You can see those pictures in any magazine.But what's the use of looking when you don't know what they mean.-Elvis CostelloWe think the New England Journal of Medicine got it wrong last year. According to the Journal of the American Medical Association,1Boutron I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar “Accurate presentation of the results of a randomized controlled trial is the cornerstone of the dissemination of the results and their implementation in clinical practice … . However, investigators usually have broad latitude in writing their articles; they can choose which data to report and how to report them. Consequently, scientific articles are not simply reports of facts, and authors have many opportunities to consciously or subconsciously shape the impression of their results for readers, that is, to add ‘spin’ to their scientific report. Spin can be defined as specific reporting that could distort the interpretation of results and mislead readers. The use of spin in scientific writing can result from ignorance of the scientific issue, unconscious bias, or willful intent to deceive.”In other words, with apologies to Elvis Costello, what's the use of looking at data if you don't know what the study results mean?Apples Versus Apples Plus OrangesOur specific concern is with the New England Journal article, “A Randomized Trial of Treatment for Acute Anterior Cruciate Ligament Tears.”2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar To be frank, we believe that, although the article should be carefully read, its conclusion should be rejected. The authors performed a carefully planned randomized controlled trial of early (within 10 weeks of injury) or delayed (later after symptomatic instability) reconstructive surgery for acute ACL tears in an active Swedish population, aged 18 to 35 years. This is a well-done Level I study that used the validated, patient-centric, Knee Injury and Osteoarthritis Outcome Score (KOOS) as the primary outcome measure. Two years after injury, there was no difference in the KOOS scores between the groups, but in fact, 39% of the patients in the delayed-surgery group had gone on to have ACL reconstruction due to instability. To be clear, 100% of the “surgery group” had surgery, but 39% of the “nonsurgical group” had surgery. The KOOS scores are limited by selection bias. The authors are not comparing apples with oranges. Rather, in our opinion, the authors are comparing apples with apples plus oranges.Jogging Is Not a Cutting SportThe article in question2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar manifests “spin”1Boutron I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar with regard to analysis of secondary outcome measures. Unfortunately, the study authors gloss over the a of to in fact, the between the study The in the surgery group was to to those in the group on limited to on the 39% in the group surgery and had is not a to is an outcome and it is to to the between to in are not that the study that the ACL patients had a R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar concern is that the authors the of ACL treatment in a conclusion that could active with ACL tears to the of the and early of is to from the of the and authors to the of in with ACL et anterior cruciate ligament A treatment J Med. PubMed Scopus Google Scholar how it is for to be to to their if a treatment for ACL is the of the patients ACL reconstruction of as a result of Is an believe that to is an outcome for patients with a However, we that is the of a the New England Journal of Medicine the of patients was in the and a of the patients not for the of patients tears at the of injury, but with the delayed-surgery of patients tears at the of injury, but by the of the of the had was a difference of patients the in the group that and the that in the delayed-surgery of the active patients with ACL tears by group that have by early the is that is of the of and of of the that is a and primary of and of at after ACL of and on and after anterior cruciate ligament A J Med. Google be authors of the New England Journal article have that, with surgery, patients of the ACL have an of of the L.S. Roos E.M. The of anterior cruciate ligament and J Med. PubMed Scopus Google Scholar However, the of the American Journal of Medicine Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar that of after anterior cruciate ligament reconstruction by have The of for with anterior cruciate ligament and a of for with Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar the study in the of Osteoarthritis after Anterior Cruciate Ligament that and in patients with anterior cruciate ligament is S. P. J. in the of after anterior cruciate ligament J Med. 2010; PubMed Scopus Google Scholar and of Knee Osteoarthritis Anterior Cruciate Ligament A 10 to of of was for those reconstruction with and I. et and of after anterior cruciate ligament A study with 10 to years of J Med. 2010; PubMed Scopus Google that the to the of is the of The New England Journal article patients randomized to the group have a of of their at a The of is to from the of the and authors to the of in their patients with ACL clinical that many of patients or to their with symptomatic and patients to a of of their the be the conclusion of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar is not the spin that is The article ACL reconstruction not result in outcome at However, the selection in their we in the article and of Randomized for I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar that we have a of the of the KOOS scores be in the of the we KOOS scores be in the of the of and the of The article that the ACL could be R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar This is spin and is the