Letters
Corresponding Author: Simon Thom, MB,BS, MD, Imperial College London, 59 N Wharf Rd, London W2 1NY, England ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Thom reported receiving reimbursement from Dr Reddy’s laboratories for travel costs related to visiting India for the UMPIRE trial start-up meetings and for trial site visits. Dr Rodgers reported receiving grants from several research charities, national funding agencies, and generic pharmaceutical companies for research on cardiovascular FDC medications. In November 2012, the George Institute for Global Health obtained an exclusive global license for the FDCs used in the UMPIRE trial. 1. Lv J, Neal B, Ehteshami P, et al. Effects of intensive blood pressure lowering on cardiovascular and renal outcomes: a systematic review and meta-analysis. PLoS Med. 2012;9(8):e1001293. 2. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ. 2009;338:b1665. 3. Thompson AM, Hu T, Eshelbrenner CL, Reynolds K, He J, Bazzano LA. Antihypertensive treatment and secondary prevention of cardiovascular disease events among persons without hypertension: a meta-analysis. JAMA. 2011;305(9):913-922.
Alexis Descatha, MD, PhD Diane Godeau, MD Zakia Mediouni, MD Author Affiliations: Occupational Health Unit, Paris Teaching Hospital, Garches, France. Corresponding Author: Alexis Descatha, MD, PhD, Unité de Pathologie Professionnelle, CHU Poincaré, 104 Bd Poincaré, 92380 Garches, France (alexis [email protected]).
4. Baigent C, Blackwell L, Emberson J, et al; Cholesterol Treatment Trialists’ (CTT) Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010;376(9753):1670-1681.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Descatha reported being the editor in chief of the Archives des Maladies Professionnelles. No other disclosures were reported.
5. O’Keefe JH Jr, Cordain L, Harris WH, Moe RM, Vogel R. Optimal low-density lipoprotein is 50 to 70 mg/dL: lower is better and physiologically normal. J Am Coll Cardiol. 2004;43(11):2142-2146.
1. Hermans J, Luime JJ, Meuffels DE, Reijman M, Simel DL, Bierma-Zeinstra SMA. Does this patient with shoulder pain have rotator cuff disease? the Rational Clinical Examination systematic review. JAMA. 2013;310(8):837-847.
6. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353(5):487-497.
2. Hanchard NCA, Lenza M, Handoll HHG, Takwoingi Y. Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum that may accompany impingement. Cochrane Database Syst Rev. 2013;4:CD007427.
Physical Tests for Shoulder Disorders To the Editor The Rational Clinical Examination meta-analysis by Dr Hermans and colleagues1 on rotator cuff disease (RCD) was published at the same time as a Cochrane review2 concerning physical tests for shoulder disorders, including RCD. Although the authors of both reviews agreed on the inapplicability of results to a primary care or nonreferred population, they came to opposite conclusions. According to Hermans et al,1 “a positive painful arc test result and a positive external rotation resistance test result were the most accurate findings for detecting RCD, whereas the presence of a positive lag test (external or internal rotation) result was most accurate for diagnosis of a full-thickness rotator cuff tear.” The Cochrane review2 stated “The large body of literature revealed extreme diversity in the performance and interpretation of tests, which hinders synthesis of the evidence and/or clinical applicability.” Both systematic reviews were based on similar databases and used similar standards,3,4 but the inclusion criteria for studies were different. For example, the Cochrane review2 included only studies in English and those published during a shorter period (ending in February 2010 vs May 2013 for the Hermans et al1 review). Twenty-eight studies were included in the Hermans et al1 review, although their conclusions were primarily based on 5 with the highest quality evidence compared with 18 for RCD in the Cochrane review.2 Hermans et al1 based their conclusions for a number of tests on only 1 study. For example, the conclusions regarding internal rotation lag was based on only 1 study of 37 patients with a sensitivity of 97%, specificity of 83%, positive likelihood ra94
tio (LR) of 5.6, and a negative LR of 0.04 but with a confidence interval ranging from 0 to 0.58.5 The authors of the Cochrane review2 found sensitivities ranging from 0% to 100% and specificities from 43% to 97% (LRs were not calculated by the authors due to the small number of studies). Thus, we think that although there were discrepancies between the inclusion criteria, results, and conclusions of the 2 reviews, both actually show that none of the clinical tests studied are sufficient to confirm RCD without imaging.
