion of a study.
      What should you do if a paper you’ve cited is later retracted. Reaction Watch.
      More bad news—“for systematic reviews that combine study findings using meta-analysis,”
      What should you do if a paper you’ve cited is later retracted. Reaction Watch.
      an error in an included citation will affect the study results. Such errors require correction, revision of the meta-analysis, and electronic attachment of the notation to the publication.

      Conclusions

      What happens to a published article if a cited article is corrected? To be clear, Dwyer et al.
      • Dwyer T.
      • Whelan D.
      • Shah P.S.
      • Ajrawat P.
      • Hoit G.
      • Chahal J.
      Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis of short-term outcomes.
      made no error. A study they cited
      • Palmer A.J.R.
      • Ayyar G.V.
      • Fernquest S.
      • et al.
      Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: Multicentre randomised controlled trial.
      was errant. The International Committee of Journal Medical Editors Recommendations
      International Committee of Journal Medical Editors. “Recommendations: Corrections, Retractions, Republications and Version Control.”.
      provide extensive guidance on how to correct errors in the primary article, but guidance regarding our specific question is limited.
      What should you do if a paper you’ve cited is later retracted. Reaction Watch.
      So what happens to a published article if a cited article is corrected or retracted? As far as we’re concerned, we prefer to never again be tasked with this question. Rather, we implore authors to do their best, and get it right the first time.
      You can’t make this stuff up!

      References

      1. What should you do if a paper you’ve cited is later retracted. Reaction Watch.
        • Ishøi L.
        • Nielsen M.F.
        • Hölmich P.
        • Thorborg K.
        Can’t see the right forest plot for the wrong trees!.
        Arthroscopy. 2020; 36: 2787-2789
        • Dwyer T.
        • Whelan D.
        • Shah P.S.
        • Ajrawat P.
        • Hoit G.
        • Chahal J.
        Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis of short-term outcomes.
        Arthroscopy. 2020; 36: 263-273
        • Kemp J.L.
        Editorial commentary: A commentary on a meta-analysis of short-term outcomes.
        Arthroscopy. 2020; 36: 274-276
        • Palmer A.J.R.
        • Ayyar G.V.
        • Fernquest S.
        • et al.
        Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: Multicentre randomised controlled trial.
        BMJ. 2019; 364: 1185
      2. Rapid responses to “Arthroscopic hip surgery compared with physiotherapy and activity modification for the treatment of symptomatic femoroacetabular impingement: multicentre randomised controlled trial.”.
        BMJ. 2019; 364: L185
        • Dwyer T.
        • Whelan D.
        • Shah P.S.
        • Ajrawat P.
        • Hoit G.
        • Chahal J.
        Author reply to “can’t see the right forest plot for the wrong trees!”.
        Arthroscopy. 2020; 36: 2789-2790
        • Kemp J.
        Regarding “Operative versus nonoperative treatment of femoroacetabular impingement syndrome: A meta-analysis.”.
        Arthroscopy. 2020; 36: 2790
        • Boon S.
        21st century science overload. CDN SciencePub.com.
      3. How is science self correcting? Quora.com.
        • Engber D.
        Is science broken? Slate.com.
      4. International Committee of Journal Medical Editors. “Recommendations: Corrections, Retractions, Republications and Version Control.”.
        • Linton J.
        All journals need to correct errors.
        Nature. 2013; 504: 33