patients surgery at and we that of the surgery in the delayed ACL R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar results in and of interpretation of the New England Journal of Medicine article results to the conclusion that after ACL be of but we know what we and we after ACL the of You with is the conclusion I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar to and to and report and there is in the authors of article for the to a controlled R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google and analysis is to the of study what's the use of for is not the to in fact, the New England Journal of Medicine a analysis by our from the Is early reconstruction for anterior cruciate ligament Engl J Med. 2010; 363: PubMed Scopus Google Scholar the KOOS is a of and the at years not the of the or the of of instability. in the of ACL are of with to the to to and the of or of early delayed ACL reconstruction have that delayed surgery is with a of to the the or the of the outcome not be due to ACL but the of or which to … . between the treatment that The group to delayed reconstruction the group to early be by the of surgery patients early ACL a concern is that have to the the of or clinical was in the delayed reconstruction not by the primary they the with or ACL that, with the New England Journal of Medicine with what we is used to the Unfortunately, we and not have a the authors of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar a not with an ACL surgery. Unfortunately, a it is not to which ACL patients surgery their of surgery our is that the of the for the of surgery for patients with ACL the the don't and not our primary we know is that treatment is a of for ACL in active patients to after ACL we know is that our as is to what is for our To we with our patients with ACL the and of of treatment and and the active that, surgery if they for treatment of ACL injury, they on have to their to on and that they have an of of or of their in the which their of over the 10 years and You can see those pictures in any magazine.But what's the use of looking when you don't know what they mean.-Elvis think the New England Journal of Medicine got it wrong last year. According to the Journal of the American Medical Association,1Boutron I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar “Accurate presentation of the results of a randomized controlled trial is the cornerstone of the dissemination of the results and their implementation in clinical practice … . However, investigators usually have broad latitude in writing their articles; they can choose which data to report and how to report them. Consequently, scientific articles are not simply reports of facts, and authors have many opportunities to consciously or subconsciously shape the impression of their results for readers, that is, to add ‘spin’ to their scientific report. Spin can be defined as specific reporting that could distort the interpretation of results and mislead readers. The use of spin in scientific writing can result from ignorance of the scientific issue, unconscious bias, or willful intent to other words, with apologies to Elvis Costello, what's the use of looking at data if you don't know what the study results Apples Versus Apples Plus OrangesOur specific concern is with the New England Journal article, “A Randomized Trial of Treatment for Acute Anterior Cruciate Ligament Tears.”2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar To be frank, we believe that, although the article should be carefully read, its conclusion should be rejected. The authors performed a carefully planned randomized controlled trial of early (within 10 weeks of injury) or delayed (later after symptomatic instability) reconstructive surgery for acute ACL tears in an active Swedish population, aged 18 to 35 years. This is a well-done Level I study that used the validated, patient-centric, Knee Injury and Osteoarthritis Outcome Score (KOOS) as the primary outcome measure. Two years after injury, there was no difference in the KOOS scores between the groups, but in fact, 39% of the patients in the delayed-surgery group had gone on to have ACL reconstruction due to instability. To be clear, 100% of the “surgery group” had surgery, but 39% of the “nonsurgical group” had surgery. The KOOS scores are limited by selection bias. The authors are not comparing apples with oranges. Rather, in our opinion, the authors are comparing apples with apples plus oranges. specific concern is with the New England Journal article, “A Randomized Trial of Treatment for Acute Anterior Cruciate Ligament Tears.”2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar To be frank, we believe that, although the article should be carefully read, its conclusion should be rejected. The authors performed a carefully planned randomized controlled trial of early (within 10 weeks of injury) or delayed (later after symptomatic instability) reconstructive surgery for acute ACL tears in an active Swedish population, aged 18 to 35 years. This is a well-done Level I study that used the validated, patient-centric, Knee Injury and Osteoarthritis Outcome Score (KOOS) as the primary outcome measure. Two years after injury, there was no difference in the KOOS scores between the groups, but in fact, 39% of the patients in the delayed-surgery group had gone on to have ACL reconstruction due to instability. To be clear, 100% of the “surgery group” had surgery, but 39% of the “nonsurgical group” had surgery. The KOOS scores are limited by selection bias. The authors are not comparing apples with oranges. Rather, in our opinion, the authors are comparing apples with apples plus oranges. Is Not a Cutting SportThe article in question2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar manifests “spin”1Boutron I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar with regard to analysis of secondary outcome measures. Unfortunately, the study authors gloss over the a of to in fact, the between the study The in the surgery group was to to those in the group on limited to on the 39% in the group surgery and had is not a to is an outcome and it is to to the between The article in question2Frobell R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar manifests “spin”1Boutron I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar with regard to analysis of secondary outcome measures. Unfortunately, the study authors gloss over the a of to in fact, the between the study The in the surgery group was to to those in the group on limited to on the 39% in the group surgery and had is not a to is an outcome and it is to to the between to in are not that the study that the ACL patients had a R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar concern is that the authors the of ACL treatment in a conclusion that could active with ACL tears to the of the and early of is to from the of the and authors to the of in with ACL et anterior cruciate ligament A treatment J Med. PubMed Scopus Google Scholar how it is for to be to to their if a treatment for ACL is the of the patients ACL reconstruction of as a result of are not that the study that the ACL patients had a R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar concern is that the authors the of ACL treatment in a conclusion that could active with ACL tears to the of the and early The of is to from the of the and authors to the of in with ACL et anterior cruciate ligament A treatment J Med. PubMed Scopus Google Scholar how it is for to be to to their if a treatment for ACL is the of the patients ACL reconstruction of as a result of Is an believe that to is an outcome for patients with a However, we that is the of a the New England Journal of Medicine the of patients was in the and a of the patients not for the of patients tears at the of injury, but with the delayed-surgery of patients tears at the of injury, but by the of the of the had was a difference of patients the in the group that and the that in the delayed-surgery of the active patients with ACL tears by group that have by early believe that to is an outcome for patients with a However, we that is the of a the New England Journal of Medicine the of patients was in the and a of the patients not for the of patients tears at the of injury, but with the delayed-surgery of patients tears at the of injury, but by the of the of the had was a difference of patients the in the group that and the that in the delayed-surgery of the active patients with ACL tears by group that have by early the is that is of the of and of of the that is a and primary of and of at after ACL of and on and after anterior cruciate ligament A J Med. Google be authors of the New England Journal article have that, with surgery, patients of the ACL have an of of the L.S. Roos E.M. The of anterior cruciate ligament and J Med. PubMed Scopus Google Scholar However, the of the American Journal of Medicine Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar that of after anterior cruciate ligament reconstruction by have The of for with anterior cruciate ligament and a of for with Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar the study in the of Osteoarthritis after Anterior Cruciate Ligament that and in patients with anterior cruciate ligament is S. P. J. in the of after anterior cruciate ligament J Med. 2010; PubMed Scopus Google Scholar and of Knee Osteoarthritis Anterior Cruciate Ligament A 10 to of of was for those reconstruction with and I. et and of after anterior cruciate ligament A study with 10 to years of J Med. 2010; PubMed Scopus Google that the to the of is the of The New England Journal article patients randomized to the group have a of of their at a The of is to from the of the and authors to the of in their patients with ACL clinical that many of patients or to their with symptomatic and patients to a of of their the The is that is of the of and of of the that is a and primary of and of at after ACL of and on and after anterior cruciate ligament A J Med. Google Scholar To be authors of the New England Journal article have that, with surgery, patients of the ACL have an of of the L.S. Roos E.M. The of anterior cruciate ligament and J Med. PubMed Scopus Google Scholar However, the of the American Journal of Medicine Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar that of after anterior cruciate ligament reconstruction by have The of for with anterior cruciate ligament and a of for with Knee after anterior cruciate ligament J Med. PubMed Scopus Google Scholar the study in the of Osteoarthritis after Anterior Cruciate Ligament that and in patients with anterior cruciate ligament is S. P. J. in the of after anterior cruciate ligament J Med. 2010; PubMed Scopus Google Scholar and of Knee Osteoarthritis Anterior Cruciate Ligament A 10 to of of was for those reconstruction with and I. et and of after anterior cruciate ligament A study with 10 to years of J Med. 2010; PubMed Scopus Google Scholar The that the to the of is the of The New England Journal article patients randomized to the group have a of of their at a The of is to from the of the and authors to the of in their patients with ACL clinical that many of patients or to their with symptomatic and patients to a of of their the be the conclusion of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar is not the spin that is The article ACL reconstruction not result in outcome at However, the selection in their we in the article and of Randomized for I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar that we