3. Whiting PF, Weswood ME, Rutjes AWS, Reitsma JB, Bossuyt PNM, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Methodol. 2006;6:9. 4. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6(7):e1000100. 5. Grimes DA, Schulz KF. Refining clinical diagnosis with likelihood ratios. Lancet. 2005;365(9469):1500-1505.
In Reply All of the studies included in our meta-analysis were conducted by specialists with referred patients; even though we did not exclude primary care studies, there were none that met our inclusion criteria, which required high quality and low bias. We do not agree with Drs Descatha and colleagues that the results are inapplicable to a primary care or nonreferred population. We used the best available evidence on the clinical examination for RCD and discussed in our article how this evidence can be applied to a nonreferred population in primary care. In total, we included 28 studies. Seven studies published since 2010 were included, whereas the Cochrane review1 included no studies from later than 2009. Twelve studies included in our review overlapped with the 20 studies on the topic of RCD in the Cochrane review. The remaining studies were either not identified or not included in both reviews likely due to differences in search strategy and inclusion criteria. In our review, 5 of the 28 included studies were assigned a level of evidence of 1 or 2 and were eligible for summary measures because they represented the highest level of evidence.2 We based our conclusions on these 5 high-quality studies. Descatha and colleagues are correct that conclusions for a number of clinical tests were based on only 1 study. We pro-
JAMA January 1, 2014 Volume 311, Number 1
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 05/20/2015
jama.com
Letters
vided the width of the estimated LR (95% CI) to express the uncertainty. Knowing this, a reader can decide whether the evidence supports using the findings or not. The Cochrane review1 based their conclusions on 18 studies without creating summary measures. For the findings that were replicated on individual signs in our review, we calculated summary measures that suggest the well-known test of Hawkins, Neer, Gerber, and the empty can test are not of high diagnostic accuracy. In addition, in the online supplemental content for our review, we elaborated on verification bias and how it might affect the estimates of sensitivity, specificity, and the accompanying LRs. We showed that when verification bias exists in published studies, the findings conducted by specialists with a high positive LR ought to be the findings that a generalist tries to emulate because it is likely that a positive test will perform even better in a broader population. Based on the best available evidence, we recommend that generalist physicians develop proficiency in the 5 tests with the best LRs and the narrowest 95% confidence intervals. The results of these tests can be combined with the presented population prevalence of RCD. In the end, this supplies the generalist physician with practical means to estimate the posterior probability of disease without the necessity of using imaging for every patient. Job Hermans, MD, MSc Jolanda J. Luime, PhD Duncan E. Meuffels, MD, PhD Author Affiliations: Department of Orthopaedic Surgery, Erasmus M C University Medical Centre, Rotterdam, the Netherlands (Hermans); Department of Rheumatology, Erasmus M C University Medical Centre (Luime); Department of Orthopaedic Surgery, Erasmus M C University Medical Centre (Meuffels). Corresponding Author: Job Hermans, MD, MSc, Erasmus M C University Medical Centre Rotterdam, PO Box 2040, 3000 CA Rotterdam, the Netherlands ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
1. Hanchard NCA, Lenza M, Handoll HHG, Takwoingi Y. Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum that may accompany impingement. Cochrane Database Syst Rev. 2013;4:CD007427. 2. Simel DL, Keitz SA. Update: a primer on the precision and accuracy of the clinical examination. In: Simel DL, Rennie D, eds. The Rational Clinical Examination: Evidence-Based Clinical Diagnosis. New York, NY: McGraw-Hill; 2009.
CORRECTION Dosage and Algorithm Updated: In the Review entitled “Conjunctivitis: A Systematic Review of Diagnosis and Treatment” published in the October 23/30, 2013, issue of JAMA (2013;310[16]:1721-1729. doi:10.1001/jama.2013.280318), information was missing from Table 2 and Figure 2. In Table 2, under “Herpes simplex virus,” the dosage for oral acyclovir should have read “400 mg: 5×/d for 7-10 d.” In Figure 2, the algorithm for treating conjunctivitis now includes a box for viral conjunctivitis. This article was corrected online.
Guidelines for Letters Letters discussing a recent JAMA article should be submitted within 4 weeks of the article's publication in print. Letters received after 4 weeks will rarely be considered. Letters should not exceed 400 words of text and 5 references and may have no more than 3 authors. Letters reporting original research should not exceed 600 words of text and 6 references and may have no more than 5 authors. They may include up to 2 tables or figures but online supplementary material is not allowed. All letters should include a word count. Letters must not duplicate other material published or submitted for publication. Letters not meeting these specifications are generally not considered. Letters being considered for publication ordinarily will be sent to the authors of the JAMA article, who will be given the opportunity to reply. Letters will be published at the discretion of the editors and are subject to abridgement and editing. Further instructions can be found at http://jama.com/public /InstructionsForAuthors.aspx. A signed statement for authorship criteria and responsibility, financial disclosure, copyright transfer, and acknowledgment and the ICMJE Form for Disclosure of Potential Conflicts of Interest are required before publication. Letters should be submitted via the JAMA online submission and review system at http: //manuscripts.jama.com. For technical assistance, please contact [email protected]. Section Editor: Jody W. Zylke, MD, Senior Editor.
jama.com
JAMA January 1, 2014 Volume 311, Number 1
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 05/20/2015
95
Corresponding Author: Simon Thom, MB,BS, MD, Imperial College London, 59 N Wharf Rd, London W2 1NY, England ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Thom reported receiving reimbursement from Dr Reddy’s laboratories for travel costs related to visiting India for the UMPIRE trial start-up meetings and for trial site visits. Dr Rodgers reported receiving grants from several research charities, national funding agencies, and generic pharmaceutical companies for research on cardiovascular FDC medications. In November 2012, the George Institute for Global Health obtained an exclusive global license for the FDCs used in the UMPIRE trial. 1. Lv J, Neal B, Ehteshami P, et al. Effects of intensive blood pressure lowering on cardiovascular and renal outcomes: a systematic review and meta-analysis. PLoS Med. 2012;9(8):e1001293. 2. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ. 2009;338:b1665. 3. Thompson AM, Hu T, Eshelbrenner CL, Reynolds K, He J, Bazzano LA. Antihypertensive treatment and secondary prevention of cardiovascular disease events among persons without hypertension: a meta-analysis. JAMA. 2011;305(9):913-922.
Alexis Descatha, MD, PhD Diane Godeau, MD Zakia Mediouni, MD Author Affiliations: Occupational Health Unit, Paris Teaching Hospital, Garches, France. Corresponding Author: Alexis Descatha, MD, PhD, Unité de Pathologie Professionnelle, CHU Poincaré, 104 Bd Poincaré, 92380 Garches, France (alexis [email protected]).
4. Baigent C, Blackwell L, Emberson J, et al; Cholesterol Treatment Trialists’ (CTT) Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010;376(9753):1670-1681.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Descatha reported being the editor in chief of the Archives des Maladies Professionnelles. No other disclosures were reported.
5. O’Keefe JH Jr, Cordain L, Harris WH, Moe RM, Vogel R. Optimal low-density lipoprotein is 50 to 70 mg/dL: lower is better and physiologically normal. J Am Coll Cardiol. 2004;43(11):2142-2146.
1. Hermans J, Luime JJ, Meuffels DE, Reijman M, Simel DL, Bierma-Zeinstra SMA. Does this patient with shoulder pain have rotator cuff disease? the Rational Clinical Examination systematic review. JAMA. 2013;310(8):837-847.
6. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353(5):487-497.
2. Hanchard NCA, Lenza M, Handoll HHG, Takwoingi Y. Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum that may accompany impingement. Cochrane Database Syst Rev. 2013;4:CD007427.
Physical Tests for Shoulder Disorders To the Editor The Rational Clinical Examination meta-analysis by Dr Hermans and colleagues1 on rotator cuff disease (RCD) was published at the same time as a Cochrane review2 concerning physical tests for shoulder disorders, including RCD. Although the authors of both reviews agreed on the inapplicability of results to a primary care or nonreferred population, they came to opposite conclusions. According to Hermans et al,1 “a positive painful arc test result and a positive external rotation resistance test result were the most accurate findings for detecting RCD, whereas the presence of a positive lag test (external or internal rotation) result was most accurate for diagnosis of a full-thickness rotator cuff tear.” The Cochrane review2 stated “The large body of literature revealed extreme diversity in the performance and interpretation of tests, which hinders synthesis of the evidence and/or clinical applicability.” Both systematic reviews were based on similar databases and used similar standards,3,4 but the inclusion criteria for studies were different. For example, the Cochrane review2 included only studies in English and those published during a shorter period (ending in February 2010 vs May 2013 for the Hermans et al1 review). Twenty-eight studies were included in the Hermans et al1 review, although their conclusions were primarily based on 5 with the highest quality evidence compared with 18 for RCD in the Cochrane review.2 Hermans et al1 based their conclusions for a number of tests on only 1 study. For example, the conclusions regarding internal rotation lag was based on only 1 study of 37 patients with a sensitivity of 97%, specificity of 83%, positive likelihood ra94
tio (LR) of 5.6, and a negative LR of 0.04 but with a confidence interval ranging from 0 to 0.58.5 The authors of the Cochrane review2 found sensitivities ranging from 0% to 100% and specificities from 43% to 97% (LRs were not calculated by the authors due to the small number of studies). Thus, we think that although there were discrepancies between the inclusion criteria, results, and conclusions of the 2 reviews, both actually show that none of the clinical tests studied are sufficient to confirm RCD without imaging.
3. Whiting PF, Weswood ME, Rutjes AWS, Reitsma JB, Bossuyt PNM, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Methodol. 2006;6:9. 4. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6(7):e1000100. 5. Grimes DA, Schulz KF. Refining clinical diagnosis with likelihood ratios. Lancet. 2005;365(9469):1500-1505.
In Reply All of the studies included in our meta-analysis were conducted by specialists with referred patients; even though we did not exclude primary care studies, there were none that met our inclusion criteria, which required high quality and low bias. We do not agree with Drs Descatha and colleagues that the results are inapplicable to a primary care or nonreferred population. We used the best available evidence on the clinical examination for RCD and discussed in our article how this evidence can be applied to a nonreferred population in primary care. In total, we included 28 studies. Seven studies published since 2010 were included, whereas the Cochrane review1 included no studies from later than 2009. Twelve studies included in our review overlapped with the 20 studies on the topic of RCD in the Cochrane review. The remaining studies were either not identified or not included in both reviews likely due to differences in search strategy and inclusion criteria. In our review, 5 of the 28 included studies were assigned a level of evidence of 1 or 2 and were eligible for summary measures because they represented the highest level of evidence.2 We based our conclusions on these 5 high-quality studies. Descatha and colleagues are correct that conclusions for a number of clinical tests were based on only 1 study. We pro-
JAMA January 1, 2014 Volume 311, Number 1
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 05/20/2015
jama.com
Letters
vided the width of the estimated LR (95% CI) to express the uncertainty. Knowing this, a reader can decide whether the evidence supports using the findings or not. The Cochrane review1 based their conclusions on 18 studies without creating summary measures. For the findings that were replicated on individual signs in our review, we calculated summary measures that suggest the well-known test of Hawkins, Neer, Gerber, and the empty can test are not of high diagnostic accuracy. In addition, in the online supplemental content for our review, we elaborated on verification bias and how it might affect the estimates of sensitivity, specificity, and the accompanying LRs. We showed that when verification bias exists in published studies, the findings conducted by specialists with a high positive LR ought to be the findings that a generalist tries to emulate because it is likely that a positive test will perform even better in a broader population. Based on the best available evidence, we recommend that generalist physicians develop proficiency in the 5 tests with the best LRs and the narrowest 95% confidence intervals. The results of these tests can be combined with the presented population prevalence of RCD. In the end, this supplies the generalist physician with practical means to estimate the posterior probability of disease without the necessity of using imaging for every patient. Job Hermans, MD, MSc Jolanda J. Luime, PhD Duncan E. Meuffels, MD, PhD Author Affiliations: Department of Orthopaedic Surgery, Erasmus M C University Medical Centre, Rotterdam, the Netherlands (Hermans); Department of Rheumatology, Erasmus M C University Medical Centre (Luime); Department of Orthopaedic Surgery, Erasmus M C University Medical Centre (Meuffels). Corresponding Author: Job Hermans, MD, MSc, Erasmus M C University Medical Centre Rotterdam, PO Box 2040, 3000 CA Rotterdam, the Netherlands ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
1. Hanchard NCA, Lenza M, Handoll HHG, Takwoingi Y. Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum that may accompany impingement. Cochrane Database Syst Rev. 2013;4:CD007427. 2. Simel DL, Keitz SA. Update: a primer on the precision and accuracy of the clinical examination. In: Simel DL, Rennie D, eds. The Rational Clinical Examination: Evidence-Based Clinical Diagnosis. New York, NY: McGraw-Hill; 2009.
CORRECTION Dosage and Algorithm Updated: In the Review entitled “Conjunctivitis: A Systematic Review of Diagnosis and Treatment” published in the October 23/30, 2013, issue of JAMA (2013;310[16]:1721-1729. doi:10.1001/jama.2013.280318), information was missing from Table 2 and Figure 2. In Table 2, under “Herpes simplex virus,” the dosage for oral acyclovir should have read “400 mg: 5×/d for 7-10 d.” In Figure 2, the algorithm for treating conjunctivitis now includes a box for viral conjunctivitis. This article was corrected online.
Guidelines for Letters Letters discussing a recent JAMA article should be submitted within 4 weeks of the article's publication in print. Letters received after 4 weeks will rarely be considered. Letters should not exceed 400 words of text and 5 references and may have no more than 3 authors. Letters reporting original research should not exceed 600 words of text and 6 references and may have no more than 5 authors. They may include up to 2 tables or figures but online supplementary material is not allowed. All letters should include a word count. Letters must not duplicate other material published or submitted for publication. Letters not meeting these specifications are generally not considered. Letters being considered for publication ordinarily will be sent to the authors of the JAMA article, who will be given the opportunity to reply. Letters will be published at the discretion of the editors and are subject to abridgement and editing. Further instructions can be found at http://jama.com/public /InstructionsForAuthors.aspx. A signed statement for authorship criteria and responsibility, financial disclosure, copyright transfer, and acknowledgment and the ICMJE Form for Disclosure of Potential Conflicts of Interest are required before publication. Letters should be submitted via the JAMA online submission and review system at http: //manuscripts.jama.com. For technical assistance, please contact [email protected]. Section Editor: Jody W. Zylke, MD, Senior Editor.
jama.com
JAMA January 1, 2014 Volume 311, Number 1
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of St. Andrews Library User on 05/20/2015
95