have a of the of the KOOS scores be in the of the we KOOS scores be in the of the of and the of The article that the ACL could be R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar This is spin and is the patients surgery at and we that of the surgery in the delayed ACL R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar results in and of interpretation of the New England Journal of Medicine article results to the conclusion that after ACL be of but we know what we and we after ACL To be the conclusion of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar is not the spin that is The article ACL reconstruction not result in outcome at However, the selection in their we in the article and of Randomized for I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar that we have a of the of the The KOOS scores be in the of the we KOOS scores be in the of the of and the of The article that the ACL could be R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar This is spin and is the patients surgery at and we that of the surgery in the delayed ACL R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar results in and of interpretation of the New England Journal of Medicine article results to the conclusion that after ACL be of but we know what we and we after ACL the of You with is the conclusion I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar to and to and report and there is in the authors of article for the to a controlled R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google and analysis is to the of study what's the use of A with is the conclusion I. Dutton S. Ravaud P. et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes.JAMA. 2010; 303: 2058-2064Crossref PubMed Scopus (480) Google Scholar to and to and report and there is in the authors of article for the to a controlled R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar However, and analysis is to the of study what's the use of for is not the to in fact, the New England Journal of Medicine a analysis by our from the Is early reconstruction for anterior cruciate ligament Engl J Med. 2010; 363: PubMed Scopus Google Scholar the KOOS is a of and the at years not the of the or the of of instability. in the of ACL are of with to the to to and the of or of early delayed ACL reconstruction have that delayed surgery is with a of to the the or the of the outcome not be due to ACL but the of or which to … . between the treatment that The group to delayed reconstruction the group to early be by the of surgery patients early ACL a concern is that have to the the of or clinical was in the delayed reconstruction not by the primary they the with or ACL that, with the New England Journal of Medicine with what we is not the to in fact, the New England Journal of Medicine a analysis by our from the Is early reconstruction for anterior cruciate ligament Engl J Med. 2010; 363: PubMed Scopus Google Scholar the KOOS is a of and the at years not the of the or the of of instability. in the of ACL are of with to the to to and the of or of early delayed ACL reconstruction have that delayed surgery is with a of to the the or the of the outcome not be due to ACL but the of or which to … . between the treatment that The group to delayed reconstruction the group to early be by the of surgery patients early ACL a concern is that have to the the of or clinical was in the delayed reconstruction not by the primary they the with or ACL that, with the New England Journal of Medicine with what we is used to the Unfortunately, we and not have a the authors of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar a not with an ACL surgery. Unfortunately, a it is not to which ACL patients surgery their of surgery our is that the of the for the of surgery for patients with ACL The is used to the Unfortunately, we and not have a the The authors of the New England Journal of Medicine R.B. Roos E.M. Roos H.P. Ranstram J. Lohmander L.S. A randomized trial of treatment for acute anterior cruciate ligament tears.N Engl J Med. 2010; 363: 331-342Crossref PubMed Scopus (553) Google Scholar a not with an ACL surgery. Unfortunately, a it is not to which ACL patients surgery their of surgery our is that the of the for the of surgery for patients with ACL the the don't and not our primary the don't and not our primary we know is that treatment is a of for ACL in active patients to after ACL we know is that our as is to what is for our To we with our patients with ACL the and of of treatment and and the active that, surgery if they for treatment of ACL injury, they on have to their to on and that they have an of of or of their in the which their of over the 10 years and we know is that treatment is a of for ACL in active patients to after ACL we know is that our as is to what is for our To we with our patients with ACL the and of of treatment and and the active that, surgery if they for treatment of ACL injury, they on have to their to on and that they have an of of or of their in the which their of over the 10 years and of writing the in the of Osteoarthritis after Knee Anterior Cruciate Ligament I the authors for the the of the article “A Randomized Trial of Treatment for Acute Anterior Cruciate Ligament in the New England Journal of should be rejected. to Anterior Cruciate Ligament as a of we the Osteoarthritis Knee Anterior Cruciate Ligament by and with the authors that scientific articles not reporting of but specific or reporting that can distort the results and mislead the readers. The a article in the New England Journal of and the of Level I study comparing surgery and for the treatment of anterior cruciate ligament
No takes yet. Share an insight, caveat, or question.
Richmond et al. (2011) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: