Functional Outcome of Supracondylar Elbow Fractures in Children Comparison of Cast Materials for the Treatment of Clubfoot Using

CLINICAL CONGRESS 2014 October 26–30, 2014 | Moscone Center | San Francisco

Vol. 57, No. 4, August/août 2014 canjsurg.ca

Functional outcome of supracondylar elbow fractures in children

Comparison of cast materials for the treatment of clubfoot using the Ponseti method: a prospective RCT

Computed tomography features associated with operative management for nonstrangulating small bowel obstruction

The Best Surgical Education All in One Place. Bridging the gap between open and minimally invasive Clinical Congress 2014. Where you want to be. pancreaticoduodenectomy: the hybrid approach

REGISTER NOW! SPONSORS Canadian Association of General Surgeons Department of Surgery, University of Alberta www.facs.org/clincon2014 Canadian Society for Vascular Surgery Department of Surgery, University of Calgary Canadian Society of Surgical Oncology Département de chirurgie, Université de Sherbrooke Canadian Association of Thoracic Surgeons Department of Surgery, McMaster University Department of Surgery, Western University Département de chirurgie, Université de Montréal Department of Surgery, Dalhousie University

The Surgeon of the Future

covercjs-aug14.indd 3 14-07-16 10:04 AM covercjs-aug14.indd 4 14-07-16 10:04 AM covercjs-aug14.indd 5 14-07-16 10:04 AM canjsurg.ca

Vol. 57, No. 4, August/août 2014 canjsurg.ca

EDITORIAL • ÉDITORIAL 237 Parathyroid hormone levels 1 hour after thyroidectomy: an early predictor of 221 No specialty alone: the Wilder Penfield strategy postoperative hypocalcemia V.C. McAlister A. AlQahtani, A. Parsyan, R. Payne, R. Tabah 223 Aucune spécialité n’est une île : la stratégie 241 Functional outcome of supracondylar elbow de Wilder Penfield fractures in children: a 3- to 5-year follow-up V.C. McAlister A.D. Isa, A. Furey, C. Stone COMMENTARY • COMMENTAIRE 247 Comparison of cast materials for the treatment of congenital idiopathic clubfoot using the 224 Enhancing medical students’ education Ponseti method: a prospective randomized and careers in global surgery controlled trial A. Gosselin-Tardif, G. Butler-Laporte, C. Hui, E. Joughin, A. Nettel-Aguirre, M. Vassiliou, K. Khwaja, G. Ntakiyiruta, S. Goldstein, J. Harder, G. Kiefer, D. Parsons, P. Kyamanywa, T. Razek, D.L. Deckelbaum C. Brauer, J. Howard 226 The validity of surgical simulation 254 Computed tomography features associated R.A. Agha, A.J. Fowler with operative management for 228 Balancing surgical innovation with cost nonstrangulating small bowel obstruction R.R. Suri, P. Vora, J.M. Kirby, L. Ruo and efficiency S. Jayaraman 260 The historic predictive value of Canadian orthopedic surgery residents’ orthopedic ONLINE COMMENTARY • COMMENTAIRE EN LIGNE in-training examination scores on their success on the RCPSC certification examination E119 #Nomoretextbooks? The impact of rapid D. Yen, G.S. Athwal, G. Cole communications technologies on medical education 263 Bridging the gap between open and minimally A. Farooq, J. White invasive pancreaticoduodenectomy: the hybrid approach Y. Wang, S. Bergman, S. Piedimonte, RESEARCH • RECHERCHE T. Vanounou 230 How to assess communication, professionalism, collaboration and the other ONLINE RESEARCH • RECHERCHE EN LIGNE intrinsic CanMEDS roles in orthopedic E121 The accuracy of the Alvarado score in residents: use of an objective structured predicting acute appendicitis in the black South clinical examination (OSCE) African population needs to be validated T. Dwyer, S. Glover Takahashi, M. Kennedy V.Y. Kong, S. Van Der Linde, C. Aldous, Hynes, J. Herold, D. Wasserstein, J.J. Handley, D.L. Clarke M. Nousiainen, P. Ferguson, V. Wadey, M.L. Murnaghan, T. Leroux, J. Semple, B. Hodges, D. Ogilvie-Harris

contentscjs-aug14.indd 218 14-07-16 1:27 PM E126 Health-related quality of life following ONLINE DISCUSSION IN SURGERY decompression compared to decompression DISCUSSION EN CHIRURGIE EN LIGNE and fusion for degenerative lumbar E146 Use of a novel energy technology for arresting spondylolisthesis: a Canadian multicentre ongoing liver surface and laceration hemorrhage study Y. Raja Rampersaud, C. Fisher, A. Yee, C.G. Ball M.F. Dvorak, J. Finkelstein, E. Wai, E. Abraham, E147 ONLINE CORRESPONDENCE S.J. Lewis, D. Alexander, W. Oxner CORRESPONDANCE EN LIGNGE E134 Oncoplastic reduction using the vertical scar superior-medial pedicle pattern technique for immediate partial breast reconstruction 270 CORRECTION Y. Barnea, A. Inbal, D. Barsuk, T. Menes, A. Zaretski, D. Leshem, J. Weiss, 287 CAREER/CLASSIFIED ADVERTISING S. Schneebaum, E. Gur ANNONCES SUR LES CARRIÈRES ET ANNONCES CLASSÉES REVIEW • REVUE

271 End-to-end ductal anastomosis in biliary © 2014 Canadian Medical Association. ISSN 0008-428X. For information on reconstruction: indications and limitations permission to reproduce material from the Canadian Journal of Surgery (CJS) B. Jabłonska see canjsurg.ca. All editorial matter in CJS represents the opinions of the authors and not ONLINE REVIEW • REVUE EN LIGNE necessarily those of the Canadian Medical Association (CMA). The CMA does not assume any responsibility or liability for damages arising from any error or omission or from the use of any information or advice con- E141 Systematic review on the inclusion of patients tained in CJS, including articles, editorials, reviews, letters and advertise- with cognitive impairment in hip fracture trials: ments. All reproduction rights are reserved. a missed opportunity? Printed by Dollco Integrated Print Solutions, Ottawa. Appears in February, S. Mundi, H. Chaudhry, M. Bhandari April, June, August, October and December. Return undeliverable Canadian copies to the CMA Member Service DISCUSSIONS IN SURGERY • DISCUSSIONS EN CHIRURGIE Centre, 1870 Alta Vista Dr., Ottawa ON K1G 6R7 (email [email protected]). © 2014 Association médicale canadienne. ISSN 0008-428X. Pour obtenir 278 Technique to achieve the symmetry of the des renseignements au sujet des permissions à obtenir afin de reproduire des extraits du Journal canadien de chirurgie (JCC), consulter canjsurg.ca. new inframammary fold Tous les articles à caractère éditorial dans le JCC représentent les opinions de M. Pozzi, G. Zoccali, E.M. Buccheri, R. de Vita leurs auteurs, qui ne sont pas nécessairement celles de l’Association médicale canadienne (AMC). L’AMC n’assume aucune responsabilité pour les 280 Users’ guide to the surgical literature: how to dommages résultant de toute erreur ou omission, ou de l’utilisation de ren- evaluate clinical practice guidelines seignements ou de conseils contenus dans le JCC, y compris les articles, C.J. Coroneos, S.H. Voineskos, éditoriaux, revues, lettres et annonces. Tous droits de reproduction réservés. S.D. Cornacchi, C.H. Goldsmith, T.A. Ignacy, La revue est imprimée par Dollco Integrated Print Solutions, Ottawa. Elle A. Thoma paraît en février, avril, juin, août, octobre et décembre. Retournez toutes copies canadiennes non livrées au Centre des services aux membres, Association médicale canadienne, 1870, prom. Alta Vista, Ottawa (Ontario) K1G 6R7 (courriel : [email protected]).

contentscjs-aug14.indd 219 14-07-16 1:27 PM The Canadian Journal of Surgery aims to contribute to the effective continuing medical education of Canadian surgical specialists, and to provide surgeons with an effective vehicle for the dissemination of observations in the areas of clinical, basic science and education research. Readers can find CJS online at canjsurg.ca. COEDITORS Submission of new manuscripts can be made at http://mc.manuscript central.com/cjs. CORÉDACTEURS Le Journal canadian de chirurgie vise à dispenser une éducation médicale continue efficace aux spécialistes en chirurgie au Canada, et fournir aux chirurgiens un mécanisme efficace pour diffuser les constatations de la Edward J. Harvey, MD, Montréal recherche clinique, fondamentale et éducative. [email protected] Les lecteurs trouveront en direct le JCC à l’adresse canjsurg.ca. Major Vivian C. McAlister, MB, London Nous favorisons l’envoi électronique de manuscrits. Veuillez visiter le http://mc.manuscriptcentral.com/cjs. [email protected] GLOBAL SURGERY SPINAL SURGERY ASSOCIATE EDITORS Dan Deckelbaum, MD, Montréal Raja Rampersaud, MD, Toronto RÉDACTEURS ASSOCIÉS GENERAL SURGERY SPORTS MEDICINE Daniel Birch, MD, Edmonton Paul Martineau, MD, Montréal BASIC SCIENCE AND SURGICAL BIOLOGY Ian McGilvray, MD, PhD, Toronto HEPATOBILIARY AND PANCREATIC SURGERY SURGICAL ONCOLOGY Timothy Daniels, MD, Toronto Shiva Jayaraman, MD, Toronto Geoff Porter, MD, Halifax

BREAST SURGERY MEDICAL EDUCATION SURGICAL ONCOLOGY, MUSCULOSKELETAL Muriel Brackstone, MD, London Carol Hutchison, MD, Calgary Frank O’Dea, MD, St. John’s

CARDIOVASCULAR SURGERY MILITARY MEDICINE SURGICAL ONCOLOGY, SOFT TISSUE Michel Carrier, MD, Montréal Homer Tien, MD, Toronto Chris de Gara, MB MS, Edmonton Michael Chu, MD, London ORTHOPEDIC FOOT AND ANKLE SURGERY THORACIC SURGERY CRITICAL CARE Karl-André Lalonde, MD, Ottawa John Yee, MD, Vancouver Raymond Kao, MD, London ORTHOPEDIC SURGERY TRAUMA SURGERY, ORTHOPEDIC ENDOCRINE SYSTEM Graham Elder, MD, Sault Ste. Marie William Dust, MD, Saskatoon Sam Wiseman, MD, Vancouver PEDIATRIC SURGERY, GENERAL TRAUMA SURGERY, SOFT TISSUE EVIDENCE-BASED MEDICINE Sigmund Ein, MD, Toronto Mary vanWijngaarden-Stephens, MD, Edmonton Michelle Ghert, MD, Hamilton Morad Hameed, MD, Vancouver PEDIATRIC SURGERY, ORTHOPEDIC GASTROINTESTINAL AND COLORECTAL SURGERY James G. Wright, MD, MPh, Toronto VASCULAR SURGERY Marcus Burnstein, MD, Toronto Kent Mackenzie, MD, Montréal Jason Park, MD, Winnipeg

PRODUCTION DESIGNERS ASSOCIATE DIRECTOR, PUBLICATIONS INFOGRAPHISTES DE LA PRODUCTION DIRECTRICE ASSOCIÉE, PUBLICATIONS Carole Lalonde, Sarah O’Neill, Jennifer Pershick, Holly Bodger Clara Walker PROGRAM MANAGER, JOURNALS Owned and published by CMA WEB PUBLISHERS GESTIONNAIRE DE PROGRAMMES, JOURNAUX 1867 prom. Alta Vista Dr. ÉDITION EN LIGNE Janis Murrey Ottawa ON K1G 5W8 James Harbinson, James Manship CMA EXECUTIVES EDITORIAL STAFF ADVERTISING DIRIGEANTS DE L’AMC ÉQUIPE DE RÉDACTION PUBLICITÉ PRESIDENT • PRÉSIDENT MANAGING EDITOR ADVERTISING Louis Hugo Francescutti DIRECTRICE DE LA RÉDACTION ANNONCES PUBLICITAIRES Wendy Carroll ([email protected]) Laurie McLeod, Susan Ritchie, David Bertrand SECRETARY GENERAL AND CEO tel 800 663-7336 (Canada, USA) 613 731-8610 x2107/2041 fax 613 565-7488 SECRÉTAIRE GÉNÉRAL ET CHEF DE LA DIRECTION or 613 731-8610 x2198; fax 613 565-5471 [email protected]; [email protected] Tim Smith EDITORIAL COORDINATOR ADVERTISING SALES VP, PROFESSIONAL SERVICES AND LEADERSHIP COORDONNATRICE DE LA RÉDACTION VENTES PUBLICITAIRES VP, SERVICES PROFESSIONNELS ET LEADERSHIP Elisa Makadi ([email protected]) Blair Graham ([email protected]) Deborah Scott-Douglas tel 800 663-7336 (Canada, USA) 416 850-1485; cell 416 560-0202; 613 731-8610 x2018; fax 613 565-5471 fax 416 850-2242

TRANSLATION • TRADUCTION Trish Sullivan ([email protected]) All prescription drug advertisements have been Marie Saumure cell 905 330-8770 cleared by the Pharmaceutical Advertising Advisory Board. Toutes les annonces de médicaments prescrits ont PRODUCTION STAFF PUBLISHING STAFF été approuvés par le Conseil consultatif de publicité pharmaceutique. ÉQUIPE DE PRODUCTION ÉQUIPE DES SERVICES D’ÉDITION

PRODUCTION MANAGER DIRECTOR AND PUBLISHER, CMA PUBLICATIONS GESTIONNAIRE DE LA PRODUCTION DIRECTRICE ET ÉDITRICE, PUBLICATIONS AMC Debbie Rupert Glenda Proctor

masthead-aug14 220 14-07-16 1:33 PM EDITORIAL • ÉDITORIAL

No specialty alone: the Wilder Penfield strategy

To gather knowledge and to find out new knowledge is the noblest The divide between physicians and surgeons is as old as occupation of the physician. To apply that knowledge with understanding and sympathy to the relief of human suffering is the loveliest occu- the science of medicine itself. While Hippocrates made phys pation; and to do both with unassuming faithfulness sets the seal on the icians promise to leave cutting for stone to surgeons who whole. — Edward Archibald, 1934.1 were practised in the art, it was not until physicians and sur- JS has posted to its website (www.canjsurg.ca) a geons collaborated in the 17th century that the mysteries of “work-in-progress” commentary as part of the CJS circulation were discovered and the development of scientific C consensus protocol.2 The authors’ goal is to medicine was begun.5 Penfield thought at first that all know develop a statement regarding gastrointestinal endoscopy ledge of a specialty could reside within one person. However training for general surgery residents.3 The setting for neuroscience expanded so rapidly, due in no small part to the their review is a growing discord between the specialties of success of his own collaborative strategy, that Penfield was gastroenterology and general surgery over endoscopy. In forced to recognize the need subspecialize.4 Canada’s Royal parts of Canada, one specialty has refused to teach trainees College was founded in 1929 by physicians and surgeons, of the other. In other jurisdictions, qualifications required who were colleagues of Penfield, as a place for all specialties. to participate in provincially sponsored screening pro- The Royal College’s recently launched “Competence by grams have been written to favour one specialty over the Design” initiative will transform education within specialties. other. These events may be aberrations as the historical It may also break down the barriers that confine each spe- partnership between these specialties continues to be cialty. The needs of patients will become the principle deter- driven by the demands of patient care. minant of the boundaries of specialties. It is clear today that Edward Archibald’s gentle challenge was contained in Canadian patients need the competent care of gastroenterol- his address at the opening of the Montreal Neurological ogists and general surgeons, just as they need the collabora- Institute in 1934.1 Almost 40 years later when the Uni- tive care of other sister medical and surgical specialties. versity Hospital in London, Ontario, was being built, the design was selected with this challenge in mind. In- Vivian C. McAlister, MB hospital and out-patient care areas were clustered by spe- Coeditor, Canadian Journal of Surgery cialty and colocated with research laboratories and doc- Competing interests: None declared. tors’ offices. There was no place for a distinction between physicians and surgeons. In 1972, the founders chose DOI: 10.1503/cjs.008414 Wilder Penfield to give the opening address. The purpose of this choice was to leave us a message. Prior to his References arrival in Canada, Penfield was aware that the restrictive practices of surgeons and physicians of related specialties 1. Penfield W. Edward Archibald; 1872–1945. Can J Surg 1958;1:167-74. had put a great limitation on the development of the sci- 2. McAlister V. Consensus ad idem: a protocol for development of ence of medicine. His personal response was to become consensus statements. Can J Surg 2013;56:365. an expert in neurology and neuropathology while he 3. Bradley NL, Bazzerelli A, Lin J, et al. Gastrointestinal endoscopy training learned neurosurgery. Later he became a pioneering in general surgery residency programs. Available: www.canjsurg.ca. 4. McAlister VC. William Harvey, Fabricius ab Acquapendente and neuroradiologist. Once in Montréal, he ceaselessly cam- the divide between medicine and surgery. Can J Surg 2007;50:7-8. paigned for an institute where physicians, surgeons and 5. Penfield W. No man alone: a neurosurgeon’s life. New York (NY): scientists would collaborate in research and patient care.4 Little, Brown and Company; 1977.

edit-aug14 221 14-07-16 1:34 PM Outcomes Up. Costs Down. Value All Around.

Canadian Surgery Forum 2014 Booth #406

Finally – an alternative to biologics! GORE® BIO-A® Tissue Reinforcement is a unique non-biologic scaffold that is gradually absorbed by the body. The open, highly interconnected 3D pore structure facilitates cell infiltration and growth. Vascularization begins quickly within one to two weeks.

• 100% synthetic, bioabsorbable tissue scaffold • Rapid cell population and vascularization • Versatile for numerous applications • Available in large sizes up to 20 cm x 30 cm With a three-year shelf life and no soaking, refrigeration or tracking required, this versatile material is the easy-to-use, performance-proven alternative that offers value for surgeons and hospitals.

Gore. Because material really does matter.

W. L. Gore & Associates, Inc. • Flagstaff, AZ 86004 • goremedical.com

Products listed may not be available in all markets. GORE®, BIO-A®, PERFORMANCE THROUGH INNOVATION, and designs are trademarks of W. L. Gore & Associates. ©2013, 2014 W. L. Gore & Associates, Inc. AS1932-EN2 JANUARY 2014

edit-aug14-fr.indd 222 14-07-16 10:17 AM Outcomes Up. Costs Down. EDITORIAL • ÉDITORIAL Aucune spécialité n’est une île : la stratégie Value All Around. de Wilder Penfield

Acquérir des connaissances et en découvrir de nouvelles est la plus noble occupation du médecin. Appliquer ces connaissances en faisant Le fossé entre les médecins et les chirurgiens est aussi Canadian Surgery preuve de compréhension et d’empathie pour le soulagement de la vieux que la science de la médecine. Hippocrate a fait Forum 2014 souffrance humaine est la plus belle profession; et faire les deux avec promettre aux médecins de laisser les opérations aux une fidélité sans prétention est le couronnement de la profession. — Edward Archibald, 1934.1 chirurgiens qualifiés, mais il aura fallu attendre une colla Booth #406 boration entre les médecins et les chirurgiens au 17e siècle e JCC a publié sur son site web (www.canjsurg.ca) un pour découvrir les mystères de la circulation et permettre commentaire « en évolution » dans le cadre du pro- l’avènement de la médecine scientifique.5 Au départ, le L tocole de consensus du JCC.2 L’objectif des auteurs Dr Penfield pensait qu’une seule personne pouvait pos- est d’élaborer un énoncé sur la formation des résidents en séder toutes les connaissances d’une spécialité. La neurochirurgie générale en endoscopie gastrointestinale.3 Le science a toutefois connu une expansion si rapide, en contexte de leur évaluation fait l’objet d’une discorde crois- grande partie en raison de la réussite de sa propre stratégie sante au sujet de l’endoscopie entre les spécialités de la de collaboration, que le Dr Penfield a dû reconnaître la gastro-entérologie et de la chirurgie générale. Dans cer- nécessité de la surspécialisation.4 Le Collège royal des taines régions du Canada, une spécialité a refusé d’en médecins et chirurgiens du Canada a été fondé en 1929 par seigner aux médecins en formation de l’autre spécialité. des médecins et des chirurgiens, qui étaient des collègues Dans d’autres provinces, les qualifications requises pour du Dr Penfield, comme un lieu de rassemblement de toutes participer aux programmes de dépistage parrainés par la les spécialités. L’initiative « La compétence par concep- province favorisent une spécialité par rapport à l’autre. Ces tion », que le Collège royal a récemment lancée, trans- événements sont peut-être des aberrations, le partenariat formera la formation au sein des spécialités. Elle pourrait existant de tout temps entre ces spécialités continuant en outre faire tomber les barrières qui délimitent chaque d’être alimenté par la demande de soins des patients. spécialité. Les besoins des patients deviendront le principal Edward Archibald avait lancé son défi à l’occasion de son déterminant des limites des spécialités. Il est évident allocution d’ouverture de l’Institut neurologique de Mon- aujourd’hui que les patients ont besoin des soins compé- tréal en 1934.1 Près de 40 ans plus tard, lors de la construc- tents des gastro-entérologues et des chirurgiens généraux, tion de l’hôpital universitaire de London (Ont.), ce défi a été tout autant qu’ils ont besoin des soins en collaboration pris en compte pour le choix du concept. Les unités de soins d’autres spécialités médicales et chirurgicales. Finally – an alternative to biologics! GORE® BIO-A® Tissue Reinforcement is a unique des patients en milieu hospitalier et de soins ambulatoires non-biologic scaffold that is gradually absorbed by the body. The open, highly étaient regroupées par spécialité et situées près des labora- Vivian C. McAlister, MB toires de recherche et des bureaux des médecins. On élimi- Co-rédacteur, Journal canadien de chirurgie interconnected 3D pore structure facilitates cell infiltration and growth. nait la possibilité de faire une différence entre les médecins Vascularization begins quickly within one to two weeks. et les chirurgiens. En 1972, les fondateurs ont choisi le Intérêts concurrents: Aucuns déclarés. r • 100% synthetic, bioabsorbable tissue scaffold D Wilder Penfield pour prononcer le discours d’ouverture. Ce choix était motivé par le désir de faire passer un message. DOI : 10.1503.cjs.008614 • Rapid cell population and vascularization Avant son arrivée au Canada, le Dr Penfield savait que les • Versatile for numerous applications pratiques restrictives des chirurgiens et des médecins de spé- Références • Available in large sizes up to 20 cm x 30 cm cialités connexes avaient limité grandement le dévelop pement de la science de la médecine. Sa réponse personnelle With a three-year shelf life and no soaking, refrigeration or tracking required, this 1. Penfield W. Edward Archibald; 1872–1945.Can J Surg 1958;1:167-74. a été de devenir un expert en neurologie et en neuropatho 2. McAlister V. Consensus ad idem: a protocol for development of versatile material is the easy-to-use, performance-proven alternative that offers logie pendant ses études de neurochirurgie. Plus tard, il a consensus statements. Can J Surg 2013;56:365. value for surgeons and hospitals. fait un travail de pionnier en qualité de neuroradiologue. 3. Bradley NL, Bazzerelli A, Lin J, et al. Gastrointestinal endoscopy training Une fois à Montréal, il a fait campagne sans relâche pour la in general surgery residency programs. Disponible : www.canjsurg.ca. Gore. Because material really does matter. 4. McAlister VC. William Harvey, Fabricius ab Acquapendente and création d’un institut où les médecins, les chirurgiens et les the divide between medicine and surgery. Can J Surg 2007;50:7-8. scientifiques pourraient collaborer à la recherche et au soin 5. Penfield W. No man alone: a neurosurgeon’s life. New York (NY): des patients.4 Little, Brown and Company; 1977.

W. L. Gore & Associates, Inc. • Flagstaff, AZ 86004 • goremedical.com

edit-aug14-fr.indd 223 14-07-16 10:17 AM COMMENTARY • COMMENTAIRE

Enhancing medical students’ education and careers in global surgery

Alexandre Gosselin-Tardif, MD* Summary Guillaume Butler-Laporte, MD* With surgical conditions being significant contributors to the global burden † of disease, efforts aimed at increasing future practitioners’ understanding, Melina Vassiliou, MD, MEd interest and participation in global surgery must be expanded. Unfortu- Kosar Khwaja, MD, MBA, MSc† nately, despite the increasing popularity of global health among medical stu- Georges Ntakiyiruta, MD‡ dents, possibilities for exposure and involvement during medical school remain limited. By evaluating student participation in the 2011 Bethune ‡ Patrick Kyamanywa, MD, MPH Round Table, we explored the role that global surgery conferences can play Tarek Razek, MD† in enhancing this neglected component of undergraduate medical education. † Study results indicate high rates of student dissatisfaction with current global Dan L. Deckelbaum, MD, MPH health teaching and opportunities, along with high indices of conference satisfaction and knowledge gain, suggesting that global health conferences can From the *Faculty of Medicine, McGill serve as important adjuncts to undergraduate medical education. University, Montréal, Que., †McGill Uni- versity Health Centre, Center for Global Surgery, Montréal, Que., and the ‡National University of Rwanda, Depart- ment of Surgery, Kigali, Rwanda espite the substantial contribution of surgical conditions to the global burden of disease, the field of global surgery remains severely under- Accepted for publication resourced, especially when compared with the well-staffed and well- Nov. 7, 2013 D funded campaigns combatting communicable and other nonsurgical diseases. One important avenue toward meeting surgical workforce needs involves Correspondence to: D.L. Deckelbaum exposing, educating and motivating future surgeons to participate in global McGill University Health Centre surgery initiatives. Over the past decade, global surgery has become increas- Centre for Global Surgery ingly popular among medical students, many of whom hope to translate their Montreal QC interest into active participation.1 [email protected] Despite these developments, current opportunities for medical student exposure and involvement in global surgery remain limited. In the absence of DOI: 10.1503/cjs.027713 uniform guidelines regarding global health teaching, experiences vary significantly, and dissatisfaction rates remain high, with 41% of graduating Ameri- can medical students finding their global health curriculum inadequate.1,2 There are also important limitations to global health electives as currently offered; these range from prohibitive cost to lack of organizational support from medical faculties.1 In this context, it becomes important to seek complementary opportunities for student exposure to global health and surgery. Based on our experience with student participation in the 2011 Bethune Round Table, we suggest that global surgery conferences can serve as key adjuncts to undergraduate medical education, helping to inform, motivate and integrate students into the global surgery community. The Bethune Round Table is an international conference dedicated to the surgical issues facing low- and middle-income countries. Organized by the Canadian Network for International Surgery (CNIS) and held annually in Canada, the conference features important participation of global surgery leaders from resource-limited settings in Africa, Eastern Europe and Asia; leaders from these settings accounted for more than half of the speak- ers at the 2011 conference.3 The conference, which was hosted in Mon- tréal, Canada on June 3–5, featured an unprecedented level of student

enhancing-goss.indd 224 14-07-16 10:21 AM COMMENTARY

involvement, with 35 first-year medical students from Interestingly, despite high rates of involvement among McGill University attending the proceedings, their medical students, only 12 of 102 respondents felt like registration having been waived in exchange for volun- legitimate members of the global health community. teering to help run the event. Considering the link established between students’ sense We took advantage of this opportunity by constructing of belonging to an academic community and the reten- a 56-item questionnaire meant to gauge student experi- tion rate within that particular discipline, this finding has ences and attitudes toward global health and surgery as important implications.4 As previous studies on higher well as to assess the outcomes and educational value of education have shown, academic conferences can be ideal such conferences. Administered electronically to all 173 forums for informal discussions between students and first-year medical students from McGill University, the experienced practitioners, contributing to the encultura- survey was completed by 102 students, including 31 of the tion process of the former.5 In this context, it is encour- 35 conference participants. aging that, on average, 40% of medical students’ inter Survey results highlight substantial interest in the disci- actions during the conference were with practising pline among medical students, with 42% of the 102 surgeons and that 55% of respondents reported an respondents reporting involvement in global health increased sense of community following the event. The research, volunteering or student initiatives and 55% fact that a now annual student-organized global surgery intending to incorporate global health into their future conference was held at McGill University less than careers. Concurrently, dissatisfaction with global health 12 months after the 2011 Bethune Round Table, with teaching as currently offered was reported by 55%, with more than 120 guests in attendance, serves to further 68% of these respondents desiring an expansion of its illustrate this point. We would therefore argue that such place in the curriculum, suggesting greater discontentment conferences favour the continuation and intensification of among this cohort than among American medical students, student involvement in global health and surgery, helping as stated previously. Although no similar information is to transform what is initially an interest among others available for medical students across Canada, these results into an issue seen as intrinsic to the practice of medicine. should be of concern, given that McGill University stands If the rising interest in global health and surgery slightly above the Canadian average in terms of global among future doctors is to be cultivated, we must health training offered.1 This unmet student demand illus- become better at offering diverse and easily accessible trates the importance of supplementing available oppor opportunities for student exposure. This will require tunities for global health exposure. enhancing the global health curriculum and increasing In terms of conference participation outcomes, major faculty support for international electives as well as fa self-perceived global surgery knowledge gain was cilitating student participation in global surgery confer- reported by 71% of the 31 attendees who completed the ences like the Bethune Round Table. By stimulating survey, while substantial increase in global health interest global health involvement early on in medical education, and in intentions to participate in global health activities such steps will help ensure that the current surge in was reported by 87% and 77%, respectively. Although interest translates into more students striving to become student participants were self-selected by their pre- tomorrow’s global health leaders. existing interests, 9 of the 31 respondents (31%) had no Competing interests: G. Ntakiyiruta and P. Kyamanywa have received prior global health experience, as measured by past par- travel support from McGill University Health Centre. T. Razek is a ticipation in global health conferences, research, vol board member (unpaid) for the Canadian Network for International Surgery. No other competing interests declared. unteering or other initiatives. This diversity among attendees did not translate into differences in perceived References outcomes, which remained highly positive regardless of previous exposure. Similarly, the Bethune Round Table 1. Izadnegahdar R, Correia S, Ohata B, et al. Global health in Canadian attracted both surgically minded (15 of 31) and medically medical education: current practices and opportunities. Acad Med minded (16 of 31) students, appealing equally to both dis- 2008;83:192-8. ciplines and leading 8 participants into newly considering 2. Association of American Medical Colleges. Medical school graduation surgical careers. These results indicate that events like the questionnaire: all schools report. Washington (DC): the Association; 2010. Bethune Round Table can be as beneficial to newcomers 3. Deckelbaum DL, Gosselin-Tardif A, Taylor R, et al. Global health as they are to more experienced participants and that they conferences: Are they truly “global”? The Bethune Round Table are equally appreciated by both groups. This versatility paradigm for promoting global surgery. Can J Surg 2011;54:422-9. suggests that global surgery conferences can be valuable 4. Chapman D, Wiessner C, Morton J, et al. Crossing scholarly divides: and easily accessible complements to the limited options barriers and bridges for doctoral students attending scholarly conferences. New horizons in adult education and human resource development currently available to medical students, functioning both 2009;23:6-24. as entry points into the discipline and as opportunities to 5. Lave J, Wenger E. Situated learning: legitimate peripheral participation. reinforce existing interest and participation. Cambridge (UK): Cambridge University Press; 1991.

Can J Surg, Vol. 57, No. 4, August 2014 225

enhancing-goss.indd 225 14-07-16 10:21 AM COMMENTARY • COMMENTAIRE

The validity of surgical simulation

Riaz A. Agha, MBBS, MSc* Summary Alexander J. Fowler† Simulation is playing an increasingly important role in training surgeons. As hours between registrar and consultant grades have decreased, trainees are From the *Department of Plastic Surgery, required to train smarter. While the majority of simulation is limited, advances Stoke Mandeville Hospital, Stoke Man- in computing and design are enabling ever more realistic, varied simulation. deville, Ayelsbury, Bucks, UK, and †Barts and the London School of Medicine and Dentistry, Queen Mary, University of Lon- don, UK imulation is an important tool in the training of juniors, but work is required to expand this training to a wider variety of surgical tech- Accepted for publication niques, not only laparoscopic ones. The very first surgical simulators Dec. 4, 2013 S were leaf and clay models used in India in 600 BC to simulate a forehead 1 Correspondence to: nasal flap reconstruction. Since then, simulation has become a highly A.J. Fowler refined training format that is used in number of high-risk industries. It has London School of Medicine & Dentistry become a key tool in the education of clinicians at all levels in a wide selec- Garrodd Building Turner Street, Whitechapel tion of specialties and is an important component in recent drives to London, UK improve patient safety. E1 2AD A large proportion of the methodological and technological develop- [email protected] ment in simulation has been in the aviation industry, where pilots have long been trained to fly before stepping into an aircraft.2 In the United King- DOI: 10.1503/cjs.032013 dom, the combination of many registrar grades into the single grade of “Specialist Registrar” (known as Calmanisation) and the European Work- ing Time Directive have reduced the period available for training. As such, the working hours between becoming a senior house officer and a consultant have estimated to have reduced by a factor of 5.3 Simulation has evolved as an effective training technique alongside this changing environment for the training of surgeons — namely the reduction in hours available for training. This drastic change in training time and practices necessitated a paradigm shift in the model of surgical education. There has been a move away from the apprenticeship model in which expertise was acquired through experience, to a more standardized, objective and competency-based approach that requires a more proactive attitude to training. Simulation has become a key part of providing this objective training and assessment, allowing mistakes to be made in a safe environment and to develop further attributes, such as understanding human factors, that exist outside the realm of pure technical ability. Much work has been carried out looking into the use of laparoscopic simulators. Seymour and colleagues4 randomized 16 surgical trainees to either a laparoscopic simulator (MIST-VR; Virtalis) training group or a control group trained traditionally. Participants then performed a cholecystectomy in an operating theatre, and the procedures were recorded for assessment. Participants in the simulator group dissected the gallbladder 29% faster and were 5 times less likely to make errors than those in the control group. These findings were supported in a similar investigation undertaken by Grantcharov and colleagues5 involving laparoscopic novices. The MIST-VR group performed significantly faster than the control group, with better economy of movement and error scores. A recent

valid-fowler.indd 226 14-07-16 10:23 AM COMMENTARY

systematic review of laparoscopic surgery simulation niques. Future research should examine the wider aspects encompassing 219 studies and 7138 trainees concluded of surgery. Simulation should be part of the learning that “simulation-based laparoscopic training of health experience but cannot replace the requisite clinical hard professionals [has] large benefits when compared with “graft” and experience a trainee surgeon needs on the no intervention and is moderately more effective than “shop floor,” supported by good trainers and mentors. 6 nonsimulation instruction.” Competing interests: None declared. Work by Kneebone and colleagues7 has taken the concept of simulation a step further: “simulated patients” force trainees to interact with real people while perform- References ing procedures. This technique has been extended to laparoscopic surgery, where tactile feedback allows train- 1. Limberg AA. The Planning of local plastic operations on the body ees to undertake the operation with a number of ana- surface: theory and practice. Lexington (MA): DC Health and Company; 1984. tomic variants and get used to the feeling of handling 2. Enochsson L, Isaksson B, Tour R, et al. Visuospatial skills and com- different tissues. Alongside these technically useful fea- puter game experience influence the performance of virtual endos- tures, authenticity is enhanced by giving the model copy. J Gastrointest Surg 2004;8:876-82. patient head and feet, artificial skin and a theatre team, 3. Chikwe J, de Souza AC, Pepper JR. No time to train the surgeons. including all those normally present for such an opera- BMJ 2004;328:418-9. tion. The quality of simulator used may impact the out- 4. Seymour NE, Gallagher AG, Roman SA, et al. Virtual reality training comes for patients — simulators without haptics can lead improves operating room performance. Ann Surg 2002;236:458-63. to distortions of pulling and pushing forces required.8 5. Grantcharov TP, Kristiansen VB, Bendix J, et al. Randomized clinical trial of virtual reality simulation for laparoscopic skills training. Br J Much work still remains to be done on transferring Surg 2004;91:146-50. teamwork and leadership skills as well as human factors 6. Zendejas B, Brydges R, Hamstra SJ, et al. State of the evidence on from the simulation suite to the operating room. simulation-based training for laparoscopic surgery: a systematic Growing evidence suggests that skills gained within review. Ann Surg 2013;257:586-93. simulated environments transition well into the real clin 7. Kneebone R, Nestel D, Wetzel C, et al. The human face of simula- ical setting. A recent review found good skill transfer in tion: patient-focused simulation training. Acad Med 2006;81:919-24. pediatric emergency situations, tracheal intubation and 8. Chmarra MK, Dankelman J, van den Dobbelsteen JJ, et al. Force feedback and basic laparoscopic skills. Surg Endosc 2008;22:2140-8. central venous catheter insertion, with reported decreases 9 9. Zendejas B, Cook DA, Bingener J, et al. Simulation-based mastery in complications and infections. Zendejas and colleagues learning improves patient outcomes in laparoscopic inguinal hernia investigated laparoscopic inguinal hernia repair in a repair: a randomized controlled trial. Ann Surg 2011;254:502-9. randomized controlled trial10 and subsequently found 10. Griswold S, Ponnuru S, Nishisaki A, et al. The emerging role of decreased procedure duration and complications. simulation education to achieve patient safety: translating deliberate Stefanidis and colleagues11 found that 71% of novices practice and debriefing to save lives.Pediatr Clin North Am 2012; trained to proficiency on a simulator retained their skills 59:1329-40. 11. Stefanidis D, Yonce TC, Korndorffer JR Jr, et al. Does the incorpora- in the operating theatre. tion of motion metrics into the existing FLS metrics lead to improved The vast majority of work pertaining to skills transla- skill acquisition on simulators? A single blinded, randomized con- tion has been undertaken in laparoscopic surgical tech- trolled trial. Ann Surg 2013;258:46-52.

Can J Surg, Vol. 57, No. 4, August 2014 227

valid-fowler.indd 227 14-07-16 10:23 AM COMMENTARY • COMMENTAIRE

Balancing surgical innovation with cost and ef ciency

Shiva Jayaraman, MD, MESc SUMMARY The standard approach to neoplasia of the pancreatic head is pancreatico- From the University of Toronto, St. duodenectomy, otherwise known as the Whipple procedure. Traditionally, Joseph’s Health Centre, Toronto, Ont. this operation is performed through an open laparotomy incision. In high- volume centres, and when performed by appropriately quali ed surgeons, See the related research paper by Wang the Whipple procedure is safe and effective management for diseases of the and colleagues on p. 263. pancreatic head. Still, this operation remains one of the most complex abdominal procedures. With the proliferation of minimally invasive sur- Accepted for publication gery, more complex operations are being performed using laparoscopy and May 20, 2014 other minimal access techniques. A group from McGill University and the Montreal Jewish General Hospital have prospectively evaluated their Correspondence to: experience with minimally invasive pancreaticoduodenectomy and have S. Jayaraman compared this experience to the open approach. This is the rst compara- University of Toronto tive series of its kind from Canada. St. Joseph’s Health Centre 30 The Queensway, Suite 240 SSW Toronto ON M6R 1B5 [email protected] he world’s first minimally invasive pancreaticoduodenectomy was reported by Canadian surgeons in 1994.1 Since then, several smaller DOI: 10.1503/cjs.006114 T reports have followed that suggest that many of the traditional advantages of minimally invasive surgery may be possible in select patients undergoing surgery for neoplasms of the pancreatic head.2 In this edition of the Canadian Journal of Surgery (CJS), Wang and colleagues3 report Canada’s rst comparative series between minimally invasive pancreaticoduodenectomy using a hybrid approach and open pancreaticoduodenectomy. In their series, they demonstrate a shorter length of stay in hospital for the minimally invasive group, with similar perioperative outcomes and complication rates between the minimally invasive and open surgery groups. They also report a trend toward lengthier operations in the minimally invasive group. This experience highlights one of the con icts we face in our health care system. As health care budgets shrink, there are increased scal pressures on hospitals and, by extension, operating room resources. Likewise, a key metric for quality in Canadian surgery is waiting times. These realities may challenge the technical progress suggested by Wang and colleagues. If advanced approaches are more broadly applied and if the observation of increased duration of surgery and the possibility of greater resource utilization persist, widespread uptake of this novel approach may have to be limited. On the other hand, if it can be demonstrated that improved perioperative outcomes are reproducible with minimally invasive pancreaticoduodenectomy and that once a suitable learning curve is overcome good outcomes can be provided to patients in a timely and ef cient manner, then more widespread, systematic uptake may be possible. Likewise, new technology may permit greater ef - ciency in such complex minimally invasive procedures, facilitating an improvement in important perioperative parameters and thereby decreasing costs. As mentioned in a previous CJS editorial, surgeons need to nd solutions to the scal and societal restraints currently in place for our health care system.4 With new technology and novel approaches to surgical diseases, we are

bal-jayaraman.indd 228 14-07-16 10:27 AM COMMENTARY

References challenged to balance our desire to improve patient outcomes and incorporate new approaches while maintaining 1. Gagner M, Pomp P. Laparoscopic pylorus-preserving pancreatico- a fiscally sound and efficient system. If more surgeons duodenectomy. Surg Endosc 1994;8:408-10. embrace minimally invasive pancreaticoduodenectomy in 2. Gumbs AA, Rodriguez Rivera AM, Milone L, et al. Laparoscopic Canada, it will be important to better understand whether pancreatoduodenectomy: a review of 285 published cases. Ann Surg the potentially increased costs of this approach are justi ed Oncol 2011;18:1335-41. by superior outcomes, new ef ciencies and improved qual- 3. Wang Y, Bergman S, Piedimonte S, et al. Bridging the gap between ity of life for patients. open and minimally invasive pancreaticoduodenectomy: the hybrid approach. Can J Surg 2014;57:263-70. Competing interests: None declared. 4. Harvey EJ. Process improvement in surgery. Can J Surg 2014;57:4.

Mark your calendar and plan to attend the 2014 Canadian Surgery Forum! September 17-21, 2014 Vancouver Convention Centre & Fairmont Waterfront Hotel

The Forum is intended for community and academic surgeons, residents in training, researchers, surgical and operating room nurses, Fellows and medical students. The 2014 Forum will offer outstanding opportunities for continuing professional development, dialogue on educational and research issues, and networking.

For more information, please visit www.cags-accg.ca and www.canadiansurgeryforum.com.

See you in Vancouver!

Can J Surg, Vol. 57, No. 4, August 2014 229

bal-jayaraman.indd 229 14-07-16 10:27 AM COMMENTARY • COMMENTAIRE

#Nomoretextbooks? The impact of rapid communications technologies on medical education

Ameer Farooq, BSc Summary Jonathan White, MD, PhD, MSc This paper was selected as the 2013 student essay winner by the Canadian Undergraduate Surgical Education Committee. The essay was in response to (MedEd) the question “How does rapid communications technology affect learning?”.

From the Department of Surgery, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, Alta. omorrow, I start my vascular surgery rotation. Before bed tonight, I will Accepted for publication watch a YouTube video of a femoral-popliteal bypass surgery, review the Dec. 4, 2013 T surgical anatomy from Zollinger’s Atlas of Surgical Operations on my iPad Correspondence to: while waiting for my car tires to be changed, listen to a podcast on peripheral A. Farooq vascular disease while riding my exercise bike and perhaps tweet about my new 8440–112 St. rotation. My day is typical of a medical student in the context of increasing use Edmonton AB T6G 2B7 of rapid communications technology: 91% of health professions students [email protected] between the ages of 18 and 25 use Facebook.1 Rapid communications technology is defined broadly in this essay as any technology that enables access to DOI: 10.1503/cjs.030913 information through an electronic device and/or permits that information to be shared in a public, social way. This includes mobile phones, laptops, tablets, online textbooks and social media. In this essay, I’ll discuss the impact of these technologies on medical education. Rapid communications technology has made access to information instantaneous and ubiquitous. Students no longer need to wait until they have access to their textbooks or handbooks to recall information — they can pull out their mobile phones and look it up instantly. There are a number of medical apps that allow users to look up drug dosing and approaches to common conditions and recall surgical anatomy.2 The ubiquity of access to information helps students to maximize their time and learning oppor tunities. In addition, this rapid access to information can help to reinforce learning in a contextual manner.3 Ho and colleagues4 found that having a personal digital assistant case log versus a paper case log enhanced student learning and reflection. Beyond access to information, rapid communications technology potentially offers exciting new ways of learning. For example, users of anatomy.tv are able to view 3-dimensional (3D) reconstructions of anatomy as opposed to the 2-dimensional images available in print textbooks. In addition, users can toggle between an abdominal computed tomography scan and the virtual 3D anatomy representation and correlate the two.5 Medical students now also regularly use podcasts and vodcasts to supplement their learning in novel ways, all while driving to clinic or exercising.6 Social media provides a means for learners to engage with their teachers and each other. Twitter, for example, could be used to involve students by creating a dialogue on the subject matter by retweeting key points and messages from lecture material.7 Through Twitter, students can continue a discussion long after the lecture is over and tie information to current events by using hashtags.7 Social media platforms provide a powerful way to integrate

medical education in a global setting. The New England nerstones of medical education. Henry David Thoreau Journal of Medicine Facebook group, for example, has said, “Men have become the tools of their tools.”15 At the more than 500 000 “likes” from users around the world. start of my new rotation and, ultimately, my new career, I Each week a new “Image Challenge” is posted, and Face- hope that we can use these rapid communications tools to book users can comment and give their diagnosis.8 A better our medical education and to become more com- recent systematic review found that social media can passionate, more empathetic and more insightful phys improve collaborative learning and engagement.9 icians of the future. However, many authors have pointed out the chal- Competing interests: None declared. lenges these new technologies present to medical education.2,10 As long as mobile phones are on, students are References bombarded by text messages, emails and phone calls, con- stantly interrupting the day.2 There is the potential for 1. Giordano C, Giordano C. Health professions students’ use of social “distracted doctoring” — mobile technology interfering media. J Allied Health 2011;40:78-81. 9 with the focus on the patient. These technologies also . 2 Dala-Ali BM, Lloyd MA, Al-Abed Y. The uses of the iPhone for have the potential to make students more superficial surgeons. Surgeon 2011;9:44-8. learners without a deeper grasp of the material.2 Instead of 3. Wallace S, Clark M, White J. ‘It’s on my iPhone’: attitudes to the reading the whole chapter on a topic, learners look up bits use of mobile computing devices in medical education, a mixed methods study. BMJ Open 2012;2:e001099. and pieces and end up with chunks of knowledge that 4. Ho K, Lauscher H, Broudo M, et al. The impact of a personal digi- aren’t integrated. In addition, the overwhelming amount tal assistant (PDA) case log in a medical student clerkship. Teach of online information can be difficult for students to sift Learn Med 2009;21:318-26. through and decipher.6 Concerns over patient confidenti- . 5 Anatomy.TV [website of anatomy.tv]. Available: www.anatomy.tv ality and student professionalism have spurred profes- (accessed 2013 Oct. 14). sional associations, such as the Canadian Medical Associa- . 6 White JS, Sharma N, Boora P. Surgery 101: evaluating the use of podcasting in a general surgery clerkship. Med Teach 2011;33:941-3. tion and Canadian Federation of Medical Students, to . 7 Forgie SE, Duff JP, Ross S. Twelve tips for using Twitter as a 11,12 release guidelines regarding social media. Other chal- learning tool in medical education. Med Teach 2013;35:8-14. lenges include blurring of professional/personal bound 8. The New England Journal of Medicine [Facebook page of NEJM]. aries, cost and technical issues.3 Available: www.facebook.com/TheNewEnglandJournalofMedicine It is difficult to assess the impact of rapid communica- (accessed 2013 Oct. 14). tions technology on medical education, largely because it . 9 Cheston CC, Flinckinger TE, Chisolm MS. Social media use in medical education: a systematic review. Acad Med 2013;88:893-901. is such a rapidly changing and diverse area.13 The technol- 10. Ross S, Forgie S. Distracted doctoring: Smartphones before ogy continues to change and progress — what was once patients? CMAJ 2012;184:1440. “hip” is now passé. The challenge for educators is to keep 11. Brasg I, Bryden P, Clouston R, et al. Canadian Federation of Medical up with the speed of innovation while limiting the prob- Students: recommendations for social media. Available: www.cfms.org lematic impacts of these new technologies.3 From a per- /attachments/article/995/CFMS%20Guide%20to%20Social%2Media sonal perspective, it is important that medical educators %20Professionalism.pdf (accessed 2013 Oct. 14). 12. Canadian Medical Association. remember the limitations of technology. Medical students Physician guidelines for online communication with patients. Available: http://policybase.cma.ca/dbtw-wpd need to learn an approach to chest pain, the branches of /PolicyPDF/PD05-03.pdf (accessed 2013 Oct. 14). the aorta and the dosing of morphine; rapid communi 13. McGee JB. Web 2.0 and medical education: It’s here. Are you ready? cations technologies provide new ways of learning that [website of the Association of American Medical Colleges]. Available: information. But beyond simply absorbing medical know www.aamc.org/members/gir/gir_resources/112066/viewpoint_may08 ledge, medical students also need to become insightful .html (acessed 2013 Oct. 14). 14. Pickering GW. The purpose of medical education. BMJ 1956;2:113-6. history-takers, keen observers and compassionate phys 15. Thoreau HD. Walden, and On The Duty of Civil Disobedience. Project 14 icians who can comfort real human beings. Excellent Gutenberg; 1995. Available: www.gutenberg.org/files/205/205 role models and genuine patient interaction are the cor- -h/205-h.htm (accessed 2013 Oct. 14).

E120 J can chir, Vol. 57, No 4, août 2014 RESEARCH • RECHERCHE

How to assess communication, professionalism, collaboration and the other intrinsic CanMEDS roles in orthopedic residents: use of an objective structured clinical examination (OSCE)

Tim Dwyer, MBBS*† Background: Assessing residents’ understanding and application of the 6 intrinsic ‡ CanMEDS roles (communicator, professional, manager, collaborator, health advocate, Susan Glover Takahashi, MA, PhD scholar) is challenging for postgraduate medical educators. We hypothesized that an Melissa Kennedy Hynes, MA‡ objective structured clinical examination (OSCE) designed to assess multiple intrinsic Jodi Herold, PhD§ CanMEDS roles would be sufficiently reliable and valid. David Wasserstein, MD§ Methods: The OSCE comprised 6 10-minute stations, each testing 2 intrinsic roles using case-based scenarios (with or without the use of standardized patients). Residents Markku Nousiainen, MD, MSc, were evaluated using 5-point scales and an overall performance rating at each station. MEd¶ Concurrent validity was sought by correlation with in-training evaluation reports (ITERs) Peter Ferguson, MD† from the last 12 months and an ordinal ranking created by program directors (PDs). Veronica Wadey, MD, MA** Results: Twenty-five residents from postgraduate years (PGY) 0, 3 and 5 participated. M. Lucas Murnaghan, MD, The interstation reliability for total test scores (percent) was 0.87, while reliability for each of the communicator, collaborator, manager and professional roles was greater than 0.8. MEd*††‡‡ Total test scores, individual station scores and individual CanMEDS role scores all showed Tim Leroux, MD§ a significant effect by PGY level. Analysis of the PD rankings of intrinsic roles demonstrated a high correlation with the OSCE role scores. A correlation was seen between * John Semple, MD, MSc ITER and OSCE for the communicator role, while the ITER medical expert and total Brian Hodges, MD, Med, PhD§§ scores highly correlated with the communicator, manager and professional OSCE scores. Darrell Ogilvie-Harris, MD, Conclusion: An OSCE designed to assess the intrinsic CanMEDS roles was suffi- MSc*‡‡¶¶ ciently valid and reliable for regular use in an orthopedic residency program.

From the *Women’s College Hospital, Contexte : Évaluer la compréhension et l’application des 6 rôles intrinsèques Can- †Mt. Sinai Hospital, ‡Postgraduate Medi- MEDS (communicateur, professionnel, gestionnaire, collaborateur, promoteur de la cine Office, University of Toronto, §Uni- versity of Toronto, ¶ Holland Orthopae- santé, érudit) chez les résidents pose un défi pour les responsables de la formation médi- dic and Arthritic Centre, **Sunnybrook cale postdoctorale. Nous avons émis l’hypothèse selon laquelle un examen clinique Health Sciences Centre, ††University of objectif structuré (ECOS) conçu pour évaluer plusieurs rôles CanMEDS intrinsèques Toronto Orthopaedics Sports Medicine, serait suffisamment fiable et valide. ‡‡The Hospital for Sick Children, §§ Méthodes : L’ECOS comportait 6 stations de 10 minutes, permettant chacune d’évaluer Toronto General Hospital, and ¶¶Toronto 2 rôles intrinsèques à l’aide de scénarios basés sur des cas (avec ou sans recours à des Western Hospital, Toronto, Ont. patients standardisés). Les résidents ont été notés au moyen d’échelles en 5 points et d’une évaluation globale de leur rendement à chacune des stations. La validité convergente a été Presented at the International Confer- vérifiée par corrélation avec les rapports d’évaluation en cours de formation (RÉF) des ence Surgical Education and Training 12 mois précédents et un classement chiffré créé par les directeurs du programme (DP). (ICOSET) 2013, Canadian Orthopaedic Association 2013, Canadian Conference Résultats: Vingt-cinq résidents des années 0, 3 et 5 y ont participé. La fiabilité intersta- Medical Education 2013. tion pour les scores totaux aux tests (en pourcentage) a été de 0,87, tandis que la fiabilité pour chacun des rôles de communicateur, collaborateur, gestionnaire et professionnel, a Accepted for publication été supérieure à 0,8. Les scores totaux aux tests, les scores aux stations individuelles et les Oct. 29, 2013 scores pour les rôles CanMEDS individuels ont tous fait état d’un effet significatif selon le niveau des résidents. L’analyse des classements établis par les DP quant aux rôles Correspondence to: intrinsèques a révélé une forte corrélation avec les scores au test ECOS pour les rôles. T. Dwyer On a observé une corrélation entre les RÉF et l’ECOS pour le rôle de communicateur, Women’s College Hospital tandis que les RÉF pour le rôle d’expert médical et les scores totaux ont été en forte cor- 76 Grenville St. rélation avec les scores de l’ECOS pour les rôles de communicateur, de gestionnaire et Toronto ON M5S 1B1 de professionnel. [email protected] Conclusion : Un ECOS conçu pour évaluer les rôles CanMEDS intrinsèques s’est révélé suffisamment valide et fiable pour un usage régulier dans un programme de DOI: 10.1503/cjs.018813 résidence en orthopédie.

how-dwyer.indd 230 14-07-16 10:29 AM RESEARCH

he 7 CanMEDS competencies (medical expert, com- test the 6 intrinsic CanMEDS roles. A focus group of aca- municator, collaborator, manager, health advocate, demic orthopedic specialists was assembled with the goal T scholar and professional) have been clearly outlined of creating clinical scenarios evaluating selected Can- in the CanMEDS 2005 Physician Competency Framework, MEDS principles for each of the roles. The focus group by the Royal College of Physicians and Surgeons of Can- relied on the source document from The Royal College of ada.1 A similar framework has been described by the Accred- Physician and Surgeons of Canada, in which each of the itation Council for Graduate Medical Education (ACGME), roles, and their key competencies, is clearly defined.1 defining 6 core competencies.2 Each of these frameworks The OSCE was 1 hour long and comprised 6 describe the principal generic abilities of physicians in health 10-minute stations. The roles of communicator, collabora- care, and are an integral component of postgraduate educa- tor, professional, manager, health advocate and scholar tion. However, despite the widespread popularity of Can- were assessed; a deliberate attempt was made to avoid test- MEDS and other competency frameworks and the mandate ing the medical expert role. Most of the 6 case-based scen to both teach and assess these competencies, the best meth- arios were designed to assess a primary and secondary role. ods of teaching them remain unknown.3,4 Stations 2–4 used standardized patients (SPs; station 2: rel- Assessment options for the intrinsic roles (those other than ative concern regarding delay in surgery; station 3: grand- the medical expert role) include in-training evaluation reports mother of child with suspected nonaccidental injury; (ITERs), structured oral examinations, 360° assessments and station 4: teenager being informed of osteosarcoma diag- objective structured clinical examinations (OSCE).4 A survey nosis), whereas station 5 used a standardized health profes- of a wide variety of medical and surgical program directors in sional (SHP; operating room manager). Two stations did Canada identified that the ITER is the most commonly used not have an SP or SHP (station 1: ethical approach to method to evaluate the CanMEDS roles, despite its acknow needlestick injuries; station 6: evidence-based medicine in ledged subjective nature.5–7 Respondents reported dissatisfac- spinal surgery). Table 1 lists the roles and associated key tion with current methods of evaluating the intrinsic roles, competencies tested in each station. especially the manager and health advocate roles. The CanMEDS OSCE development was facilitated by The “OSCE” is a term used to describe a variety of multi- an exam blueprint and case development guides. A mem- station examinations and is a format currently favoured at ber of the focus group was assigned as the lead to design orthopedic certification examinations in Canada and other each station, which was then reviewed by the entire focus countries. Studies using OSCEs to assess the role of medical group. Any discrepancies or ambiguities were addressed or expert have demonstrated reliability and validity in postgradu- removed. A number of 5-point performance rating scales ate physician training,8–11 with some literature demonstrating were developed for each of the intrinsic roles. The ratings the ability of the OSCE to assess communication skills.12–17 In were anchored by descriptions of performance to be dem- fact, there is evidence that improved interpersonal skills are onstrated by the residents for each role. An overall 5-point linked to better overall OSCE performance.18–20 global rating was also assigned for each resident at the end There is also some evidence that an OSCE can be used of the station. The SPs and SHPs were selected from an to assess other CanMEDS roles, including the scholar role established standardized patient bank at the University of (application of evidence-base medicine or demonstration Toronto. For the OSCE, 2 SPs/SHPs were trained for of teaching skills)21,22 and the professional role (cultural each of the stations by an experienced SP trainer; each awareness and the application of ethical principles).23,24 An received a minimum of 3 hours’ training for each role. No OSCE has also been used to assess multiple CanMEDS SP or SHP assessment of performance was used. competencies in other fields of postgraduate training, such as radiology and neonatology.6,25 Study design To our knowledge, no research exists regarding methods of assessing the intrinsic roles in orthopedic postgradu- Convenience sampling was used to recruit residents from ate training. We hypothesized that an OSCE designed to specific postgraduate years (PGY) of training: PGY0 assess multiple intrinsic CanMEDS roles would have suffi- (incoming residents who had not yet begun training), cient reliability and validity to distinguish between differ- PGY3 residents and PGY5 residents (volunteers who ent years of postgraduate training in orthopedic residents. asked to sit the OSCE). The PGY5 residents had all recently passed their orthopedic certification examina- Methods tions, and were included as the “gold standard.” Members of the orthopedic faculty at each station Exam development evaluated residents independently. It was not possible to blind examiners from the year of training of the residents, The orthopaedic residency program at the University of as many of the residents were familiar to the staff surgeons. Toronto, in collaboration with the Postgraduate Medical However, examiners were asked to disregard the year of Education Department, designed an orthopedic OSCE to training when making assessments. The OSCE was

Can J Surg, Vol. 57, No. 4, August 2014 231

how-dwyer.indd 231 14-07-16 10:29 AM RECHERCHE

Table 1. Breakdown of each station by primary and secondary conducted in 4 1-hour sessions over the course of a single roles and key competencies tested day. Each candidate signed a consent form permitting the Station; role Key competencies tested* use of exam results for research purposes. On completion of Needlestick the exam, residents were invited to provide feedback using a Professional • Bioethical principles 5-point Likert scale. Summative and formative feedback • Informed consent/confidentiality was given to each resident at the end of the OSCE. • Commitment to professional standards Concurrent validity was sought in 2 ways. First, we Manager • Priority setting obtained the ITERs from the preceding 12 months for the • Time management PGY3 and PGY5 residents, and the results on the 6 intrin- Trauma list sic roles correlated with the OSCE total score and role Manager • Task prioritization and time management • Leadership scores. Second, the 2 program directors (PDs) formed an • Appropriate use of resources in hospital ordinal ranking of the residents in PGY3 and PGY5 and Communicator • Effective listening rated each resident’s ability in each of the CanMEDS roles • Empathy using a 5-point scale (1 = needs significant improvement; • Patient-centred approach to communication 2 = below expectations; 3 = solid, competent performance; Nonaccidental injury 4 = exceeds expectations; 5 = superb). The overall ranking Health advocate • Recognition of patient risk factors and the rating for each role were also correlated to the • Modification of risk factors total OSCE score and role scores. • Patient safety Manager • Leadership Approval for this study was obtained from the Research • Negotiation Ethics Board, University of Toronto. Each resident signed Breaking bad news a consent form to permit the use of the OSCE results and Communicator • Breaking bad news ITER results for research purposes. • Addressing end of life issues • Empathy Statistical analysis Professional • Autonomy in decision making • Bioethical principles All data were deidentified, and residents were assigned a Interacting with OR team study-specific number. Raw scores from the individual sta- Collaborator • Conflict resolution • Respect for members of the health care team tion scores and role scores were entered into a spreadsheet • Recognizing one’s own roles and limitations and analyzed using SPSS version 19. All scores were con- Spinal evidence verted into a percentage, with results reported as mean ± Scholar • Evidence-based medicine standard deviation. Reliability was established using the • Critical appraisal of evidence interstation α coefficient of reliability (Cronbachα ) for • Translating knowledge into practice each of the scoring tools. We evaluated scores from the Communicator • Interactive process different rating scales using regression analysis. The effect • Efficiency and accuracy of PGY on total test scores (%), overall ratings of per OR = operating room. *Based on the CanMEDS 2005 Physician Competency Framework.1 formance, individual station scores and role scores were evaluated using 1-way analysis of variance (ANOVA). We Table 2. Resident distribution, by considered results to be significant atp < 0.05. We used PGY level the Scheffe test for post hoc analysis to understand differences in scores between each possible pair of years of PGY No. of residents training. We determined the correlation between total PGY0 6 OSCE scores and role scores with ITER role scores and PGY3 13 PGY5 6 PD rankings using Pearson correlation and Spearman 2 Total 25 Rho (R ), and the Student t test was used to compare the

PGY = postgraduate year. PD ratings of resident performance between PGY levels.

Results Table 3. Total test scores and overall rating of performance Group; mean ± SD Twenty-five residents from PGY0, 3 and 5 took part in the Test score Grand PGY0 PGY3 PGY5 OSCE (Table 2). The roles of communicator, manager Total, out of 100 75% ± 12.9 62.4% ± 5 73.4% ± 9 91.1% ± 8.3 and professional were assessed in multiple stations; collab- Overall rating of 3.55 ± 0.78 2.75 ± 0.35 3.45 ± 0.50 4.56 ± 0.44 orator, health advocate and scholar were assessed in 1 sta- performance, tion each. The total test scores (converted to a percentage) out of 5 and the mean overall rating of performance are shown PGY = postgraduate year; SD = standard deviation. in Table 3; ANOVA testing demonstrated a significant

232 J can chir, Vol. 57, No 4, août 2014

how-dwyer.indd 232 14-07-16 10:29 AM RESEARCH

difference of the effect of PGY on both scores (p < 0.001). 100 PGY0 We found a significant difference between PGY0 and 90 PGY3 PGY3 (p = 0.039), PGY3 and PGY5 (p = 0.001), and 80 PGY5 PGY0 and PGY5 (p < 0.001). 70 The interstation reliability (percent) was 0.87 for total % correct 60 test scores and 0.83 for overall ratings of performance. The internal consistency for 4 of the 6 role scores is shown in 50 Table 4; the consistency for each of these 4 roles was very 40 high (> 0.80). We were not able to compute internal con 30 sistency coefficients for the scholar and advocate roles, as Station score, 20 only 1 rating scale was used for each of these roles. 10 The total test scores for the individual stations by PGY 0

are displayed in Figure 1. The effect of PGY on the indi NAI

Bad news OR team vidual station scores was significant (stations 1, 5 and 6, all Needlestick Trauma list p < 0.01; stations 2 and 4, both p < 0.05) with the exception Spinal evidence of station 3 (p = 0.07). Post hoc analysis demonstrated a sig Fig. 1. Total station scores (% correct) by postgraduate year (PGY) nificant difference in station scores between the PGY5 and for each of the stations. Each station showed a significant differ- PGY0 groups and between the PGY5 and PGY3 groups ence by PGY (p < 0.05) except for station 3 (nonaccidental injury for all stations except station 3. No significant difference in [NAI], p = 0.07). Error bars represent standard error of the mean. scores was seen between the PGY0 and PGY3 groups, but there was a trend for increased scores in the PGY3 group. The total test scores for each of the intrinsic roles by PGY 100 PGY0 are shown in Figure 2. The ANOVA testing for the effect of 90

t PGY3 PGY on each of the role scores was significant (communica 80 PGY5 tor, collaborator, manager and professional roles, all p < 70 0.001; health advocate and scholar roles, both p < 0.05). For

% correc 60 each of the role scores, the PGY0 and PGY3 groups differed 50 significantly from the PGY5 group, but not from each other. Analysis of the PD ratings of intrinsic roles demon 40 strated a good correlation between these and the corres 30 ponding OSCE role scores (Table 5). The ITERs from Station score, 20 12 months before the OSCE were available for the PGY3 10 and PGY5 residents. There was no correlation between 0 ITERs and OSCE scores within any role except for com municator (0.64); however, the ITER overall scores correl ated with the communicator (0.58), manager (0.51) and professional (0.56) OSCE role scores. Fig. 2. Total test scores (% correct) for each of the CanMEDS roles There was a 64% (16 of 25) response rate to the resi by postgraduate year (PGY). Analysis of variance showed signifi- dent survey. Overall, 87.6% of respondents agreed or cant differences for all roles (communicator, collaborator, manager, professional, all p < 0.001; advocate, scholar, both p < 0.05). strongly agreed that the scenarios reflected encounters that an orthopedic surgeon would have to deal with in general practice, and 81.3% agreed or strongly agreed that partici pating in the OSCE would help prepare them for their Table 5. Correlation between the program directors’ ratings of final Royal College examination. However, only 56.3% resident ability in each of the agreed or strongly agreed that the OSCE was an effective intrinsic roles and the corresponding way to assess their understanding of each of the Can OSCE role score MEDS roles. PD role rating OSCE role score Communicator 0.79 Table 4. Internal consistency for the 4 role scores with more than 1 rating scale Collaborator 0.65 Manager 0.66 Role Coefficient Item numbers α Health advocate 0.74 Communicator 0.91 17 items in 5 of 6 stations Scholar 0.70 Collaborator 0.96 3 items in 1 of 6 stations Professional 0.72

Manager 0.83 5 items in 3 of 6 stations OSCE = objective structured clinical examination; Professional 0.84 3 items in 2 of 6 stations PD = program director.

Can J Surg, Vol. 57, No. 4, August 2014 233

how-dwyer.indd 233 14-07-16 10:29 AM RECHERCHE

Discussion Jefferies and colleagues6 recently demonstrated that the OSCE may be a valid and reliable method of simul This orthopedic OSCE, designed to test the 6 intrinsic taneously assessing multiple competencies in neonatal/ CanMEDS roles, has shown excellent overall reliability, as perinatal medicine. Subspecialty trainees were assessed well as excellent reliability for the communicator, collab using a combination of binary checklists, 5-point Can orator, professional and manager roles. Furthermore, MEDS ratings, as well as SPs’ and SHPs’ assessments of using role-specific global ratings, we demonstrated an interpersonal and communication skills. Interstation reli ability to distinguish between orthopedic residents at dif ability was acceptable to excellent for 6 of the 7 roles, with ferent levels of training. To our knowledge, this is the first the scholar role being the exception. Only the teaching time formal assessment of intrinsic roles has been studied component of the scholar role was assessed; the authors in the field of postgraduate orthopedic surgical training. recommended creating a single station to assess the com With regards to total test scores, each of the stations petencies inherent to the scholar role, including the ability were able to demonstrate a statistically significant differ to understand and evaluate research. We applied this tech ence by year of training, with the exception of station 3. nique with success; our scholar station, designed to assess Station 3 was a case-based scenario using a standardized application of evidence-based medicine, was able to distin patient — a grandmother who has brought in a child guish between residents with different levels of training. thought to have sustained a nonaccidental injury. Residents Jefferies and colleagues3 also studied the use of the were asked to take a focused history regarding the home structured oral examination in assessing the 7 CanMEDs situation and explain to the grandmother the need to alert roles, including the medical expert role. Interstation reli the appropriate authorities and admit the child. Despite ability was acceptable for the roles of medical expert, the fact that the PGY effect on station 3 scores was not sta scholar and professional (0.6–0.8), but not for the roles of tistically significant p( = 0.07), the scores demonstrate the communicator, collaborator or health advocate (0.4–0.6) or same general trend as all other stations: PGY5 (mean for the role of manager (0.19). In comparison to their pre 90.5%) > PGY3 (mean 74.3%) > PGY1 (mean 69.5%). It vious OSCE study, interstation reliability was lower for all may be that this station did not achieve significance roles except scholar. However, costs were reduced signifi because the PGY0 and PGY3 residents performed well, cantly by not using standardized patients. Although we felt suggesting that these competencies may have been covered that SPs were an important component of our OSCE, the in undergraduate medical programs. costs (in the region of $3000) were not insignificant, equat Careful blueprinting was used in this CanMEDS OSCE ing to a cost of $250 per resident. It may be possible to to avoid redundancy. Roles were spread among stations, substitute orthopedic fellows or staff in place of SPs in and the stations that assessed the same roles focused on future iterations. However, given the importance of estab different competencies within that role, as outlined in the lishing competence in these areas by both the Royal Col CanMEDS 2005 Framework.1 For example, in this OSCE, lege and the ACGME, this could be seen as a reasonable 2 stations (needlestick and trauma list) both examined the cost for training programs to bear on an annual basis. roles of communicator and manager, with the needlestick The Royal College has a published handbook detailing station additionally examining resident understanding of assessment methods for the CanMEDS competencies.27 It the professional role (bioethical principles and informed states that oral examinations and OSCEs are not well consent). However, the trauma list station focused on the suited to evaluate the roles of manager and scholar. Other competencies of priority setting and time management documents attest to the perceived difficulty with assessing within the manager role, while the needlestick station the intrinsic roles, especially the role of health advo sought to examine the competency of managing practice cate.5,28,29 However, the reliability of the manager role in and career effectively. our study was high enough to be used in a high-stakes The OSCE has been previously been shown to be a examination. While we cannot attest to the reliability of valid and reliable tool for the assessment of the medical the health advocate and scholar roles owing to insufficient expert role, with some evidence for its use in assessing the items, ANOVA testing demonstrated a significant ability communicator role.12–14 Improved communication skills to distinguish between residents of different PGY levels for have previously been linked to both advanced year of train both of these roles. We believe that the OSCE is a very ing18 and to increased clinical competence.19 The OSCE appropriate means of assessment, as clinical scenarios that has also been adapted to assess competencies within the mimic real life encounters can be used. roles of professional23,24,26 and scholar,21,22 with varying An advantage of this type of OSCE is that both teaching amounts of success. For example, Singer and colleagues,24 (formative evaluation) and assessment (summative evalua in an OSCE designed to assess clinical ethics, found a low tion) can be incorporated. As noted by Zuckerman and reliability with only 4 stations; it was felt that increasing colleagues,4 assessment motivates residents to learn impor the number of stations would be required to obtain accept tant skills and is therefore a form of learning in itself. We able reliability. believe that by exposing very junior residents (PGY0) to

234 J can chir, Vol. 57, No 4, août 2014

how-dwyer.indd 234 14-07-16 10:29 AM RESEARCH

scenarios they will likely soon encounter (complaints of physicians.15,19,31 Importantly, the examiners will have delayed surgery, difficult interaction with operating room known some residents and their PGY of training, raising staff), learning opportunities can be created in an environ- the potential for bias. Examiners were asked to disregard ment suitable for feedback and coaching.30 Furthermore, the PGY level of the resident; however, it may be that the by retesting mid-rank residents (PGY3), an assessment use of SPs’ ratings will help to offset this risk. In this of their skills in each of the CanMEDS roles can be re- OSCE, neither station nor role weighting was used, as it evaluated, and appropriate feedback can be provided. At was felt that each of the CanMEDS roles was equally our institution, a bank of multiple CanMEDS scenarios important. Finally, it is not possible to know how this has been created; we believe that all residents will benefit CanMEDS OSCE compares to a more traditional OSCE from exposure to a CanMEDs OSCE twice in their train- with incorporated assessment of CanMEDS roles within ing, once as a junior and once as a senior resident. those stations; however, we have demonstrated a high We are not aware of any OSCE designed to test only degree of reliability or internal consistency, a measure that the intrinsic CanMEDS roles. While it is difficult to indicates an exam is performing well. It may be that the remove the medical expert role from such an examination, high degree of reliability seen in this CanMEDS OSCE every effort was made to minimize scenarios dependent on may be a result of its narrow focus. orthopedic knowledge. For example, in the station focusing on the role of manager, residents were asked to man- Conclusion age an overbooked trauma list; some degree of orthopedic knowledge was required to know the urgency of each case, An OSCE designed to assess the intrinsic CanMEDS but residents were graded on their reasoning and on their roles proved to be sufficiently valid and reliable for regular ability to handle a phone call from a disgruntled relative. use in an orthopedic residency program. In the scholar station, residents were expected to know Competing interests: None declared. levels of evidence and how to perform database searches; in Contributors: T. Dwyer, P Ferguson, M.L. Murnaghan, B. Hodges the needlestick case (professional role) residents were and D. Ogilvie-Harris designed the study. T. Dwyer, D. Wasserstein, expected to know the immediate and delayed management M. Nousiainen, P. Ferguson, V. Wadey, T. Leroux and D. Ogilvie- of such an occurrence as well as the ethical principles Harris acquired the data, which T. Dwyer, S. Glover-Takahashi, M. Hynes, J. Herold, J.L. Semple, B. Hodges and D. Ogilvie-Harris involved regarding patient consent and notification of the analyzed. T. Dwyer, S. Glover-Takahashi, M. Hynes, J. Herold, appropriate monitoring bodies. For this reason, we do not B. Hodges and D. Ogilvie-Harris wrote the article, which all authors believe that there were any major qualitative differences reviewed and approved for publication. regarding the degree of core knowledge assessed in each station. References We were interested in obtaining concurrent validity; for 1. Frank JR. The CanMEDS 2005 physician competency framework. example how could the station creator be certain that com- Ottawa (ON): The Royal College of Physicians and Surgeons of munication stations were truly assessing the communicator Canada; 2005. role. All case scenarios were based on real-life clinical situ- 2. ACGME. Common program requirements: general competencies ations, and adherent to role descriptions provided by the [website of ACGME]. Available: www.acgme.org/acgmeweb Royal College of Physicians and Surgeons Canada.1,6 /Portals/0/PFAssets/ProgramRequirements/CPRs2013.pdf (acces sed 2013 Mar. 25). Interestingly, no correlation was seen between the ITER 3. Jefferies A, Simmons B, Ng E, et al. Assessment of multiple physician role scores and the OSCE role scores, but a good correla- competencies in postgraduate training: utility of the structured oral tion was seen with program director ratings of the roles. examination. Adv Health Sci Educ Theory Pract 2011;16:569-77. This suggests that ITERs are not a particularly effective 4. Zuckerman JD, Holder JP, Mercuri JJ, et al. Teaching professional- form of assessment for the intrinsic roles. ism in orthopaedic surgery residency programs. J Bone Joint Surg Am 2012;94:e51. 5. Chou S, Cole G, McLaughlin K, et al. CanMEDS evaluation in Limitations Canadian postgraduate training programmes: tools used and programme director satisfaction. Med Educ 2008;42:879-86. Limitations included our inability to comment on the reli- 6. Jefferies A, Simmons B, Tabak D, et al. Using an objective struc- ability of the roles of scholar and health advocate, as only tured clinical examination (OSCE) to assess multiple physician competencies in postgraduate training. Med Teach 2007;29:183-91. a single global rating was used for each of these roles — 7. Catton PTS, Rothman A. A guide to resident assessment for pro- this will be remedied in the future. However, each of these gram directors. Ann Roy Coll Phys Surg Can 1997;403-9. roles was useful in distinguishing between different years 8. Schwartz RW, Witzke DB, Donnelly MB, et al. Assessing residents’ of training. Objectivity may have been increased by the clinical performance: cumulative results of a four-year study with use of SPs or SHPs to provide global ratings of the resi- the Objective Structured Clinical Examination. Surgery 1998;124: 307-12. dents, a method that has been used to good effect in the 9. Hodges B, Regehr G, Hanson M, et al. Validation of an objective medical education literature, with evidence of good cor structured clinical examination in psychiatry. Acad Med 1998;73: relation between ratings completed by SPs and faculty 910-2.

Can J Surg, Vol. 57, No. 4, August 2014 235

how-dwyer.indd 235 14-07-16 10:29 AM RECHERCHE

10. Cohen R, Reznick RK, Taylor BR, et al. Reliability and validity of 21. Fliegel JE, Frohna JG, Mangrulkar RS. A computer-based OSCE the objective structured clinical examination in assessing surgical station to measure competence in evidence-based medicine skills in residents. Am J Surg 1990;160:302-5. medical students. Acad Med 2002;77:1157-8. 11. O’Sullivan P, Chao S, Russell M, et al. Development and imple- 22. Schol S. A multiple-station test of the teaching skills of general mentation of an objective structured clinical examination to provide practice preceptors in Flanders, Belgium. Acad Med 2001;76:176-80. formative feedback on communication and interpersonal skills in 23. Altshuler L, Kachur E. A culture OSCE: teaching residents to geriatric training. J Am Geriatr Soc 2008;56:1730-5. bridge different worlds. Acad Med 2001;76:514. 12. Srinivasan J. Observing communication skills for informed consent: 24. Singer PA, Pellegrino ED, Siegler M. Clinical ethics revisited. an examiner’s experience. Ann R Coll Physicians Surg Can 1999;32: BMC Med Ethics 2001;2:E1. 437-40. 25. Probyn LF, Finley K. The CanMEDS objective structured clinical 13. Keely E, Myers K, Dojeiji S. Can written communication skills be examination (OSCE): Evaluating outside of the box — an example tested in an objective structured clinical examination format? Acad from diagnostic radiology. Ottawa (ON): Royal College of Phys Med 2002;77:82-6. icians and Surgeons of Canada; 2010. Available: www.royalcollege 14. Yudkowsky R, Alseidi A, Cintron J. Beyond fulfilling the core com- .ca/portal/page/portal/rc/common/documents/canmeds/whatworks petencies: an objective structured clinical examination to assess /canmeds_probyn_finlay_osce_e.pdf. communication and interpersonal skills in a surgical residency. Curr 26. Hilliard RIT, Tabak S.E. D. Use of an objective structured clinical Surg 2004;61:499-503. examination as a certifying examination in pediatrics. Ann R Coll 15. Donnelly MB, Sloan D, Plymale M, et al. Assessment of residents’ Physicians Surg Can 2000;33:222-8. interpersonal skills by faculty proctors and standardized patients: a 27. Bandiera GS, Sherbino J, Frank JR. The CanMEDS assessment psychometric analysis. Acad Med 2000;75:S93-5. tools handbook: an introductory guide to assessment methods for the 16. Hodges B, McIlroy JH. Analytic global OSCE ratings are sensitive CanMEDS competencies. Ottawa (ON): Royal College of Physicians to level of training. Med Educ 2003;37:1012-6. and Surgeons of Canada; 2006. 17. Hodges B, Turnbull J, Cohen R, et al. Evaluating communication 28. Verma S, Flynn L, Seguin R. Faculty’s and residents’ perceptions of skills in the OSCE format: reliability and generalizability. Med Educ teaching and evaluating the role of health advocate: a study at one 1996;30:38-43. Canadian university. Acad Med 2005;80:103-8. 18. Warf BC, Donnelly MB, Schwartz RW, et al. The relative contri- 29. Murphy DJ, Bruce D, Eva KW. Workplace-based assessment for butions of interpersonal and specific clinical skills to the perception general practitioners: using stakeholder perception to aid blueprint- of global clinical competence. J Surg Res 1999;86:17-23. ing of an assessment battery. Med Educ 2008;42:96-103. 19. Colliver JA, Swartz MH, Robbs RS, et al. Relationship between 30. Duffy FD, Gordon GH, Whelan G, et al. Assessing competence in clinical competence and interpersonal and communication skills in communication and interpersonal skills: the Kalamazoo II report. standardized-patient assessment. Acad Med 1999;74:271-4. Acad Med 2004;79:495-507. 20. Sloan DA, Donnelly MB, Johnson SB, et al. Assessing surgical resi- 31. Cooper C, Mira M. Who should assess medical students’ communi- dents’ and medical students’ interpersonal skills. J Surg Res 1994; cation skills: their academic teachers or their patients? Med Educ 57:613-8. 1998;32:419-21.

Online manuscript submission and peer review

AVAILABLE at http://mc.manuscriptcentral.com/cjs

236 J can chir, Vol. 57, No 4, août 2014

how-dwyer.indd 236 14-07-16 10:29 AM RESEARCH • RECHERCHE

Parathyroid hormone levels 1 hour after thyroidectomy: an early predictor of postoperative hypocalcemia

Awad AlQahtani, MD, MSc*§ Background: Parathyroid dysfunction leading to symptomatic hypocalcemia is not Armen Parsyan, MD† uncommon following a total or completion thyroidectomy and is often associated with significant patient morbidity and a prolonged hospital stay. A simple, reliable indicator Richard Payne, MD‡ to identify patients at risk would permit earlier pharmacologic prophylaxis to avoid Roger Tabah, MD*† these adverse outcomes. We examined the role of intact parathormone (PTH) levels 1 hour after surgery as a predictor of post-thyroidectomy hypocalcemia. From the *Division of Surgical Oncology, Methods: We prospectively reviewed the cases of consecutive patients undergoing †Division of General Surgery, ‡Department total or completion thyroidectomy. Ionized calcium (Ca2+) and intact PTH levels of Otolaryngology, McGill University, were measured preoperatively and at 1-, 6- and 24-hour intervals postoperatively. Montréal, Que., and the §Division of The specificity, sensitivity, negative and positive predictive values of the 1-hour General Surgery, King Saud University, PTH serum levels (PTH-1) in predicting 24-hour post-thyroidectomy hypocalcemia Riyadh, Saudi Arabia and eucalcemia were determined. Accepted for publication Results: We reviewed the cases of 149 patients. Biochemical hypocalcaemia (Ca2+ < Nov. 11, 2013 1.1 mmol/L) developed in 38 of 149 (25.7%) patients 24 hours after thyroidectomy. The sensitivity, specificity, positive and negative predictive values of a low PTH-1 Correspondence to: were 89%, 100%, 97% and 100%, respectively. A. AlQahtani 1603–350 Maisonneuve West Conclusion: We found that PTH-1 levels were predictive of symptomatic hypo- Montréal QC H3A 0B4 calcemia 24 hours after thyroidectomy. Routine use of this assay should be con [email protected] sidered, as it could prompt the early administration of calcitriol in patients at risk of hypocalcemia and allow for the safe and timely discharge of patients expected to remain eucalcemic. DOI: 10.1503/cjs.008013 Contexte : Il n’est pas rare qu’un dysfonctionnement des glandes parathyroïdes entraînant une hypocalcémie symptomatique s’observe après une thyroïdectomie totale ou de complétion et il est souvent associé à une importante morbidité chez les patients et à un séjour hospitalier prolongé. Un indicateur simple et fiable permettant de reconnaître les patients à risque pourrait favoriser une prophylaxie pharmacologique précoce afin d’éviter ces complications. Nous avons examiné le rôle des taux de parathormone (PTH) intacte une heure après la chirurgie comme prédicteurs de l’hypocalcémie post-thyroïdectomie. Méthodes : Nous avons passé en revue de manière prospective des cas consécutifs de patients soumis à une thyroïdectomie totale ou de complétion. Les taux de calcium ionisé (Ca2+) et de PTH intacte ont été mesurés avant l’intervention, puis 1 heure, 6 heures et 24 heures après. Il a ainsi été possible de déterminer la spécifi cité, la sensibilité, la valeur prédictive négative et positive des taux sériques de PTH 1 heure après l’intervention (PTH-1) pour ce qui est de prédire l’hypocalcémie et l’eucalcémie 24 heures après la thyroïdectomie. Résultats : Nous avons analysé 149 cas. L’hypocalcémie biochimique (Ca2+ < 1,1 mmol/L) a été observée chez 38 patients sur 149 (25,7 %) 24 heures après la thy- roïdectomie. La sensibilité, la spécificité, la valeur prédictive positive et négative d’un taux de PTH-1 faible ont été respectivement de 89 %, 100 %, 97 % et 100 %. Conclusion : Nous avons noté que les taux de PTH-1 étaient prédictifs d’une hypocalcémie symptomatique 24 heures après la thyroïdectomie. L’utilisation d’emblée de ce test est à envisager puisqu’elle permettrait l’administration pré- coce de calcitriol chez les patients exposés à un risque d’hypocalcémie et un congé sécuritaire et rapide chez les patients dont on s’attend à ce qu’ils demeur- ent eucalcémiques.

para-alqahtani.indd 237 14-07-16 10:38 AM RECHERCHE

arathyroid dysfunction leading to hypocalcemia is not surgery (benign v. malignant lesion) and type of surgery uncommon after total or completion thyroidectomy (completion v. total thyroidectomy). Preoperative serum ion- and, if symptomatic, is often associated with significant ized calcium (Ca2+) levels and preoperative PTH (PTH-P) P 1 patient morbidity and a prolonged hospital stay. Often, para- levels were recorded; we also recorded PTH levels deter- thyroid tissue is deliberately or inadvertently resected with the mined 1, 6 and 24 hours postoperatively (PTH-1, PTH-6 surgical specimen. When recognized intraoperatively, nor- and PTH-24, respectively). Ionized calcium was measured mal, devitalized glands should be autotransplanted into the on a blood gas machine, the ABL800 Flex (Radiometer). surrounding skeletal muscle. More often though, the Normal Ca2+ ranged from 1.10–1.32 mmol/L. Serum intact observed fall in parathormone (PTH) levels is a result of trau- PTH was measured using the Roche Elecsys 2010 System matic injury to the parathyroids during thyroidectomy lead- electrochemiluminescence immunoassay (Roche Diagnotics). ing to either transient ischemia or even frank infarction of the The normal PTH level in our laboratory ranges from 1.5– glands. Thus, the hyposecretion of PTH may be relatively 6.9 pmol/L. Hypocalcaemia was defined as at least 1 ionized transient (1.6%–68%) or may be permanent (0.4%–33%).2,3 serum calcium measurement below 1.10 mmol/L (normal Intact PTH is secreted by the parathyroid glands in range: 1.1–1.32 mmol/L). Oral calcium supplementation response to serum ionized calcium levels. Intact PTH in an with or without calcitriol was given to patients in whom 84-amino acid protein and control of its release is under direct symptomatic hypocalcemia developed or when the serum feedback from extracellular calcium ions, as mediated through calcium level was less than 1.0 mmol/L. Patients with acral the cell surface calcium receptor of the parathyrocyte.4 The numbness, paresthesias, a positive Chvostek or Trousseau half-life of intact PTH is measured in minutes and can be reli- sign, cardiac arrhythmias or muscular spasms with or without ably assayed.5,6 It is degraded into several smaller proteins with stridor were considered symptomatic.5,6 Severe hypocalcemia variable half-lives and biologic activity.6 The appearance of was defined as Ca2+ of 0.9 mmol/L or less.An intravenous postoperative hypocalcemia may be delayed up to 48 hours calcium gluconate (10%) infusion was reserved for patients depending on the levels of such biologically active peptides as with severe, symptomatic hypocalcaemia. well as the patient’s vitamin D and electrolyte status and the presence or absence of “hungry” bone, among other factors.4,5 Statistical analysis Because of the lag between the hyposecretion of PTH and the development of symptomatic hypocalcemia, early We calculated Pearson correlation coefficients. All statistical pharmacologic support with calcium, magnesium and cal- analyses were performed using Microsoft Excel 2011. citriol can potentially keep high-risk patients eucalcemic and asymptomatic, thereby avoiding morbidity. Early prophylac- Results tic administration of calcitriol is important in maintaining eucalcemia in high-risk patients because, as a fat-soluble vita- We reviewed the cases of 149 patients (14 [9.4%] men and min, its pharmacokinetics are such that an increase in serum 135 [90.6%] women). The mean age of the entire cohort calcium may take 24–48 hours. Thus, the early identification was 57.8 years and the median age was 58 years. Of the of post-thyroidectomy patients at risk of hypoparathyroidism entire cohort, 136 (91.3%) patients underwent total thyroid- and hypocalcemia would allow for the early introduction ectomy and 13 (8.7%) underwent completion thyroidec- of calcitriol.7 Conversely, patients at low risk of post- tomy for malignancy. On final histologic review, 140 (94%) thyroidectomy hypocalcemia can be spared the discomfort of patients had diagnoses of primary thyroid cancer, whereas excessive blood tests and can reliably be discharged without the 9 (6%) remaining patients had benign lesions. fear of symptomatic hypcocalcemia. A significant number of patients had hypoparathormon The present study correlates preoperative serum PTH emia (as defined by a PTH < 1.5 pmol/L (normal range: 1.5– and ionized calcium (Ca2+) levels with those obtained at 1, 6.9 pmol/L). The incidence at 1, 6 and 24 hours postoperative 6 and 24 hours after total thyroidectomy with an emphasis was 34 (22.8%), 36 (24.2%) and 38 (25.5%), respectively on the 1-hour PTH (PTH-1) and the subsequent develop- (Table 1). The incidence of hypoparathormonemia at 1, 6 and ment of hypocalcemia.

Table 1. Incidence of hypoparathomonemia and Methods hypocalcemia after total or completion thyroidectomy, n = 149

We retrospectively reviewed the cases of consecutive patients Group; % undergoing total or completion thyroidectomy at a single 2+ institution (Montreal General Hospital) between July 2009 Post-thyroidectomy PTH < 1.5 pmol/L Ca < 1.1 mmol/L and February 2011. No patients with coexisting parathyroid 1 h 22.8 2 or renal pathology were included. The data were generated 6 h 24.2 5 24 h 25.5 25.5 as part of our adopted thyroidectomy care pathway. We 2+ recorded the following information: age, sex, indication for Ca = Ionized calcium; PTH = parathormone.

238 J can chir, Vol. 57, No 4, août 2014

para-alqahtani.indd 238 14-07-16 10:38 AM RESEARCH

24 hours postoperative seemed to plateau early and changed Discussion little. Furthermore, the mean PTH levels in hypocalcemic patients also attained an early plateau (Fig. 1). In eucalcemic Hypocalcemia post-thyroidectomy is the most common post- patients at 24 hours postoperative, the mean PTH levels, operative complication, with a reported incidence of although significantly higher, tended to decrease slightly by 1%–50%.1,6 In the present study, the percentage of hypocalce- PTH-24 (Fig. 1). The Pearson r coefficients comparing mic patients reached 25.5% at 24 hours postoperative. Early PTH-1 and PTH-6 with PTH-24 were 0.80 and 0.95, identification of patients at risk of hypocalcemia would allow respectively, thus reflecting the relative stability of PTH meas for prophylactic treatment, thus avoiding the development of urements within the first 24 hours postoperative. symptomatic hypocalcemia.8 Conversely, patients who are not Hypocalcemia, as defined by Ca2+ less than 1.1 mmol/L at risk can be spared repeated blood work and can safely be (normal range: 1.1–1.32 mmol/L) at 1, 6 and 24 hours discharged without fear of returning with hypocalcemia.9 postoperative was documented in 3 (3.5%), 8 (5.4%) and Early postoperative calcium monitoring, although impor- 38 (25.5%) patients, respectively (Table 1). To assess the tant, is a poor predictor of subsequent symptomatic hypo- ability of Ca2+-1 measurements to predict hypocalcemia calcemia. Despite the fact that the slope of postoperative 24 hours postoperative, we performed sensitivity and serum calcium levels correlates with the development of specificity analyses. Of 149 patients, only 3 (3.5%) had symptomatic hypocalcemia, its utility is limited as the results low Ca2+-1; Ca2+-24 was low in 38 (25.5%) patients. Thus, are not available until 24–48 hours post-thyroidectomy.8,9 the sensitivity of a low Ca2+-1 in predicting a low Ca2+-24 Combined with the lag between oral calcitriol and its onset was only 11%. The specificity, however, was 100%. The of action, the delays in achieving calcium homeostasis negative (NPV) and positive predictive values (PPV) were become excessive. The routine administration of calcitriol 76% and 80%, respectively. Thus, Ca2+-1 is of limited use and/or calcium to all patients either pre- or postoperatively in predicting hypocalcemia 24 hours post-thyroidectomy. are strategies that have been advocated.10,11 This is reflected in the Pearsonr coefficients comparing Other approaches to monitor and predict postoperative Ca2+-1 and Ca2+-6 with Ca2+-24 of only 0.33 and 0.41, parathyroid function and subsequent hypocalcemia have respectively. Unless they are low, Ca2+-1 levels are a poor been described in the literature. These include PTH assayed predictor of hypocalcemia at 24 hours postoperative. either intraoperatively or in the early postoperative Of the 38 patients in whom hypocalcemia developed period.10–13 These approaches are based on the fact that intact 24 hours post-thyroidectomy, PTH-1 levels less than PTH has a short half-life of 1–4 minutes,14,15 thus allowing 1.15 pmol/L (normal range: 1.5–6.9 pmol/L) in 34. Thus, the detection of its fall early in the perioperative period. Early sensitivity of the PTH-1 in predicting the incidence of hypo- diagnosis of hypocalcemia and the relevance of PTH levels calcemia at 24 hours postoperative is 89%. The specificity of for enabling an earlier diagnosis after total thyroidectomy PTH-1 less than 1.5 pmol/L in predicting hypocalcemia at have been the subject of numerous studies and several 24 hours postoperative is 100%. The NPV and PPV were authors.14–19 These studies vary widely in terms of trial type, 97% and 100%, respectively (Table 2). Thus, PTH-1 is highly methodology, measurement technique, timing of measure- reliable in predicting which patients are at risk of hypocalcemia ment, protocols and thresholds maintained. and which will remain eucalcemic the day after surgery. Pattou and colleagues3 found that a postoperative PTH Of our entire cohort, 30 patients were discharged on level of 12 pg/mL or less was a good predictor of hypocalce- oral medication consisting of combinations calcium, cal- mia, but they did not state how long after surgery PTH citriol and magnesium.

Table 2. Sensitivity and specificity analysis of 6 Hypocalcemic PTH-1 measurement comparing Ca2+-24 and Eucalcemic hypocalcemia 5 Factor Ca2+ < 1.1 mmol/L Ca2+ ≥ 1.1 mmol/L 4 PTH-1 > 1.5 pmol/L, 4 111 no. 3 PTH PTH-1 ≤ 1.5 pmol/L, 34 0 no. 2 Total no. 38 111 1 Sensitivity 89% Specificity 100% 0 Preoperative 1 h6 h24 h NPV 97% Time PPV 100% Ca2+ = Ionized calcium; NPV = negative predictive value; PPV = positive predictive value; PTH = parathormone. Fig. 1. Mean postoperative parathormone (PTH) in hypocalcemic and eucalcemic patients.

Can J Surg, Vol. 57, No. 4, August 2014 239

para-alqahtani.indd 239 14-07-16 10:38 AM RECHERCHE

15 15 Contributors: A. Alqahtani and R.J. Tabah participated in all manu values were obtained. Lombardi and colleagues found script preparation activities. A. Parsyan and R. Payne designed the study greater precision with measurements taken at 4 and 6 hours, and reviewed and approved the final version for publication. with an overall accuracy of 98%. Lam and Kerr16 reported that all patients with a PTH level less than 8 pg/mL meas References ured 1 hour after the surgery became hypocalcemic, and all patients with a PTH level greater than 9 pg/mL did not. 1. Reeve T, Thompson NW. Related complications of thyroid surgery: Higgins and colleagues17 demonstrated that 64% of those how to avoid them, how to manage them, and observations on their possible effect on the whole patient. World J Surg 2000;24:971-5. patients who subsequently required calcium supplementation 2. Thomusch O, Machens A, Sekulla C, et al. The impact of surgical had a decrease in PTH levels greater than 75% from base technique on postoperative hypoparathyroidism in bilateral thyroid line 20 minutes after surgery, and 74% of those who did not surgery: a multivariate analysis of 5846 consecutive patients. Surgery need calcium supplementation demonstrated a decrease of 2003;133:180-5. less than 75% from baseline. The Australian Endocrine 3. Pattou F, Combemale F, Fabre S, et al. Hypocalcemia following thy Guidelines,18 published in 2007, adopted the recom roid surgery: incidence and prediction of outcome. World J Surg 1998;22:718-24. 15 mendations of Lombardi and colleagues to standardize 4. Mundy GR, Guise TA. Hormonal control of calcium homeostasis. obtaining a PTH level 4 hours after a thyroidectomy. The Clin Chem 1999;45:1347-52. wide variability of the predictors for the development of 5. Lindblom P, Westerdahl J, Bergenfelz A. Low parathyroid hor hypocalcemia across centres suggests that the measurement mone levels after thyroid surgery: a feasible predictor of hypocalce of PTH at any time in the postoperative period may be a mia. Surgery 2002;131:515-20. 19 6. Falk SA. Metabolic complication of thyroid surgery: hypocalcemia reliable predictor of hypocalcemia. We believe that obtain and hypoparathyroidisim; hypocalcetonemia; and hypothyroidism ing a PTH level 1 hour after an operation is the optimal time and hyperthyroidism. In: Falk SA, editor. Thyroid disease. 2nd ed. to predict the need for calcium and/or calcitriol supplemen Philadelphia (PA): Lippincott-Raven; 1997:717-745. tation after discharge from the hospital. 7. Quiros RM, Pesce CE, Wilhelm SM, et al. Intraoperative parathy The present study demonstrates that PTH-1, PTH-6, roid hormone levels in thyroid surgery are predictive of postopera and PTH-24 assays are good predictors of hypocalcemia tive hypoparathyroidism and need for vitamin D supplementation. Am J Surg 2005;189:306-9. 24 hours post-thyroidectomy. A PTH-1 assay alone is 8. Fahmy FF, Gillett D, Lolen Y, et al. Management of serum calcium predictive of the development of postoperative hypocal levels in post-thyroidectomy patients. Clin Otolaryngol 2004;29:735-9. cemia at 24 hours, with a sensitivity of 89% and an NPV 9. Grodski S, Farrell S. Early postoperative PTH levels as a predictor of 97%. Practically, we found no significant advantage in of hypocalcemia and facilitating safe early discharge after total thy determining PTH-6 and PTH-24 in asymptomatic roidectomy. Asian J Surg 2007;30:178-82. 10. Lindblom P, Westerdahl J, Bergenfelz A. Low parathyroid hormone patients with a PTH-1 greater than 1.5 pmol/L. levels after thyroid surgery: a feasible predictor of hypocalcemia. Should PTH-1 be greater than 1.5 pmol/L, asymptom Surgery 2002;131:515-20. atic patients can be discharged without the need for further 11. Scurry WC Jr, Beus KS, Hollenbeak CS, et al. Perioperative para routine calcium monitoring. Should PTH-1 be less than thyroid hormone assay for diagnosis and management of post 1.5 pmol/L, patients can be administered a “loading dose” of thyroidectomy hypocalcemia. Laryngoscope 2005;115:1362-6. calcitriol early in the recovery room. This is of practical 12. Richards ML, Bingener-Casey J, Pierce D, et al. Intraoperative parathyroid hormone assay — an accurate predictor of symptomatic importance given a lag of 24–48 hours before calcitriol hypocalcemia following thyroidectomy. Arch Surg 2003;138:632-5. exhibits its clinical effects. 13. Wiseman JE, Mossanen M, Ituarte PH, et al. An algorithm The purpose of the present study was to identify a simple informed by the parathyroid hormone level reduces hypocalcemic predictor of early postoperative hypocalcemia, which may complications of thyroidectomy. World J Surg 2010;34:532-7. either be transient (lasting weeks–months) or may persist. 14. Sywak MS, Palazzo FF, Yeh M, et al. Parathyroid hormone assay predicts hypocalcaemia after total thyroidectomy. ANZ J Surg 2007;77:667-70. Factors other than PTH-1, such as extent of dissection, 15. Lombardi CP, Raffaelli M, Princi P. Early prediction of postthyroid number of parathyroids visualized intraoperatively, number ectomy hypocalcemia by one single iPTH measurement. Surgery of parathyroids identified in the pathologic specimen, and 2004;136:1236-41. vitamin D levels, should be taken into account to predict 16. Lam A, Kerr P. Parathyroid hormone: an early predictor of post chronic hypoparathyroidism and the need for calcitriol and thyroidectomy hypocalcemia. Laryngoscope 2003;113:2196-200. calcium for longer than 6 months after surgery. 17. Higgins KM, Mandell DL, Govindaraj S, et al. The role of intra operative rapid parathyroid hormone monitoring for predicting thyroidectomy-related hypocalcemia. Arch Otolaryngol Head Neck Surg Conclusion 2004;130:63-7. 18. AES Guidelines G. Australian Endocrine Surgeons Guidelines Our results suggest that PTH-1 is an excellent predictor of AES06. 01. Postoperative parathyroid hormone measurement and patients who are at risk for hypocalcemia 24 hours postop early discharge after total thyroidectomy: analysis of Australian data and management recommendations. ANZ J Surg 2007;77:199-202. eratively. Should PTH-1 be less than 1.5 pmol/L, prophy 19. Quiros RM, Pesce CE, Wihelm SM, et al. Intraoperative parathyroid lactic pharmacotherapy with calcitriol should be started to hormone levels in thyroid surgery are predictive of postoperative avoid the development of symptomatic hypocalcemia. hypoparathyroidism and need for vitamin D supplementation. Am J Competing interests: None declared. Surg 2005;189:306-9.

240 J can chir, Vol. 57, No 4, août 2014

para-alqahtani.indd 240 14-07-16 10:38 AM RESEARCH • RECHERCHE

Functional outcome of supracondylar elbow fractures in children: a 3- to 5-year follow-up

Ahaoiza Diana Isa, MD Background: Long-term functional outcomes of supracondylar elbow fractures Andrew Furey, MD, MSc (SCEF) have not been well documented in the literature. We retrospectively evaluated functional outcomes of pediatric SCEF using the Disabilities of the Arm, Shoulder and Craig Stone, MD, MSc Hand (DASH) questionnaire.

From the Discipline of Surgery, Faculty Methods: We retrospectively reviewed the outcomes of patients who presented to of Medicine, Memorial University of our tertiary care pediatric emergency department with SCEF between January 2005 Newfoundland, St. John’s, Nfld. and December 2009. We reviewed their charts to assess several clinical parameters, including age, sex, Gartland classification of SCEF, weight, comorbidities, treatment Presented as a poster at the annual intervention, physiotherapy and the extremity involved. The DASH questionnaire meetings of the Canadian Orthopaedic was administered in 2012. We performed a multiple linear regression analysis to Association 2013, Controversies in Rheu- determine the significance of these clinical parameters as they related to the DASH matology & Autoimmunity 2013 and the score for functional outcome. Orthopaedic Trauma Association 2013. Results: We included 94 patients with SCEF in our review. Pediatric SCEF had good functional outcomes based on the DASH questionnaire (mean score 0.77 ± Accepted for publication 2.10). We obtained the following DASH scores: 0.45 ± 2.20 for type I, 1.09 ± 1.70 for Jan. 6, 2014 type II and 1.43 ± 2.40 for type III fractures. There was no statistical difference in functional outcome, regardless of sex (p = 0.07), age at injury (p = 0.96), fracture type Correspondence to: (p = 0.14), weight (p = 0.59), right/left extremity (p = 0.26) or surgery (p = 0.52). A. Isa Discipline of Surgery Conclusion: Our results demonstrate that good functional outcomes can be expected Faculty of Medicine with pediatric SCEF based on the DASH questionnaire, regardless of age at injury, Memorial University of Newfoundland sex, weight, right/left extremity or surgical/nonsurgical intervention, provided satis- Health Sciences Centre factory reduction is achieved and maintained. St. John’s NL A1B 3V6 [email protected] Contexte : Les répercussions fonctionnelles à long terme des fractures du coude supracondyliennes (FCSC) n’ont pas été bien documentées dans la littérature. Nous DOI: 10.1503/cjs.019513 avons évalué de manière rétrospective les résultats fonctionnels des FCSC pédiatriques à l’aide du questionnaire DASH (Disabilities of the Arm, Shoulder and Hand). Méthodes : Nous avons passé en revue rétrospectivement les résultats chez les patients amenés pour une FCSC au service d’urgence pédiatrique de notre établissement de soins tertiaires entre janvier 2005 et décembre 2009. Nous avons examiné leurs dossiers pour évaluer plusieurs paramètres cliniques, dont l’âge, le sexe, la classification de Gartland pour les FCSC, le poids, les comorbidités, l’intervention thérapeutique, la physiothérapie et la latéralité du membre affecté. Le questionnaire DASH a été administré en 2012. Nous avons procédé à une analyse de régression linéaire multiple pour déterminer la significa- tion de ces paramètres cliniques en regard du score DASH d’évaluation fonctionnelle. Résultats : Nous avons inclus 94 patients ayant subi une FCSC dans notre analyse. La FCSC pédiatrique évolue bien au plan fonctionnel selon le questionnaire DASH (score moyen 0,77 ± 2,10). Nous avons obtenu les scores DASH suivants : 0,45 ± 2,20 pour les fractures de type I, 1,09 ± 1,70 pour les fractures de type II et 1,43 ± 2,40 pour les fractures de type III. On n’a noté aucune différence statistique quant aux résultats fonctionnels, indépendamment du sexe (p = 0,07), de l’âge au moment de la fracture (p = 0,96), du type de fracture (p = 0,14), du poids (p = 0,59), de la latéralité (p = 0,26) ou de la chirurgie (p = 0,52). Conclusion : Nos observations démontrent qu’on peut s’attendre à de bons résultats fonctionnels dans les cas de FCSC en se fondant sur le questionnaire DASH, indépendamment de l’âge au moment de la fracture, du sexe, du poids, de la latéralité ou de l’intervention chirurgicale ou non chirurgicale, à la condition d’obtenir et de maintenir une réduction satisfaisante.

fun-isa.indd 241 14-07-16 10:34 AM RECHERCHE

ediatric supracondylar elbow fractures (SCEF) are the questionnaire themselves under a parent’s supervision. a common occurrence in children. These fractures We chose to remove the sex-related question to make the P are commonly extra-articular, unlike adult SCEF. questionnaire more age appropriate. The optional work Return of elbow motion after treatment of supracondy- module was not used, but the optional sports/performing lar humeral fracture in children has been well docu- arts module was. mented in the literature.1 However, the return of elbow The DASH score is scaled between 0 and 100. Higher range of motion and function is usually measured as an scores indicate worse function, and lower scores indicate objective parameter, such as a return of normal range of better function relating to upper-extremity disability. arc motion in the sagittal plane (flexion and extension).1 Our inclusion criteria for joining the study were isolated Long-term functional outcome using standardized tools extension SCEF in patients younger than 13 years at the has not been well documented in the literature. One time of injury, closed injury and consent to join the study study analyzed the correlation between the values of a (obtained from parents). The exclusion criteria were poly- modified Disabilities of the Arm, Shoulder and Hand trauma, flexion type SCEF, ipsilateral injury requiring sur- (DASH) questionnaire and change of elbow function gery, iatrogenic neurologic injury, reinjury to the same after SCEF of the humerus.2 However, the study elbow within the study period, metabolic bone disease, addressed only flexion SCEF, which represents a small condyle, epicondyle fractures and transphyseal fractures percentage of pediatric SCEF. The authors concluded that had been reported as SCEF. We chose to exclude the that the value of a modified DASH questionnaire correl flexion type SCEF owing to the small number (1.2%) of ates with objective indicators of elbow function after patients with this condition. flexion SCEF in children. A standardized functional Three fellowship-trained pediatric orthopedic surgeons outcome tool, such as the well-validated DASH ques- independently reviewed the radiographs and grouped the tionnaire, has been instrumental in measuring func- patients according to the Gartland classification,4 which is tional disability. widely used in the literature. In the Gartland classification Bot and colleagues3 evaluated the clinimetric quality of system, type I fractures are essentially nondisplaced. Type 16 self-administered shoulder disability questionnaires, II fractures are displaced with a variable amount of angula- including DASH, the Shoulder Pain and Disability Index tion, but more importantly, the posterior cortex of the and the American Shoulder and Elbow Surgeons Stan- humerus is intact. Type III fractures are completely dis- dardised Shoulder Assessment Form. For clinimetric pur- placed, with no cortical continuity (Fig. 1). A κ value for poses, the DASH questionnaire received the best ratings.3 interobserver agreement was then calculated. Thus, we chose the DASH questionnaire, as it is a The pediatric orthopedic surgeons in our tertiary care region-specific standardized functional outcome tool that centre treated patients either operatively or nonoperatively is well validated has been important in measuring func- based on the degree of angulation and displacement (dis- tional disability. placement of the anterior humeral line [Fig. 2] and altera- The purpose of the present study was to provide a ret- tion of the Baumann angle [Fig. 3]). rospective, longitudinal evaluation of functional outcome The anterior humeral line should intersect the middle in a large population of children treated for pediatric third of the capitellum on lateral radiographs. A Baumann SCEF, using the DASH for standardized measure of angle within 5° of the uninjured side was considered accept- outcome. A secondary goal was to determine whether able. This was the radiographic measure of coronal plane factors such as age at injury, sex, weight, right or left deformity. The normal physiologic tilt of the capitellum on extremity, Gartland fracture type and surgical versus the humerus is about 30° anteriorly; reduction was not nonsurgical intervention could predict long-term func- required if this physiologic angulation was 20° or more and if tional outcomes. the anterior humeral line intersected the middle third of the capitellum on lateral radiographs in extension type SCEF. Methods Type I fractures were treated nonoperatively. Type II fractures were treated operatively or nonoperatively depending We retrospectively reviewed the cases of children with on the amount of angulation and displacement. Type II frac- SCEF who presented to our tertiary care pediatric emer- tures that failed closed reduction based on the radiographic gency department between January 2005 and December parameters mentioned previously were treated operatively. All 2009. We reviewed their charts for several parameters, type III fractures were treated operatively. Patients who were including age, sex, classification of fracture severity, managed operatively underwent closed reduction and percu- weight, comorbidities, operative or nonoperative treat- taneous pinning using Kirschner wires under fluoroscopy. If ment intervention, postoperative physiotherapy and asso- reduction could not be obtained with closed reduction, open ciated nerve injury. A DASH questionnaire was adminis- reduction and percutaneous pinning was performed. tered in 2012 by the parents of the patients with the child Patients were seen by the pediatric orthopedic sur- present; if they were old enough, the patients completed geons for scheduled follow-up visits. They were seen

242 J can chir, Vol. 57, No 4, août 2014

fun-isa.indd 242 14-07-16 10:34 AM RESEARCH

10–14 days postoperatively for a cast change, wound not requiring closed reduction were casted and seen in check and radiographs to ensure displacement had not clinic at 2 weeks postinjury for a cast check and radio- occurred. Displacement was measured using the anterior graphs and at 4 weeks for cast removal, radiographs and humeral line and the relationship with the capitellum. range of motion check. Alignment was considered acceptable if the anterior Three fellowship-trained pediatric orthopedic sur- humeral line intersected the middle third of the capitel- geons followed all patients to ensure fracture healing, lum on lateral radiographs. Patients were seen for pin as seen radiographically. One of us (A.I.) independently removal and cast change at 4 weeks postoperative and reviewed all the radiographs to ensure that reductions again for final cast removal, radiographs and range of were maintained and that fractures had healed. motion check at 6 weeks postoperative. At the 3-month follow-up, patients were seen to assess the need for phys- Statistical analysis iotherapy and to check range of motion. The patients who received closed reduction or who Mean DASH scores were then calculated based on sex, sustained a nondisplaced or minimally displaced SCEF right or left extremity, weight, intervention (operative v.

A B C

Fig. 1. Types of supracondylar fractures. A) type I supracondylar fracture with a posterior fat pad sign, B) type II supracondylar fracture with an intact posterior hinge, C) type III supracondylar fracture with no cortical continuity. Arrows delineate fracture lines.

Fig. 2. Anterior humeral line, drawn in line with the anterior Fig. 3. Baumann angle: (A) Angle between long axis of humeral humeral shaft, should intersect the middle of capitellum. shaft and (B) growth plate of lateral humeral condyle.

Can J Surg, Vol. 57, No. 4, August 2014 243

fun-isa.indd 243 14-07-16 1:55 PM RECHERCHE

nonoperative) and age at time of injury. A multiple linear Results regression analysis was then performed, and we con sidered results to be significant atp < 0.05. A total of 158 patients with 161 SCEF were eligible to par ticipate in our study. Ten patients were excluded, as they did not meet the inclusion criteria. Fifty-four patients did not respond because they had moved away or changed 161 pediatric SCEF phone numbers without providing an update in the medical record system, and 1 patient declined participation, leaving Excluded, n = 12 94 patients for analysis (Fig. 4). Patients who did not reply or declined participation had a similar distribution in terms 149 SCEF of type of SCEF compared with our study group (Table 1). Of the 94 patients included in the study (Table 2), 54 No reply, n = 54 Declined participation, n = 1 (57%) were male and 40 (43%) were female. We subdivided patients by fracture type: 53 (56.4%) had type I, 26 (27.7%) 94 SCEF used had type II and 15 (15.9%) had type III fractures; 2 (2.1%) in study had flexion type SCEF. Patients were further subdivided by sex, right or left extremity, weight, intervention (operative v. Fig. 4. Enrolment of eligible patients. SCEF = supracondylar nonoperative) and age at time of injury (Table 2). There was a elbow fractures. higher incidence of SCEF in boys (57%) than in girls, and the left elbow was more commonly injured than the right elbow (61.7%). The mean age of patients with type I SCEF was Table 1. Distribution of supracondylar 70 months, type II was 74 months, type III was 83 months elbow fracture and Flexion type SCEF was 73.5 months (Table 2). Group; no. (%) A multiple linear regression analysis was performed to Type Study group No reply determine the significance of the clinical parameters as they 1 53 (56.4) 32 (58.2) related to the DASH score for functional outcome. There 2 26 (27.7) 14 (25.5) was no statistical difference in functional outcome using the 3 15 (15.9) 9 (16.3) DASH score regardless of sex (p = 0.07), age at injury (p = Total 94 55

Table 3. Mean dash scores and standard deviation

Table 2. Patient demographics Group; mean ± SD Group; no. (%) Type I, Type II, Type III, Type 1, Type 2, Type 3, Characteristic n = 58 n = 18 n = 16 Characteristic n = 53 n = 26 n = 15 Sex Sex Male 0.03 ± 0.17 0.78 ± 1.28 1.69 ± 2.42 Male 29 (55) 16 (62) 9 (60) Female 0.96 ± 3.22 1.57 ± 2.19 1.04 ± 2.55 Female 24 (45) 10 (28) 6 (40) Extremity Extremity Right elbow 0.81 ± 3.21 0.76 ± 1.26 2.17 ± 2.57 Right elbow 18 (34) 11 (42) 7 (47) Left elbow 0.27 ± 1.46 1.32 ± 1.96 0.78 ± 2.21 Left elbow 35 (66) 15 (58) 8 (53) Weight, kg Weight, kg ≤ 15 0.11 ± 0.32 1.98 ± 1.93 6.25* ≤ 15 9 (17) 3 (12) 1 (7) 16–30 0.65 ± 2.72 0.90 ± 1.64 0.80 ± 2.07 16–30 34 (64) 19 (73) 9 (60) 31–45 — 1.29 ± 2.05 — 31–45 8 (15) 4 (15) 2 (13) > 45 0.45 ± 0.63 — 2.68 ± 2.36 > 45 2 (4) 0 (0) 3 (20) Intervention Intervention Operative — 1.39 ± 1.99 0.83 ± 1.91 Operative 0 (0) 11 (42) 13 (87) Nonoperative 0.45 ± 2.20 0.86 ± 1.48 5.35 ± 1.27 Nonoperative 53 (100) 15 (58) 2 (13) Age at injury, mo Age at injury, mo ≤ 30 — — — < 30 2 (4) 1 (4) 0 (0) 31–60 0.70 ± 2.97 0.66 ± 1.02 1.56 ± 3.13 31–60 21 (39) 6 (23) 4 (27) 61–90 0.51 ± 2.09 1.21 ± 1.63 1.44 ± 2.72 61–90 17 (32) 12 (46) 5 (33) 91–120 0.09 ± 0.28 1.66 ± 2.58 — 91–120 10 (19) 6 (23) 3 (20) > 120 — 0.86* 2.68 ± 2.36

> 120 3 (6) 1 (4) 3 (20) SD = standard deviation. Mean age, mo 70 74 83 *There was only 1 patient in this category, so the SD could not be calculated.

244 J can chir, Vol. 57, No 4, août 2014

fun-isa.indd 244 14-07-16 10:34 AM RESEARCH

0.96), type of fracture (p = 0.14), weight (p = 0.59), right or tion system should be both prognostic and provide a guide left extremity (p = 0.26) or surgery (p = 0.52; Tables 3 and 4). to clinical management. Our study demonstrated no sta Using the optional module (sports/performing arts), tistical difference in functional outcomes across all pediat there was no statistical difference in functional outcome ric SCEF despite Gartland classification. There were no regardless of sex (p = 0.33), age at injury (p = 0.90, type of significant differences between other parameters, such as fracture (p = 0.62), weight (p = 0.99), or right or left sex, right or left extremity, weight, intervention (operative extremity (p = 0.28; Table 5). v. nonoperative) and age at time of injury and their correl Our interobserver agreement to indicate the reprodu ation with functional outcomes. These types of fractures, cibility of the classification was calculated using weighted regardless of several parameters, tend to have good func Fleiss κ, with κ representing the proportion of agreement tional outcomes based on the DASH score, provided that among the orthopedic surgeons beyond that expected by satisfactory reduction is maintained by either surgical or chance. A value of 0 indicates what would be expected by nonsurgical (cast) means and that the technique of reduc chance and a value of 1 indicates perfect agreement. A value tion and subsequent treatment course is uncomplicated. less than 0 is an indication that agreement is less than what The pediatric orthopedic surgeons at our institution is expected by chance.5 Our calculated κ score was 0.76, still use the Gartland classification; however, for clinical which represents good interobserver reliability (Table 6). decision making, degree of displacement is used. For example, not all Gartland type II SCEF were treated oper Discussion atively; based on degree of displacement/intersection of the anterior humeral line and capitellum (Fig. 3) and the Supracondylar elbow fractures are common in children. Baumann angle (Fig. 4) as well as on failure of closed The Gartland classification has been widely used for clas reduction, patients were treated operatively or nonopera sification of pediatric SCEF. Ideally, a fracture classifica tively. There was no significant difference in DASH scores (p = 0.52) between patients treated operatively and those treated nonoperatively. This does not mean that surgery is Table 4. Multiple linear regression analysis not important in the management of pediatric SCEF; ade quate reduction is important to the functional outcome. Model p value Inadequate reduction can lead to potential functional long- Constant 0.25 term problems.6 Garland type II SCEF is not an indication Class 0.16 for surgery. The degree of displacement should guide Sex 0.07 Weight 0.59 management, as SCEF tend to do well if adequate reduc Age 0.96 tion is obtained and maintained and if it follows an uncom 7 Operative v. nonoperative 0.52 plicated course. A study by Heal and colleagues further Right or left extremity 0.26 supports this, as they also conclude that pediatric SCEF should be treated based on the degree of displacement rather than the Gartland classification. Table 5. Multiple linear regression analysis of sports/performing arts In the same study, based on a calculated κ score of 0.54, module Heal and colleagues7 concluded that there was a moderate

Model p value interobserver agreement with the Gartland classification, with poor agreement over type I extension SCEF, fair to Constant 0.48 Class 0.62 moderate agreement with type II and good to very good Sex 0.33 agreement with type III using κ scores. Our interobserver Weight 0.99 agreement to indicate the reproducibility of the classifica Age 0.90 tion was calculated using weighted Fleiss κ, which repre Right or left 0.28 sents the proportion of agreement among the orthopedic extremity surgeons beyond that expected by chance. In contrast to Heal and colleagues,7 our calculated score was 0.76, which Table 6. Interpretation of κ, Landis indicates substantial or good agreement among the ortho and Koch (1997) pedic surgeons, bearing in mind that the calculated κ κ Strength of agreement scores could also be due to chance. < 0 Poor Our study showed no statistical difference in the func 0.01–0.20 Slight tional outcome based on the Gartland classification. 0.21–0.40 Fair Pediatric SCEF, if treated appropriately based on degree 0.41–0.60 Moderate of displacement and adequacy/maintenance of closed 0.61–0.80 Substantial/good reduction tend to do well despite the classification of 0.81–1.00 Almost perfect fracture with little to no functional limitation of day-

Can J Surg, Vol. 57, No. 4, August 2014 245

fun-isa.indd 245 14-07-16 10:34 AM RECHERCHE

to-day activities or associated pain. In the sports/ Conclusion performing arts module, there was also no statistical difference in mean DASH score among the groups Return of range of motion after an isolated pediatric (Table 4), regardless of age, weight, right or left extrem- SCEF has been well documented in the literature. ity, sex, intervention (operative v. nonoperative) or We found that overall, most parents and patients Gartland classification of severity. These patientsalso reported no functional interference with normal social tend to function well with no significant limitation in activities, sports or performing arts, activities of daily sports or performing arts (Table 5). living (including self-care), and no functionally limiting Spencer and colleagues1 demonstrated the effect of age symptoms, regardless of age at injury, sex, weight, right and severity of fracture (determined by those requiring or left extremity, operative or nonoperative intervention operative intervention) on the recovery of elbow motion, or Gartland classification, provided that satisfactory with patients older than 5 years demonstrating a 3%–9% reduction is maintained by either surgical or nonsurgical lower relative arc of motion at the follow-up points than (cast) means and that the means of reduction and treat- younger patients and a slower recovery in motion in ment course are not complicated. Perhaps adequate those who had more severe fractures requiring surgical reduction is more important than simply treatment of intervention. We studied whether there were any func- fracture type. tional differences between these groups using the DASH Acknowledegments: We thank Minnie Parsons for her help in the questionnaire and found that despite the findings of assembling of patient data, ensuring we had the necessary documenta- Spencer and colleagues,1 there was no statistical differ- tion to meet the ethics board requirements and printing and distribut- ing the DASH questionnaire. ence in functional outcome regardless of age or operative or nonoperative intervention (p = 0.52) based on multiple Competing interests: None declared. linear regression analysis. Contributors: All authors designed the study. A.D. Isa acquired the data, which all authors analyzed. A.D. Isa and A. Furey wrote the article, which all authors reviewed and approved for publication. Limitations References The limitations of our study include the fact that it was retrospective and, in some cases, the parents filled out the DASH 1. Spencer HT, Wong M, Fong YJ, et al. Prospective longitudinal questionnaire based on their perceptions of their children’s evaluation of elbow motion following pediatric supracondylar functioning. Another limitation is the validity of this ques- humeral fractures. J Bone Joint Surg Am 2010;92:904-10. tionnaire in this age group. Although the DASH question- 2. Colovic H, Stankovic I, Dimitrijevic L, et al. The value of modified naire has not yet been formally validated in this age group, DASH questionnaire for evaluation of elbow function after supra- multiple studies have used the DASH questionnaire in pedi- condylar fractures in children. Vojnosanit Pregl 2008;65:27-32. atric populations.8–11 There is also a possibility for skewed 3. Bot SD, Terwee CB, van der Windt DA, et al. Clinimetric evaluation of shoulder disability questionnaires: a systematic review of the data given that 54 patients did not respond and 1 declined literature. Ann Rheum Dis 2004;63:335-41. participation. Thus, given our small sample size, there is a 4. Gartland JJ. Management of supracondylar fractures of the chance that we missed the difference in DASH scores among humerus in children. Surg Gynecol Obstet 1959;109:145-54. the groups. We did not perform a subgroup analysis on the 5. Viera AJ, Garrett JM. Understanding interobserver agreement: the fractures treated with open reduction, as the numbers would kappa statistic. Fam Med 2005;37:360-3. be small and we would have therefore been unable to make 6. O’Driscoll SW, Spinner RJ, McKee MD, et al. Tardy posterolateral firm statistical conclusions; however, this was not the primary rotatory instability of the elbow due to cubitus varus. J Bone Joint focus of our study and represents a potential future area of Surg Am 2001;83-A:1358-69. research. Despite the small number at follow-up, 7. Heal J, Bould M, Livingstone J, et al. Reproducibility of the we reviewed the charts of all 149 patients who were eligible Gartland classificationfor supracondylar humeral fractures in children. J Orthop Surg (Hong Kong) 2007;15:12-4. for participation in the study to identify any adverse out- 8. Behr B, Heffinger C, Hirche C, et al. Scaphoid nonunions in skele- comes or the need for further therapy or intervention. There tally immature adolescents. J Hand Surg Eur Vol 2013 [Epub ahead were none identified from our chart review. We also of print]. reviewed the surgeon’s notes at subsequent follow-ups. 9. Al-Aubaidi Z, Pedersen NW, Nielsen KD, et al. Radial neck frac- To our knowledge, our study is the first to attempt to tures in children treated with the centromedullary Métaizeau tech- identify risk factors for poor functional outcomes using a nique. Injury 2012;43:301-5. standardized measure of outcome in a large group of chil- 10. Colovic H, Stankovic I, Dimitrijevic L, et al. The value of modified dren with all types of extension SCEF. Our study further DASH questionnaire for evaluation of elbow function after supracondylar fractures in children. Vojnosanit Pregl 2008;65:27-32. provides a longitudinal evaluation of functional outcome 11. Lawrence JT, Patel NM, Macknin J, et al. Return to competitive in a large population of children treated for pediatric sports after medial epicondyle fractures in adolescent athletes: results SCEF fracture using the DASH for standardized measure of operative and nonoperative treatment. Am J Sports Med 2013; of outcome. 41:1152-7.

246 J can chir, Vol. 57, No 4, août 2014

fun-isa.indd 246 14-07-16 10:34 AM RESEARCH / RECHERCHE

Comparison of cast materials for the treatment of congenital idiopathic clubfoot using the Ponseti method: a prospective randomized controlled trial

Catherine Hui, MD* Background: The Ponseti method of congenital idiopathic clubfoot correction has Elaine Joughin, MD† traditionally specified plaster of Paris (POP) as the cast material of choice; however, there are negative aspects to using POP. We sought to determine the influence of cast Alberto Nettel-Aguirre, PhD‡ material (POP v. semirigid fibreglass [SRF]) on clubfoot correction using the Ponseti Simon Goldstein, MD† method. James Harder, MD† Methods: Patients were randomized to POP or SRF before undergoing the Ponseti Gerhard Kiefer, MD† method. The primary outcome measure was the number of casts required for clubfoot correction. Secondary outcome measures included the number of casts by severity, † David Parsons, MD ease of cast removal, need for Achilles tenotomy, brace compliance, deformity relapse, Carmen Brauer, MD† need for repeat casting and need for ancillary surgical procedures. Jason Howard, MD§ Results: We enrolled 30 patients: 12 randomized to POP and 18 to SRF. There was no difference in the number of casts required for clubfoot correction between the From the *Division of Orthopaedic Sur- groups (p = 0.13). According to parents, removal of POP was more difficult gery, Department of Surgery, University (p < 0.001), more time consuming (p < 0.001) and required more than 1 method of Alberta, Edmonton, Alta., †Division of (p < 0.001). At a final follow-up of 30.8 months, the mean times to deformity relapse Orthopaedic Surgery, Department of requiring repeat casting, surgery or both were 18.7 and 16.4 months for the SRF and Surgery, Alberta Children’s Hospital, Cal- POP groups, respectively. gary, Alta., ‡Departments of Paediatrics and Community Health Sciences, Alberta Conclusion: There was no significant difference in the number of casts required for Children’s Hospital, Calgary, Alta., §Divi- correction of clubfoot between the 2 materials, but SRF resulted in a more favourable sion of Orthopaedic Surgery, Depart- parental experience, which cannot be ignored as it may have a positive impact on ment of Surgery, Sidra Medical and psychological well-being despite the increased cost associated. Research Center, Doha, Qatar

Previously presented at the following Contexte : La méthode de Ponseti pour la correction du pied bot congénital meetings: International Clubfoot Sym- idiopathique a de tout temps spécifié l’utilisation du plâtre de Paris comme matériau posium, Iowa City, IA, Oct 2012 and the de choix; il y a toutefois certains inconvénients associés au plâtre de Paris. Nous avons Canadian Orthopaedic Association Meet- voulu déterminer l’influence du matériau utilisé (plâtre de Paris c. fibre de verre semi- ing, Edmonton, Alta., June 2010. rigide) sur la correction du pied bot selon la méthode de Ponseti. Méthodes : Les patients ont été assignés aléatoirement soit au plâtre de Paris soit à Accepted for publication la fibre de verre semi-rigide en vue de l’intervention de Ponseti. Le principal Jan. 15, 2014 paramètre mesuré était le nombre de plâtres requis pour corriger le pied bot. Les paramètres secondaires incluaient le nombre de plâtres en fonction de la gravité, la Correspondence to: facilité de retrait du plâtre, la nécessité de sectionner le tendon d’Achille, le port J.J. Howard assidu de l’attelle, le retour de la difformité, la nécessité d’autres plâtres et interven- Division of Orthopaedic Surgery tions chirurgicales auxiliaires. Department of Surgery Sidra Medical and Research Center Résultats : Nous avons inscrit 30 patients : 12 ont été assignés au plâtre de Paris et P.O. Box 26999 18 à la fibre de verre. On n’a noté aucune différence entre les groupes quant au nom- Doha, Qatar bre de plâtres requis pour la correction du pied bot (p = 0,13). Selon les parents, le [email protected] retrait du plâtre de Paris était plus difficile (p < 0,001), prenait plus de temps (p < 0,001) et nécessitait le recours à plus d’une méthode (p < 0,001). Au moment du DOI: 10.1503/cjs.025613 dernier suivi à 30,8 mois, les intervalles moyens avant un retour de la difformité nécessitant la pose d’un autre plâtre et/ou une chirurgie ont été de 18,7 et 16,4 mois dans les groupes traités au moyen de la fibre de verre semi-rigide et du plâtre de Paris, respectivement. Conclusion : On n’a noté aucune différence significative entre les 2 matériaux quant au nombre de plâtres requis pour corriger le pied bot, mais la fibre de verre a donné lieu à une expérience plus agréable pour les parents, ce qui ne peut être ignoré en raison de l’impact potentiellement positif sur le bien-être psychologique, et ce, malgré un coût plus élevé.

compar-howard.indd 247 14-07-16 10:41 AM RECHERCHE

ongenital idiopathic clubfoot is a 3-dimensional ously used for clubfoot correction and the treatment of deformity that includes cavus and adductus of the resistant metatarsus adductus with some success.21,22 Scotch- midfoot, combined with hindfoot varus and equi- cast Soft Cast Casting Tape (3M) is a popular fibreglass C1–13 nus. The goal of treatment is to correct all components casting material that was originally developed for extremity of the deformity, such that a pain-free, plantigrade foot with injuries not requiring rigid immobilization. This material is good mobility is achieved for the long term. Initiation of semirigid when dry and has the benefit of not tightly adher- timely and appropriate treatment is paramount to achieve ing to itself, thus allowing easy removal by unwrapping.21,22 these successful long-term outcomes. Though nonoperative Many centres ask parents to remove their children’s Ponseti management of clubfoot had been the standard for centur casts just before their clinic visits to avoid injury during ies, modern treatment of clubfoot has, until recently, been removal with a cast knife or saw, which may give SRF an primarily surgical.1,14 The resurgence of the Ponseti method advantage over POP. In a related study investigating paren- in recent years has been punctuated by less than favourable tal satisfaction and clubfoot casting, SRF was preferred to long-term outcomes for surgically treated feet.4,7,9,10,12,15–17 POP owing to improved durability, performance, ease of The Ponseti method consists of weekly serial manipulations removal, ulcer prevention, weight, appearance, ease of and above-knee plaster casting.1,3,4,9,11,14,16,17 With more cleaning and water resistance.22 Despite these advantages, a recent studies confirming its long-term success, it is the cur- recent study by Zmurko and colleagues21 demonstrated that rent gold standard of treatment.5,9,15,16,18 After cast correction SRF costs about 7 times more than POP and is biomechan of the cavus, adductus and varus components of the deform ically inferior to both POP and traditional rigid fibreglass ity, a percutaneous achilles tenotomy is required for defini- material. They suggested the need for a prospective trial to tive equinus correction in more than 70% of cases.7 evaluate these materials for clinical significance. The 2 most common casting materials currently used in Our goal was to determine whether the choice of cast the treatment of idiopathic clubfoot are plaster of Paris material influenced the number of casts required for correc- (POP) and semirigid fibreglass (SRF). The Ponseti method tion of clubfoot deformity using the Ponseti method. We of clubfoot correction has traditionally specified POP as also assessed the parents’ experience with the cast material, the cast material of choice. It is a cheaper and stiffer mater particularly with respect to ease of removal. ial than SRF and is easily mouldable. Some negative aspects associated with its use, however, may include a Methods small risk of injury associated with the exothermic reaction that occurs during curing, more difficult cast removal and Study design and patient selection the potential for cast saw accidents (Fig. 1).19–21 Fibreglass casting materials were introduced in the 1970s We conducted a prospective randomized controlled trial, and have the advantages of radiolucency, lighter weight, completed in a tertiary-level children’s hospital. We improved durability, faster curing time, lower risk of ther- enrolled consecutive patients with congenital idiopathic mal burn, cleaner application and potentially easier clubfoot presenting to the regional tertiary-level chil- removal.19–23 Semirigid fibreglass materials have been previ- dren’s hospital between July 2007 and December 2008. Patient referrals were screened through a central intake within the orthopedic clinic and were distributed equally and sequentially among the 7 pediatric orthopedic surgeons participating in the study. Following ethics approval from our institutional review board, we obtained written informed consent from the parents of all patients included in our study. Clubfoot casting was initiated at the first clinical visit and subsequently at weekly intervals using serial manipulation and above-knee casting according to the Ponseti method.4,7,9,12,13,24–27 Clubfoot etiology was determined by a thorough history and physical examination (and additional tests as necessary) performed by the treating surgeon. Once the diagnosis of congenital idiopathic clubfoot was made, the patient was randomly assigned to receive either POP or SRF casts. Patients were excluded from this study Fig. 1. Infant with substantial skin injury following removal of a if the cause of clubfoot was nonidiopathic (e.g., arthrogry- Ponseti plaster cast with oscillating saw. This incident caused substantial parental anxiety, such that subsequent casts were posis), or if they had been previously treated for clubfoot. removed using prolonged soaking in warm water and unwrap- Patients with positional clubfoot deformities were also ping of the plaster roll. excluded.

248 J can chir, Vol. 57, No 4, août 2014

compar-howard.indd 248 14-07-16 10:41 AM RESEARCH

Ponseti method and Pirani classification The primary outcome variable in this study was the number of casts required for correction of the clubfoot Each of the participating pediatric orthopedic surgeons had deformity to the point where the foot was ready for a per- considerable previous experience and specialized training in cutaneous tendo-Achilles tenotomy, if necessary, or when the Ponseti method. To ensure that the indications for ces- dorsiflexion of the ankle greater than or equal to 15° was sation of cast treatment were reasonably uniform, each sur- achieved. A percutaneous tendo-Achilles tenotomy was per- geon was required to attend a refresher training session in formed when there was sufficient abduction of the foot, the Ponseti method and the Pirani classification system. The verified by palpation of the anterior process of the calcaneus Pirani classification was used to measure initial clubfoot as it externally rotates from beneath the talus; foot abduc- severity and allowed for surveillance during treatment. This tion of approximately 60° in relation to the frontal plane of 6-grade ordinal system is scored based on the status of the the tibia; and neutral or slight valgus of the calcaneus. midfoot and hindfoot during correction and has been shown According to the Ponseti method, the foot should be casted to have excellent intra- and interobserver reliability.28 A in 15° of dorsiflexion and abducted to 70° for 3 weeks after Pirani score of 6 is the most severe grading, and a score of 0 tenotomy. This cast was not included in the analysis, as represents a fully corrected foot (Fig. 2). A poster outlining each foot was fully corrected at the time of its application. the Pirani classification and the indications for cessation of Secondary outcome variables included the need for percu- casting and/or tenotomy was displayed for reference in the taneous tendo-Achilles tenotomy, total time in casts (weeks), clubfoot casting room for the duration of the study. ease of cast removal, duration of cast removal (minutes), method(s) of cast removal, complications relating to the cast- Assessment and outcomes ing material, compliance with postcorrection foot-abduction orthosis (FAO), deformity relapse, the need for repeat Ponseti Photographs were taken before initiation of casting and at the casting and the need for ancillary surgical procedures. end of casting during foot-abduction orthosis fitting. At each visit, a Pirani grade was given and tabulated using standardized Sample size and randomization data collection forms. The parents were told to remove the cast at home before each clinic visit. A clinic nurse provided Based on the results of a pilot study of SRF and POP instructions for cast removal specific to each material. After the materials performed at our institution involving 10 patients first cast and fourth casts were removed, the parents were with idiopathic clubfoot, we determined that a sample size asked to complete a questionnaire (see the Appendix, available of 30 was required. Our calculation was based on a desired at canjsurg.ca) relating to their experience with the selected assessment of the primary outcome variable with a clinically casting material. The questions were primarily related to the significant difference of 2 casts and an equal standard devia- ease of cast removal, the time needed for removal and the tion of 1.88 (from pilot data) for a power of 80% based on a number of methods required. 2 sample t test at a significance level ofα = 0.05.

1 0 0.5 1 Empty heel Easily palpable, 1 Palpable deep, 0.5 0 0.5 Not palpable, 1

Rigid equinus Empty heel Posterior crease

Talar head 0 0.5 1 None, 0 Partial, 0.5 Full 1 0 0.5 1

Curved lateral border Medial crease Lateral head of talus (LHT)

Fig. 2. Pirani scoring system for clubfeet. (A) Hindfoot score (HS); (B) midfoot score (MS). Total score = (HS + MS) ÷ 6. Repro- dued with permission from Global HELP organization.10

Can J Surg, Vol. 57, No. 4, August 2014 249

compar-howard.indd 249 14-07-16 10:41 AM RECHERCHE

Randomization of patients was performed using con- were 2.0 (range 1–11.7) and 2.3 (range 0.7–5.7) weeks, cealed number tracked envelopes according to a computer- respectively. In the SRF group, a unilateral clubfoot was generated randomization list. The envelope remained sealed present in 10 of 18 patients (56%), and bilateral clubfeet and was opened by the surgeon just before the initiation of were present in the remaining 8 patients, for a total of cast treatment. Only 1 type of cast material was used for 26 clubfeet. In the POP group, a unilateral clubfoot was each patient to prevent crossover (i.e., randomization was by present in 6 of 12 patients (50%), and bilateral clubfeet were patient, not by foot). Block randomization was not applied. present in the remaining 6 patients, for a total of 18 clubfeet. Whenever bilateral clubfeet were present, the primary out- Statistical analysis come (number of casts) was taken from the more severe foot (i.e., higher Pirani score at initial assessment). The mean Collected data are reported as descriptive statistics (mean ± initial Pirani score was 5.3 (range 2–6) and 4.9 (range 3–6) in standard deviation) for continuous variables and percentages the SRF and POP groups, respectively. In addition, patients for categorical variables. We generated box plots for the pri- were grouped according to clubfoot severity, with more mary variable. Confidence intervals (CIs) were determined severe deformities having Pirani scores of 5 or more and less where appropriate. We used a Student t test at a 5% signifi- severe deformities having Pirani scores less than 5. Assign- cance level to determine if there was a significant difference ing levels of severity using the Pirani score has been sug- between the means of the number of casts needed per gested previously by other authors.29 The number of more material. Other tests for analysis of secondary outcomes were severe clubfeet was 22 of 26 feet (85%) in the SRF group χ2 or Fisher exact test (as appropriate) for categorical variables. and 12 of 18 feet (67%) in the POP group. For bilateral A PhD statistician (A.N.-A.) performed the data analysis. cases, the most severe clubfoot was analyzed for consistency. A tendo-Achilles tenotomy was performed for 15 of 26 club- Results feet (58%) in the SRF group and 14 of 18 clubfeet (78%) in the POP group. Forty-five patients with clubfoot were initially assessed for There was no significant difference in the mean number eligibility in the study; 15 were excluded for various reasons of casts required for clubfoot correction between the (Fig. 3). Of the 30 patients identified for inclusion, 18 (60%) groups (SRF: 5.7 ± 2.8 casts; POP: 4.4 ± 1.6 casts, p = were randomized to SRF and 12 (40%) to POP. No patients 0.13). The distributions for the groups are displayed as box were lost to follow-up during the casting phase of this study. plots in Figure 4. The 95% CI for the difference in the

The mean ages at first visit for the SRF and POP groups mean number of casts (µSRF-µPOP) was (–0.41 to 3.0). When

Assessed for eligibility, n = 45 Excluded, n = 15 Possessed exclusion criteria, n = 8 • Positional foot deformity, n = 3 • Teratologic clubfoot, n = 4 • Rx started elsewhere rst, n = 1 Enrollment Refused, n = 5 • Wanted POP, n = 1 • Wanted SRF, n = 4

Randomized, n = 30 Other, n = 2 • Did not understand English, n = 1

POP, n = 12 Allocation SRF, n = 18

Lost to follow-up, n = 0 Follow-up Lost to follow-up, n = 0

Analyzed, n = 12 Analysis Analyzed, n = 18

Fig. 3. CONSORT diagram demonstrating the flow of participants through the initial casting phase of the trial. POP = plaster of Paris; SRF = semirigid fibreglass.

250 J can chir, Vol. 57, No 4, août 2014

compar-howard.indd 250 14-07-16 10:41 AM RESEARCH

analyzed by clubfoot severity, the mean number of casts for was (0.317–0.916). Plaster of Paris casts required more than both materials in the less severe group was 3. In the more 1 method for removal in 9 of 12 (75%) patients compared severe group, the mean number of casts was 6.4 in the SRF with 2 of 12 (17%) patients in the SRF group (p < 0.001). The group and 4.7 in the POP group. 95% CI for the difference in the proportion of removals that

Twenty-four of 30 (80%) parental questionnaires were needed 1 method (ρSRF-ρPOP) was (0.341–0.926). completed after the first visit and subsequently analyzed. The Data for secondary outcome measures, including compli- response rate for the fourth cast questionnaire was too low to ance with FAO, deformity relapse, need for repeat Ponseti provide useful results and thus they were not included in the casting and need for ancillary surgical procedures following analysis. According to parents, POP removal was rated as successful initial clubfoot correction by the Ponseti method, “manageable” or “difficult” by 8 of 12 (67%) parents com- were collected at a mean final follow-up of 30.8 ± pared with 1 of 12 (8%) parents in the SRF group (p < 0.001). 14.2 months. A summary of these results is provided in The remaining parents in each group rated cast removal as Table 1. The mean final follow-up for the SRF and POP “easy” or “very easy.” The 95% CI for the difference in the groups was 35.8 ± 11.3 months and 23.7 ± 14.4 months, proportion of “easy/very easy” removals between groups respectively. Two of 30 patients (1 in each treatment group)

(ρSRF-ρPOP) was (0.317–0.916). Plaster of Paris took longer than were lost to final follow-up. The mean times to deformity 30 minutes for removal in 8 of 12 (67%) patients compared relapse requiring repeat Ponseti casting, surgery or both with 1 of 12 (8%) patients in the SRF group (p < 0.001). The were 18.7 ± 15.0 and 16.4 ± 21.1 months for the SRF and remaining patients in each group had removal durations of POP groups, respectively. Surgical interventions were varied, 0–29 minutes. The 95% CI for the difference in the propor- but included posterior release, posteromedial release, tibialis

tion of “0–29 minutes” removals between groups (PSRF-PPOP) anterior tendon transfer and tibialis posterior recession.

Discussion

n = 18 n = 12 The Ponseti method of clubfoot management has revolu- tionized the treatment of this common condition through 12 the reduction in extensive surgical procedures and improved long-term outcomes.17 Despite this, there are important emotional and psychological impacts associated 10 with the execution of this treatment regimen that may have an impact on parental compliance with the Ponseti proto- 8 col. As such, measures that serve to shorten treatment duration and improve parental satisfaction while still achieving

No. of casts clinical success should be sought. The present study was 6 designed to determine whether the choice of cast material influenced the number of casts required for correction of 4 clubfoot deformity using the Ponseti method. Parental experience with the cast material, particularly with respect to ease of removal, was also investigated to determine if 2 there was a preference for one material over the other. SRF POP Successful treatment of idiopathic clubfoot through Cast material serial manipulation and casting by the Ponseti method requires strict adherence to the ordered reduction of the Fig. 4. Number of casts necessary for clubfoot correction, by components of the deformity, followed by subsequent material. POP = plaster of Paris; SRF = semirigid fibreglass. immobilization in the corrected position for a defined time

Table 1. Deformity relapse and need for repeat Ponseti casting and/or late surgical intervention according to cast material and at final follow-up

Group; no. (%)

Need for repeat Need for surgery Cast material FAO compliance Deformity relapse Ponseti casting after casting Semirigid fibreglass 12 (70.6) 8 (47.1) 5 (29.4) 7 (41.2) Plaster of Paris 11 (91.7) 3 (25.0) 2 (16.7) 2 (16.7) Total 23 (74.2) 11 (35.0) 7 (23.3) 9 (29.0)

FAO = foot abduction orthosis.

Can J Surg, Vol. 57, No. 4, August 2014 251

compar-howard.indd 251 14-07-16 10:41 AM RECHERCHE

period. This method has been purported to allow for grad- This preference was supported by our study, in which a ual ligamentous and muscular lengthening through creep higher proportion of parents whose children had SRF and stress relaxation in keeping with the viscoelastic prop- reported positive outcomes with respect to ease and time erties of the tissues involved.25 Theoretically, a more rigid of removal of casts. Semirigid fibreglass can be quickly casting material (e.g., POP) would allow for a more rapid removed by simply unwrapping the cast tape, whereas pro- correction, given the increased stretch imposed on the tar- longed soaking in warm water and/or other agents (e.g., get tissues. In our study this theory appeared to have some vinegar) was required to soften POP to facilitate its merit, given the indication of a reduction in casts required removal. In the present study, the poor response rate for for more severe clubfeet (Pirani ≥ 5) when POP was used the parental questionnaire after the fourth cast may indi- than when SRF was used. For clubfeet with a Pirani score cate a decreasing learning curve with successive cast less than 5, SRF seemed to perform as well as POP, sug- removals, which might diminish the importance of gesting that the stiffness of this material was sufficient for material choice overall. One could surmise, however, that less severe cases. Although it seems that POP provided a the emotional stress associated with having a child born more rapid correction for severe clubfeet, our study was not with clubfoot might be compounded by the need for more designed to have the power to statistically test this result. onerous parental involvement with POP — especially for To verify whether the superior material properties of POP the initial few casts. A recent study showed that the would be advantageous for the treatment of more severe psychological well-being and coping strategies for mothers clubfeet would require a larger sample size. of children with clubfoot are negatively impacted.31 This Several technical points concerning casting during club- situation might be further exacerbated by difficulties with foot correction have been emphasized by Ponseti.25 Given cast handling and removal. Interestingly, in the study by that the talonavicular joint is the fulcrum about which mid- Pittner and colleagues,30 there was no difference in paren- foot and hindfoot correction is achieved, cast moulding tal satisfaction between the 2 casting groups. Further study over the lateral aspect of the talar head is one of the tenets using validated questionnaires is required to definitively of this procedure. Stabilization of the talar head seems to answer the question relating psychological well-being to be more effectively achieved with POP, given the stiffness ease of cast removal and the relative importance of a of the material and the reported difficulties with moulding parental preference in clubfoot casting material. SRF casts.19,22 In addition, Ponseti also suggested providing Despite some clear disadvantages with respect to paren- adequate posterior moulding superior to the calcaneus to tal satisfaction, POP has been shown to be more econom help prevent cast slippage; this is more difficult to perform ical than SRF, although this was not investigated in the with SRF than with POP. Despite these theoretical advan- present study. Zmurko and colleagues21 showed that the tages, POP was not shown to be superior to SRF for cor- cost of SRF was purported to be up to 7 times that of rection of idiopathic clubfoot (p = 0.13), and cast slippage POP. The question remains whether the advantages in was not a significant problem in the present study. parental experience warrant the increased cost of SRF Since the commencement of the present study, Pittner given the lack of improvement in clinical outcomes com- and colleagues30 have reported the results of the first ran- pared with the substantially cheaper POP. domized trial comparing POP to SRF. As in the present In the present study, more patients in the SRF than the study, there was no significant difference in the mean num- POP group had a deformity relapse, requiring repeat ber of casts required for Ponseti correction between the Ponseti casting, surgical intervention or both. There may be 2 groups (6.1 in the SRF group v. 5.2 in the POP group, several reasons for this unrelated to the choice of cast p = 0.20). They did, however, note a statistically significant material used for initial clubfoot correction. The mean difference in the final severity scores (according to the duration of final follow-up for the SRF group was signifi- Dimeglio system) post-Ponseti casting, with the SRF and cantly longer than for the POP group (35.8 v. 23.7 months, POP groups each having residual scores of 6.4 (moderate) respectively), allowing more time for the deformity to and 4.1 (benign), respectively.30 This suggests incomplete relapse. Despite this, the mean times to deformity relapse clubfoot correction on average (at least for the SRF group). and initiation of further treatment were similar for the SRF As such, it is unclear whether further casting would have and POP groups (18.7 v. 16.4 mo, respectively). More reduced the deformity to a more benign Dimeglio score, in importantly, FAO compliance post-Ponseti casting was turn increasing the number of casts to final correction even markedly reduced in the SRF group compared with the further. In the present study, the indications for cessation of POP group (70.6% v. 91.7%, respectively). Noncompliance clubfoot casting and/or tentomy were clearly defined and, with the standard Ponseti bracing protocol (FAO worn as such, we were satisfied that the number of casts reported 3 months full-time, then at night and naptime for 3 years) for each treatment group was accurate. has been shown to be the factor most related to the risk of A previous study investigating cast treatment for club- relapse in several previous studies and may be the most foot and metatarsus adductus reported that 94% of parents likely reason why the SRF group in the present study had an had a definite preference for SRF-type casting over POP.22 increased prevalence of repeat casting and surgery.29,32

252 J can chir, Vol. 57, No 4, août 2014

compar-howard.indd 252 14-07-16 10:41 AM RESEARCH

7. Ponseti IV. Treatment of congenital club foot. J Bone Joint Surg Am The strength of this study lies in its design. It is a pro- 1992;74:448-54. spective, randomized controlled trial, with sample size and 8. Ippolito E, Fraracci L, Farsetti P, et al. The influence of treatment power calculations determined from the results of a pilot on the pathology of club foot. CT study at maturity. J Bone Joint study conducted before the commencement of data collec- Surg Br 2004;86:574-80. tion. Applying block randomization techniques would have 9. Global HELP. Clubfoot: Ponseti Management. 2nd edition. Global- Help Publication; 2005. Available: www.global-help.org/publications resulted in a more even distribution of patients between /books/help_cfponseti.pdf (accessed 2010 Feb. 1). the treatment groups but would not likely have had an 10. Carroll NC. Clubfoot: What have we learned in the last quarter cen- effect on the results obtained with respect to number of tury? J Pediatr Orthop 1997;17:1-2. casts. Our sample size was determined based on pilot data 11. Colburn M, Williams M. Evaluation of the treatment of idiopathic with a standard deviation of 1.88 casts and a power of 0.8. clubfoot by using the ponseti method. J Foot Ankle Surg 2003; 42:259-67. Prestudy calculations using a standard deviation of 1 cast 12. Ponseti IV. Clubfoot management. J Pediatr Orthop 2000;20: called for 7 patients in each group. As such, the current 699-700. treatment group numbers were adequate for the desired 13. Ponseti IV. Common errors in the treatment of congenital clubfoot. study power. The Ponseti technique and Pirani classifica- Int Orthop 1997;21:137-41. tion was reviewed before commencing the study with all 14. Noonan KJ, Richards BS. Nonsurgical management of idiopathic participating surgeons, to control the casting technique. clubfoot. J Am Acad Orthop Surg 2003;11:392-402. 15. Ippolito E, Farsetti P, Caterini R, et al. Long-term comparative Despite this, the sample size was not large enough for sub- results in patients with congenital clubfoot treated with two different group analysis according to clubfoot severity or deformity protocols. J Bone Joint Surg Am 2003;85-A:1286-94. relapse. The main weakness of the study was the use of a 16. Herzenberg JE, Radler C, Bor N. Ponseti versus traditional methods nonvalidated questionnaire to evaluate parental experience. of casting for idiopathic clubfoot. J Pediatr Orthop 2002;22:517-21. 17. Laaveg SJ, Ponseti IV. Long-term results of treatment of congenital club foot. J Bone Joint Surg Am 1980;62:23-31. Conclusion 18. Cooper DM, Dietz FR. Treatment of idiopathic clubfoot. A thirty- year follow-up note. J Bone Joint Surg Am 1995;77:1477-89. There was no significant difference in the number of casts 19. Silfverskiold JP. Fiberglass versus plaster casts. how to choose required for correction of clubfoot between the 2 materi- between them. Postgrad Med 1989;86:71-2, 74. als, SRF and POP. There may be an advantage in using 20. Mihalko WM, Beaudoin AJ, Krause WR. Mechanical properties and material characteristics of orthopaedic casting material. J Orthop POP both economically and in the correction of more Trauma 1989;3:57-63. severe clubfeet (Pirani score ≥ 5), but our study was not 21. Zmurko MG, Belkoff SM, Herzenberg JE. Mechanical evaluation of powered or designed to determine these aspects. In addi- a soft cast material. Orthopedics 1997;20:693-8. tion, the significant improvement in parental experience 22. Coss HS, Hennrikus WL. Parent satisfaction comparing two ban- with SRF determined in this study cannot be ignored, as it dage materials used during serial casting in infants. Foot Ankle Int 1996;17:483-6. may have a positive impact on psychological well-being 23. Berman AT, Parks BG. A comparison of the mechanical properties despite the increased cost associated. of fiberglass cast materials and their clinical relevance. J Orthop Competing interests: None declared. Trauma 1990;4:85-92. 24. Morcuende JA, Abbasi D, Dolan LA, et al. Results of an accelerated Contributors: C. Hui, A. Nettel-Aguirre and J.J. Howard designed the Ponseti protocol for clubfoot. J Pediatr Orthop 2005;25:623-6. study. C. Hui, V. Joughin, S. Goldstein, G. Kiefer, D. Parsons, 25. Ponseti IV. The Ponseti technique for correction of congenital C. Brauer and J.J. Howard acquired the data, which C. Hui, V. Joughin, clubfoot. [author reply 1890-1] J Bone Joint Surg Am 2002;84- A. Nettel-Aguirre and J.J. Howard analyzed. C. Hui and J.J. Howard . A:1889-90. wrote the article, which all authors reviewed and approved for publication. 26. Ponseti IV. Relapsing clubfoot: Causes, prevention, and treatment. Iowa Orthop J 2002;22:55-6. References 27. Ponseti IV. Congenital clubfoot: fundamentals of treatment. New York (NY): Oxford University Press; 1996. 1. Dobbs MB, Morcuende JA, Gurnett CA, et al. Treatment of idio- 28. Flynn JM, Donohoe M, Mackenzie WG. An independent assess- pathic clubfoot: an historical review. Iowa Orthop J 2000;20:59-64. ment of two clubfoot-classification systems. J Pediatr Orthop 1998; 2. Pirani S, Zeznik L, Hodges D. Magnetic resonance imaging study of 18:323-7. the congenital clubfoot treated with the Ponseti method. J Pediatr 29. Haft GF, Walker CG, Crawford HA. Early clubfoot recurrence after Orthop 2001;21:719-26. use of the Ponseti method in a New Zealand population. J Bone Joint 3. Roye DP Jr, Roye BD. Idiopathic congenital talipes equinovarus. J Am Surg Am 2007;89:487-93. Acad Orthop Surg 2002;10:239-48. 30. Pittner DE, Klingele KE, Beebe AC. Treatment of clubfoot with the 4. Morcuende JA, Dolan LA, Dietz FR, et al. Radical reduction in the ponseti method: a comparison of casting materials. J Pediatr Orthop rate of extensive corrective surgery for clubfoot using the Ponseti 2008;28:250-3. method. Pediatrics 2004;113:376-80. 31. Coppola G, Costantini A, Tedone R, et al. The impact of the baby’s 5. Tindall AJ, Steinlechner CW, Lavy CB, et al. Results of manipula- congenital malformation on the mother’s psychological well-being: tion of idiopathic clubfoot deformity in malawi by orthopaedic clin an empirical contribution on the clubfoot. J Pediatr Orthop ical officers using the ponseti method: A realistic alternative for the 2012;32:521-6. developing world? J Pediatr Orthop 2005;25:627-9. 32. Dobbs MB, Rudzki JR, Purcell DB, et al. Factors predictive of out- 6. Ponseti IV, Campos J. Observations on pathogenesis and treatment come after use of the Ponseti method for the treatment of idiopathic of congenital clubfoot. Clin Orthop Relat Res 1972; (84):50-60. clubfeet. J Bone Joint Surg Am 2004;86-A:22-7.

Can J Surg, Vol. 57, No. 4, August 2014 253

compar-howard.indd 253 14-07-16 10:41 AM RESEARCH • RECHERCHE

Computed tomography features associated with operative management for nonstrangulating small bowel obstruction

Rakesh R. Suri, MD* Background: The management of nonstrangulating small bowel obstruction (SBO) Parag Vora, MD† may require surgery, but the need for and timing of surgical intervention isn’t always apparent. We sought to determine whether specific features on computed tomog John M. Kirby, MD† raphy (CT) can predict the necessity for operative management. * Leyo Ruo, MD Methods: Two radiologists independently reviewed CT scans from all patients admitted to hospital with SBO between 2004 and 2006. We examined the associ From the *Division of General Surgery, ation between radiographic features and operative management by univariate Department of Surgery, McMaster Univer- analysis using the χ2 or Fisher exact test. Significant factors with high concor sity, and the †Department of Diagnostic dance between radiologists were entered into a multivariable stepwise logistic Imaging, McMaster University, Hamilton, regression model. Ont. Results: There were 228 patients with SBO, 63 of whom met our inclusion criteria Accepted for publication and had CT scans available for review. Three CT features were frequently associated Feb. 5, 2014 with operative management and had good concordance between radiologists: com plete bowel obstruction, small bowel dilation greater than 4 cm and transition point. Correspondence to: Transition point was the only significant factor predictive of operative management L. Ruo for SBO on multivariable logistic regression analysis (OR 19, 95% confidence interval Division of General Surgery, Department 1.8–201, p = 0.014). of Surgery Juravinski Hospital Conclusion: In patients with nonstrangulating SBO, the presence of a transition 711 Concession St. point on CT scan should alert the surgeon to the increased likelihood that operative Hamilton ON L8V 1C3 management may be required. [email protected] Contexte : La prise en charge de l’occlusion du grêle sans étranglement peut néces DOI: 10.1503/cjs.008613 siter une chirurgie, mais il n’y a pas de règles claires pour déterminer le bien-fondé et le moment de l’intervention. Nous avons voulu déterminer si certaines caractéris tiques spécifiques observées à la tomodensitométrie (TDM) permettent de prédire la nécessité d’une prise en charge chirurgicale. Méhodes : Deux radiologistes ont passé en revue de manière indépendante les TDM de tous les patients hospitalisés pour obstruction du grêle entre 2004 et 2006. Nous avons analysé le lien entre les caractéristiques radiographiques et la prise en charge chirurgicale par analyse univariée à l’aide du test du χ2 ou du test exact de Fisher. Les facteurs importants assortis d’une étroite concordance entre les radiologistes ont été intégrés à un modèle de régression logistique multivariée séquentielle. Résultats : On a dénombré 228 patients atteints d’une occlusion du grêle, dont 63 répondaient à nos critères d’inclusion et pour lesquels on disposait de résultats de TDM à soumettre à l’examen des radiologistes. Trois caractéristiques à la TDM ont fréquemment été associées à la prise en charge chirurgicale, en plus de faire l’objet d’une bonne concordance entre les radiologistes : obstruction intesti nale complète, dilatation du grêle de plus de 4 cm et point de transition (ou saut de calibre). Le point de transition a été le seul facteur prédictif important à l’égard de la prise en charge chirurgicale de l’occlusion du grêle à l’analyse de régression logistique multivariée (rapport de cotes 19; intervalle de confiance de 95 %, 1,8–201; p = 0,014). Conclusion : Chez les patients qui présentent une occlusion du grêle sans étrangle ment, la présence d’un point de transition à la TDM devrait indiquer au chirurgien une plus grande probabilité de traitement chirurgical nécessaire.

comp-ruo.indd 254 14-07-16 12:12 PM RESEARCH

he management of nonstrangulating small Methods bowel obstruction (SBO) may require surgical T intervention. The goal of operative manage- Patients ment is to avoid the increased morbidity and mortality associated with intestinal strangulation while rec- We identified patients discharged with a diagnosis of SBO ognizing the potential for surgical morbidity and between June 2004 and March 2006 from 3 tertiary care hos- mortality. Unfortunately, both the requirement for pitals with joint academic affiliation. We included those who surgery and the timing of surgical intervention may had a CT scan performed within 48 hours of admission. not always be readily apparent, which continues to Exclusion criteria were history of intra-abdominal cancer, challenge surgeons.1 inflammatory bowel disease, abdominal surgery within Various imaging modalities help the surgeon diag- 30 days, previous abdominal or pelvic radiation, comorbid nose SBO.2 Most radiographic methods are currently ities precluding surgical intervention, immediate surgical unable to predict which patients will benefit from early intervention based on clinical evaluation and transfer from surgery; rather, they may illustrate strangulation once outside hospitals. From the hospital’s electronic database and this has occurred.3 One exception is the administration written patient records, we collected information on patient of oral gastrograffin, as its appearance in the colon demographics, clinical and laboratory data, operative findings 24 hours after administration has been shown to suc- and pathological specimens when a resection was performed. cessfully predict the nonoperative resolution of SBO.4 Recurrence of SBO within 2 years was documented. The role of computed tomography (CT) in predict- We acquired CT data from the level of the diaphragm to ing the need for surgical intervention in patients with the lesser trochanters. Oral contrast consisted of 20 mL of nonstrangulating SBO is currently under active investi- ioxithalamate mixed in 1000 mL of water and was adminis- gation. Two studies have shown the small bowel feces tered 1 hour before the study. Rectal contrast, when given, sign to be predictive of nonoperative resolution of consisted of 3 mL of ioxithalamate in 150 mL of water, SBO.5,6 However, there are conflicting reports on the 10 mL of iohexol-300 in 250 mL of water or 15 mL of association between other radiographic features, such as iodixanol in 500 mL of water at hospitals 1, 2 and 3, respect the presence of a transition point or ascites, and the ively. Intravenous contrast was iodixanol, administered as need for surgical intervention.5–10 Furthermore, it is cur- 2 mL/kg to a maximum of 150 mL at a rate of 3 mL/second, rently unknown whether the aforementioned CT find- and the data were acquired in the portal venous phase with a ings are reliably interpreted by independent radiologists 60- to 70-second delay. Data were obtained on multidetec- in the setting of nonstrangulating SBO. Ideally, radio- tor CT scanners: a single- or 16-slice scanner at hospital 1, a graphic features with both good interobserver correla- 4- or 64-slice scanner at hospital 2, and a 4- or 16-slice scan- tion and a strong association with operative manage- ner at hospital 3. Based on a review of the literature for ment will enable the surgeon to monitor selected commonly described CT radiographic features in patients patients who warrant careful observation for the with SBO and at the recommendation of a body radiologist, increased likelihood of surgical intervention without we evaluated 9 CT features: ascites, beak sign, complete progressing to intestinal strangulation. The purpose of bowel obstruction, internal hernia, diameter at point of this study was to determine whether specific features on maximal small bowel dilation (in centimetres), small bowel CT scans exhibiting good interobserver correlation can feces sign, target sign, transition point and whirl sign.1,3,11,12 predict the necessity for operative management in Definitions of these features are provided in Table 1, with patients with nonstrangulating SBO. selected illustrations in Figure 1. Two radiologists (P.V. and

Table 1. Definitions of CT features seen in patients with small bowel obstruction2,12,13

CT feature Definition Ascites Presence of excess peritoneal fluid Beak sign Tapering of the dilated bowel to form what resembles a bird’s beak at the point of obstruction Complete bowel obstruction Lack of oral contrast distal to the point of obstruction Internal hernia Presence of a mesenteric defect through which intestinal loops traverse Maximal small bowel dilation Measurement of the largest small bowel diameter from 1 outer wall to the opposite outer wall Small bowel feces sign Intraluminal particulate matter containing gas bubbles identified in the dilated small bowel segment Target sign Thickened enhancing bowel wall with submucosal edema giving the appearance of 3 concentric rings, with inner and outer rings displaying high attenuation and a middle ring displaying low attenuation Transition point A discrete, focal change in calibre from dilated bowel proximally to collapsed bowel distally Whirl sign Stretched mesenteric vessels converging to a point of intestinal torsion

CT = computed tomography.

Can J Surg, Vol. 57, No. 4, August 2014 255

comp-ruo.indd 255 14-07-16 2:33 PM RECHERCHE

J.K.), blinded to both clinical outcome and prior CT Baseline characteristics between patients undergoing reports, independently analyzed the CT scans. Consensus operative management and those managed nonoperatively was achieved through joint consultation. The local research were compared using the χ2 test or Student t test as appropri ethics board approved our study protocol. ate. We tested associations between each of the radiographic features and the primary outcome of surgical intervention Statistical analysis using either the χ2 test or Fisher exact test for sample sizes of fewer than 6 patients. Concordance was calculated between The primary outcome was operative management for the 2 independent radiologists using κ for those features with SBO. The secondary outcome was recurrence of SBO a significance ofp < 0.05 on univariate analysis. Features with within 2 years of discharge from hospital. both p < 0.05 and κ > 0.5 on univariate analysis were entered stepwise into a multivariable logistic regression model to obtain adjusted odds ratios (OR) with 95% confidence inter A vals (CI). Finally, we compared recurrence of SBO in the surgical and nonsurgical groups using the χ2 test. We con sidered results to be significant atp < 0.05.

Results

There were 228 patients with a diagnosis of SBO during the specified time interval. Of these, 104 patients were excluded: history of intra-abdominal cancer (n = 43), inflammatory bowel disease (n = 20), abdominal surgery within 30 days (n = 17), abdominal or pelvic radiation (n = 3), comorbidities precluding surgical intervention (n = 4), clinical parameters to mandate immediate surgical intervention (n = 8), and transfer from other hospitals (n = 9). Of the remaining 124 patients, 63 had CT images available for review. Of these 63 patients, 27 (43%) underwent operative manage ment and 36 (57%) were managed nonoperatively (Fig. 2).

228 patients with small B bowel obstruction

104 patients excluded • 43 abdominal cancer • 20 in ammatory bowel disease • 17 postoperative less than 30 days • 3 abdominal or pelvic radiation • 8 immediate surgery • 4 comorbid conditions • 9 transferred from another hospital

124 patients

63 patients

Operative management, Nonoperative management, n = 27 (43%) n = 36 (57%) Fig. 1. Illustrations of computed tomography (CT) scan features. (A) Diameter at point of maximal small bowel dilation (B) transition point. Fig. 2. Selection of patients with small bowel obstruction.

256 J can chir, Vol. 57, No 4, août 2014

comp-ruo.indd 256 14-07-16 12:12 PM RESEARCH

All patients undergoing operative management were there were concomitant diagnoses of internal hernia (n = 5) confirmed to have SBO at surgery. The etiologies for and small bowel volvulus (n = 1). Seven patients (26%) obstruction included adhesions (n = 21), incisional hernias required small bowel resection; small bowel ischemia was (n = 3), mesenteric mass (n = 1), appendicitis (n = 1) and confirmed pathologically in 6 of these patients. peristomal hernia (n = 1). In 6 patients with adhesions, There was no significant difference in baseline demo graphic characteristics between the operative and nonoper Table 2. Demographic and clinical characteristics of patients ative groups (Table 2). The presence of abdominal ten with small bowel obstruction managed operatively and derness on examination and white blood cell count on nonoperatively admission were similar between the groups. Patients with a Group; no. (%)* history of multiple abdominal procedures were more likely Surgery, No surgery, to require surgical intervention (OR 2.8, 95% CI 0.90–8.8, Clinical factor n = 27 n = 36 p value p = 0.08), although this was not significant. Sex 0.82 Of the 9 radiographic features studied, 5 were signifi Female 15 (56) 21 (58) cantly associated with surgical intervention: beak sign (OR Male 12 (44) 15 (42) 10, 95% CI 3.1–32, p < 0.001), complete bowel obstruc Mean age, yr 64 63 0.87 tion (OR 8.5, 95% CI 2.6–28, p < 0.001), maximal small No. of previous surgeries 0.08 bowel dilation greater than 4 cm (OR 5.1, 95% CI 1.5– > 1 19 (70) 18 (50) 7.9, p = 0.010), small bowel feces sign (OR 3.6, 95% CI 1 6 (22) 16 (44) 0 2 (7) 2 (6) 1.1–12, p = 0.039) and transition point (OR 32, 95% CI Abdominal tenderness 0.17 4.0–270, p < 0.001). Those features not achieving statisti Yes 22 (82) 34 (94) cal significance were the presence of ascites p( = 0.14) and No 3 (11) 2 (6) target sign (p = 0.71). Internal hernia and whirl sign were Unknown 2 (7) 0 (0) found in 6 (22%) and 5 (19%) patients, respectively, and Mean WBC x 103/mm3 13 12 0.36 all 11 patients were submitted to operative management. WBC = white blood cell count. The type of CT contrast used was not significant between *Unless otherwise indicated. the operative and nonoperative groups (Table 3).

Table 3. Univariate analysis of CT features in operative and nonoperative patients, and κ values for features with p < 0.05

Group; no. (%)

Surgery, No surgery, CT feature n = 27 n = 36 OR (95% CI) p value κ Ascites 14 (52) 12 (33) 2.2 (0.77–6.0) 0.14 Beak sign 20 (74) 8 (22) 10 (3.1–32) < 0.001 0.43 Complete bowel obstruction 17 (63) 6 (17) 8.5 (2.6–28) < 0.001 0.52 Internal hernia 6 (22) 0 (0) SB dilation > 4 cm 23 (85) 19 (53) 5.1 (1.5–7.9) 0.010 0.63 SB feces sign 10 (37) 5 (14) 3.6 (1.1–12) 0.039 0.30 Target sign 3 (11) 3 (8) 1.4 (0.26–7.4) 0.71 Transition point 26 (96) 16 (44) 32 (4.0, 266) < 0.001 0.66 Whirl sign 5 (19) 0 (0) Intravenous contrast 22 (82) 32 (89) 0.55 (0.13–2.28) 0.41 Oral contrast 24 (89) 33 (92) 0.73 (0.13–3.9) 0.71 Rectal contrast 6 (22) 6 (17) 1.43 (0.40–5.0) 0.58

CI = confidence interval; CT = computed tomography; OR = odds ratio; SB = small bowel.

Table 4. Multivariable model of CT features associated with operative management for small bowel obstruction

CT feature OR (95% CI) p value Complete bowel obstruction 3.2 (0.15, 13) 0.09 SB dilation > 4 cm 0.87 (0.15, 4.9) 0.88 Transition point 19 (1.8, > 200) 0.014

CI = confidence interval; CT = computed tomography; OR = odds ratio; SB = small bowel.

Can J Surg, Vol. 57, No. 4, August 2014 257

comp-ruo.indd 257 14-07-16 2:33 PM RECHERCHE

Of the 5 radiographic features achieving statistical sig toward operative management in patients having under nificance, 3 showed good correlation between radiologists gone more than 1 previous abdominal surgery; this find with κ values greater than 0.5: complete bowel obstruction ing may be explained by the development of extensive (κ = 0.52), maximal small bowel dilation greater than 4 cm adhesions often anticipated in patients with multiple (κ = 0.63) and transition point (κ = 0.66; Table 3). These prior surgeries. 3 features were entered into a multivariable stepwise logis In comparison with other studies to date that have tic regression model. Transition point retained statistical explored the association between CT radiographic fea significance (OR 19, 95% CI 1.8–201,p = 0.014), while tures and the need for subsequent surgical intervention complete bowel obstruction (p = 0.09) and maximal small in patients with SBO, the present study consists of a bowel dilation greater than 4 cm (p = 0.88) did not; includ strictly defined cohort. The study population was ing complete bowel obstruction and small bowel dilation selected to consist only of patients with suspected adhe did not significantly improve the fit of the model (Table 4). sive nonstrangulating SBO through predefined exclu A transition point was identified in all 7 patients requiring a sion criteria. Patients were excluded for clinical suspi small bowel resection. cion of strangulation that would require immediate Recurrence of SBO did not differ significantly p( = 0.75) operative management. Patients were also excluded if between the groups, occurring in 3 (12%) patients managed there was the potential for favouring nonoperative man operatively and 5 (14%) patients managed nonoperatively. agement owing to other medical circumstances, such as in patients with incurable intra-abdominal malignancy, Discussion inflammatory bowel disease, recent abdominal surgery, prior abdominal or pelvic radiation, or severe comorbid The current role of CT in the management of SBO lies illnesses.18,19 By using 2 expert radiologists blinded to in its ability to diagnose obstruction, to define the etiol each other’s interpretations and to patient outcomes, ogy and probable location of the obstruction, and to dif only CT findings with good interobserver correlation ferentiate nonstrangulating from strangulating obstruc were included in the multivariable logistic regression tion. Computed tomography assessment is effective, model. Our study was specifically designed to achieve with a sensitivity of 83%–100% and specificity of 61%– results that may be more readily extrapolated to the sur 93%.3 In patients with nonstrangulating SBO, there are gical management of patients with nonstrangulating limited data on whether CT may help predict which SBO at other centres. patients will require surgical intervention. While a sur An identifiable transition point on CT was most geon’s decision to operate ultimately depends on the significantly associated with the need for operation in patient’s clinical condition, CT features predictive for patients with nonstrangulating SBO, both on univari operative management of patients with nonstrangulating ate and multivariate analyses. A transition point was SBO may facilitate care. Such CT findings could target also the only consistent CT finding in all patients who a subset of these patients for heightened vigilance in required a small bowel resection. The association is an effort to minimize operative delay, thereby reducing plausible, given the discrete and localized change in the increased morbidity and mortality from intestinal intestinal calibre seen in a transition point. Our results ischemia and associated complications. suggest that a transition point represents a fixed rather The cohort of patients in this study with nonstrangulat than a transient point of intestinal obstruction unlikely ing, adhesive SBO is representative of analogous popula to resolve without operative intervention. Four tions in comparable studies. The proportion of patients studies6,8–10 to date have evaluated the clinical relevance having had multiple, 1 or no prior abdominal surgeries is of a transition point; however, only the study by corroborated by previous studies, and the operative rate of Hwang and colleagues8 supports the finding of an 43% lies within the widely reported range of 27%–66%.14 increased likelihood of operative management. All 4 of The rates of small bowel resection and small bowel these studies were subject to less stringent inclusion/ ischemia are comparable to those reported in another exclusion criteria and may not represent the population recent study.7 Furthermore, the similar recurrence rates of of patients targeted in the present study. Furthermore, SBO in operative and nonoperative patients are also sup CT interpretation in these studies may have been sub ported by current literature.14,15 ject to observer bias, as analysis was performed by Other studies have shown that clinical findings and either a single radiologist or without blinding, or was laboratory measurements at initial presentation are based on findings extracted from the original CT inadequate to predict the need for surgical interven reports. tion.16 Consistent with previous reports,17,18 the presence In the present study, complete bowel obstruction was of abdominal tenderness or leukocytosis at admission in significantly associated with surgical intervention on the present study had no predictive value for requiring univariate analysis, clearly a sound and probably antici surgical intervention. There was, however, a trend pated clinical decision. Although not significant on

258 J can chir, Vol. 57, No 4, août 2014

comp-ruo.indd 258 14-07-16 12:12 PM RESEARCH

Contributors: All authors designed the study, acquired and analyzed multivariate analysis and perhaps a function of the small the data, wrote and reviewed the article and approved the final version number of patients involved, there was a clear associa- for publication. tion with operative management in patients with this CT finding on univariate analysis (OR 8.5, 95% CI 2.6– References 28). Other studies have reported a similar association on univariate analysis.8,10 The absence of orally adminis- 1. Hayanga AJ, Bass-Wilkins K, Bulkley GB. Current management of tered contrast beyond a fixed point of obstruction (i.e., small-bowel obstruction. Adv Surg 2005;39:1-33. complete bowel obstruction) may be a sufficient indica- 2. Suri S, Gupta S, Sudhakar PJ, et al. Comparative evaluation of plain tion for surgical intervention. films, ultrasound and CT in the diagnosis of intestinal obstruction. Acta Radiol 1999;40:422-8. The positive correlation between small bowel feces sign and operative management found in the present 3. Burkill G, Bell J, Healy J. Small bowel obstruction: the role of com- 5,6 puted tomography in its diagnosis and management with reference study is discordant with the findings of 2 recent studies. to other imaging modalities. Eur Radiol 2001;11:1405-22. That said, the small bowel feces sign was previously 4. Abbas SM, Bissett IP, Parry BR. Meta-analysis of oral water-soluble reported to occur more frequently in patients with contrast agent in the management of adhesive small bowel obstruction. 11 moderate and high-grade SBO. While further pro- Br J Surg 2007;94:404-11. spective studies are required to clarify this discrepancy, 5. Deshmukh SD, Shin DS, Willmann JK, et al. Non-emergency modest interobserver agreement among reporting radi- small bowel obstruction: assessment of CT findings that predict ologists, as demonstrated in this study, may limit the need for surgery. Eur Radiol 2011;21:982-6. usefulness of the small bowel feces sign in guiding clin- 6. Zielinski MD, Eiken PW, Bannon MP, et al. Small bowel obstruction ical decision making. — who needs an operation? A multivariate prediction model. World J Additional CT features, which have not been exten- Surg 2010;34:910-9. sively reported in the literature, were evaluated in the 7. O’Daly BJ, Ridgway PF, Keenan N, et al. Detected peritoneal fluid present study. Identification of an internal hernia and in small bowel obstruction is associated with the need for surgical whirl sign, although infrequent, was found only in the intervention. Can J Surg 2009;52:201-6. group of surgically managed patients. In the setting of 8. Hwang JY, Lee JK, Lee JE, et al. Value of multidetector CT in SBO, an internal hernia may be similar in clinical behav- decision making regarding surgery in patients with small-bowel iour to an incarcerated external hernia. Entrapped small obstruction due to adhesion. Eur Radiol 2009;19:2425-31. bowel is unlikely to reduce spontaneously, prompting 9. Colon MJ, Telem DA, Wong D, et al. The relevance of transition operative intervention. The whirl sign suggests stretch- zones on computed tomography in the management of small bowel ing of the mesenteric vessels toward a point of intestinal obstruction. Surgery 2010;147:373-7. torsion; such tension on the small bowel mesentery may 10. Jones K, Mangram AJ, Lebron RA, et al. Can a computed tomography signify an irreversible consequence of intestinal obstruc- scoring system predict the need for surgery in small-bowel obstruction? Am J Surg 2007;194:780-3, discussion 783-4. tion that requires surgical correction. Although it was not possible to calculate an OR for these 2 CT features, 11. Lazarus DE, Slywotsky C, Bennett GL, et al. Frequency and relevance of the “small-bowel feces” sign on CT in patients with small-bowel internal hernia and whirl sign appear to represent find- obstruction. Am J Roentgenol 2004;183:1361-6. ings with a physiologic basis and clinical rationale for 12. Torreggiani WC, Harris AC, Lyburn ID, et al. Computed tomog surgical intervention. raphy of acute small bowel obstruction: pictorial essay. Can Assoc Radiol J 2003;54:93-9. Conclusion 13. Ahualli J. The target sign: bowel wall. Radiology 2005;234:549-50. The management of patients with nonstrangulating SBO 14. Miller G, Boman J, Shrier I, et al. Natural history of patients with adhesive small bowel obstruction. Br J Surg 2000;87:1240-7. remains a clinical challenge. While the timing and need for surgery ultimately depends on the surgeon’s assess- 15. Miller G, Boman J, Shrier I, et al. Etiology of small bowel obstruction. Am J Surg 2000;180:33-6. ment of the patient’s condition and course in hospital, the presence of a transition point on early CT scan 16. Maglinte DD, Kelvin FM, Sandrasegaran K, et al. Radiology of small bowel obstruction: contemporary approach and controversies. should alert the surgeon to an increased likelihood that Abdom Imaging 2005;30:160-78. operative management will be required to resolve the SBO. Heightened awareness driven by CT findings 17. Sarr MG, Bulkley GB, Zuidema GD. Preoperative recognition of intestinal strangulation obstruction. Prospective evaluation of diag- should prompt close patient monitoring to minimize nostic capability. Am J Surg 1983;145:176-82. delay in surgical intervention and thereby reduce the 18. Shatila AH, Chamberlain BE, Webb WR. Current status of diagnosis potential risk for intestinal ischemia and its conse- and management of strangulation obstruction of the small bowel. Am quences in this population. J Surg 1976;132:299-303. Acknowledgements: We thank Kevin Chan, MD, for his help with 19. Cox MR, Gunn IF, Eastman MC, et al. The safety and duration of data entry. non-operative treatment for adhesive small bowel obstruction. Aust N Competing interests: None declared. Z J Surg 1993;63:367-71.

Can J Surg, Vol. 57, No. 4, August 2014 259

comp-ruo.indd 259 14-07-16 12:12 PM RESEARCH • RECHERCHE

The historic predictive value of Canadian orthopedic surgery residents’ orthopedic in-training examination scores on their success on the RCPSC certification examination

David Yen, MD* Background: Positive correlation between the orthopedic in-training examination George S. Athwal, MD† (OITE) and success in the American Board of Orthopaedic Surgery examination has been reported. Canadian training programs in internal medicine, anesthesiology and urology ‡ Gary Cole, PhD have found a positive correlation between in-training examination scores and performance on the Royal College of Physicians and Surgeons of Canada (RCPSC) certification exam From the *Department of Surgery, ination. We sought to determine the potential predictive value of the OITE scores of Queen’s University, Kingston, Ont., Canadian orthopedic surgery residents on their success on their RCPSC examinations. †Department of Surgery, University of Western Ontario, London, Ont., ‡Royal Methods: A total of 118 Canadian orthopedic surgery residents had their annual College of Physicians and Surgeons of OITE scores during their 5 years of training matched to the RCPSC examination oral Canada Office of Education, Ottawa, Ont. and multiple-choice questions and to overall examination pass/fail scores. We calculated Pearson correlations between the in-training examination for each postgraduate year Accepted for publication and the certification oral and multiple-choice questions and pass/fail marks. Feb. 5, 2014 Results: There was a predictive association between the OITE and success on the RCPSC examination. The association was strongest between the OITE and the written Correspondence to: multiple-choice examination and weakest between the OITE and the overall examination D. Yen pass/fail marks. Douglas 5, Kingston General Hospital 76 Stuart St. Conclusion: Overall, the OITE was able to provide useful feedback to Canadian Kingston ON K7L 2V7 orthopedic surgery residents and their training programs in preparing them for their [email protected] RCPSC examinations. However, when these data were collected, truly normative data based on a Canadian sample were not available. Further study is warranted based on a more refined analysis of the OITE, which is now being produced and includes norma- DOI: 10.1503/cjs.014913 tive percentile data based on Canadian residents.

Contexte : On a signalé une corrélation positive entre l’examen intermédiaire en orthopédie (EIO) et la réussite aux examens de l’American Board of Orthopaedic Sur- gery. Les programmes canadiens de formation en médecine interne, en anesthésiolo- gie et en urologie ont constaté une corrélation positive entre les notes aux EIO et les résultats aux examens du Collège royal des médecins et chirurgiens du Canada (CRMCC). Nous avons cherché à déterminer la valeur prédictive potentielle des notes des résidents en chirurgie orthopédique à l’EIO pour ce qui est de leur réussite aux examens du CRMCC. Méthodes : Les notes de 118 résidents en chirurgie orthopédique du Canada aux EIO pendant leurs 5 années de formation ont été comparées à leurs résultats aux examens oraux et à choix multiples, ainsi qu’à leur note globale de passage ou d’échec. Nous avons calculé les corrélations de Pearson entre les résultats à l’EIO pour chaque année de formation et les notes aux examens de certification oraux et à choix multiples, et les notes de passage ou d’échec. Résultats : Il y avait un rapport prédictif entre le résultat à l’EIO et la réussite de l’examen du CRMCC. Le rapport était le plus étroit entre les résultats à l’EIO et les résultats à l’examen écrit à choix multiples, et il était le plus faible entre les résultats à l’EIO et les notes globales de passage ou d’échec. Conclusion : Dans l’ensemble, l’EIO a produit une rétroaction utile pour les résidents en chirurgie orthopédique et leurs programmes de formation pour les préparer aux examens du CRMCC. Toutefois, lorsque ces données ont été recueillies, de véritables données normatives fondées sur un échantillon canadien n’étaient pas disponibles. Une étude plus pous- sée s’impose à partir d’une analyse plus approfondie de l’EIO; cette analyse est en voie de réalisation et comprend des données normatives percentiles sur les résidents canadiens.

historic-yen.indd 260 14-07-16 10:49 AM RESEARCH

anadian orthopedic surgery residency training pro- dic surgery residents for their success on their RCPSC cer- grams require residents to write the orthopedic in- tification examinations. The RCPSC certification in C training examination (OITE) annually. This is a orthopedic surgery is based on a compensatory examina- multiple-choice examination overseen by the Committee tion combining oral and written components to yield an on Examinations and Evaluation of the American Academy overall mark that determines whether the candidates pass of Orthopaedic Surgeons (AAOS), administered continu- or fail. ously since 1963.1 It covers all aspects of orthopedic surgery with questions designed to test recall, comprehension, Methods application, problem solving, evaluation and synthesis (tax- onomy levels 1 through 6, respectively). Thirteen English-speaking residency training programs Correlation between the OITE and success in the American elected to participate in this study. The study cohort con- Board of Orthopaedic Surgery (ABOS) examination has been sisted of 118 Canadian orthopedic surgery residents (38 in studied. One study reported a high risk of failure (63%; 5 of 8 2000–2001, 44 in 2001–2002, 36 in 2002–2003). The annual failed) on the ABOS Part-I examination when a resident scored OITE scores obtained by each resident during their 5 years below the 29th percentile for postgraduate year (PGY)-3 and of training were collected by their program directors and below the 20th percentile for PGY5. No failures occurred (50 matched to their corresponding residents’ RCPSC ID num- passed) when either the PGY3 score was above the 32nd per- bers in a nonidentifying blinded fashion on a data sheet. Sim- centile or the PGY4 score was above the 27th percentile.2 ilarly, a staff member of the Educational Research Unit of Another study reported that those who averaged in the 27th the RCPSC Office of Education entered the residents’ percentile or lower on the OITE had a 57% chance of failing RCPSC certification examination scores on a data sheet the ABOS Part-I examination.3 Crawford and colleagues4 matched to their corresponding residents’ RCPSC ID num- reported that PGY3 OITE percentile scores predicted ABOS bers, with no reference to resident names. The data sheets Part-I and Part-II passage, with residents who scored in the were sent to another individual at the RCPSC to match the lower quartile having a 5.2 times greater risk of failure on certification oral and multiple-choice examinations and over- ABOS Part-I. Herndon and colleagues5 reported that the all pass/fail examination marks using the nonidentifying resi- OITE percentile score in the final year in training was a predic- dent RCPSC ID numbers. No resident names were attached tor of success on the ABOS Part-I and Part-II examinations. to any of the data, and their confidentiality was preserved. Canadian training programs in internal medicine, anesthesiology and urology make use of examinations similar to the Statistical analysis OITE and have studied the correlation between resident performance on these examinations and their subsequent success The Pearson correlations between the OITE raw and percen- on their Royal College of Physicians and Surgeons of Canada tile scores for each of the candidates’ final PGY and the certi- (RCPSC) certification examinations.6–8 All 3 specialties have fication scores on the oral and written multiple-choice exami- found a positive correlation between in-training examination nations and on the overall pass/fail marks were calculated. scores and performance on the RCPSC certification examination. Specifically, in internal medicine, it has been reported Results that there is a high correlation between the results of the in- training examination and the written component of the The correlation between the OITE percentile and RCPSC RCPSC certification examination. In-training examination oral and written multiple-choice examination marks was sig- scores above the 50th percentile were predictive of a low fail- nificant at the 0.01 level in each of the 3 final academic years. ure rate (< 1.5%), and scores below the 10th percentile were The correlation between the OITE raw score and the associated with a high failure rate (24%) on the written com RCPSC pass/fail marks was significant at the 0.01 level in 2 ponent of the RCPSC examination6. In anesthesiology, in- of the 3 years; the calculation in the third year was not possi- training scores above the 50th percentile were highly predic- ble because all the candidates passed. There was a stronger tive of success on the written component of the RCPSC association between the RCPSC oral and written multiple- examination, and scores above the 60th percentile were highly choice examinations and overall pass/fail marks and the predictive of success on the oral component of the examina- OITE raw score than with the OITE percentile in 2 of the tion. In-training scores below the 20th percentile were pre- 3 years. The strongest OITE association was with the dictive of failure on both the written and oral components.7 RCPSC multiple-choice examination, and the weakest asso- This provides useful feedback to the residents and their train- ciation was with the overall pass/fail mark (Table 1). ing programs in preparing them for RCPSC certification examinations. To our knowledge, no similar investigation has Discussion been done for orthopedic surgery. The purpose of this study was to determine the poten- The RCPSC certification examination in orthopedic surgery tial predictive value of OITE scores of Canadian orthope- was a compensatory examination combining oral and written

Can J Surg, Vol. 57, No. 4, August 2014 261

historic-yen.indd 261 14-07-16 10:49 AM RECHERCHE

components. The OITE was a multiple-choice examination tion. Since 2009, substantial improvements have been made representing the spectrum of clinical orthopedics. It is not sur- to the OITE reports. Prior to 2009 there was no standard- prising that the strongest correlation was with the RCPSC ization of scores on the OITE and no breakdown of norms multiple-choice examination, which is a similar assessment into different groups. The OITE reports now contain measure. There was also a significant correlation with the Canadian norms (i.e., percentiles) as well as reports for the RCPSC oral examination marks, indicating that the OITE was different content domains in orthopedics. This presents the a useful tool for preparing residents for this component as well. possibility that greater prediction may be possible not only There was a significant correlation between the OITE raw based on the overall percentile scores, but also based on a scores and the RCPSC overall pass/fail marks in 2 of the 3 regression analysis representing the content domains. In years; the calculation in the third year was not possible because addition, an objective structured clinical examination for- all the candidates were successful. These results indicate that mat has been added to the oral component of the RCPSC the OITE was able to provide useful feedback to the residents certification examination. Overall, further study into the and their training programs concerning their acquisition of current association between the OITE and the RCPSC appropriate knowledge in preparation for RCPSC certification. certification examination is warranted. There was a stronger association between the RCPSC Acknowledgements: We acknowledge the orthopedic surgery training oral and written multiple-choice examinations and the over- program directors who offered support and encouragement to this pro all examination pass/fail marks with the OITE raw scores ject. This work was supported by the Wigle Award from the Office of Postgraduate Medical Education, Queen’s University. than with the OITE percentiles. The percentile scores provided were not an accurate reflection of the competence of Competing interests: None declared. the Canadian residents. The percentile is an individual’s raw Contributors: D. Yen and G. Athwal designed the study. D. Yen and G. Cole score compared with their peers in the same year in training acquired the data, which G. Cole analyzed. D. Yen wrote the article, which G. thwal and G. Cole reviewed. All authors approved the paper for publication. (YIT) with a resident in YIT-1 defined as having completed 6 months of orthopedic training. In Canada, individual train- References ing programs vary in terms of the amount of time spent in orthopedics during the first 2 years of core surgery training, 1. Mankin HJ. The orthopaedic in-training examination (OITE). thus resulting in the possibility of different YIT assignment Clin Orthop Relat Res 1971; (75):108-16. for the OITE between programs of residents in the same PGY. We believe that this inconsistent association between 2. Klein GR, Austin MS, Randolph S, et al. Passing the boards: Can PGY and YIT in different programs is one reason that we USMLE and orthopaedic in-training examination scores predict the ABOS part-1 examination? J Bone Joint Surg Am 2004;86-A:1092-5. could not find an OITE threshold percentile for those passing or failing the RCPSC multiple-choice and oral examina- 3. Dougherty PJ, Walter N., Schilling P., et al. Do scores of the tions and the overall examination. A second reason that the USMLE Step 1 and OITE correlate with the ABOS Part I certify- percentile scores were inaccurate is that they were based on ing examination?: a multicenter study. Clin Orthop Relat Res the entire sample of residents taking the OITE, most of 2010;468:2797-802. whom were American candidates. 4. Crawford CH III, Nyland J, Roberts CS, et al. Relationship among United States Medical Licensing Step 1, orthopaedic in-training, Conclusion subjective clinical performance evaluations, and American Board of Orthopaedic Surgery examination scores: a 12-year review of an orthopedic surgery residency program. J Surg Educ 2010;67:71-8. This study has demonstrated that the OITE had the potential to predict the success of Canadian residents in the years 5. Herndon JH, Allan BJ, Dyer G, et al. Predictors of success on the 2000 through 2003 on the RCPSC certification examina- American Board of Orthopaedic Surgery examination. Clin Orthop Relat Res 2009;467:2436-45.

Table 1. Pearson correlations between the orthopedic in-training examination 6. Brill-Edwards P, Couture L, Evans G, et al. Pre- for each postgraduate year and the certification oral and written multiple-choice dicting performance on the Royal College of Phy- examinations and the overall examination pass/fail marks sicians and Surgeons of Canada internal medicine Group; Pearson correlation written examination. CMAJ 2001;165:1305-7.

2002–03 2001–02 2000–01 7. Kearney RA, Sullivan P, Skakun E. Performance on ABA-ASA in-training examination predicts suc- Raw, Percentile, Raw, Percentile, Raw, Percentile, Factor n = 31 n = 36 n = 38 n = 38 n = 29 n = 38 cess for RCPSC certification. Can J Anaesth 2000;47:914-8. Oral 0.72* 0.63* 0.57* 0.44* 0.36 0.42* Multiple-choice 0.76* 0.67* 0.74* 0.61* 0.37† 0.65* 8. Baverstock RJ, MacNeily AE, Cole G. The Ameri- Pass/fail 0.61* 0.56* 0.50* 0.26 ‡ ‡ can Urological Association In-Service Examina- *Correlation is significant at the 0.01 level (2-tailed). tion: performance correlates with Canadian and †Correlation is significant at the 0.05 level (2-tailed). American specialty examinations. J Urol 2003; ‡Could not be computed, as all candidates in the sample passed. 170:527-9.

262 J can chir, Vol. 57, No 4, août 2014

historic-yen.indd 262 14-07-16 2:31 PM RESEARCH • RECHERCHE

Bridging the gap between open and minimally invasive pancreaticoduodenectomy: the hybrid approach

Yifan Wang Background: Minimally invasive pancreatic surgery has evolved rapidly, but total laparo- Simon Bergman, MD, MSc scopic pancreaticoduodenectomy has not been widely adopted owing to its technical complexity. Hybrid laparoscopy-assisted pancreaticoduodenectomy (HLAPD) combines the Sabrina Piedimonte, MSc relative ease of open surgery with the benefits of a minimally invasive approach. This study Tsafrir Vanounou, MD, MBA evaluates the safety and effectiveness of the hybrid approach compared with open surgery. Methods: We retrospectively analyzed data of consecutive patients undergoing From the Division of General Surgery, either hybrid or open pancreaticoduodenectomy (OPD) at our institution between Jewish General Hospital, Montréal, Que. September 2009 and December 2013. Demographic, operative and oncologic data were collected to compare outcomes between HLAPD and OPD. See the related commentary by Jayaraman on p. 228. Results: Our analysis included 33 patients (HLAPD: n = 13; OPD: n = 20). There were no differences in patient demographics, comorbidities or surgical indications. The Abstract presented at the Society of HLAPD group had significantly lower intraoperative blood loss (450 mL v. 1000 mL, American Gastrointestinal and Endo- p = 0.023) and shorter length of hospital stay (8 v. 12 d, p = 0.025) than the OPD group. scopic Surgeons (SAGES) meeting, Bal- Duration of surgery did not differ significantly between the groups. There were no dif- timore, MD, April 2013 and the Canadian ferences in postoperative analgesic requirements, Clavien grade I/II or grade III/IV Association of General Surgeons (CAGS) complications or 90-day mortality. Oncologic outcomes showed no significant differ- meeting, Ottawa, ON, September 2013. ences in tumour size, R1 resection rate or number of lymph nodes harvested. Conclusion: In select patients, HLAPD is a safe and effective procedure with compar Accepted for publication able outcomes to conventional open surgery. Wider adoption of the hybrid approach Feb. 7, 2014 will allow a greater number of patients to benefit from a less invasive procedure while facilitating the transition toward purely minimally invasive pancreaticoduodenectomy. Correspondence to: T. Vanounou Division of General Surgery Contexte : La chirurgie pancréatique minimalement effractive a rapidement évolué, Jewish General Hospital mais la pancréatoduodénectomie laparoscopique totale n’a pas été largement adoptée 3755 Cote-Sainte-Catherine en raison de sa complexité technique. La pancréatoduodénectomie hybride sous lapa- Montréal QC H3T 1E2 roscopie (PDHL) allie la relative facilité de la chirurgie ouverte aux avantages d’une [email protected] approche minimalement effractive. Cette étude compare l’innocuité et l’efficacité de l’approche hybride à celles de la chirurgie ouverte. DOI: 10.1503/cjs.026713 Méthodes : Nous avons analysé de manière rétrospective les données concernant des patients consécutifs soumis à une pancréatoduodénectomie hybride ou ouverte (PDO) dans notre établissement entre septembre 2009 et décembre 2013. Les données démographiques, opératoires et oncologiques ont été recueillies pour comparer les résultats entre la PDHL et la PDO. Résultats : Notre analyse a inclus 33 patients (PDHL : n = 13; PDO : n = 20). Il n’y avait aucune différence quant aux caractéristiques démographiques, comorbidités ou indications chirurgicales chez les patients. Le groupe soumis à la PDHL a connu des pertes sanguines peropératoires significativement moindres (450 mL c. 1000 mL,p = 0,023) et un séjour hospitalier significativement plus bref (8 j c. 12 j,p = 0,025) comparativement au groupe soumis à la PDO. La durée de la chirurgie n’a pas significativement différé entre les groupes. On n’a noté aucune différence sur le plan des besoins en anal- gésiques postopératoires, des complications de grade I/II ou III/IV sur l’échelle de Clavien ou de la mortalité à 90 jours. Quant aux paramètres oncologiques, aucune dif- férence significative n’a été notée pour ce qui est de la taille de la tumeur, du taux de résection R1 ou du nombre de ganglions recueillis. Conclusion : Pour certains patients, la PDHL est une intervention sécuritaire et efficace qui donne des résultats comparables à la chirurgie ouverte classique. L’adoption à plus grande échelle de l’approche hybride permettra à plus de patients de bénéficier d’une intervention moins effractive et facilitera la transition complète vers la pancréa- toduodénectomie minimalement effractive.

bridging-vanou.indd 263 14-07-16 10:51 AM RECHERCHE

ecent advances in laparoscopic techniques have led to evidence of major vascular involvement. Patients whose an increased interest in minimally invasive pancreatic lesions were at high risk of a positive margin or of abutment R surgery. Compared with conventional open surgery, of major vessels were excluded from the HLAPD group and minimally invasive procedures allow for decreased postopera- underwent OPD. Patients were not excluded on the basis of tive pain, shorter hospital stay and improved cosmesis.1–3 demographic factors, such as age, body mass index (BMI), Despite these benefits, the adoption of total laparoscopic pan- Charlson Index and American Society of Anesthesiologists creaticoduodenectomy (TLPD) has been hindered by con- (ASA) grade. Prior to surgery, all patients were informed of cerns regarding the technical complexity of laparoscopic the potential advantages and complications of both tech- reconstruction. Since the first report by Gagner and Pomp in niques, and they provided written informed consent. 1994,4 only a few centres worldwide have published large In our initial institutional experience, we imposed a low TLPD patient series.4–8 A direct transition from open surgery threshold to convert to an open procedure. To accurately to TLPD may constitute a hazardous and imprudent leap for interpret the benefits and shortcomings of the hybrid surgeons without extensive prior laparoscopic experience. In approach, we conducted a non-intent-to-treat analysis, defin- light of the steep learning curve, the transition toward TLPD ing procedures as HLAPD only when all 3 resections (antrec- may be more safely and effectively achieved as a multistep tomy, choledochectomy and pancreatectomy) had been per- progression using a spectrum of minimally invasive tech- formed laparoscopically. Patients whose cases began niques. In this report, we describe hybrid laparoscopy-assisted laparoscopically but were converted to open surgery before pancreaticoduodenectomy (HLAPD): a hybrid laparoscopic– completion of the resections were included in the OPD group. open approach in which pancreaticoduodenal resection is performed laparoscopically, while reconstruction is completed Operative technique via a small upper midline minilaparotomy.9 The hybrid method combines the relative ease of conventional open sur- The patient is placed in the supine position on a split-leg

gery with the benefits of a minimally invasive approach. table, and CO2 pneumoperitoneum is established via a Potentially, HLAPD may serve as a valuable stepping stone to 12 mm infraumbilical trocar inserted using an open facilitate the transition from open to purely minimally inva- Hasson technique. A 30º camera is used to assess for any sive pancreaticoduodenectomy without incurring additional evidence of metastatic disease. If no contraindications to risk to the patient. Although the feasibility of HLAPD has resection are found, 6 additional trocars are inserted along been described, the current literature mainly comprises small a semicircle centred on the head of the pancreas (Fig. 1). patient series lacking comparison groups.7,10–12 To our know The operation is begun by dividing the gastrocolic liga- ledge, only 3 reports have compared the outcomes of patients ment with a LigaSure (Valleylab). With the stomach and undergoing HLAPD versus open pancreaticoduodenectomy left lateral segment of the liver retracted against the anter (OPD).13–15 In the studies by Cho and colleagues13 and Lee ior abdominal wall using a miniretractor (Mediflex), the and colleagues,15 patients with preoperatively diagnosed peri- gastroepiploic omentum is separated off the transverse ampullary carcinoma automatically underwent OPD; patients mesocolon. The right gastroepiploic vein is followed to its who underwent HLAPD displayed only benign or low-grade junction with the infrapancreatic superior mesenteric vein lesions. To our knowledge, we report the first Canadian study (SMV) and divided. The right colon is mobilized, and a evaluating the safety, feasibility and operative outcomes of laparoscopic Kocher manoeuvre is performed to the level HLAPD compared with OPD. of the ligament of Treitz. The gastric antrum is transected using serial purple loads of the EndoGIA stapler (US Sur- Methods gical Corp.). The common hepatic node is identified and resected. The gastrohepatic ligament is opened to expose With institutional review board approval, we performed a the common hepatic artery, from which the gastroduo retrospective chart review on all patients undergoing denal artery can be traced down. Flow within the common HLAPD or OPD at a single institution between Septem- hepatic artery is verified using a laparoscopic ultrasound ber 2009 and December 2013. Demographic, operative probe before transecting the gastroduodenal artery using and outcome data were collected from a prospectively the white load of the EndoGIA stapler. A retropancreatic maintained database. All surgeries were performed by a tunnel is created by dissecting between the posterior sur- single experienced pancreatic surgeon (T.V.), with another face of the pancreas and the anterior plane of the SMV in attending surgeon (S.B.) as first assistant. a cephalad direction. The tunnelled pancreas is then encircled using a Penrose tape. A complete hilar lympha Patient selection denectomy is undertaken to harvest periportal and peri- pancreatic lymph nodes. A retrograde cholecystectomy is Preoperatively, all patients underwent appropriate imaging performed, leaving the gallbladder attached to the com- studies to assess tumour resectability. The selection criteria mon bile duct for traction. The hepatoduodenal ligament for HLAPD were tumours of any size without preoperative is dissected to isolate the underlying portal vein. The

264 J can chir, Vol. 57, No 4, août 2014

bridging-vanou.indd 264 14-07-16 10:51 AM RESEARCH

common bile duct is encircled with an umbilical tape and 3–0 silk stitches.16 Next, an end-to-side hepaticojejunos- transected above the junction with the cystic duct using tomy is performed using interrupted 5–0 PDS, and a side- the white load of the EndoGIA stapler. The proximal jeju- to-side retrogastric antecolic loop gastrojejunostomy is num is brought back to the right side of the abdomen and completed using the blue load of the EndoGIA stapler. transected 10 cm distal to the ligament of Treitz using the Two Jackson–Pratt drains (Allegiance Healthcare Corpo- white load of the EndoGIA stapler. The pancreas paren- ration) are placed near the biliary and pancreatic anasto- chyma is then divided across the neck using the LigaSure moses at the end of the procedure. starting inferiorly and moving toward the superior border anterior to the portal vein and mesenteric vessels, making Outcomes sure to immediately identify the pancreatic duct after transection. The uncinate process dissection is performed The preoperative variables we examined included age, sex, by dividing the SMV and jejunal branches along the BMI, Charlson Index and ASA grade. Operative data adventitial layers of the superior mesenteric vessels to included duration of surgery, intraoperative blood loss and ensure adequate clearance of the uncinate margin. blood transfusions. We also examined oncologic outcomes, The reconstruction is begun by creating a 5–6 cm verti- such as tumour size and histopathology, margin status and cal upper midline minilaparotomy incision through which number of lymph nodes harvested. The R1 resection rate the en bloc resected specimen is retrieved in an endobag. reflects the number of patients who had a positive margin The transected end of the proximal jejunum is brought up out of the total number of patients with a malignant pathol- to the right upper quadrant through a defect in the trans- ogy. Seven-day analgesic use consisted of the total amount verse mesocolon. A 2-layer duct-to-mucosa pancreatico of narcotics administered over the first 7 postoperative jejunostomy is constructed in Blumgart fashion using 5–0 days. Analgesic requirement data were collected from the polydioxanone sutures (PDS) and through-and-through medical administration record, which documents daily scheduled medications and those administered when necessary (PRN) for each patient. Each medication that is actu- TV ally taken by the patient is subsequently signed off by the nursing staff. We calculated daily epidural and patient- controlled analgesia rates from specific documentation sheets. All routes of opioid administration (i.e., epidural, oral, intravenous, intramuscular, transdermal) were tabulated and subsequently converted into intravenous (IV) morphine equivalents. Nonopioid analgesics, such as acetamino- phen and ibuprofen, were not included in the analysis. Postoperatively, we analyzed length of hospital stay, and morbidity and mortality were recorded up to 90 days after surgery. We classified complications according to the Clavien system, which grades severity according to the invasiveness of the required treatment.17 For patients with S FA * MI 6 cm multiple complications, only the most severe one was regis- tered. Pancreatic fistula was defined, according to Interna- 5 mm 5 mm C tional Study Group of Pancreatic Fistula (ISGPF) criteria, 5 mm 12 mm 5 mm as any measurable drain output on or after postoperative 10 mm 10 mm day 3, with an amylase content greater than 3 times the normal serum level.18 Cases were divided into 4 categories: no fistula; biochemical fistula without clinical sequelae (grade A), fistula requiring any therapeutic intervention (grade B) and fistula with severe clinical sequelae (grade C).

SA Statistical analysis

Continuous variables were expressed as medians with Fig. 1. Hybrid laparoscopic pancreaticoduodenectomy trocar ranges and compared using the Mann–Whitney U test. placement and operative setup. Black dots indicate standard tro- Categorical variables were compared using the χ2 test or cars, and grey dots represent optional trocars. The asterisk rep- Fisher exact test. We considered results to be significant resents the liver retractor port. C = camera port; FA = first assistant; MI = mini-laparotomy incision; S = surgeon; SA = second at p < 0.05, 2-tailed. We performed all statistical analyses assistant; TV = television monitor. using SPSS version 17.0.

Can J Surg, Vol. 57, No. 4, August 2014 265

bridging-vanou.indd 265 14-07-16 10:51 AM RECHERCHE

Results pared with 9 (45%) patients in the OPD group (Table 3). One patient in the HLAPD group died due to refractory Between September 2009 and December 2013, we per- sepsis following a leak at a gastric staple site, which required formed HLAPD and OPD on 13 and 20 patients, respec- surgical repair and drainage. In the OPD group, 4 deaths tively. Of the 22 cases begun laparoscopically, 9 were con- occurred within 90 days. One patient had acute hepatic and verted to open surgery before completion of the resections renal failure after 2 subsequent surgeries for portal vein (5 patients had extensive abdominal adhesions, 4 had thrombosis; 1 had an acute myocardial infarction; 1 had tumours showing vascular abutment or involvement); they hemorrhagic shock due to bleeding from the portal vein, were included in the OPD group. There were no signifi- which required surgical intervention; and 1 succumbed to cant differences in age, sex, BMI, ASA grade or Charlson abdominal sepsis following operative repair of hepaticojeju- Index between the groups (Table 1). nostomy and gastrojejunostomy leaks. The HLAPD group had a significantly lower estimated intraoperative blood loss (450 mL v. 1000 mL, p = 0.023) Discussion and a shorter length of hospital stay (8 v. 12 d, p = 0.025) than the OPD group. There were no significant differ- The advent of minimally invasive surgery has resulted in ences in duration of surgery or intraoperative blood trans- increased use of laparoscopic techniques to pancreatic fusion rates between the groups. There were no intraoper- surgery. The benefits of a minimally invasive approach ative deaths. Twelve (92%) patients in the HLAPD group include reduced incisional pain, decreased postoperative used an epidural for postoperative pain control compared complications, shortened hospital stay and improved cos- with 19 (95%) patients in the OPD group. Mean 7-day mesis. Although some surgeons have advocated a direct analgesic requirements were lower in patients who under- transition from an open to a purely laparoscopic approach, went HLAPD than those who underwent OPD (174 mg v. such a shift requires extensive prior laparoscopic experi- 288 mg), but this trend did not achieve significance p( = ence and has been successfully accomplished in only a few 0.08; Fig. 2). Ninety-day mortality was similar between the centres. Concerns regarding the complexity of laparo- HLAPD and OPD groups (8% v. 20%). scopic reconstruction and the adequacy of oncologic Pathology findings are summarized in Table 2. Malig- resection have hindered the adoption of TLPD.19 This nant lesions were found in 10 (77%) patients in the HLAPD report describes the value of HLAPD as a pragmatic group compared with 15 (75%) patients in the OPD group. stepping stone in the transition from open to purely min- Median tumour size, R1 resection rate and lymph node har- imally invasive pancreaticoduodenectomy at our institu- vest did not differ significantly between the groups. tion. The hybrid method combines the safety and famil- Within 90 days postoperative, major complications iarity of conventional open surgery with the benefits of a (Clavien grade III/IV) occurred in 2 (15%) patients in the minimally invasive approach. Given its favourable learn- HLAPD group compared with 8 (40%) patients in the ing curve, it may be more realistically and widely adopted OPD group. Six (46%) patients in the HLAPD group by hepatobiliary surgeons, even those without extensive experienced minor complications (Clavien grade I/II) com- laparoscopic experience. The adoption of a multistep

Table 1. Demographic and outcome data

Group; median (range)*

Characteristic HLAPD OPD p value No. of patients 13 20 Age, yr 69 (49–88) 67 (33–78) 0.45 Sex, male:female, % 85%:15% 65%:35% 0.26 BMI 24.2 (20.6–32.0) 25.0 (16.4–33.3) 0.87 Charlson Comorbidity Index 1 (0–4) 2 (0–6) 0.32 ASA score 2 (2–3) 3 (2–3) 0.36 Operative time, min 594 (407–779) 553 (303–892) 0.06 Estimated blood loss, mL 450 (100–4000) 1000 (300–6500) 0.023 Intraoperative blood transfusion, no. 5 (38%) 10 (50%) 0.72 (%) Total 7-day analgesic use, mg IV, mean 174 ± 117 288 ± 226 0.08 ± SD Length of stay, d 8 (6–14) 12 (6–26) 0.025 90-day mortality, no. (%) 1 (8%) 4 (20%) 0.63

ASA = American Society of Anesthesiologists; BMI = body mass index; HLAPD = hybrid laparoscopy-assisted pancreaticoduodenectomy; IV = Intravenous; OPD = open pancreaticoduodenectomy; SD = standard deviation. *Unless otherwise indicated.

266 J can chir, Vol. 57, No 4, août 2014

bridging-vanou.indd 266 14-07-16 10:51 AM RESEARCH

approach using a spectrum of minimal access procedures experience, duration of surgery did not differ significantly may allow more institutions to successfully implement between the HLAPD and OPD groups (594 v. 553 min., minimally invasive pancreatic surgery programs. p = 0.6. Figure 3 depicts the duration of surgery of The main concerns regarding HLAPD are whether HLAPD and OPD in chronological order. Initially, we smaller incisions are achieved at the expense of the qual observed significantly longer surgery with HLAPD, as ity of oncologic resection and whether any tangible expected during the initial learning phase.24,25 An analysis patient benefit is achieved.20 For pancreaticoduodenec conducted after 2 years of institutional experience, tomy, positive margin rates of 20%–40% have been including 7 HLAPD and 12 OPD procedures, revealed reported in the literature.21,22 Recently, a systematic review significantly longer surgeries in the HLAPD group than of 707 patients undergoing laparoscopic pancreaticoduo in the OPD group (703 v. 572 min.; p = 0.035). How denectomy reported an R1 resection rate of 42.5%.23 Our ever, the duration of HLAPD decreased from a median oncological outcomes with HLAPD compare favourably 703 minutes in the first 7 patients to 582 minutes in the to these standards (R1 resection rate 30%) and confirm last 6 patients (p = 0.003), whereas the duration of OPD the oncologic soundness of the hybrid method. In our remained relatively stable. Our results project the con study, lymph node retrieval and R1 resection rates did tinued convergence of the 2 trendlines with increasing not differ between the OPD and HLAPD groups, further operative experience. Importantly, the learning curve corroborating the adequacy of laparoscopic resection. appears to affect the duration of the procedure, but is not Certain groups who perform minimally invasive pancre associated with increased morbidity or compromise of aticoduodenectomy only for benign or low malignant oncologic outcomes. Tseng and colleagues26 reported potential disease have reported much lower R1 resection that surgeons typically achieved significantly decreased rates. However, these rates are not comparable to those estimated blood loss, duration of surgery, length of stay found in our study, in which 77% of HLAPD procedures and R1 resection rates after performing approximately were performed for malignant indications.13,15 Although 60 OPD procedures. In light of the important learning certain patients with complex tumours were inherently curve, preference should be given to a hybrid approach selected to the open group (including conversions), we before transitioning to total laparoscopic pancreatico nonetheless achieved acceptable oncologic outcomes in duodenectomy to acquire sufficient experience and HLAPD patients with malignant disease, as compared ensure patient safety. with values reported in the literature. Our results dem Despite technical advancements and increased surgeon onstrate that oncologic principles are not compromised experience, pancreaticoduodenectomy remains associated by the use of the hybrid approach, provided careful with high morbidity. We stratified adverse events by patient selection. severity of the clinical treatment required. Our 90-day Long learning curves and increased duration of sur Clavien III/IV complication rates for the HLAPD and gery are often invoked as drawbacks of minimally inva OPD groups were 15% and 40%, respectively, which sive pancreaticoduodenectomy. An advantage of our compare acceptably to the 40% morbidity reported in study is that all operations were performed by the same surgeons, allowing for a more accurate assessment of Table 2. Pathology and oncologic outcomes

progression along the learning curve. In our cumulative Group; no. (%)*

Factor HLAPD, n = 13 OPD, n = 20 p value Tumour size, cm 3.5 (1.8–4.2) 3.5 (1.5–6.5) 0.71 70 HLAPD R1 resection margin 3/10 (30) 7/15 (47) 0.68 OPD Lymph node harvest 22 (14–56) 20 (7–45) 0.09 60 Positive lymph nodes 9 (69) 9 (53) 0.47 50 Pathology Malignant 10 (77) 15 (75) — 40 Pancreatic adenocarcinoma 9 11 — 30 Ampullary adenocarcinoma 1 3 — Neuroendocrine tumour 0 1 — 20 Benign 3 (23) 5 (25) —

10 Intraductal papillary 1 3 — mucinous neoplasm

Median morphine equivalents (mg IV) 0 Duodenal polyp 1 1 — 1234567 Autoimmune pancreatitis 1 0 — Postoperative day Perforated gastric ulcer 0 1 — HLAPD = hybrid laparoscopy-assisted pancreaticoduodenectomy; OPD = open Fig. 2. Seven-day analgesic use. HLAPD = hybrid laparoscopy-assisted pancreaticoduodenectomy *Unless otherwise indicated. pancreaticoduodenectomy; OPD = open pancreaticoduodenectomy.

Can J Surg, Vol. 57, No. 4, August 2014 267

bridging-vanou.indd 267 14-07-16 10:51 AM RECHERCHE

previously published studies.8,27 Pancreatic fistula remains cohorts. Importantly, a very large proportion of patients the most important morbidity after pancreaticoduodenec in both study groups had malignant pathology, which may tomy. In our study, pancreatic fistula rates in the HLAPD explain the higher mortality in our study than other and OPD groups were similar: 31% and 30%, respect studies focusing on benign disease. In addition, our sam ively. Although the pancreatic fistula rate was not reduced ple size was small, and any calculated rates should be taken with the hybrid approach, our data suggest that, even in the context of these limited patient numbers. within the initial learning curve, complication rates with Patients who underwent HLAPD had a significantly HLAPD are acceptable and consistent with those shorter length of hospital stay than those who underwent reported in large open and TLPD series.6,28 Ninety-day OPD. Larger series with longer patient follow-up will be mortality was comparable between the HLAPD and OPD required to assess for any tangible benefits, such as quicker return to baseline function. Total analgesic use Table 3. Ninety-day complications during the first 7 postoperative days was consistently Group; no. lower and tapered off faster in the HLAPD group than in the OPD group, but this trend did not achieve statis OPD, Complication HLAPD, n = 13 n = 20 p value tical significance (Fig. 2). The decreased analgesic Clavien I/II* 6 9 > 0.99 requirements following a minilaparotomy versus a stan Wound infection 1 2 dard subcostal incision likely reflect the correlation Hypotension 1 2 between postoperative pain and incision length. Further Intra-abdominal abscess 1 0 more, while TLPD constitutes the least invasive proced Delayed gastric emptying 1 0 ure, it classically requires a 5 cm Pfannenstiel incision for Thrombocytopenia 1 0 specimen extraction. The difference in morbidity from a Urinary retention 1 0 Pfannenstiel versus a minilaparotomy incision may be of Anemia 0 3 limited clinical importance, thus attenuating some bene Pneumonia 0 2 fits of directly transitioning to a purely minimally inva Clavien III/IV† 2 8 0.25 sive approach. Intra-abdominal abscess 1 3 Our study’s small sample size does not allow for defini Anastomotic breakdown 1 2 Portal vein thrombosis 0 1 tive conclusions to be drawn regarding the comparative Postoperative hemorrhage 0 1 effectiveness of either technique. However, our objective Acute myocardial infarction 0 1 was not to define the better procedure, but rather to Pancreatic fistula 4 6 > 0.99 assess whether the hybrid procedure is feasible and Grade A 3 3 effective without incurring additional risk to the patient. Grade B 1 3 Our study is limited methodologically by its nonran HLAPD = hybrid laparoscopy-assisted pancreaticoduodenectomy; OPD = open domized and retrospective design. Selection bias is inher pancreaticoduodenectomy. *Not necessitating radiological, endoscopic or operative intervention and not causing ent, given that patients with major vascular involvement, organ failure. which poses additional technical challenges, were †Necessitating radiological, endoscopic or operative intervention and/or causing organ failure. excluded from the HLAPD group. We had an important conversion rate in our study, as 9 of 22 (41%) cases begun laparoscopically were converted to laparotomy. It is 1000 important to highlight that these conversions largely HLAPD occurred early during the procedure: 4 cases were con 900 OPD verted before any resection, 4 after gastrectomy only and 800 1 after choledochectomy only. As such, the surgery per formed in these converted cases is more comparable to an 700 open than to a hybrid procedure, and the associated out 600 comes are more representative when included in the OPD group. A subanalysis of strictly open versus converted 500

Operative time, min patients was undertaken to compare patient outcomes 400 (Table 4). Oncologic outcomes, such as R1 resection rate and lymph node harvest, were similar between the open 300 1 234567891011121314151617181920 and converted groups. The duration of surgery in the con Operations (in chronological order) verted group was also longer. Although this difference did not achieve statistic al significance in our study, the dura tion of surgery is undoubtedly affected by the process of Fig. 3. Operative times in chronological order. HLAPD = hybrid laparoscopy-assisted pancreaticoduodenectomy; OPD = open converting from laparoscopic to open surgery. Patients in pancreaticoduodenectomy. the converted group had significantly higher estimated

268 J can chir, Vol. 57, No 4, août 2014

bridging-vanou.indd 268 14-07-16 10:51 AM RESEARCH

Table 4. Comparison of cases begun open versus converted cases

Factor Open Converted p value No. of patients 11 9 Age, yr 62 (33–76) 72 (47–78) 0.06 Sex, male/female 55%/45% 78%/22% 0.37 BMI 26.2 (18.5–33.3) 23.5 (16.4–31.6) 0.53 ASA score 2 (2–3) 3 (2–3) 0.36 Operative time, min. 518 (303–665) 613 (470–892) 0.07 Estimated blood loss, mL 700 (300–6500) 1650 (800–5050) 0.020 Tumour size, cm 3.1 (1.8–4.5) 3.7 (2.8–6.5) 0.27 R0 resection margin, no. (%) 3/8 (38%) 4/7 (57%) 0.62 Lymph node harvest 20 (7–30) 19 (15–45) 0.56 Length of stay, d 11 (6–23) 15.5 (7–26) 0.31 Clavien I/II, no. (%) 5 (45%) 4 (44%) > 0.99 Clavien III/IV, no. (%) 4 (36%) 4 (44%) > 0.99 90-day mortality, no. (%) 1 (10%) 3 (33%) 0.30

ASA = American Society of Anesthesiologists; BMI = body mass index.

Funding: This work was supported by a medical student research bur- blood loss (1650 mL v. 700 mL; p = 0.020) than those who sary from the McGill University Faculty of Medicine. began with open surgery. Importantly, however, since no Competing interests: None declared. case was converted due to excessive bleeding, this difference in blood loss likely reflects inherently difficult Contributors: Y. Wang, S. Bergman and T. Vanounou designed the study. Y. Wang acquired the data, which all authors analyzed. All authors pathology and surgical complexity rather than complica- wrote and reviewed the article and approved it for publication. tions of laparoscopic resection or the act of conversion. As such, these outcomes may have remained largely References unchanged even if they had initially been begun by lapa- 1. Huscher CG, Mingoli A, Sgarzini G, et al. Laparoscopic versus rotomy. Because similar conversion rates have been open subtotal gastrectomy for distal gastric cancer: five-year results reported in the literature, further studies evaluating the of a randomized prospective trial. Ann Surg 2005;241:232-7. validity of more rigorous selection criteria are warranted 2. Lacy AM, García-Valdecasas JC, Delgado S, et al. Laparoscopy- to reduce conversion rates going forward.19 assisted colectomy versus open colectomy for treatment of non- metastatic colon cancer: a randomised trial. Lancet 2002;359:2224-9. Robotic-assisted pancreaticoduodenectomy (RAPD) 3. Veldkamp R, Kuhry E, Hop WC, et al. Laparoscopic surgery versus has gained increasing acceptance because it offers open surgery for colon cancer: short-term outcomes of a randomised 3-dimensional visualization, superior ergonomics and trial. Lancet Oncol 2005;6:477-84. enhanced suturing capabilities.29 Since 2011, our centre 4. Gagner M, Pomp A. Laparoscopic pylorus-preserving pancreatico- has progressed from an open to a robotic reconstruction duodenectomy. Surg Endosc 1994;8:408-10. 5. Palanivelu C, Jani K, Senthilnathan P, et al. Laparoscopic pancrea with favourable results, and the laparoscopic experience ticoduodenectomy: technique and outcomes. J Am Coll Surg 2007; initially acquired with HLAPD has been valuable in this 205:222-30. transition. All patients eligible for a minimally invasive 6. Kendrick ML, Cusati D. Total laparoscopic lancreaticoduodenec- procedure now undergo RAPD, provided robot availabil- tomy; feasibility and outcome in an early experience. Arch Surg 2010; ity. We reserve HLAPD for those patients for whom the 145:19-23. 7. Dulucq JL, Wintringer P, Mahajna A. Laparoscopic pancreatico robotic platform is unavailable for logistic reasons. duodenectomy for benign and malignant diseases. Surg Endosc 2006; 20:1045-50. Conclusion 8. Gumbs AA, Rodriguez Rivera AM, Milone L, et al. Laparoscopic pancreatoduodenectomy: a review of 285 published cases. Ann Surg Hybrid laparoscopy-assisted pancreaticoduodenectomy is Oncol 2011;18:1335-41. 9. Uyama I, Ogiwara H, Iida S, et al. Laparoscopic minilaparotomy a safe, feasible and effective procedure with comparable pancreaticoduodenectomy with lymphadenectomy using an abdom- outcomes to OPD in select patients. The favourable inal wall-lift method. Surg Laparosc Endosc 1996;6:405-10. learning curve makes HLAPD a pragmatic procedure that 10. Suzuki O, Kondo S, Hirano S, et al. Laparoscopic pancreaticoduo may allow a greater number of patients to benefit from a denectomy combined with minilaparotomy. Surg Today 2012;42:509-13. minimally invasive approach. The transition from open to 11. Staudacher C, Orsenigo E, Baccari P, et al. Laparoscopic assisted duodenopancreatectomy. Surg Endosc 2005;19:352-6. purely minimally invasive pancreaticoduodenectomy may 12. Pugliese R, Scandroglio I, Sansonna F, et al. Laparoscopic be more effectively achieved as a multistep process using pancreaticoduodenectomy: a retrospective review of 19 cases. Surg the hybrid approach as a valuable stepping stone. Laparosc Endosc Percutan Tech 2008;18:13-8.

Can J Surg, Vol. 57, No. 4, August 2014 269

bridging-vanou.indd 269 14-07-16 10:51 AM RECHERCHE

13. Cho A, Yamamoto H, Nagata M, et al. Comparison of laparoscopy- lymph node ratio following pancreaticoduodenectomy for pancreatic assisted and open pylorus-preserving pancreaticoduodenectomy for cancer. Surgery 2007;141:610-8. periampullary disease. Am J Surg 2009;198:445-9. 22. Verbeke CS. Resection margins and R1 rates in pancreatic can- 14. Kuroki T, Adachi T, Okamoto T, et al. A non-randomized compara- cer — are we there yet? Histopathology 2008;52:787-96. tive study of laparoscopy-assisted pancreaticoduodenectomy and open 23. Nakamura M, Nakashima H. Laparoscopic distal pancreatectomy pancreaticoduodenectomy. Hepatogastroenterology 2012;59:570-3. and pancreatoduodenectomy: Is it worthwhile? A meta-analysis of 15. Lee JS, Han JH, Na GH, et al. Laparoscopic pancreaticoduodenec- laparoscopic pancreatectomy. J Hepatobiliary Pancreatic Sci 2013; tomy assisted by mini-laparotomy. Surg Laparosc Endosc Percutan 20:421-8. Tech 2013;23:e98-102. 24. Fisher WE, Hodges SE, Wu MF, et al. Assessment of the learning 16. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic curve for pancreaticoduodenectomy. Am J Surg 2012;203:684-90. fistula: an international study group (ISGPF) definition. Surgery 25. Hardacre JM. Is there a learning curve for pancreaticoduodenectomy 2005;138:8-13. after fellowship training? HPB Surg 2010;2010:230287. 17. Grobmyer SR, Kooby D, Blumgart LH, et al. Novel pancreatico 26. Tseng JF, Pisters PW, Lee JE, et al. The learning curve in pancreatic jejunostomy with a low rate of anastomotic failure-related complica- surgery. Surgery 2007;141:694-701. tions. J Am Coll Surg 2010;210:54-9. 27. Zureikat AH, Breaux JA, Steel JL, et al. Can laparoscopic pancrea 18. Dindo D, Demartines N, Clavien PA. Classification of surgical ticoduodenectomy be safely implemented? J Gastrointest Surg 2011; complications: a new proposal with evaluation in a cohort of 6,336 15:1151-7. patients and result of a survey. Ann Surg 2004;240:205-13. 28. Cullen JJ, Sarr MG, Ilstrup DM. Pancreatic anastomotic leak after 19. Gagner M, Pomp A. Laparoscopic pancreatic resection: Is it worth- pancreaticoduodenectomy: incidence, significance, and management. while? J Gastrointest Surg 1997;1:20-6. Am J Surg 1994;168:295-8. 20. Park A, Schwartz R, Tandan V, et al. Laparoscopic pancreatic sur- 29. Giulianotti PC, Coratti A, Angelini M, et al. Robotics in general surgery. Am J Surg 1999;177:158-63. gery: personal experience in a large community hospital. Arch Surg 21. Pawlik TM, Gleisner AL, Cameron JL, et al. Prognostic relevance of 2003;138:777-84.

Correction

Trauma Association of Canada abstracts

DOI: 10.1503/cjs.005714

The Trauma Association of Canada (TAC) 2014 Annual Scientific Meeting abstract supplement published in May 2014 contained an error in the abstract on page S48, entitled “Chest drain insertion in the emergency department: compliance with guidelines in a university hospital emergency department.” The author list should read K.T.D. Yeung, J. Ollerton. We apologize for this error.

270 J can chir, Vol. 57, No 4, août 2014

bridging-vanou.indd 270 14-07-16 10:51 AM RESEARCH • RECHERCHE

The accuracy of the Alvarado score in predicting acute appendicitis in the black South African population needs to be validated

Victor Y. Kong, MSc* Background: The Alvarado score is the most widely used clinical prediction tool to Stefan Van Der Linde, MSc* facilitate decision-making in patients with acute appendicitis, but it has not been validated in the black South African population, which has much wider differential diag- Colleen Aldous, PhD* nosis than developed world populations. We investigated the applicability of this Jonathan J. Handley, FCA(SA)† score to our local population and sought to introduce a checklist for rural doctors to facilitate early referral. Damian L. Clarke, M Med Sci, MBA, M Phil* Methods: We analyzed patients with proven appendicitis for the period January 2008 to December 2012. Alvarado scores were retrospectively assigned based on patients’ admission charts. We generated a clinical probability score (1–4 = low, 5–6 = From the *Department of Surgery, intermediate, 7–10 = high). †Department of Anaesthetics and Critical Care, Nelson R. Mandela School of Medi- Results: We studied 1000 patients (54% male, median age 21 yr). Forty percent had cine, University of KwaZulu-Natal, Dur- inflamed, nonperforated appendices and 60% had perforated appendices. Alvarado ban, South Africa scores were 1–4 in 20.9%, 5–6 in 35.7% and 7–10 in 43.4%, indicating low, inter mediate and high clincial probability, respectively. In our subgroup analysis of Accepted for publication 510 patients without generalized peritonitis, Alvarado scores were 1–4 in 5.5%, 5–6 in Nov. 7, 2013 18.1% and 7–10 in 76.4%, indicating low, intermediate and high clinical probability, respectively. Correspondence to: D.L. Clarke Conclusion: The widespread use of the Alvarado score has its merits, but its applicability Department of Surgery in the black South African population is unclear, with a significant proportion of patients Nelson R. Mandela School of Medicine with the disease being potentially missed. Further prospective validation of the Alvarado University of KwaZulu-Natal score and possible modification is needed to increase its relevance in our setting. Private Bag 7 Congella, 4013 Durban, South Africa Contexte : Le score d’Alvarado est l’outil de prédiction clinique le plus couramment [email protected] utilisé pour faciliter la prise de décision chez les patients présentant une appendicite aiguë, mais il n’a pas été validé dans la population noire sud-africaine chez qui le diagnostic différentiel est beaucoup plus vaste que dans les populations des pays industria DOI: 10.1503/cjs.023013 lisés. Nous avons exploré l’applicabilité de ce score à notre population locale et tenté de présenter une liste de vérification aux médecins ruraux pour accélérer les demandes de consultation. Méthodes : Nous avons analysé les dossiers de patients atteints d’une appendicite avérée pendant la période allant de janvier 2008 à décembre 2012. Les scores d’Alvarado ont été assignés rétrospectivement selon les dossiers d’admission des patients. Nous avons généré un score de probabilité clinique (1–4 = faible, 5–6 = inter- médiaire, 7–10 = élevé). Résultats : Nous avons ainsi étudié 1000 patients (54 % de sexe masculin, âge médian 21 ans). Quarante pour cent présentaient des appendices enflammés non per- forés et 60 % des appendices perforés. Les scores d’Alvarado se situaient à 1–4 chez 20,9 %, à 5–6 chez 35,7 % et à 7–10 chez 43,4 %, correspondant à une probabilité clinique faible, intermédiaire et élevée, respectivement. Dans notre analyse de sous- groupes sur 510 patients indemnes de péritonite généralisée, les scores d’Alvarado se situaient à 1–4 chez 5,5 %, à 5–6 chez 18,1 % et à 7–10 chez 76,4 %, correspondant à une probabilité clinique faible, intermédiaire et élevée, respectivement. Conclusion : L’utilisation répandue du score d’Alvarado a ses mérites, mais son applicabilité dans la population noire d’Afrique du Sud est indéterminée, la maladie risquant de passer inaperçue chez une proportion significative de patients. Il faudra procéder à une validation prospective plus approfondie du score d’Alvarado et le modifier peut-être si l’on veut en accroître la pertinence dans notre contexte.

t is increasingly accepted that the omission of surgical appendicitis, the score is less reliable; the same review care from the Millennium Development Goals was a stated that “the pooled diagnostic accuracy in terms of I serious oversight, and over the last decade there has been ‘ruling in’ appendicitis at a cut-point of 7 points is not suffi- an increased awareness of the important role that surgery ciently specific in any patient group to proceed directly to plays in global health.1,2 Disparities in access to surgical care surgery.” The score is well calibrated in men, but tends to result in major discrepancies in the outcomes of patients overpredict the presence of acute appendicitis in women.10 with common surgical conditions, and our group has In children, the score has also been shown to be inaccurate.7 studied the outcomes of acute appendicitis in our setting.3–5 The applicability of the Alvarado score in South Africa is We have demonstrated that acute appendicitis in rural unclear, and there is evidence to suggest that the clinical South Africa has a very different disease profile to that seen presentation of acute appendicitis is different to that in the in the developed world.3 It is associated with prolonged developed world.3,11 Furthermore, the differential diagnosis delays to definitive surgical care and significant morbidity of abdominal pain in South Africa is much broader than in due to intra-abdominal sepsis.4,5 We proceeded to investi- the developed world. There is a high incidence of child- gate the reasons behind these lengthy delays in presentation hood diarrheal illness; HIV; and tropical diseases, such as and identified rural origin as an independent risk factor for amoebiasis, abdominal tuberculosis and typhoid, which may poor outcome from this disease.5 It would appear that rural all present with acute abdominal symptoms.12 Prior to patients in South Africa experience delays before presenting designing a possible tick-box-style sheet for abdominal pain to district hospitals, and once they have presented to these to be used in our rural hospitals, we set out to establish the district facilities they experience further delays owing to validity of the Alvarado score at our institution. failure of staff to diagnose the condition and refer them through to regional centres with surgical capacity.5 There is Methods a causal relationship between delay to definitive surgery and poor outcome in the management of acute appendicitis, We obtained ethics approval to audit acute appendicitis and strategies to reduce these delays are urgently required.6 from the Umgungundlovu Health Ethics review board and One of the suggested strategies aimed at facilitating the from the Biomedical Research Committee of the Univer- diagnosis of acute appendicitis is the introduction of tick- sity of KwaZulu-Natal. This study was conducted at Eden- box-style clerking sheets to facilitate clinical decision- dale Hospital, a large regional hospital in Pietermaritzburg, making among junior doctors working in relatively the capital city of KwaZulu-Natal, South Africa. Edendale unsupervised, resource-constrained environments. A Hospital drains a predominantly black African population number of authors have advocated the use of clinical pre- from the urban areas around Pietermaritzburg and from diction rules (CPRs) to assist with clinical decision- the deep rural areas of Sisonke Health District (SHD), a making in cases of acute appendicitis.7,8 These CPRs rural area in southwestern KwaZulu-Natal with a popula- attempt to quantify the possibility of a disease being pres- tion of half a million people and 4 district hospitals. This ent based on key symptoms, signs and the results of spe- study was conducted from January 2008 to December 2012. cial investigations and to generate a score that predicts For the period from January 2008 to December 2009, we the probability of the disease being present.8 We sought retrospectively reviewed the records of all patients with to generate a tick-box-style sheet with a CPR that would acute appendicitis and entered the data into an Excel data- allow junior staff working in relatively unsupervised dis- base. From January 2010 onwards, data from all patients trict hospitals to triage patients with abdominal pain into with acute appendicitis were entered prospectively into the those who require urgent referral and those who can be same database. Individual Alvarado scores were generated discharged home. for all patients using data from their charts, and a score was The Alvarado score is the most widely used CPR for acute appendicitis and sums up 3 symptoms and 3 signs as well as the results of standard blood tests to give an overall Box 1. The Alvarado score score out of 10 (Box 1).9 On the basis of this score, 3 groups Feature Score of patients are identified.9 Patients with a score of 1–4 can Migration of pain 1 be discharged home, those with a score 5–6 should be Anorexia 1 admitted and those with a score of 7–10 should be con Nausea 1 Right lower quadrant 2 sidered candidates for surgery. A recent review of the pub- tenderness lished data on the Alvarado score reported that it is most Rebound pain 1 useful in predicting the absence of appendicitis, and an Elevated temperature 1 Alvarado score below 5 has a sensitivity of 94%–99% for > 37.5° C appendicitis not being present.10 The authors concluded Leucocytosis 2 that a score of 5 or less rules out appendicitis.10 When it Left shift of white cell count 1 comes to positively establishing the presence of acute Total 10

E122 J can chir, Vol. 57, No 4, août 2014 RESEARCH assigned to each patient. On the basis of each individual Table 1 compares the outcomes of acute appendicitis score a clinical probability score was generated, as previ- at our institution with those in institutions in the de- ously described.9 veloped world.11

Statistical analysis Alvarado score

We entered all data into an Excel spreadsheet for process- For the entire cohort of 1000 patients, Alvarado scores ing. All statistical analysis was performed using SPSS version were 1–4 in 20.9%, 5–6 in 35.7% and 7–10 in 43.4%, 19 (IBM Corp). indicating low, intermediate and high clincial probability, respectively. The frequency of occurrence of each Results item on the Alvarado score and relative clinical probabilities are shown in Tables 2 and 3. Figure 1 provides a Our study sample comprised 1000 patients (54% male, summary of the Alvarado scores for all patients with 46% female, median age 21 [range 12–26] yr) with acute acute appendicitis. appendicitis confirmed both intraoperatively and with histology during the 5-year period from January 2008 to Subgroup analysis December 2012. Medical care was sought on average 4.2 days after the onset of symptoms. Half of the patients For the purpose of subgroup analysis, a total of presented from rural areas and the other half from urban 510 patients (65.5% male, 34.5% female, median age 19 areas. A total of 490 patients were considered to have [range 11–25] yr) who did not have generalized peri generalized peritonitis at presentation, and the remaining tonitis on presentation were analyzed separately. A total 510 patients presented with localized peritonitis or non- of 393 of 510 (77.1%) patients had inflamed, nonper specific abdominal pain. Intraoperative findings were as forated appendices and 117 (22.9%) had perforated follows: 405 (40.5%) had inflamed, nonperforated appen- appendices associated with localized intra-abdominal dices and 595 (59.5%) had perforated appendices. Of the sepsis. cohort with perforated appendicitis 177 (29.7%) had The Alvarado scores of all 510 patients were 1–4 in perforation-associated localized intra-abdominal sepsis, 5.5%, 5–6 in 18.1% and 7–10 in 76.4%, indicating low, and 418 (70.2%) had perforation-associated generalized intermediate and high clinical probability, respectively. intra-abdominal sepsis. In all, 234 (23.4%) patients The frequency of occurrence of each item on the required temporary abdominal closure, and 406 (40.6%) Alvarado score and relative clinical probabilities are patients required revision laparotomy for residual sepsis. Ninety-five (9.5%) patients required postoperative intensive care admission owing to perforation and generalized sepsis. The mean length of stay in intensive care was Table 2. Alvarado score for all patients with acute appendicitis in, 6 days. The remaining patients were admitted to the gen- n = 1000 eral surgical wards. Overall complications were as follows: 82 (8.2%) patients had hospital-acquired pneumo- Alvarado score No. (%) nia, 57 (5.7%) had acute kidney injury, 142 (14.2%) had 1 20 (2.0) wound sepsis, and 20 (2.0%) experienced other complica- 2 25 (2.5) 3 44 (4.4) tions. Overall mortality was 1.3%. 4 120 (12.0) 5 155 (15.5) Table 1. Comparative data between the US Department of 6 202 (20.2) Defense and our institution 7 110 (11.0) US Department 8 120 (12.0) Comparative data of Defense Edendale Hospital 9 135 (13.5) Year 1997 2008–2012 10 69 (6.9) Patients, no. 4950 1000 Centres, no. 147 1 Patients/centre/yr, no. 25 200 Perforation rate, % 24 60 Table 3. Clinical probability according to Alvarado Mortality, % 0.08 1 score, n = 1000 Intensive care unit, % NA 10 Score Clinical probability No. (%) Reoperation rate, % 0.5 23 1–4 Low 209 (20.9) Temporary abdominal closure, % NA 41 5–6 Intermediate 357 (35.7) NA = not available. 7–10 High 434 (43.4)

Can J Surg, Vol. 57, No. 4, August 2014 E123 RECHERCHE shown in Tables 4 and 5. Figure 1 provides a summary and high clinical probability, respectively. The frequency of the Alvarado score with separate subgroup analysis. of occurrence of each item on the Alvarado score and rel- The Alvarado scores of the 393 patients with inflamed, ative clinical probabilities are shown in Tables 6 and 7. nonperforated appendices were 1–4 in 6.9%, 5–6 in The Alvarado scores of the 117 patients with perfor 21.9% and 7–10 in 71.2%, indicating low, intermediate ated appendices (localized sepsis) were 1–4 in 0.9%, 5–6 in 5.1% and 7–10 in 94.0%, indicating low, intermediate and high clinical probability, respectively. The frequency Table 4. Alvarado score for all of occurrence of each item on the Alvarado score and rel- patients without generalized peritonitis on presentation, ative clinical probabilities were shown in Tables 6 and 7. n = 510 iscussion Alvarado score No. (%) D 1 0 (0) 2 0 (0) Acute appendicitis is an important clinical problem in 3 9 (1.8) South Africa, and the incidence appears to be increasing 1,13 4 19 (3.7) among the general population. It is associated with long 5 31 (6.1) delays to definitive surgery, major morbidity and high 6 61 (12.0) cost.3–5 While there is evidence to suggest that patients do 7 87 (17.0) not present early and that a great deal of the morbidity is 8 114 (22.4) related to the presence of barriers to care, there is a con- 9 124 (24.3) cern that even once contact with the health system has 10 65 (12.7) been made, clinical failure to recognize the condition exacerbates the delays.5 There are a number of structural reasons for the high incidence of clinical failure that Table 5. Clinical probability score according revolve around junior staff working in areas of limited to Alvarado score, n = 510 resources with inadequate supervision.14 However, it has Score Clinical probability No. (%) been suggested that the clinical presentation of the disease 1–4 Low 28 (5.5) in South Africa is also different to that in the developed 5–6 Intermediate 92 (18.0) world.3,11 Abdominal tuberculosis; HIV; and other tropical 7–10 High 390 (76.5) diseases, such as typhoid, amoebiasis and pediatric diar- rhea, may all mimic acute appendicitis.12 In our previous study on acute appendicitis, only a small proportion of our Table 6. Alvarado score for subgroups of patients without patients presented with the classic migratory abdominal generalized peritonitis pain.3 The most common symptoms encountered were all

Group; no. (%) nonspecific, and these findings were similar to those previously reported in Durban, South Africa.15 The nonspe- Inflamed Perforation, local sepsis Alvarado score n = 393 n = 117 cific nature of these symptoms has implications for the clinical assessment of black African patients. The present 1 0 (0) 0 (0) 2 0 (0) 0 (0) results seem to support our suspicion that the presentation 3 9 (2.3) 0 (0) of acute appendicitis among the South African population 3,16 4 18 (4.6) 1 (0.9) is different to that in the developed world. 5 29 (7.4) 2 (1.7) 6 57 (14.5) 4 (3.4) Limitations 7 69 (17.6) 18 (15.4) 8 81 (20.6) 33 (28.2) There are a number of limitations to our study. As the 9 85 (21.6) 39 (33.3) Alvarado score was applied retrospectively to patients 10 45 (11.4) 20 (17.1) already known to have the disease, there is a significant

Table 7. Clinical probability score, subgroup

Group; no. (%)

Perforation, local sepsis, Score Clinical probability Inflamed,n = 393 n = 117 1–4 Low 27 (6.9) 1 (0.9) 5–6 Intermediate 86 (21.9) 6 (5.1) 7–10 High 280 (71.2) 110 (94.0)

E124 J can chir, Vol. 57, No 4, août 2014 RESEARCH potential for selection bias, and it is quite possible that the This is likely to be related to a much wider range of average Alvarado score of patients in our study is higher pathologies and atypical clinical presentations. Future than that of patients presenting to our institutions with prospective research must be undertaken to validate the nonspecific abdominal pain who did not receive surgery. Alvarado score, with a possible modification, in order to We are interested in developing a triage tool for rural improve its relevance in our environment. hospitals. The concept would be to create tick-box-style Competing interests: None declared. clerking sheets in district hospitals that would enable Contributors: All authors designed the study. V. Kong acquired the junior doctors to score each patient presenting with data, which V. Kong, S. van der Linde, J. Handley and D. Clarke ana- abdominal pain. Patients meeting a specific score could lyzed. V. Kong and J. Handley wrote the article, which all authors then be triaged for urgent referral to a regional institution reviewed and approved for publication. with surgical capacity. However, before the widespread introduction of the use of the Alvarado score in our set- References ting, we need to prospectively investigate its applicability in our institutions. We have increasingly used tick-box- 1. Clarke DL, Kong VY, Handley J, et al. A concept paper: using the style clerking sheets to improve the quality of care in our outcomes of common surgical conditions as quality metrics to setting. This is taken directly from the aviation industry, benchmark district surgical services in South Africa as part of a sys- which makes frequent use of tick-box-style checklists to temic quality improvement programme. S Afr J Surg 2013;51:84-6. 17 . 2 Spiegel DA, Gosselin RA. Surgical services in low income and middle- improve safety. The assessment of abdominal pain may income countries. Lancet 2007;370:1013-5. be amenable to such an intervention, and a major attrac- . 3 Kong VY, Bulajic B, Allorto NL, et al. Acute appendicitis in a tions of the Alvarado Score is that it can be tabulated into a developing country. World J Surg 2012;36:2068-73. routine clerking sheet.18,19 However, our study has shown 4. Kong V, Aldous C, Handley JJ, et al. The cost effectiveness of the that using the Alvarado score, more than one-quarter of all early management of acute appendicitis underlies the importance of curative surgical services to a primary health care program. Ann R patients with proven acute appendicitis would have been Coll Surg Engl 2013;95:280-4. classified as having a low to intermediate probability of the 5. Kong VY, van der Linde S, Handley J, et al. Quantifying the dis- disease being present and that slightly less than 5% of parity in outcome between urban and rural patients with acute these patients would have been discharged home despite appendicitis in South Africa. S Afr Med J 2013;103:742-5. having the disease. The implications of this finding for 6. Bickell NA, Aufses JAH, Rojas M, et al. How time affects the risk of rupture in appendicitis. J Am Coll Surg 2006;202:401-6. staff in rural district hospitals are unclear. These individ 7. Kulik DM, Uleryk EM, Maguire JL. Does this child have appendi- uals are usually busy generalists with limited access to citis? A systematic review of clinical prediction rules for children advanced imaging who are unable to undertake the opera- with acute abdominal pain. J Clin Epidemiol 2013;66:95-104. tion themselves.14 There appear to be 3 options available to 8. Andersson RE. Meta-analysis of the clinical and laboratory diagno- them: discharge, admit or transfer the patient. Our results sis of appendicitis. Br J Surg 2004;91:28-37. 9. Alvarado A. A practical score for the early diagnosis of acute appen- suggest that approximately 20% of patients who have the dicitis. Ann Emerg Med 1986;15:557-64. disease may have been admitted to a district hospital for 10. Ohle R, O’Reilly F, O’Brien KK, et al. The Alvarado score for pre- ongoing observations. Yet we know from our previous dicting acute appendicitis: a systematic review. BMC Med 2011;9:139. research that there is already a delay in transferring 11. Hale DA, Molloy M, Pearl RH, et al. Appendectomy: a contempo- patients from district to regional hospitals, so this may sim- rary appraisal. Ann Surg 1997;225:252-61. 5 12. Clarke DL, Thomson SR, Bissetty T, et al. A single surgical unit’s ply exacerbate the problem. A further 5% of patients with experience with abdominal tuberculosis in the HIV/AIDS era. the disease would have been sent home. Similarly, we World J Surg 2007;31:1087-96, discussion 1097-8. know that a substantial number of patients are in fact 13. Walker AR, Segal I. Appendicitis: an African perspective. J R Soc incorrectly sent home from a district-level facility despite Med 1995;88:616-9. the presence of the disease.5 The concern with the 14. De Villiers MR. The knowledge and skills gap of medical practi tioners delivering district hospital services in the Western Cape, Alvarado score remains that in our under-resourced hospi- South Africa. S Afr Fam Pract 2006;48:16. tals its use may exacerbate rather than improve the current 15. Chamisa I. A Clinicopathological review of 324 appendices situation. removed for acute appendicitis in Durban, South Africa: a retrospective analysis. Ann R Coll Surg Engl 2009;91:688-92. Conclusion 16. Rogers AD, Hampton MI, Bunting M, et al. Audit of appendicecto- mies at Frere Hospital, Eastern Cape. S Afr J Surg 2008;46:74-7. 17. Reason J. Human error: models and management. BMJ 2000; Acute appendicitis remains a common clinical diagnostic 320:768-70. problem, and in our environment it is associated with sig- 18. Shreef KS, Waly AH, Abd-Elrahman S, et al. Alvarado score as an nificant delays and poor clinical outcomes. The wide- admission criterion in children with pain in right iliac fossa. Afr J spread use of the Alvarado score as a clinical prediction Paediatr Surg 2010;7:163-5. 19. Denizbasi A, Unluer EE. The role of the emergency medicine resi- tool has its merits, but its applicability in the black South dent using the Alvarado score in the diagnosis of acute appendicitis African population is unclear, with a significant propor- compared with the general surgery resident. Eur J Emerg Med tion of patients with the disease being potentially missed. 2003;10:296-301.

Can J Surg, Vol. 57, No. 4, August 2014 E125 RESEARCH • RECHERCHE

Health-related quality of life following decompression compared to decompression and fusion for degenerative lumbar spondylolisthesis: a Canadian multicentre study

Y. Raja Rampersaud, MD*† Background: Decompression alone (D) is a well-accepted treatment for patients Charles Fisher, MD, MHSc‡ with lumbar spinal stenosis (LSS) causing neurogenic claudication; however, D is controversial in patients with LSS who have degenerative spondylolisthesis (DLS). Our § Albert Yee, MD, MSc goal was to compare the outcome of anatomy-preserving D with decompression and Marcel F. Dvorak, MD‡ fusion (DF) for patients with grade I DLS. We compared patients with DLS who had § elective primary 1–2 level spinal D at 1 centre with a cohort who had 1–2 level spinal Joel Finkelstein, MD, MSc DF at 5 other centres. Eugene Wai, MD, MSc¶** Methods: Patients followed for at least 2 years were included. Primary analysis †† Edward Abraham, MD included comparison of change in SF-36 physical component summary (PCS) scores Stephen J. Lewis, MD, MSc*† and the proportion of patients achieving minimal clinically important difference David Alexander, MD‡‡ (MCID) and substantial clinical benefit (SCB). William Oxner, MD‡‡ Results: There was no significant difference in baseline SF-36 scores between the groups. The average change in PCS score was 10.4 versus 11.4 (p = 0.61) for the D and From the *Division of Orthopaedic Sur- DF groups, respectively. Sixty-seven percent of the D group and 71% of the DF group gery, Department of Surgery, University of attained MCID, while 64% of both D and DF groups attained SCB. There was no sig- Toronto, Toronto, Ont., †Spinal Program, nificant difference between D and DF for change in PCS score (p = 0.74) or likelihood Krembil Neuroscience Center, Toronto of reaching MCID (p = 0.81) or SCB (p = 0.85) after adjusting for other variables. Western Hospital University Health Net- Conclusion: In select patients with DLS, the outcome of D is comparable to DF at a work, University of Toronto, Toronto, Ont., minimum of 2 years. ‡Combined Neurosurgical and Orthopae- dic Spine Program (CNOSP), Department of Orthopaedics, University of British Contexte : La décompression seule est un traitement bien accepté pour la sténose Columbia, Vancouver Coastal Health lombaire (SL) causant une claudication neurogène. Son utilisation ne fait cependant Research Institute, Vancouver General Hos- pital, Vancouver, BC, §Sunnybrook Health pas l’unanimité chez les patients atteints de SL qui souffrent d’un spondylolisthésis Sciences Centre, Toronto, Ont., ¶Division dégénératif (SLD). Notre objectif était de comparer l’issue d’une décompression avec of Orthopaedic Surgery, Department of préservation anatomique à celle d’une décompression-arthrodèse (DA) chez des Surgery, University of Ottawa, The Ottawa patients atteints de SLD de grade 1. Nous avons comparé les patients atteints de SLD Hospital, Ottawa, Ont., **Clinical Epidemi- ayant subi une décompression élective primaire de niveau 1–2 dans un centre à une ology Program, Ottawa Hospital Research cohorte ayant subi une DA de niveau 1–2 effectuée dans 5 autres centres. Institute, Ottawa, Ont., ††Atlantic Health Méthodes : Nous avons inclus les patients qui ont été suivis pendant au moins 2 ans. Science Corporation, St. John, NB, and ‡‡Dalhousie University, QEII Health Sci- L’analyse primaire comportait une comparaison des changements aux scores som- ences Centre, Halifax, NS maires pour la composante physique (CP) du questionnaire SF-36 et la proportion de patients ayant obtenu une différence minimale cliniquement importante (DMCI) et en ayant tiré un bienfait clinique substantiel (BCS). Accepted for publication Jan. 27, 2014 Résultats : Il n’y avait pas de différence significative entre les scores SF-36 des 2 groupes au départ. Le changement moyen du score pour la CP a été de 10,4 c. 11,4 Correspondence to: (p = 0,61) dans les groupes soumis à la décompression et à la DA, respectivement. Y.R. Rampersaud Soixante-sept pour cent des patients du groupe soumis à la décompression et 71 % du Division of Orthopaedic Surgery and groupe soumis à la DA ont obtenu une DMCI, tandis que 64 % des 2 groupes ont Neurosurgery obtenu un BCS. On n’a noté aucune différence significative entre les groupes soumis Toronto Western Hospital à la décompression et à la DA pour ce qui est du changement du score pour la CP (p = 399 Bathurst St. East Wing, 1-441 Toronto ON M5T 2S8 0,74) ou de la probabilité d’obtenir une DMCI (p = 0,81) ou un BCS (p = 0,85) après [email protected] ajustement pour tenir compte d’autres variables. Conclusion : Chez certains patients souffrant de SLD, l’issue de la décompression DOI: 10.1503/cjs.032213 est comparable à celle de la DA après une période minimale de 2 ans.

egenerative lumbar spondylolisthesis (DLS) is a Methods common spinal disorder that can lead to substan- D tial back and/or leg pain. It is also a very common We conducted a Canadian multicentre ambispective (ret- reason for spinal surgery in individuals older than rospective review of prospectively collected data) cohort 65 years.1 The estimated incidence of DLS is 12.7% with study. We sought to determine whether the 2-year post- an overall prevalence of 6% that increases between the operative improvement in health-related quality of life fifth and eighth decades of life.2–4 For symptomatic (HRQoL) outcomes for D was equivalent to that of DF patients, the recent Spine Patient Outcomes Research for the management of focal (1–2 level) stenosis and asso- Trial (SPORT) — DLS study has demonstrated that sur- ciated DLS with similar clinical presentation. The study gical management is superior to conservative care at 2 and was approved by each institution’s research ethics board. 4 years post-intervention.5,6 From a surgical perspective, since the controlled study Patient population by Herkowitz and Kurz7 demonstrated a high failure rate following decompression (conventional midline laminec- Inclusion and exclusion criteria were applied to prospective tomy) alone (D), decompression and fusion (DF) has surgical databases collecting HRQoL outcome measures become the surgical treatment of choice for patients with from 6 academic spine centres across Canada. We included DLS. A recent systematic review by Martin and col- patients with DLS who had 1- or 2-level surgery for whom leagues8 concluded that “decompression and spinal fusion baseline and 2-year primary outcome data were available. All may lead to better clinical outcome compared to decom- patients had failed at least 6 months of nonoperative care. pression alone.” The contemporary management of DLS Exclusion criteria were other causes of spinal stenosis (e.g., is reflected in the SPORT — DLS study in which 95% of congenital, post-traumatic, degenerative scoliosis), multi- patients underwent fusion, the majority of which (74%) level surgery (> 2 levels), previous surgeries (at the symp- were instrumented fusions.5 tomatic or adjacent level; a prior discectomy was allowed) or Degenerative spondylolisthesis, however, represents a multilevel coronal and/or sagittal plane deformity. spectrum of pathology from very stable collapsed discs to maintained disc height with significant translation on load- Surgical technique ing dynamic imaging studies. Clinical symptoms also vary, with the patient experiencing either classical bilateral Indication for surgery and type of surgery was as per the neuro genic claudication symptoms and/or unilateral/ individual surgeons’ practices. bilateral lumbar radiculopathy/sciatica. Physicians and sur- Decompression alone was performed at 1 centre only geons experienced in the treatment of patients with this (Toronto Western Hospital [TWH]). This technique was condition recognize this broad clinical presentation, which chosen for patients with neurogenic claudication/mechanical is a consideration given recent randomized controlled trials radiculopathy (i.e., leg-dominant symptoms that were (RCTs) evaluating “similar patients” from an experimental relieved by postural change and/or rest), no (or tolerable) and control perspective. By carefully delineating these mechanical back pain, facet anatomy favourable to facet- potential subgroups the question arises as to whether all sparing (i.e., undercutting) decompression, up to a 25% cases of DLS require fusion and, if so, whether instrumen- (grade I) spondylolisthesis, and no obvious dynamic instabil- tation may or may not be required as an adjunct to fusion. ity on imaging. Radiographic dynamic instability was The importance of this question is further amplified when defined as an increase in spondylolisthesis by 4–5 mm one considers the additional morbidity associated with or more demonstrated on supine to standing or flexion- instrumented spinal fusion in elderly patients and the scar- extension imaging.14 Preoperative disc height was not con- city of health care resources for this growing segment of sidered in the decision for D. It entailed a midline-sparing, the population.9–13 bilateral decompression from a unilateral approach using a The development of less destructive midline anatomy- tubular retractor system (METRx Medtronic) that has been sparing decompressive techniques have created renewed previously described by Kelleher and colleagues.14 interest in D rather than DF for certain patients with At the time of surgery (2000–2006) all 8 surgeons from “stable” DLS.14 The literature to date has demonstrated the 5 other academic centres performed DF for all patients good efficacy of the less invasive decompressive tech- with symptomatic DLS. This group represents the broader niques in treating simple lumbar spinal stenosis (LSS),15–25 structural presentations of DLS, including the more “stable” but to our knowledge no comparative study in a pure patients amenable to D as well as those with more complex cohort of patients with DLS has been conducted. The structural pathology (e.g., grade II or greater listhesis and/or purpose of the present study was to assess the outcomes of more complex coronal or sagittal plane spinal alignment) for anatomy-preserving D in a select subgroup of patients which DF may be indicated. The primary indications for with DLS compared with those of a multicentre cohort of surgery were leg-dominant pain and, to a much lesser patients with DLS who underwent DF. extent, back and leg pain. Fusion for back-dominant pain

Can J Surg, Vol. 57, No. 4, August 2014 E127 RECHERCHE

was rarely performed by any of the surgeons. All fusions Results were instrumented using pedicle screws with posterolateral and/or interbody fusion. A total of 179 patients underwent surgery for the diagnosis of spinal stenosis with DLS. Decompression alone Data collection was performed in 46 patients (57% single-level), whereas DF was performed in 133 patients (64% single-level). Data included the patient characteristics of age and sex. The The baseline demographic and clinical characteristics of preoperative and postoperative (2 yr minimum) Medical the groups are presented in Table 1. The D group was Outcomes Study Short-Form General Health Survey (SF- on average 5 years older (p = 0.003) and had 15% fewer 36) was administered. Data were obtained from site-specific women (p = 0.044) than the DF group. The mean time prospective surgical registries collecting patient-reported from surgery was equivalent between the groups (p = HRQoL (SF-36) data. Varying definitions of what consti- 0.69). The D group had slightly more 2-level procedures tuted an adverse event and different methods for reporting than the DF group (43% vs. 36%, respectively). Baseline all or selected events precluded comparison of adverse events. SF-36 values are presented in Table 2. There was no significant difference in baseline SF-36 scores between the Outcome measures groups; however, 3 SF-36 components nearly reached significance: MCS, general health (GH) and mental The SF-36 physical component summary (PCS) score was health (MH; all p = 0.06). With the exception of GH, the primary outcome measure. Primary analysis included there was significant improvement pre- to postopera- comparison of the degree of change between pre- and tively in all SF-36 subscales and summary scores for both postoperative PCS scores and the proportion of patients the D and DF groups (Table 3). from each cohort reaching minimal clinically important Comparison between the subgroups of D and DF difference (MCID) and substantial clinical benefit (SCB) patients from the only centre performing D (TWH) are for PCS, as defined for degenerative spinal surgery.26 Our shown in Table 4. There was no significant difference secondary analysis compared the 2-year postoperative between the D and DF groups’ baseline and 2 year SF-36 change in scores on the 8 SF-36 subscales and the mental scores (data presented for only the PCS, physical func- component summary (MCS) score. tioning [PF] and bodily pain [BP] scores; no difference was noted for any other subscales). The results of patients Statistical analysis who underwent DF at TWH were also compared with those of patients who underwent DF at the other centres. We performed univariate analysis using an unpaired Student There was no significant difference between the D and t test for continuous variables and a Pearson χ2 test for cate- DF groups’ baseline and 2 year SF-36 scores (data pre- gorical variables. Multivariate analysis was performed to con- sented for only the PCS, PF and BP scores; no difference trol for any significant baseline difference between cohorts. was noted for any other subscales).

A priori power analysis Primary outcome

Using historical standard deviations for PCS in this popu- With regard to the numeric mean change in overall phys lation, with α (type I error rate) set 0.05 and power at ical HRQoL (PCS) there was no significant difference in 80%, we determined that 50 patients per group would be the mean change in PCS for D and DF (10.4 v. 11.4, p = required to detect an MCID for PCS between groups. 0.61). Similarly, the number of patients reaching MCID

Table 1. Demographic and clinical characteristics of the study sample

Group; mean ± SD or no. (%)

Decompression alone, Decompression and fusion, Characteristic n = 46 n = 133 p value* Age, yr 67.80 ± 8.66 62.47 ± 10.83 0.003 Sex, female 27 (59) 98 (74) 0.044 % with 1-level surgery 26 (57) 85 (64) 0.35 Time from surgery 29.95 ± 14.34 29.17 ± 10.36 0.69 Baseline PCS score 28.90 ± 7.90 30.00 ± 7.00 0.39 Baseline MCS score 42.90 ± 12.70 46.80 ± 11.80 0.06

MCS = mental component summary; PCS = physical component summary; SD = standard deviation. *Two sample Student t test (mean) or Pearson χ2 test (percentage).

E128 J can chir, Vol. 57, No 4, août 2014 RESEARCH

(4.9 point change in PCS) was 68% for D and 73% for was no significant difference between the D and DF DF (p = 0.58). The number of patients reaching SCB groups in the pre- to postoperative change in any of the (6.2 point change in PCS) was 64% for D and 66% for subscales or the MCS (all p > 0.19). DF (p = 0.81). The results of multivariate analysis are shown in Tables 5–7. The multivariate analysis demon- Discussion strated that baseline age (p = 0.039) and PCS (p < 0.003) were independent predictors of change in PCS and likeli- The results of our study demonstrate that in a select sub- hood of reaching MCID and SCB for PCS. Older patients group of patients with DLS (i.e., those with leg-dominant and those with higher baseline PCS scores (i.e., better symptoms and what is typically termed a “stable DLS”) D physical HRQoL) had less change in PCS and were less can achieve the same significant improvement in HRQoL likely to reach MCID and SCB for PCS. There was no as DF. significant difference in change in PCS or likelihood of The generally accepted clinical belief that DF is super reaching MCID or SCB for PCS between the D and DF ior to D for the surgical management of DLS was recently groups when adjusted for other variables (all p > 0.74). supported by a systematic review by Martin and colleagues.8 Historically superior outcomes of DF versus D Secondary outcome are demonstrated in patient-reported outcomes,7,8,27–30 postoperative increase in listhesis (instability)7,27,30,31 and The mean change in SF-36 subscale and component sum- reoperation rates.8,28,32 However, in contrast to the present mary scale scores are presented in Table 3. Overall, there study, a distinct “stable cohort” of patients with DLS was

Table 2. Baseline and postoperative SF-36 subcomponent scores

Group; mean ± SD

Decompression alone, n = 46 Decompression and fusion, n = 133 Overall SF-36 component Baseline Postoperative p value* Baseline Postoperative p value* p value* Physical component 28.89 ± 7.95 39.02 ± 11.69 < 0.001 29.97 ± 7.00 41.39 ± 10.59 < 0.001 0.39 summary Mental component 42.91 ± 12.69 50.23 ± 10.62 0.004 46.78 ± 11.83 50.94 ± 10.66 0.003 0.06 summary Physical functioning 24.82 ± 20.64 52.16 ± 30.66 < 0.001 30.43 ± 22.09 60.81 ± 27.26 < 0.001 0.13 Role-physical 13.72 ± 20.35 46.60 ± 35.51 < 0.001 16.71 ± 25.65 54.01 ± 36.26 < 0.001 0.48 Bodily pain 26.15 ± 22.15 57.29 ± 25.16 < 0.001 27.52 ± 14.57 56.38 ± 23.88 < 0.001 0.62 General health 59.87 ± 22.87 61.00 ± 22.99 0.81 66.39 ± 19.55 68.50 ± 19.18 0.37 0.06 Vitality 37.69 ± 19.01 51.69 ± 22.09 0.002 38.36 ± 20.28 56.43 ± 21.14 < 0.001 0.84 Social functioning 44.57 ± 26.57 76.67 ± 23.47 < 0.001 47.92 ± 26.58 75.28 ± 26.58 < 0.001 0.46 Role-emotional 46.37 ± 39.97 70.46 ± 35.73 0.003 56.57 ± 40.98 74.00 ± 34.57 < 0.001 0.15 Mental health 61.13 ± 20.28 73.12 ± 18.07 0.004 68.21 ± 18.29 76.55 ± 34.57 < 0.001 0.06

SD = standard deviation. *Two-sample Student t test comparing pre- and postoperative values.

Table 3. Two-year change in health-related quality of life, SF-36 components

Group; mean ± SD Δ quality of life

Decompression alone, Decompression and fusion, SF-36 component n = 46 n = 133 p value* Physical component summary 10.43 ± 10.77 11.36 ± 10.21 0.61 Mental component summary 7.36 ± 14.01 4.26 ± 13.35 0.19 Physical functioning 27.34 ± 31.16 30.37 ± 27.06 0.53 Role-physical 32.88 ± 36.41 37.71 ± 38.62 0.46 Bodily pain 32.78 ± 23.57 28.53 ± 27.04 0.35 General health 1.06 ± 22.53 2.01 ± 21.70 0.80 Vitality 14.28 ± 25.34 17.88 ± 25.35 0.41 Social functioning 32.50 ± 32.57 27.18 ± 34.28 0.36 Role-emotional 23.06 ± 48.43 17.99 ± 45.28 0.53 Mental health 12.67 ± 20.06 8.35 ± 19.23 0.20

SD = standard deviation. *Two-sample Student t test.

Can J Surg, Vol. 57, No. 4, August 2014 E129 RECHERCHE not identified and a midline anatomy-sparing minimally lished SPORT — DLS study, 19 patients underwent D and invasive approach was not used.7,8,27,28,33 A traditional lami- 344 patients underwent DF.36 As reported by Tosteson and nectomy does not preserve any of the midline structures colleagues,36 the quality-adjusted life years (QALY) gained and also may not be facet-preserving. Consequently, a tra- by the 19 patients who had D was the same as that in the ditional laminectomy has a higher likelihood of increased DF cohort 2 years post-surgery.36 Unfortunately, no details postoperative instability, clinical failure and revision rate regarding selection criteria for those undergoing D are pro- over time, particularly in the DLS patient population.7,34 vided in these studies. However, several small series in which facet-preserving It is our belief that, from a structural and clinical perspec- techniques were used also suggest that DF was still super tive, all patients with asymptomatic DLS are not equal. ior for this patient population.28,30,32,35 Symptomatic patients typically present with 3 clinical scen The findings of the present study are contrary to those arios (back-dominant pain, leg-dominant pain and equal in most of the literature and to surgeon belief. back and leg pain) and a stable or unstable (i.e., mobile) low- Although limited in number, there are a few published grade (I-II) listhesis. Regardless of outcome, it would appear studies that contradict the studies favouring fusion and that that the 2 main selection criteria used in this study are con- support the findings of the present study. Matsudaira and sistent with those of other contemporary studies where D colleagues31 demonstrated no difference in outcome was applied in the DLS population: leg-dominant symptoms between midline-sparing (bilateral laminotomy), facet- and stable (< 3–5mm of movement) grade 1 spondylolisthe- preserving decompression (n = 18) and decompression and sis.14,31,32,37 Essentially, these patients present with unilateral instrumented posterolateral fusion (n = 19) 2 years after or bilateral neurogenic claudication symptoms, much like surgery in patients with grade 1 DLS.31 In the recently pub- patients with LSS.38 Two studies using these selection

Table 4. SF-36 component scores: decompression alone versus decompression and fusion in Toronto Western Hospital patients, and decompression and fusion in Toronto Western Hospital patients versus all sites

TWH groups; mean ± SD* Group; mean ± SD*

TWH Decompression decompression All sites Decompression and fusion, and fusion, depression and Factor alone, n = 46 n = 25 p value† n = 25 fusion, n = 108 p value† Baseline PCS 28.9 ± 8.0 31.2 ± 7.6 0.25 31.2 ± 7.6 29.7 ± 6.9 0.39 2 year PCS 39.0 ± 11.7 42.8 ± 9.7 0.16 42.8 ± 9.7 41.1 ± 10.8 0.45 Change in PCS 10.4 ± 10.8 12.1 ± 9.4 0.51 12.1 ± 9.4 11.2 ± 10.4 0.69 PCS MCID,‡ % 65 76 76 70 PCS SCB,§ % 61 72 72 63 Baseline PF 24.8 ± 20.6 31.1 ± 23.7 0.27 31.1 ± 23.7 30.3 ± 21.8 0.87 2-year PF 52.2 ± 30.7 62.2 ± 25.0 0.14 62.2 ± 25.0 60.5 ± 27.9 0.77 Change in PF 27.3 ± 31.2 31.1 ± 24.7 0.58 31.1 ± 24.7 30.2 ± 27.7 0.88 Baseline BP 26.2 ± 19.5 30.3 ± 18.4 0.38 30.3 ± 18.4 26.9 ± 13.6 0.39 2-year BP 57.3 ± 25.2 59.2 ± 20.7 0.73 59.2 ± 20.7 55.7 ± 24.6 0.46 Change in BP 32.8 ± 23.6 29.0 ± 25.4 0.54 29.0 ± 25.4 28.4 ± 27.5 0.93

BP = bodily pain; MCID = minimal clinically important difference; PCS = physical component summary; PF = physical functioning; SCB = substantial clinical benefit; SD = standard deviation; TWH = Toronto Western Hospital. *Unless otherwise indicated. †Two-sample Student t test. ‡MCID for PCS = 4.9. §SCB for PCS = 6.2.

Table 5. Multiple linear regression results for change in Table 6. Logistic regression results for MCID on physical physical component summary score, n = 175 component summary score,* n = 125

Variable Coefficient (95% CI) p value Variable Odds ratio (95% CI) p value Age, yr –0.20 (–0.34 to –0.05) 0.009 Age, yr 0.96 (0.93–1.00) 0.039 Sex, female –1.21 (–4.66 to 2.24) 0.49 Sex, female 0.80 (0.36–1.82) 0.60 Baseline PCS –0.47 (–0.70 to –0.25) < 0.001 Baseline PCS score 0.91 (0.87–0.96) 0.001 Baseline MCS 0.06 (–0.07 to 0.19) 0.35 Baseline MCS score 1.02 (0.99–1.05) 0.13 Decompression and fusion 0.60 (–2.97 to 4.17) 0.74 Decompression and fusion 1.11 (0.48–2.55) 0.81

CI = confidence interval; MCS = mental component summary; PCS = physical CI = confidence interval; MCID = minimal clinically important difference; MCS = mental component summary; SD = standard deviation. component summary; PCS = physical component summary; SD = standard deviation. *MCID for PCS = 4.9.

E130 J can chir, Vol. 57, No 4, août 2014 RESEARCH

criter ia have directly assessed the outcome of D alone for (4% at 48 month follow-up14,37) are comparable to that DLS compared with LSS patients without DLS. Sasai and reported in the literature for contemporary DF in this popu- colleagues37 demonstrated that the outcomes of midline lation.5,6 Regardless, the clinical and economic impact of any facet-preserving D in select patients with DLS (n = 23) was potential difference in the long-term revision rate of these similar to those of patients with LSS (n = 25) without spon- cohorts requires further investigation. Although not part of dylolisthesis at a minimum of 2 years (mean follow-up was the present study, the TWH patients reported in Table 4 are 4 yr). Most recently, Kelleher and colleagues14 demonstrated part of an ongoing observational study with follow-up comparable outcomes at 2 years with D in the same sub ranging from 5 to 13 years. In this group, the longer term population of DLS patients as those in our study compared revision rate for those who underwent D was 11% (n = 5 [3 to patients with LSS without spondylolisthesis. with same site and 2 adjacent segment procedures]; 3 of these By applying the aforementioned selection criteria (see patients required a subsequent DF and 2 had repeat D; mean methods), D for this defined subset of patients with DLS has time to revision was 61.2 months) and 36% for those with several obvious advantages. From the perspective of elderly DF (n = 9 [2 with same site and 7 with adjacent segment pro- patients, reduced surgical morbidity and recovery time with cedures]; all had a repeat DF; mean time to revision was similar clinical outcomes are clearly desirable.6,39,40 From a 62.1 months). It must be emphasized that the primary DF health care system perspective, the reduced duration of sur- group at this centre would represent the more unstable and gery, length of hospital stay and cost of D versus DF trans- complex anatomic presentations of DLS. Given the possibil- late to cost savings or increased service delivery (i.e., more ity of therapeutic equipoise, the question of D versus DF for patients treated) for the same cost. However, before wide a defined subpopulation of patients with DLS lends itself adoption of D for a subpopulation of patients with DLS can ideally to an RCT. However, as demonstrated by the be considered, the generalizability and sustainability of the ongoing controversy of instrumented versus noninstru- alternative technique must be demonstrated. Although a key- mented fusion for DLS, an RCT demonstrating minimal hole technique was used in our study (i.e., the preferred difference without long-term follow up is unlikely to change access of the specific surgeon using this approach), others the established practice of fusion for most — if not all — have demonstrated similar findings with more conventional patients with DLS who require surgical intervention.9,35,41–44 access and a bilateral technique.31,37 The postoperative increase in radiographic listhesis demonstrated in the studies Limitations of Kelleher and colleagues,14 Matsudaira and colleagues31 and Sasai and colleagues37 (1.7% to 8.4%) is concerning regard- The major strength of our study is that it assesses an alter- ing long-term sustainability. However, these studies all noted native surgical management strategy (D) in a highly that an increase in listhesis did not correlate with an inferior selected subpopulation of DLS patients compared with a clinical outcome or higher reoperation rate 2–4 years postop- generalizable multicentre cohort of DLS patients with sim- eratively. Furthermore, the revision rates in these studies ilar clinical presentation in whom this selection criteria was not applied and who all received DF. To our knowledge, this study also represents the largest comparative study of Table 7. Logistic regression results for SCB on physical its kind and presents clearly defined selection criteria and component summary score,* n = 114 surgical principle for D in the DLS population. The meth- Variable Odds ratio (95% CI) p value odological limitations of this study are related to the retro- Age, yr 0.96 (0.92–0.99) 0.010 spective nature of our data abstraction from prospective Sex, female 1.07 (0.50–2.28) 0.87 databases. The potential confounding effects of patient and Baseline PCS score 0.93 (0.88–0.97) 0.002 surgeon selection biases, differential complication rates and Baseline MCS score 1.01 (0.98–1.04) 0.41 differences in surgical technique (mix of posterolateral or Decompression and fusion 0.93 (0.42–2.05) 0.85 interbody instrumented fusion) for the DF cohort cannot CI = confidence interval; MCS = mental component summary; PCS = physical be accounted for, but may reinforce the generalizability of component summary; SCB = substantial clinical benefit; SD = standard deviation. *SCB for PCS = 6.2. our control group. In addition, all the other participating surgeons in this study performed DF for DLS patients. Table 8. Comparison of degenerative spondylolisthesis Patients who received DF at the centre performing selec- patients in SPORT trial sample versus present sample: change tive D demonstrated similar results compared with the rest in SF-36 physical functioning and bodily pain scores of the DF cohort as well as compared with the D cohort,

Group; Δ score suggesting a similar treatment effect can be achieved for the selected subgroup from within the same centre. The experi- SF-36 component SPORT, n = 324 Present, n = 179 mental group (D) was a highly selected subpopulation of Physical functioning 26.6 29.7 patients with DLS and was thus not generalizable to the Bodily pain 29.9 30.1 current literature. In addition, these patients underwent a SPORT = Spine Patient Outcomes Research Trial. specific minimally invasive decompression technique that

Can J Surg, Vol. 57, No. 4, August 2014 E131 RECHERCHE

and has received grants from Medtronic, AO Spine and the Orthopae- has not been assessed for generalizability, thus introducing dic Research and Education Foundation. M. Dvorak and E. Abraham the possibility of a technique-based surgeon and procedural are paid consultants of and have received speaker fees and travel assis- bias. Furthermore, the baseline demographic characteristics tance from Medtronic. W. Oxner has received honoraria and speaker fees from Medtronic. No other competing interests declared. were not equal between cohorts. The D group was on average 5 years older, included fewer women and had more Contributors: All authors designed the study and acquired the data. Y.R. Rampersaud and C. Fisher analyzed the data. Y.R. Rampersaud wrote 2-level procedures than the DF group. Furthermore, the D the article, which all authors reviewed and approved for publication. group had a trend toward lower baseline MCS, GH and MH scores (Table 2, p = 0.06). However, these differences References would more likely bias against the D group.45–47 As noted in the results, multivariate analysis controlling for age, sex and 1. Deyo RA, Gray DT, Kreuter W, et al. United States trends in lum- baseline PCS and MCS scores did not alter the outcome bar fusion surgery for degenerative conditions. Spine 2005;30:1441-7. between those with D and DF. However, we did not con- . 2 Aono K, Kobayashi T, Jimbo S, et al. Radiographic analysis of newly developed degenerative spondylolisthesis in a mean twelve- trol for other potential counfounders, such as medical year prospective study. Spine 2010;35:887-91. comorbidities, smoking status, fusion techniques (i.e., pos- 3. Kalichman L, Kim DH, Li L, et al. Spondylolysis and spondylolis- terolateral v. interbody fusion) or percent of spondylolisthesis: prevalence and association with low back pain in the adult thesis. Although we cannot comment on all possible con- community-based population. Spine 2009;34:199-205. founders, the older age and less intensive procedure 4. Jacobsen S, Sonne-Holm S, Rovsing H, et al. Degenerative lumbar spondylolisthesis: an epidemiological perspective: the Copenhagen performed in the D cohort would suggest they were prob Osteoarthritis Study. Spine 2007;32:120-5. ably more likely to have other medical comorbidities, which . 5 Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical compared would again cause bias toward a lower SF-36 outcome in with nonoperative treatment for lumbar degenerative spondylolis- that group.46 Finally, it is possible that a superior result thesis. Four-year results in the Spine Patient Outcomes Research could have been obtained for the DF cohort if all patients Trial (SPORT) randomized and observational cohorts. J Bone Joint Surg Am 2009;91:1295-304. underwent more contemporary interbody fusion using less 6. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical versus noninvasive techniques. To date, however, studies comparing surgical treatment for lumbar degenerative spondylolisthesis. minimally invasive surgery to open fusion for spondylolis- N Engl J Med 2007;356:2257-70. thesis at 2 years or greater have demonstrated equivalence . 7 Herkowitz HN, Kurz LT. Degenerative lumbar spondylolisthesis in clinical outcome.48–50 Furthermore, if we compare our with spinal stenosis. A prospective study comparing decompression with decompression and intertransverse process arthrodesis. J Bone overall DLS cohort to the as-treated surgical cohort from Joint Surg Am 1991;73:802-8. the SPORT — DLS study, the mean age, sex, preoperative 8. Martin CR, Gruszczynski AT, Braunsfurth HA, et al. The surgical and 2-year postoperative PF and BP scores between our management of degenerative lumbar spondylolisthesis: a systematic studies are very similar (Table 8). Consequently, with the review. Spine 2007;32:1791-8. aforementioned limitations considered, it seems that our . 9 Buckwalter JA, Heckman JD, Petrie DP. AOA. An AOA critical issue: aging of the North American population: new challenges for cohorts and overall outcomes are consistent with those of a orthopaedics. J Bone Joint Surg Am 2003;85-A:748-58. 5,6 contemporary surgical DLS population. 10. Agency for Healthcare Research and Quality. Healthcare Cost and Utilization Project. 2013. Available: http://hcup-us.ahrq.gov/ Conclusion (accessed 2013 Aug. 27). 11. Lafortune G, Balestat G. Trends in severe disability among elderly people. Paris (FR): Organisation for Economic Co-operation and The present study demonstrates that for a specific subpopu- Development; 2007. Available: www.oecd-ilibrary.org/content lation of patients with DLS (i.e., those with leg-dominant /workingpaper/217072070078 (accessed 2013 Aug. 27). symptoms and a radiographically stable grade I spondylolis- 12. Deyo RA, Mirza SK, Martin BI, et al. Trends, major medical com- thesis), undergoing an anatomic midline-sparing micro plications, and charges associated with surgery for lumbar spinal decompression alone can achieve the same improvement in stenosis in older adults. JAMA 2010;303:1259-65. 13. Institute of Medicine of the National Academies. Initial national HRQoL as that of DF for the overall DLS population at priorities for comparative effectiveness research [report]. 2009 June 2 years postoperatively. The routine implementation of D 30. Available: www.iom.edu/Reports/2009/ComparativeEffectiveness for this defined subpopulation of patients with DLS could ResearchPriorities.aspx. (accessed 2013 Aug. 27). result in fewer surgical complications, improved reactivation 14. Kelleher MO, Timlin M, Persaud O, et al. Success and failure of and potentially less health care utilization in a growing seg- minimally invasive decompression for focal lumbar spinal stenosis in patients with and without deformity. Spine 2010;35:E981-7. ment of society. Therefore, further multicentre prospective 15. Khoo LT, Fessler RG. Microendoscopic decompressive laminot- evaluation and longer term follow up is probably warranted. omy for the treatment of lumbar stenosis. Neurosurgery 2002; Acknowledgements: This study was supported by a grant from the 51(Suppl):S146-54. Canadian Spine Society through unrestricted research funds from Depuy, 16. Podichetty VK, Spears J, Isaacs RE, et al. Complications associated Medtronic, Stryker, and Synthes. Additional Funding was also received with minimally invasive decompression for lumbar spinal stenosis. from the Minimal Access Ambulatory Surgical Spine Research and Edu- J Spinal Disord Tech 2006;19:161-6. cation Fund, Toronto General and Western Hospital Foundation. 17. Palmer S, Turner R, Palmer R. Bilateral decompression of lumbar spinal stenosis involving a unilateral approach with microscope and Competing interests: Y.R. Rampersaud is a paid consultant of Medtronic. C. Fisher is a paid consultant of Medtronic and Nuvasive tubular retractor system. J Neurosurg 2002;97(Suppl):213-7.

E132 J can chir, Vol. 57, No 4, août 2014 RESEARCH

18. Palmer S, Turner R, Palmer R. Bilateral decompressive surgery in 36. Tosteson ANA, Tosteson TD, Lurie JD, et al. Comparative effect lumbar spinal stenosis associated with spondylolisthesis: unilateral iveness evidence from the spine patient outcomes research trial: approach and use of a microscope and tubular retractor system. surgical versus nonoperative care for spinal stenosis, degenerative Neurosurg Focus 2002;13:E4. spondylolisthesis, and intervertebral disc herniation. Spine 2011; 19. Guiot BH, Khoo LT, Fessler RG. A minimally invasive technique 36:2061-8. for decompression of the lumbar spine. Spine 2002;27:432-8. 37. Sasai K, Umeda M, Maruyama T, et al. Microsurgical bilateral 20. Cavusoglu H, Kaya RA, Türkmenoglu ON, et al. Midterm out- decompression via a unilateral approach for lumbar spinal canal come after unilateral approach for bilateral decompression of lum- stenosis including degenerative spondylolisthesis. J Neurosurg Spine bar spinal stenosis: 5-year prospective study. Eur Spine J 2007;16: 2008;9:554-9. 2133-42. 38. Suri P, Rainville J, Kalichman L, et al. Does this older adult with 21. Thomé C, Zevgaridis D, Leheta O, et al. Outcome after less- lower extremity pain have the clinical syndrome of lumbar spinal invasive decompression of lumbar spinal stenosis: a randomized stenosis? JAMA 2010;304:2628-36. comparison of unilateral laminotomy, bilateral laminotomy, and 39. Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical versus laminectomy. J Neurosurg Spine 2005;3:129-41. nonsurgical therapy for lumbar spinal stenosis. N Engl J Med 22. Mackay DC, Wheelwright EF. Unilateral fenestration in the treat- 2008;358:794-810. ment of lumbar spinal stenosis. Br J Neurosurg 1998;12:556-8. 40. Wang MY, Green BA, Shah S, et al. Complications associated with 23. Weiner BK, Walker M, Brower RS, et al. Microdecompression for lumbar stenosis surgery in patients older than 75 years of age. lumbar spinal canal stenosis. Spine 1999;24:2268-72. Neurosurg Focus 2003;14:e7. 24. McCulloch JA. Microdecompression and uninstrumented single- 41. Tsutsumimoto T, Shimogata M, Yoshimura Y, et al. Union versus level fusion for spinal canal stenosis with degenerative spondylolis- nonunion after posterolateral lumbar fusion: a comparison of long- thesis. Spine 1998;23:2243-52. term surgical outcomes in patients with degenerative lumbar spon- 25. Ikuta K, Tono O, Oga M. Clinical outcome of microendoscopic dylolisthesis. Eur Spine J 2008;17:1107-12. posterior decompression for spinal stenosis associated with degen- 42. Kornblum MB, Fischgrund JS, Herkowitz HN, et al. Degenerative erative spondylolisthesis — minimum 2-year outcome of 37 lumbar spondylolisthesis with spinal stenosis: a prospective long- patients. Minim Invasive Neurosurg 2008;51:267-71. term study comparing fusion and pseudarthrosis. Spine 2004;29: 26. Glassman SD, Copay AG, Berven SH, et al. Defining substantial 726-34. clinical benefit following lumbar spine arthrodesis.J Bone Joint 43. Kimura I, Shingu H, Murata M, et al. Lumbar posterolateral fusion Surg Am 2008;90:1839-47. alone or with transpedicular instrumentation in L4–L5 degenera- 27. Feffer HL, Wiesel SW, Cuckler JM, et al. Degenerative spondylo- tive spondylolisthesis. J Spinal Disord 2001;14:301-10. listhesis. To fuse or not to fuse. Spine 1985;10:287-9. 44. Fischgrund JS, Mackay M, Herkowitz HN, et al. 1997 Volvo award 28. Lombardi JS, Wiltse LL, Reynolds J, et al. Treatment of degenera- winner in clinical studies. Degenerative lumbar spondylolisthesis tive spondylolisthesis. Spine 1985;10:821-7. with spinal stenosis: a prospective, randomized study comparing 29. Satomi K, Hirabayashi K, Toyama Y, et al. A clinical study of decompressive laminectomy and arthrodesis with and without spi- degenerative spondylolisthesis. Radiographic analysis and choice of nal instrumentation. Spine 1997;22:2807-12. treatment. Spine 1992;17:1329-36. 45. Hopman WM, Towheed T, Anastassiades T, et al. Canadian nor- 30. Bridwell KH, Sedgewick TA, O’Brien MF, et al. The role of fusion mative data for the SF-36 health survey. Canadian Multicentre and instrumentation in the treatment of degenerative spondylolis- Osteoporosis Study Research Group. CMAJ 2000;163:265-71. thesis with spinal stenosis. J Spinal Disord 1993;6:461-72. 46. Slover J, Abdu WA, Hanscom B, et al. The impact of comorbidities 31. Matsudaira K, Yamazaki T, Seichi A, et al. Spinal stenosis in grade I on the change in Short-Form 36 and Oswestry scores following degenerative lumbar spondylolisthesis: a comparative study of out- lumbar spine surgery. Spine 2006;31:1974-80. comes following laminoplasty and laminectomy with instrumented 47. Aalto TJ, Malmivaara A, Kovacs F, et al. Preoperative predictors for spinal fusion. J Orthop Sci 2005;10:270-6. postoperative clinical outcome in lumbar spinal stenosis: systematic 32. Ghogawala Z, Benzel EC, Amin-Hanjani S, et al. Prospective out- review. Spine 2006;31:E648-63. comes evaluation after decompression with or without instru- 48. Dhall SS, Wang MY, Mummaneni PV. Clinical and radiographic mented fusion for lumbar stenosis and degenerative Grade I spon- comparison of mini-open transforaminal lumbar interbody fusion dylolisthesis. J Neurosurg Spine 2004;1:267-72. with open transforaminal lumbar interbody fusion in 42 patients 33. Mardjetko SM, Connolly PJ, Shott S. Degenerative lumbar spondy- with long-term follow-up. J Neurosurg Spine 2008;9:560-5. lolisthesis. A meta-analysis of literature 1970-1993. Spine 1994; 49. Tsutsumimoto T, Shimogata M, Ohta H, et al. Mini-open versus 19(Suppl):2256S-65S. conventional open posterior lumbar interbody fusion for the 34. Martin BI, Mirza SK, Comstock BA, et al. Reoperation rates fol- treatment of lumbar degenerative spondylolisthesis: comparison lowing lumbar spine surgery and the influence of spinal fusion proof paraspinal muscle damage and slip reduction. Spine 2009;34: cedures. Spine 2007;32:382-7. 1923-8. 35. Yone K, Sakou T, Kawauchi Y, et al. Indication of fusion for lum- 50. Peng CWB, Yue WM, Poh SY, et al. Clinical and radiological out- bar spinal stenosis in elderly patients and its significance.Spine comes of minimally invasive versus open transforaminal lumbar 1996;21:242-8. interbody fusion. Spine 2009;34:1385-9.

Can J Surg, Vol. 57, No. 4, August 2014 E133 RESEARCH • RECHERCHE

Oncoplastic reduction using the vertical scar superior-medial pedicle pattern technique for immediate partial breast reconstruction

Yoav Barnea, MD* Background: Oncoplastic breast reduction in women with medium to large breasts Amir Inbal, MD* has reportedly benefitted them both oncologically and cosmetically. We present our experience with an oncoplastic breast reduction technique using a vertical scar Daphna Barsuk, MD† superior-medial pedicle pattern for immediate partial breast reconstruction. † Tehila Menes, MD Methods: All patients with breast tumours who underwent vertical scar superior- Arik Zaretski, MD* medial pedicle reduction pattern oncoplastic surgery at our centre between Septem- David Leshem, MD* ber 2006 and June 2010 were retrospectively studied. Follow-up continued from 12 months to 6 years. Jerry Weiss, MD* Schlomo Schneebaum, MD† Results: Twenty women (age 28–72 yr) were enrolled: 16 with invasive carcinoma and 4 with benign tumours. They all had tumour-free surgical margins, and no fur- Eyal Gur, MD* ther oncological operations were required. The patients expressed a high degree of satisfaction from the surgical outcome in terms of improved quality of life and a good From the Departments of *Plastic and cosmetic result. Reconstructive Surgery and the †Breast Health Center, Tel Aviv Sourasky Medical Conclusion: The vertical scar superior-medial pedicle reduction pattern is a versatile Center, affiliated with the Sackler Faculty oncoplastic technique that allows breast tissue rearrangement for various tumour of Medicine, Tel Aviv University, Tel locations. It is oncologically beneficial and is associated with high patient satisfaction. Aviv, Israel Contexte : Chez des femmes qui avaient une poitrine de moyenne à volumineuse, la The first and second authors both had equal contribution to the manuscript. réduction mammaire oncoplastique aurait exercé des bienfaits, tant au plan oncologique que cosmétique. Nous présentons notre expérience d’une technique de réduction mammaire oncoplastique à cicatrice verticale et pédicule supéromédian Accepted for publication pour une reconstruction mammaire partielle immédiate. Jan. 31, 2014 Méthodes : Tous les cas de tumeurs mammaires soumis à la réduction à cicatrice Correspondence to: verticale et pédicule supéromédian en chirurgie oncoplastique dans notre centre entre Y. Barnea septembre 2006 et juin 2010 ont été passés en revue rétrospectivement. Le suivi s’est Department of Plastic and Reconstructive échelonné sur 1 à 6 ans. Surgery Tel-Aviv Sourasky Medical Center Résultats : Vingt femmes (âgées de 28 à 72 ans) ont été inscrites : 16 étaient atteintes 6 Weizman St. d’un cancer envahissant et 4 de tumeurs bénignes. Elles présentaient toutes des mar- Tel Aviv, Israel, 64239 ges chirurgicales libres de tumeur et aucune autre intervention oncologique n’a été [email protected] nécessaire. Les patientes ont exprimé un degré élevé de satisfaction à l’endroit des résultats de la chirurgie pour ce qui est de l’amélioration de leur qualité de vie et de l’effet cosmétique positif. DOI: 10.1503/cjs.031213 Conclusion : La technique de réduction à cicatrice verticale et pédicule supéromé- dian est une technique oncoplastique flexible qui permet un réarrangement des tissus mammaires en fonction de la localisation des tumeurs. Au plan oncologique, elle est bénéfique et associée à un degré élevé de satisfaction chez les patientes.

reast cancer is the most common malignancy affecting women in the western world.1,2 The surgical treatment for breast cancer has continu- B ously undergone profound changes over the past 3 decades, and the medical community currently endorses breast-conserving therapy (BCT) as the gold standard approach for most women with early-stage breast cancer.1,2 The combination of partial mastectomy and postsurgical radiation therapy has sometimes resulted in poor cosmetic results, characterized by deformation and

E134 J can chir, Vol. 57, No 4, août 2014 © 2014 Association médicale canadienne RESEARCH noticeable asymmetry of the shape and size of the operated Each patient was closely followed postoperatively by her breast.3–7 This occurs more often when the tumour:breast plastic and general surgeons as well as by her oncologist. size ratio is high. Several studies have shown a direct cor Cosmetic outcome was determined based on patient satis- relation between the magnitude of parenchymal and cuta- faction and by grading from 5 independent reviewers, all of neous excision and cosmetic outcome.3–7 whom were plastic surgeons. Categories for evaluation Improvements in diagnostic technology and mammo- included breast shape, nipple position and breast/nipple graphic screening as well as increased use of preoperative local symmetry. Each were given a score on a scale of 1 to 4 (1 = or systemic therapies have extended the indications for poor, 2 = satisfactory, 3 = good, 4 = very good). The patients BCT.1,2 Several oncoplastic breast surgery techniques have graded their satisfaction on a scale of 1 to 4 (1 = regret the been introduced in an attempt to optimize the balance decision, 2 = disappointed, 3 = satisfied, 4 = very satisfied). between the risk of local recurrence and the cosmetic out- Data were collected retrospectively from outpatient charts. come of BCT.3–14 The combined plastic surgery techniques of tissue replacement or rearrangement provide a wider local Surgical technique excision while achieving better breast shape and symmetry.3–23 Although macromastia has been considered a contraindica- The patients were seen preoperatively by a multidisci- tion for BCT owing to difficulties in administering radiation plinary breast surgery team, and a plastic surgeon was con- therapy at the surgical site, it has become standard procedure sulted because of large breast volume, ptosis or tumour size for a select group of patients with both breast cancer and and location. Patients with macromastia and tumours not breast hypertrophy.15–23 The combination of a tumour resec- located in the superior-medial pole of the breast were con- tion in a reduction pattern with a contralateral breast reduc- sidered candidates for an oncoplastic breast reduction tech- tion was first developed in the late 1980s and has been nique using the vertical scar superior-medial pedicle pattern reported by many authors.15–23 An increasing number of for immediate reconstruction. Tumour size and location, reports have stated that bilateral breast reduction in conjunc- surgical scars, resection area and axillary dissection were tion with tumour-directed breast-conserving therapy is a sur- planned and discussed among the participating specialists gical technique that can potentially improve the effectiveness after reviewing all relevant breast imaging. of radiation therapy, alleviate neuropathic symptoms that may Tumour location was marked on the breast skin. Nipple accompany macromastia and enhance the patients’ perceptions location was spotted 1–2 cm below the inframammary fold of their bodies after surgery.15–23 Numerous surgical tech- on the central meridian of the breast. A mosque pattern was niques for oncologic breast reduction have been described in marked around the new nipple location, and the medial and the literature, and many of them have variably overlapping lateral margins of the skin resection were patterned using technical details, all of which can cause some confusion when the Lassus manoeuvre.24–26 The superior-medial pedicle was evaluating and comparing the published results.15–23 marked with a width of 6–8 cm, depending on the volume of Our objective was to describe our experience with the the breast and planned tissue resection. The base width of vertical scar superior-medial pedicle reduction pattern the pedicle included 5 cm of the medial pillar on the vertical approach for immediate oncoplastic reconstruction surgery limb 1–3 cm from the medial part of the mosque. The pedi- on BCT deformities. The indications, advantages and limi- cle was oriented more medially than superiorly when the tations of the technique are discussed, and the simplicity of distance between the new and old nipple was shorter. a single reduction technique for accommodating the differ- Tumour resection was achieved through the skin resection ent breast tumour regions is emphasized. markings of both the general surgeon and the plastic surgeon. Skin undermining beside the tumour bed was per- Methods formed in order to permit wide glandular resection. After resection, tissue extensions were taken from all tumour bed All patients who underwent BCT and immediate reconstruc- dimensions, and the tumour bed margins were marked by tion using the vertical scar superior-medial pedicle reduction surgical clips to facilitate locating the original tumour bed pattern oncoplastic surgery at the Tel-Aviv Sourasky Medical for the expected radiation boost. The tumour specimens Center and a private clinic between September 2006 and were marked and weighed. Axillary dissection, when needed, June 2010 were included in this series. We retrospectively was performed through a separate axillary incision. collected and evaluated data on their demographic character- Once the tumour had been removed, its location dictated istics, oncologic findings, hospital admissions and postopera- the reduction pattern, resection and insetting. For tumour tive outcomes. Oncologic information included tumour type, defects located in the inferior pole of the breast, the remaining size and location; axillary lymph-node surgery; and adjuvant skin and glandular tissue were resected according to the previ- chemo- and radiotherapy. All breast specimens had been ous reduction pattern markings, and the pedicle was rotated marked and weighed, and surgical margins were assessed by superiorly to its new position. The medial and lateral pillars pathology to determine if the tumour had been fully excised were then shaped and plicated. For tumours that were located and whether the margins were tumour-free. in the lateral pole of the breast, the pedicle was harvested with

Can J Surg, Vol. 57, No. 4, August 2014 E135 RECHERCHE additional glandular tissue from the inferior pole that filled the were done immediately by the plastic surgeon after defect in the lateral area once the pedicle had been rotated tumour removal by the general surgeon. The participants’ superiorly (Fig. 1). Centrally located tumours that required demographic and clinical characteristics are summarized resection of the nipple/areola complex (NAC) were marked by in Table 1. Sixteen patients (80%) had invasive carcin an inverted “V” instead of the mosque design, positioning the omas and the other 4 patients (20%) had benign breast tip of the “V” at the planned new nipple position. The pedicle tumours (2 fibroadenoma, 1 cystosarcoma phylloides and stub (resected NAC) was harvested with additional glandular 1 lipoma). Eighteen patients (90%) underwent unilateral tissue from the inferior-medial pole, which was rotated to the oncoplastic reconstruction and 2 patients had bilateral central area of the breast for better breast projection (Fig. 2). oncoplastic reconstruction. Contralateral breast surgery Lateral fullness was addressed by thorough undermining and (n = 18) included 16 with reduction/mastopexy, 1 with emptying of the lower-lateral triangles during the procedure. contralateral mastectomy and immediate tissue expander No lateral liposuction was performed. reconstruction and 1 with no surgery, yielding a total of The resected breast tissue was added to the weight of the 39 operated breasts. The tumour locations are displayed tumour specimen for determining total tissue removal. The in Figure 3. One patient had unilateral NAC resection contralateral breast underwent reduction in the superior- because of tumour location. Duration of surgery averaged medial pattern. We usually tried to reduce the normal 3 hours and 30 minutes (range 2.5–5.5 h), and hospital breast by 10% more than the affected breast in order to stay averaged 3.5 (range 2–8) days. achieve better symmetry after radiation therapy. The oncological data are summarized in Table 2. Eleven The patient was then positioned upright for final assess- of the 16 patients with malignant disease underwent axillary ment of symmetry, flap moulding and breast shape. Three lymph node dissection, and 5 patients underwent sentinel drains were inserted (1 in each breast and 1 in the axilla lymph node biopsy that was negative for metastasis, both after dissection) and the incisions were sutured. The sur performed from a separate axillary incision. Five patients gical scars were protected with gauze pads, and a sports bra had chemotherapy before surgery (Table 2) and 8 had it was fitted comfortably over the entire surgical field. after surgery. All patients with malignant tumours received postoperative radiation therapy. Radiotherapy was adminis- Results tered after chemotherapy and included daily fractionated doses up to a total of 45–50 Gy and an additional boost of A total of 20 women with breast tumours underwent sur- 10 Gy to the primary tumour bed. gery by means of the vertical scar superior-medial pedicle All patients had tumour-free surgical margins, and no reduction pattern oncoplastic surgery technique. Their further oncological operations were required. One patient mean age was 46.6 (range 28–72) years. Reconstructions had invasive ductal carcinoma that reached 0.5 cm from

A B

C D

Fig. 1. A 45-year-old patient with infiltrating ductal carcinoma (IDC) of her left breast located at the lateral pole (A, B). She previously underwent left lumpectomy with a periareolar incision twice, with involved tumour margins on both occasions. She underwent left relumpectomy with vertical scar superior-medial pedicle reduction pattern oncoplastic surgery and right breast reduction. Postoperative pictures (C, D) 1.5 years after surgery and radiation therapy to the left breast.

E136 J can chir, Vol. 57, No 4, août 2014 RESEARCH the surgical margin; extension biopsies taken from that The postoperative follow-up period averaged 34.7 area during surgery were negative. No residual tumour was (range 12–72) months. One patient was lost to long-term seen on the pathological specimens of 2 patients after they follow-up because she lived abroad. Postoperative compli- had undergone preoperative neoadjuvant chemotherapy cations included dehiscence of the upper vertical scar and (complete pathological response). lateral fat necrosis in 1 patient, who was successfully

A B

C D

E F

Fig. 2. A 46-year-old patient with infiltrating ductal carcinoma (IDC) of her right breast located under the nipple/areola complex (NAC) (A, B). The patient underwent right lumpectomy, including the NAC, leaving a large central defect (C). Glandular tissue from the inferior-medial pole based medially was rotated to the central area of the breast to fill in the defect (D). The left breast was reduced simultaneously. Postoperative pictures (E, F) 1 year after surgery and radiation therapy to the right breast.

Table 1. Patient characteristics

No. (%) or mean Characteristic [range] Age, yr 46.6 [28–72] Comorbidities Hyperlipidemia 3 (15) Hypertension 2 (10) Asthma 1 (5) Body mass index 29 [22–36] Smoker 2 (10) Premenopausal 15 (75) Previous breast surgery 1 (5)

Can J Surg, Vol. 57, No. 4, August 2014 E137 RECHERCHE treated conservatively. There were no surgical complica- Figure 5 shows a patient who was considered to have a good tions in the contralateral healthy breast. Two patients to satisfactory outcome based on the evaluation by the underwent revision surgery unrelated to radiation therapy: independent observers. 1 for improved areolar symmetry and the other for rereduction of the oncoplastic breast owing to asymmetry of Discussion the implant-reconstructed contralateral breast (Fig. 4). The patients reported a high degree of satisfaction with Breast conservative surgery in combination with postopera- the surgical outcome in terms of improved breast shape, vol- tive radiation therapy has become the gold standard for ume and position, all of which were retained after radiation early-stage breast cancer.1,2 In select patients, the lumpec- therapy. Eighteen patients were either very satisfied or satis- tomy defect and adjuvant radiation therapy can cause sub- fied with their results, while 2 patients were disappointed (1 stantial breast deformity in shape, size and NAC position. owing to breast asymmetry between the oncoplastic-reduced Poor cosmetic results of BCT have been reported in breast and an implant-reconstructed contralateral breast 5%–40% of patients.10–14 The management of secondary [Fig. 4] and 1 owing to hypertrophic scarring on the healthy breast deformities from partial mastectomies can be chal- breast). None of the 18 women regretted having undergone lenging, particularly when operating in a radiated field, and the surgery. The independent observers’ evaluation of the increasing attention is being paid to long-term cosmetic 19 patients who completed follow-up was that most of the results. Immediate breast repair before adjuvant radiotherapy patients had a very good to good surgical outcome regarding has been shown by many studies to be oncologically safe and breast shape, NAC position and breast symmetry (Table 3). esthetically beneficial.3–7,10,14 The breast reduction pattern

Fig. 3. Schematic illustration of patients’ breast tumour locations (22 tumours in 20 patients). B

Table 2. Oncological data

Factor No. (%) or mean [range] Tumour type, n = 22 Malignant IDC 18 (82) Nonmalignant 4 (18) Tumour location, n = 22 Inferior pole 13 (59) Lateral pole 8 (36) Nipple/areola complex 1 (5) C Tumour size, cm; n = 22 3.5 [1–12] Tumour specimen weight mean, g 255 [50–600] Total tumour + reduction resection, g 534 [50–1265] Total contralateral breast reduction, g; n = 16 642 [50–1146] Neoadjuvant chemotherapy, n = 16 5 (31) Adjuvant chemotherapy, n = 16 8 (50) Malignant tumour staging, n = 16 T1, < 2 cm 8 (50) T2, 2–5 cm 6 (37.5) T3, > 5 cm 2 (12.5) Fig. 4. A 56-year-old patient with a history of left breast cancer and Tumour receptors, n = 16 current diagnosis of bilateral breast cancer (A). She underwent left Estrogen positive 10 mastectomy and immediate reconstruction with implant and right HER2-positive 5 lumpectomy with vertical scar superior-medial pedicle reduction pattern oncoplastic surgery (B). She underwent right revision sur- HER2 = human epidermal growth factor receptor 2; IDC = infiltrating ductal carcinoma. gery owing to breast asymmetry and nipple position (C).

E138 J can chir, Vol. 57, No 4, août 2014 RESEARCH technique has been described in various series as having a imaging is technically easier and more precise in reduced high success rate and good patient satisfaction, especially for breasts.10–23 Furthermore, removal of additional breast tissue those with macromastia and carcinoma of the breast.15–23 through reduction techniques allows examination of contra- The oncoplastic breast reduction technique for partial lateral breast tissue for occult breast lesions and theoretically mastectomy reportedly has numerous advantages: it permits makes sense in terms of reducing the risk of breast cancer. wider resection margins with a higher probability of nega- The reduction of the normal contralateral breast in paral- tive tumour margins; breast tissue rearrangement is done lel with the oncoplastic reduction pattern reconstruction using local tissue, with no other donor sites or foreign body results in smaller-sized breasts that are aesthetically more materials, obliterating the lumpectomy tissue dead-space; pleasing, have better symmetry and provide relief from back the reduced breast has better radiation-field efficiency and and neck pain for patients with large, heavy and pendulous less radiation fibrosis during radiation therapy compared breasts.10–23 There are various management algorithms and with larger breasts; and long-term breast surveillance approaches for reduction pattern oncoplastic surgery, including different skin reduction patterns, NAC pedicles 15–23 Table 3. Outcome evaluation by independent observers* and breast tissue rearrangement. Our experience with the vertical scar superior-medial pedicle reduction pattern No. of Outcome parameter; score range patients Mean (range) approach to oncoplastic breast surgery was highly reward- Breast shape 3.1 (2.3–3.8) ing. This technique is based on the Hall-Findlay vertical 24 Very good (n = 4) to good (n = 3) 14 reduction mammaplasty and has several advantages over Good (n = 3) to satisfactory (n = 2) 5 other reduction pattern oncoplastic techniques. It is rela- Satisfactory (n = 2) to poor (n = 1) 0 tively simple and has a short learning curve. It involves a Nipple/areola complex position 3.1 (2.4–3.6) straightforward glandular resection and shorter skin inci- Very good (n = 4) to good (n = 3) 15 sions, resulting in a shorter duration of surgery.24 Further- Good (n = 3) to satisfactory (n = 2) 4 more, the superior-medial pedicle offers a reliable NAC for Satisfactory (n = 2) to poor (n = 1) 0 different breast sizes as well as versatility for different Breast symmetry 2.9 (2.0–3.8) tumour locations and tissue rearrangement.24–26 The tumour Very good (n = 4) to good (n = 3) 14 location in this series was in the inferior and lateral poles of Good (n = 3) to satisfactory (n = 2) 4 the breast, with 1 patient having a tumour under the NAC Satisfactory (n = 2) to poor (n = 1) 1 (Fig. 1). The vertical scar we used has low skin vascular *5 observers. compromise, sparing the long inframammary horizontal scar of the traditional wise pattern. Long-term follow-up A findings demonstrated high patient satisfaction as well as high scoring for breast shape, NAC position and breast symmetry by independent observers (Table 3). Local recurrence is an important consideration in oncoplastic surgery. In our series, the average tumour specimen weighed 255 g, compared with institutional norms of about 40–50 g with the nononcoplastic approach, thus reducing the risk for local recurrence.4 All our patients had tumour-free surgical margins; there was no need to widen the margins in any of them. However, in B the event that there had been positive margins, they could have been managed either by completion mastectomy and reconstruction or re-excision, depending on a variety of patient- and surgeon-related factors and preferences as well as pathological findings. There is minimal downside to conversion to mastectomy and reconstruction. The skin is spared and immediate reconstruction is performed. Early postoperative complications were limited to partial dehiscence of the vertical scar and lateral fat necrosis in 1 patient that was successfully treated conservatively. The Fig. 5. A 39-year-old patient with infiltrating ductal carcinoma vertical scar was closed secondarily before radiation therapy (IDC) of her right breast located at the upper-lateral pole (A). She was initiated. There was no incidence in which adjuvant underwent right lumpectomy with vertical scar superior-medial pedicle reduction pattern oncoplastic surgery and left breast treatment was delayed due to surgical complications. Two reduction. Postoperative picture (B) 1 year after surgery and radi- patients underwent revision surgery: 1 for periareolar scars ation therapy to the right breast. and another for rereduction of the oncoplastic breast. No

Can J Surg, Vol. 57, No. 4, August 2014 E139 RECHERCHE

5. Giacalone PL, Roger P, Dubon O, et al. Comparative study of the revision surgery was needed as a consequence of postopera- accuracy of breast resection in oncoplastic surgery and quadrantec- tive radiation therapy or owing to breast shape (e.g., fat tomy in breast cancer. Ann Surg Oncol 2007;14:605-14. necrosis, breast fibrosis) or symmetry. Because the shape of . 6 Kaur N, Petit JY, Rietjens M, et al. Comparative study of surgical the breast is generally preserved, cases of postradiation margins in oncoplastic surgery and quadrantectomy in breast cancer. asymmetry can be treated with minor adjustments to the Ann Surg Oncol 2005;12:539-45. . 7 Asgeirsson KS, Rasheed T, McCulley SJ, et al. Oncological and contralateral nonradiated breast rather than by reconstruct- cosmetic outcomes of oncoplastic breast conserving surgery. Eur J ing a deformity in the radiated breast. Surg Oncol 2005;31:817-23. . 8 Losken A, Hamdi M. Partial breast reconstruction: current per- Limitations spectives. Plast Reconstr Surg 2009;124:722-36. . 9 Kronowitz SJ, Kuerer HM, Buchholz TA, et al. A management There are several limitations associated with the vertical algorithm and practical oncoplastic surgical techniques for repairing partial mastectomy defects. Plast Reconstr Surg 2008;122:1631-47. scar superior-medial pedicle reduction pattern technique. 10. Clough KB, Lewis JS, Couturaud B, et al. Oncoplastic techniques Tumours located in the superior and medial poles require allow extensive resections for breast-conserving therapy of breast modifying the NAC pedicle to an inferiorly or laterally carcinomas. Ann Surg 2003;237:26-34. based pedicle. Furthermore, the vertical scar limits skin 11. Kronowitz SJ, Feledy JA, Hunt KK, et al. Determining the optimal resection in the vertical aspect, leaving skin puckers, rip- approach to breast reconstruction after partial mastectomy. Plast Reconstr Surg 2006;117:1-11. pling and a mid-inferior dog-ear that can potentially cause 12. Anderson BO, Masetti R, Silverstein MJ. Oncoplastic approaches to scar dehiscence and takes weeks to straighten out and partial mastectomy: an overview of volume-displacement tech- improve. Breast tissue rearrangement can cause internal niques. Lancet Oncol 2005;6:145-57. scar tissue and fat necrosis that sometimes require tissue 13. Huemer GM, Schrenk P, Moser F, et al. Oncoplastic techniques sampling to rule out recurrent cancer. However, overall allow breast-conserving treatment in centrally located breast can- cers. Plast Reconstr Surg 2007;120:390-8. we found the technique to be safe and effective without 14. Rietjens M, Urban CA, Rey PC, et al. Long-term oncological results 15 significantly affecting postoperative cancer surveillance. of breast conservative treatment with oncoplastic surgery. Breast 2007;16:387-95. Conclusion 15. Losken A, Schaefer TG, Newell M, et al. The impact of partial breast reconstruction using reduction techniques on postoperative cancer surveillance. Plast Reconstr Surg 2009;124:9-17. Our experience was that the vertical scar superior-medial 16. Losken A, Styblo TM, Carlson GW, et al. Management algorithm pedicle reduction pattern was a simple, reliable and highly and outcome evaluation of partial mastectomy defects treated using versatile technique with the other recognized benefits of reduction or mastopexy techniques. Ann Plast Surg 2007;59:235-42. reduction pattern oncoplastic surgery. It was associated 17. Kronowitz SJ, Hunt KK, Kuerer HM, et al. Practical guidelines for with tumour-free oncological margins, high patient satis- repair of partial mastectomy defects using the breast reduction technique in patients undergoing breast conservation therapy. faction and pleasing aesthetic results. Judicious patient Plast Reconstr Surg 2007;120:1755-68. selection, coordinated planning and meticulous intraopera- 18. Losken A, Elwood ET, Styblo TM, et al. The role of reduction tive management are the keys to favourable surgical out- mammaplasty in reconstructing partial mastectomy defects. Plast come. We encourage the education of patients and phys Reconstr Surg 2002;109:968-75. icians about the benefits of oncoplastic surgery for BCT. 19. Shestak KC, Johnson RR, Greco RJ, et al. Partial mastectomy and breast reduction as a valuable treatment option for patients with mac- Acknowledgements: We thank Esther Eshkol for editorial assistance. romastia and carcinoma of the breast. Surg Gynecol Obstet 1993; Competing interests: None declared. 177:54-6. 20. Newman LA, Kuerer HM, McNeese MD, et al. Reduction mam- Contributors: Y. Barnea, D. Barsuk, T. Menes, A. Zaretski, D. Leshem, moplasty improves breast conservation therapy in patients with J. Weiss, S. Schneebaum and E. Gur designed the study. Y. Barnea and macromastia. Am J Surg 2001;181:215-20. A. Inbal acquired the data, which Y. Barnea analyzed. Y. Barnea and A. Inbal 21. Spear SL, Pelletiere CV, Wolfe AJ, et al. Experience with reduction wrote the article, which all authors reviewed and approved for publication. mammaplasty combined with breast conservation therapy in the treatment of breast cancer. Plast Reconstr Surg 2003;111:1102-9. References 22. Jones JA, Pu LL. Oncoplastic approach to early breast cancer in women with macromastia. Ann Plast Surg 2007;58:34-8. 1. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a 23. Munhoz AM, Montag E, Arruda EG, et al. Critical analysis of reduc- randomized trial comparing total mastectomy, lumpectomy, and tion mammaplasty techniques in combination with conservative breast lumpectomy plus irradiation for the treatment of invasive breast surgery for early breast cancer treatment. Plast Reconstr Surg 2006;117: cancer. N Engl J Med 2002;347:1233-41. 1091-103. 2. Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up of 24. Hall-Findlay EJ. A simplified vertical reduction mammaplasty: a randomized study comparing breast conserving surgery with radical shortening the learning curve. Plast Reconstr Surg 1999;104:748-59. mastectomy for early breast cancer. N Engl J Med 2002;347:1227-32. 25. Orlando JC, Guthrie RH Jr. The superomedial dermal pedicle for 3. Baildam AD. Oncoplastic surgery of the breast. Br J Surg 2002;89: nipple transposition. Br J Plast Surg 1975;28:42-5. 532-3. 26. Munhoz AM, Montag E, Arruda EG, et al. Superior-medial dermo- . 4 Masetti R, Di Leone A, Franceschini G, et al. Oncoplastic tech- glandular pedicle reduction mammaplasty for immediate conserva- niques in the conservative surgical treatment of breast cancer: an tive breast surgery reconstruction: technical aspects and outcome. overview. Breast J 2006;12(Suppl 2):S174-80. Ann Plast Surg 2006;57:502-8.

E140 J can chir, Vol. 57, No 4, août 2014 REVIEW • REVUE

End-to-end ductal anastomosis in biliary reconstruction: indications and limitations

Beata Jabłonska, MD, PhD End-to-end ductal anastomosis is a physiologic biliary reconstruction that is commonly used in liver transplantation and less frequently in the surgical treatment of iatrogenic bile From the Department of Digestive Tract duct injuries. Currently, end-to-end ductal anastomosis is the biliary reconstruction of Surgery, Medical University of Silesia, choice for liver transplantation in most adult patients. In recent years, it has also been per- Katowice, Poland formed for liver transplantation in children and in select patients with primary sclerosing cholangitis. The procedure is also performed in some patients with iatrogenic bile duct Accepted for publication injuries, as it establishes physiologic bile flow. Proper digestion and absorption as well as Oct. 21, 2013 postoperative endoscopic access are possible in patients who undergo end-to-end ductal anastomosis. It allows endoscopic diagnostic and therapeutic procedures in patients fol- Correspondence to: lowing surgery. This anastomosis is technically simple and associated with fewer early B. Jabłonska postoperative complications than the Roux-en-Y hepaticojejunostomy; however, end-to- Department of Digestive Tract Surgery end ductal anastomosis is not possible to perform in all patients. This review discusses the University Hospital of the Medical indications for and limitations of this biliary reconstruction, the technique used in liver University of Silesia transplantation and surgical repair of injured bile ducts, suture types and use of a T-tube. Medyków 14 St, 40-752 Katowice, Poland [email protected] L’anastomose termino-terminale du canal biliaire est la technique de reconstruction biliaire physiologique la plus couramment utilisée lors de la greffe du foie; elle est moins souvent utilisée pour le traitement chirurgical des blessures iatrogènes affectant le canal DOI: 10.1503/cjs.016613 biliaire. À l’heure actuelle, l’anastomose termino-terminale est la reconstruction biliaire privilégiée lors d’une transplantation hépatique chez la plupart des patients adultes. Ces dernières années, on y a également eu recours pour la greffe hépatique chez les enfants et dans certains cas de cholangite sclérosante. L’intervention est également effectuée chez certains patients présentant des traumatismes iatrogènes affectant le canal biliaire, puisqu’elle permet la circulation physiologique de la bile. Une digestion et une absorption adéquates, de même qu’un accès endoscopique postopératoire sont donc possibles chez les patients qui subissent une anastomose termino-terminale. Elle facilite les interventions diagnostiques et thérapeutiques endoscopiques chez les patients après la chirurgie. Cette anastomose est simple au plan technique et associée à moins de complications durant la période postopératoire immédiate comparativement à l’hépa ticojéjunostomie Roux en Y. Toutefois, l’anastomose termino-terminale n’est pas réali sable chez tous les patients. La présente analyse aborde les indications et les limites de cette reconstruction biliaire, la technique utilisée lors de la greffe hépatique et lors de la réparation chirurgicale des canaux biliaires lésés, les types de sutures et l’utilisation d’un tube en T.

nd-to-end ductal anastomosis is a physiologic biliary reconstruction that is commonly used in liver transplantation and in general surgery, including E the treatment of iatrogenic bile duct injuries (IBDI). End-to-end ductal anastomosis and Roux-en-Y hepaticojejunostomy (HJ) are the 2 most common biliary reconstructions, and the former is the most common in patients who have had liver transplantation, including those with primary sclerosing cholangitis (PSC). In recent years, the traditional method of HJ has been challenged by end- to-end biliary reconstruction in these patient groups; however, in patients with IBDI, HJ is performed most frequently. End-to-end ductal anastomosis has many advantages: it is physiologically simpler and associated with fewer early postoperative complications than HJ. End-to-end ductal anastomosis establishes physiologic bile flow; therefore, proper digestion and absorption are possible following this procedure. Postop- erative endoscopic access is also possible, facillitating different diagnostic and therapeutic procedures. Despite its advantages, it is not possible to perform end-to-end ductal anastomosis in all patients.1–5

end-jablon.indd 271 14-07-16 10:53 AM REVUE

The aim of this paper was to present the use of end-to- significantly lower incidence of leakage and a higher inci- end ductal anastomosis in patients undergoing liver trans- dence of stricture. However, 74.5% of the stricture was plantation and gastrointestinal surgery. This review also managed with endoscopic treatment. It should be empha- discusses the limitations in using this biliary reconstruction sized that in recent years, the traditional method of HJ has method and describes a surgical technique of end-to-end been challenged by end-to-end ductal anastomosis biliary ductal anastomosis. reconstruction even in patients undergoing liver transplantation owing to PSC. Damrah and colleagues9 compared HJ End-to-end ductal anastomosis in liver and end-to-end ductal anastomosis after liver transplanta- transplantation tion in patients who had PSC. They used end-to-end ductal anastomosis when the recipient’s common bile duct was free Biliary anastomosis is referred to as the Achilles’ heel of liver of gross disease. Morbidity, mortality, disease recurrence transplantation. The noted incidence of biliary complica- and graft and patient survival were comparable between the tions is 5%–15% after deceased donor liver transplantation groups. Based on these results, the authors recommended (DDLT) and 20%–34% after right-lobe live donor liver end-to-end ductal anastomosis for select patients with PSC transplantation (LDLT).6 Nowadays, different methods of as the first option for reconstruction. Similar results have biliary reconstruction are used: Roux-en-Y HJ, end-to-end been presented in other studies.10–12 ductal anastomosis and side-to-side ductal anastomosis. Other biliary reconstructions, such as cholecystoduodenos- End-to-end ductal anastomosis in pediatric liver tomy, cholecystojejunostomy or the gallbladder conduit transplantation technique, were used in the early experience of liver transplantation, but they were associated with a high risk (up to Currently, end-to-end ductal anastomosis is the biliary 70%) of septic complications. Also, many gallbladders had reconstruction of choice in adults. Its pediatric feasibility has to be removed because of cystic duct obstruction. Subse- rarely been reported. Tanaka and colleagues13 compared quently, HJ was introduced as the standard technique in 14 patients who underwent end-to-end ductal anastomosis liver transplantation, and it was a common method of biliary and 46 patients who underwent HJ; the incidence of biliary reconstruction for a long time. End-to-end ductal anasto- leakage was 7.1% and 8.7%, respectively, and that of stric- mosis was performed as the standard biliary reconstruction ture was 28.6% and 10.9%, respectively, but the differences after DDLT, whereas HJ was the standard technique per- were not significant. The authors observed that, compared formed after LDLT. This trend has changed because of the with the HJ group, biliary stricture in the end-to-end ductal disadvantages of HJ: longer duration of surgery and higher anastomosis group tended to require revision surgery with risk of bacterial contamination due to construction of the HJ and longer treatment with percutaneous transhepatic Roux-Y limb. Moreover, the re-established bilioenteric con- biliary drainage. Based on these results, the authors recom- tinuity is not physiologic and does not allow endoscopic mended HJ as the preferable reconstruction in children. access after liver transplantation. Currently, end-to-end duc- They recommended that end-to-end ductal anastomosis tal anastomosis is the standard biliary reconstruction for should be considered when making a new Roux-Y limb is both DLDT and LDLT in adults. This method is prefera- impossible or troublesome owing to abdominal dense adhe- ble because of an intact sphincter Oddi that can prevent sep- sion or short bowel syndrome. Liu and colleagues14 analyzed tic cholangitis due to ascending infections. Moreover, the results of end-to-end ductal anastomosis in 7 children procedure facillitates subsequent endoscopic diagnostic and undergoing LDLT using a left-lobe graft. The authors con- therapeutic procedures in patients with biliary complications cluded that end-to-end ductal anastomosis biliary recon- after liver transplantation. However, end-to-end ductal struction without external stent tube in patients undergoing anastomisis with a small duct (< 4 mm in diameter) is associ- left-lobe LDLT was feasible in a select group of children ated with a higher risk of biliary strictures than HJ.4,7,8 with normal extrahepatic bile ducts. In smaller recipients with larger grafts, the use of a transanastomotic biliary tube Hepaticojejunostomy versus end-to-end ductal could prevent anastomotic kinking, although the authors anastomosis in liver transplantation suggested HJ as a better method of biliary reconstruction for this condition. Other studies have also confirmed the There are a number of studies comparing HJ and end-to- usefulness of end-to-end ductal anastomosis for liver trans- end ductal anastomosis in liver transplantation in the litera- plantation in select pediatric patients.15–18 ture. Kasahara and colleages5 compared different biliary reconstructions in 321 recipients of right lobe LDLT. Bili- End-to-end ductal anastomosis in the surgical ary reconstruction was performed with HJ in 121 patients, treatment of iatrogenic bile duct injuries end-to-end ductal anastomosis in 192 patients, and combined HJ and end-to-end ductal anastomosis in 8 patients. Iatrogenic bile duct injuries are still an important problem They found that end-to-end ductal anastomosis showed a in gastrointestinal surgery. Noninvasive, percutaneous

272 J can chir, Vol. 57, No 4, août 2014

end-jablon.indd 272 14-07-16 10:53 AM REVIEW

radiological end endoscopic techniques are recommended available to repair the damage. Reuver and colleagues21 rec- as initial treatment of IBDI. When these techniques are not ommendeded end-to-end ductal anastomosis in patients effective, surgical management is considered. The goal of with injuries detected preoperatively when there was not surgical treatment is to reconstruct the proper bile flow to extensive tissue loss. In patients with extensive tissue loss, the alimentary tract. The long-term results depend on the particularly in those with more proximal injuries within the type of biliary reconstruction performed. Different biliary hepatic bifurcation or intrahepatic lesions, the authors rec- reconstructions have been reported in the surgical treat- ommend no primary repair. Kohneh and colleagues25 ment of IBDI: Roux-en-Y HJ, end-to-end ductal biliary achieved better results with end-to-end ductal anastomosis anastomosis, choledochoduodenostomy, Lahey HJ, jejunal (100%) than with HJ (71.4%) during early repair procedures interposition hepaticoduodenostomy, Blumgart (Hepp) (< 30 d after the initial trauma). They performed end-to-end anastomosis, Heinecke–Mikulicz biliary plastic reconstruc- ductal anastomosis in patients with bile duct injuries classi- tion and Smith mucosal graft.2,19,20 fied as type II (Bismuth) or E2 (Strasberg). In the Depart- ment of Digestive Tract Surgery, Katowice, Poland, end-to- Hepaticojejunostomy versus end-to-end ductal end ductal anastomosis reconstruction was performed when anastomosis in the surgical treatment of IBDI bile duct loss was 0.5–4 cm. Excision of the bile duct stricture, dissection and refreshing of the proximal and distal Currently, Roux-en-Y HJ is the most common surgical stumps as far as the tissues are healthy and without in- reconstruction of IBDI.1,2 Most authors have reported a pref- flammation, andthe use of nontraumatic, monofilament- erence for HJ owing to the lower number of postoperative interrupted sutures 5–0 yielded good long-term results com- anastomosis strictures with HJ than with end-to-end ductal parable to the results achieved with HJ. Recurrent stricture anastomosis. The latter procedure is seldom performed in was observed in 5.3% of patients after HJ and 9.6% after patients with IBDI because of a higher incidence of postop- end-to-end ductal anastomosis.1 Another study revealed that erative anastomosis strictures (up to 80%) compared with quality of life was also comparable after HJ and end-to-end HJ.21 However, after HJ, bile flow into the alimentary tract is ductal anastomosis. Moreover, it should be emphasized that not physiologic because the duodenum and upper part of the physical functioning was significantly better in patients who jejunum are excluded from bile passage. Roux-en-Y HJ is underwent end-to-end ductal anastomosis than in those who associated with different disturbances in the release of gastro- underwent HJ.3 Another essential advantage of end-to-end intestinal hormones leading to maldigestion and malabsorp- ductal anastomosis is the possibility of control endoscopic tion.1,2,22,23 Significantly lower weight gain in patients who examination and therapeutic procedures in patients after bili- had HJ than in those who had end-to-end ductal anastomosis ary reconstruction. End-to-end ductal anastomosis strictures was observed in a previous study.1 Moreover, a higher num- can be easily dilated endoscopically in contrast to HJ. Fewer ber of duodenal ulcers has been observed in patients under- early complications have been observed after end-to-end going HJ, and this may be associated with a loss of the neu- ductal anastomosis than HJ; the complications were associ- tralizing effect of the bile, including bicarbonates and the ated with opening of the alimentary tract and a higher num- secondary gastric hypersecretion. Control endoscopic exam ber of performed anastomoses (biliary-enteric and entero- ination and endoscopic dilatation of strictured biliary anasto- enteric) in patients who underwent HJ.1 mosis is not possible after HJ.1,2 End-to-end ductal anasto- It should be noted that end-to-end ductal anastomosis has mosis should be considered the treatment of choice in select some limitations and cannot be performed in patients with all patients with IBDI because it is a more physiologic proced bile duct injuries; it is not possible to perform the procedure ure than HJ; however, HJ should be considered in patients in in patients with complex vasculobiliary injuries. According to whom end-to-end ductal anastomosis is not possible.1 Strasberg and Helton,26 a vasculobiliary injury (VBI) is an It has been shown that good long-term results can be injury to both a bile duct and a hepatic artery and/or portal achieved in a select group of patients following end-to-end vein; the bile duct injury can be caused by surgical trauma, be ductal anastomosis. Gazzaniga and colleagues24 performed ischemic in origin or both, and can or cannot be accompa- end-to-end ductal anastomosis in the immediate repair pro- nied by various degrees of hepatic ischemia. Injury of a right cedures only when the injury did not exceed one-third of the hepatic artery (RHA) is the most frequent type of VBI. duct circumference and was not located more than 2 cm There are contradictory reports regarding the association below the ductal confluence (Strasberg E2), or when injury between the outcome of bile duct injuries and RHA injuries was detected during the primary operation. In this series, in the literature. Strasberg and Helton26 reviewed studies on injuries were type E2 in 18 patients, type E3 in 29 patients, VBI. Koffron and colleagues27 reported an associated injury and type E4 in 15 patients. Direct repair is not recom- of the artery in 61% of patients with recurrent strictures after mended when more than one-third of the bile duct circum- primary bile duct repair. Schmidt and colleagues28 reported ference is injured. It cannot be carried out when the lesion that the presence of combined vascular and bile duct injuries involves the bifurcation of 1 or both hepatic ducts (Strasberg and injury at or above the level of the biliary bifurcation were E3/E4). In such cases a Roux-en-Y HJ is the only procedure significant independent predictors of poor outcome in

Can J Surg, Vol. 57, No. 4, August 2014 273

end-jablon.indd 273 14-07-16 10:53 AM REVUE

patients undergoing Roux-en-Y HJ. Madariaga and col- to be comparable. In the Department of Digestive Tract leagues29 described early necrosis of a biliary anastomosis Surgery, if there was a difference between a diameter of anas- requiring right hepatic lobectomy in the presence of an RHA tomosed ends, the narrower end was incised longitudinally in injury. Sarno and colleagues30 noted that patients with con- the anterior surface to extend it. End-to-end ductal anasto- comittant VBI had worse outcomes after bile duct injury mosis repair was not carried out in bile ducts that were too repair. In contrast to the aforementioned studies, Alves and narrow (diameter < 4 mm). The approximating of both ends colleagues31 reported comparable incidence of postoperative is possible because of a wide Kocher manoeuvre (mobiliza- complications in patients with and without arterial injury. tion of the pancreatic head with the descending, horizontal Stewart and colleagues32 did not report any influence of and ascending part of the duodenum out of the peritoneum). RHA injury on long-term results following biliary recon- Patients undergoing a first or, exceptionally, second bile duct struction, but RHA injury was associated with a higher inci- repair can be a candidate for end-to-end ductal anastomosis. dence of postoperative abscess, bleeding, hemobilia, hepatic Hepaticojejunostomy should be performed in patients who ischemia, and the need for hepatic resection. Results of RHA do not satisfy the aforementioned criteria.1 injury and vasculobiliary injury involving both RHA and bile duct are different. It is associated with the arterial blood sup- Suture type ply of the extrahepatic biliary tract. In an injury to the RHA without biliary injury, occlusion Both continuous (CS) or interrupted (IS) and absorbable of the RHA results in ischemia of the right liver, but blood (polydioxanone) or nonabsorbable (prolene or polypropy flow is restored by preformed collateral arterial shunts. In a lene), 5–0, 6–0 or 7–0 sutures are used for end-to-end duc- combined vasculobiliary injury involving the RHA, E1–3 tal anastomosis in patients undergoing liver transplanta- injuries leave the hilar shunt (hilar plexus) open but obstruct tion.4,14,15,35,36 Initially, IS was the standard for these the longitudinal shunt (axial arteries at 3, 9 and 12 o’ clock) patients; CS was not adopted for end-to-end ductal anasto- and may induce greater hepatic ischemia than RHA occlusion mosis owing to concern for higher stricture rates than IS. only, and E4 injuries induce greater ischemia than right Continuous sutures are quicker to perform than IS. hepatic injuries alone by obstructing the important hilar shunt Castaldo and colleagues35 compared CS and IS for end-to- and the longitudinal shunt. Therefore, it is not possible to end ductal anastomosis in patients undergoing liver trans- perform end-to-end ductal anastomosis in patients with com- plantation. The authors reported comparable results with plex vasculobiliary injuries that require Roux-en-Y HJ and, both surgical techniques. There was no difference in biliary frequently, hepatic hepatectomy or liver transplantation.26,30 complications, graft survival or patient survival between the analyzed groups. The overall biliary complication rate was Technique of end-to-end ductal anastomosis 15%. There was no difference in the proportion of leaks (CS 7.3% v. IS 8.5%) or strictures (CS 9.8% v. IS 5.1%) General principles between groups. The nontraumatic, monofilament- interrupted 5–0 suture is the technique of choice for end- Two main conditions must be met for proper healing of to-end ductal anastomosis in patients with IBDI.1 each biliary anastomosis. The anastomosed edges should be healthy; there should be no inflammation, ischemia or T-tube use fibrosis; and the anastomosis should be tension-free and properly vascularized.33 Dissection and refreshing of the The use of a T-tube in end-to-end ductal anastomosis proximal and distal stumps as far as the tissues are healthy remains controversial. There are contradictory reports in and without inflammation should be performed. However, the literature regarding the feasibility of biliary drainage for careful dissection is required to save intact axial arteries end-to-end ductal anastomosis in patients undergoing liver within a wall of the common bile and hepatic ducts.34 Bili- transplantation and those undergoing IBDI repair. The ary reconstruction should be performed when no active advantage of biliary drainage is to limit the inflammation inflammation process is present, particularly in patients and fibrosis that occur after the surgical procedure. There- with IBDI, who frequently have ischemia, fibrosis and fore, some authors believe that the presence of the biliary inflammation within the bile ducts.1 Ischemia, either associ- tube prevents anastomosis stricture.1,28 The disadvantage is ated with graft preservation injury or inflammation due to the higher risk of postoperative complications.1 Scatton and rejection, has also been observed during liver transplanta- colleagues37 compared the incidence of biliary complication.35 Both proximal and distal ductal stumps should be tions after liver transplantation in patients undergoing end- dissected and approximated without tension. End-to-end to-end ductal anastomosis with or without T-tube in a large ductal anastomosis could be recommended for patients multicentre, prospective, randomized trial. The study when the maximal length loss of the bile duct is 4 cm. The included 108 patients divided into 2 groups: patients with sutured ends have to be healthy and without inflammation (n = 90) or without (n = 90) a T-tube who underwent sur- and ischemia. The diameter of both anastomosed ends has gery in 6 French liver transplantation centres. The authors

274 J can chir, Vol. 57, No 4, août 2014

end-jablon.indd 274 14-07-16 10:53 AM REVIEW

reported an increased biliary complication rate in the that longer duration of biliary drainage did not provide a T-tube group, that was linked to minor complications. The greater advantage.2,43 The 2 main types of biliary drainage incidence of biliary fistula was 10% in the T-tube group using T-tube can be distinguished: external T-drainage and 2.2% in the group without a T-tube. Therefore, the (Fig. 1A) and internal Y-drainage (Fig. 1B). External authors did not recommend the performance of end-to-end T-drainage involves using a typical T-tube with insertion of ductal anastomosis with a T-tube in patients undergoing its short branches into the bile duct and conducting of its long liver transplantation. Recently, López-Andújar and col- branch through the abdominal wall outside. It can be leagues38 compared the incidence and severity of biliary removed percutaneously after healing of the end-to-end duc- complications due to liver transplantation after end-to-end tal anastomosis. Internal Y-drainage involves insertion of ductal anastomosis with or without a T-tube in a single- short branches of the T-tube into both the right and left centre, prospective, randomized trial. The study involved hepatic ducts, splinting of the anastomosis and conducting of 95 patients with a T-tube and 92 patients without a T-tube. its long branch into the duodenum by the papilla of Vater. Significantly fewer anastomotic strictures were reported in This drainage can be removed endoscopically after healing of the T-tube group (n = 2 [2.1%]) than in the non-T-tube the end-to-end ductal anastomosis. It should be emphasized group (n = 13 [14.1%]). No difference in anastomotic bili- that the internal Y-drainage is less traumatic (does not involve ary leakage was observed between the groups. The authors additional incision of the bile duct wall) than the external concluded that complications in the T-tube group were less T-drainage. Therefore, it should be recommended as the severe and required less aggressive treatment than those in drainage of choice in end-to-end ductal anastomosis.1,2,24 the non-T-tube group. The incidence of anastomotic strictures was higher in patients without T-tubes. The authors Complications of end-to-end ductal anastomosis recommended using a rubber T-tube for end-to-end ductal anastomosis during liver transplantation in risky anastomo- An anastomostic fistula and stenosis are the 2 common sis and when the bile duct diameter is less than 7 mm. Con- postoperative complications following end-to-end ductal tradictory meta-analyses regarding the usefulness of a anastomosis. In patients who have had end-to-end ductal T-tube in end-to-end ductal anastomosis can also be found anastomosis, endoscopic control and treatment of these in the literature. Sotiropoulos and colleagues39 pooled the complications are possible. In anastomotic leakages and outcomes of 1027 patients undergoing end-to-end ductal strictures, endoscopic retrograde cholangiopancreatog anastomosis with or without T-tube in 9 of 46 screened raphy (ERCP) with stenting or stricture balloon dilatation trials by means of fixed- or random-effects models. In this is the first-line treatment. Percutaneous transhepatic bili- meta-analysis, the patients without T-tubes had fewer epi- ary drainage can also be performed.6 Yoshiya and col- sodes of cholangitis and peritonitis, and they demonstrated leagues44 described the use of rendezvous ductoplasty to a favourable trend for fewer overall biliary complications. treat biliary anastomotic stricture after LDLT. Biliary Anastomotic bile leaks or fistulas, end-to-end ductal anasto- anastomotic stricture was classified according to ERCP mosis revisions, dilatation and stenting, hepatic artery findings after normal pressure contrast injection: type I n( = thromboses, retransplantation and death due to biliary 32) in which the stricture was visualized; type II (n = 13) in complications were comparable in between the groups. which the common hepatic duct and graft intrahepatic Therefore, the authors did not recommend the use of a T-tube for end-to-end ductal anastomosis in patients undergoing liver transplantation. In contrast, Huang and A B colleagues40 reviewed 5 randomized control trials (RCTs) and 8 comparative studies. They suggested that the insertion of a T-tube reduced the incidence of biliary stenosis without increasing the incidence of other biliary complications. Based on these results, the use of a T-tube for end- to-end ductal anastomosis in patients undergoing liver transplantation could be recommended. The use and duration of biliary drainage in patients with IBDI is controversial. The advantage of biliary drainage is limitation of the inflammation and fibrosis occurring after the surgical procedure. In some authors’ opinions, the presence of the biliary tube prevents anastomosis stricture.41 The disadvantage of biliary drainage is a higher risk of postoperative complications.2,42 The duration of drainage is also controversial. According to most authors, the optimal duration Fig. 1: Types of biliary drainage using T-tube. (A) External for biliary drainage is about 3 months. Investigations showed T-drainage. (B) Internal Y-drainage.

Can J Surg, Vol. 57, No. 4, August 2014 275

end-jablon.indd 275 14-07-16 10:53 AM REVUE

9. Damrah O, Sharma D, Burroughs A, et al. Duct-to-duct biliary ducts were visualized, but the stricture was not visualized; reconstruction in orthotopic liver transplantation for primary scleros- or type III (n = 8) in which the stricture and graft intrahe- ing cholangitis: a viable and safe alternative. Transpl Int 2012;25:64-8. patic ducts were not visualized. The number of attempts to 10. Heffron TG, Smallwood GA, Ramcharan T, et al. Duct-to-duct pass the guidewire through the stricture was significantly biliary anastomosis for patients with sclerosing cholangitis under- lower in type I than type II or type III. The treatment suc- going liver transplantation. Transplant Proc 2003;35:3006-7. cess rate was 78.1% for type I, 38.5% for type II, and 11. Esfeh JM, Eghtesad B, Hodgkinson P, et al. Duct-to-duct biliary reconstruction in patients with primary sclerosing cholangitis 50.0% for type III. Rendezvous ductoplasty was the first undergoing liver transplantation. HPB (Oxford) 2011;13:651-5. successful treatment in a higher proportion of types II and 12. Schmitz V, Neumann UP, Puhl G, et al. Surgical complications and III patients than type I patients (66.7% vs. 6.3%). Cumula- long-term outcome of different biliary reconstructions in liver transplan- tive treatment success rates were not significantly different tation for primary sclerosing cholangitis-choledochoduodenostomy ver- between the rendezvous ductoplasty and the non–rendez- sus choledochojejunostomy. Am J Transplant 2006;6:379-85. vous ductoplasty groups. Hsieh and colleagues45 described 13. Tanaka H, Fukuda A, Shigeta T, et al. Biliary reconstruction in pediatric live donor liver transplantation: duct-to-duct or Roux-en-Y aggressive endoscopy-based treatment with maximal stent hepaticojejunostomy. J Pediatr Surg 2010;45:1668-75. placement that allowed 100% resolution of all biliary anas- 14. Liu C, Loong CC, Hsia CY, et al. Duct-to-duct biliary reconstruc- tomotic strictures after LDLT without the need for sur tion in selected cases in pediatric living-donor left-lobe liver trans- gical intervention or retransplantation. When less invasive plantation. Pediatr Transplant 2009;13:693-6. (using endoscopy and interventional radiology) treatment 15. Haberal M, Sevmis S, Emiroglu R, et al. Duct-to-duct biliary recon- is not successful, surgery is needed.36 struction in pediatric liver transplantation: one center’s results. Transplant Proc 2007;39:1161-3. 16. Shirouzu Y, Okajima H, Ogata S, et al. Biliary reconstruction for Conclusion infantile living donor liver transplantation: Roux-en-Y hepaticojejunostomy or duct-to-duct choledochocholedochostomy? Liver Transpl End-to-end ductal anastomosis is used for biliary recon- 2008;14:1761-5. struction in patients undergoing liver transplantation and 17. Okajima H, Inomata Y, Asonuma K, et al. Duct-to-duct biliary surgical repair of IBDI. The use of end-to-end ductal anas- reconstruction in pediatric living donor liver transplantation. Pediatr Transplant 2005;9:531-3. tomosis in patients undergoing liver transplantation is 18. Kimura T, Hasegawa T, Ihara Y, et al. Feasibility of duct-to-duct more common than in those undergoing surgical treat- biliary reconstruction in pediatric living related liver transplanta- ment of IBDI. The achievement of good long-term results tion: report of three cases. Pediatr Transplant 2006;10:248-51. is possible in patients undergoing both treatments. End-to- 19. Jabłonska B, Lampe P, Olakowski M, et al. Surgical treatment of iatro- end ductal anastomosis should be considered as the biliary genic bile duct injuries — early complications. Pol J Surg 2008;80: reconstruction of choice because it is more physiologic 299-305. than HJ and it is associated with fewer early postoperative 20. Jabłonska B, Lampe P, Olakowski M, et al. Long-term results in the surgical treatment of iatrogenic bile duct injuries. Pol J Surg complications. 2010;82:354-61. Competing interests: None declared. 21. de Reuver PR, Bush ORC, Rauws EA, et al. Long-term results of a primary end-to-end anastomosis in peroperative detected bile duct References injury. J Gastrointest Surg 2007;11:296-302. 22. Nielsen ML, Jensen SL, Malstrom J, et al. Gastryn and gastric acid 1. Jabłonska B, Lampe P, Olakowski M, et al. Hepaticojejunostomy vs. secretion in hepaticojejunostomy Roux-en-Y. Surg Gynecol Obstet end-to-end biliary reconstructions in the treatment of iatrogenic 1980;150:61-4. bile duct injuries. J Gastrointest Surg 2009;13:1084-93. 23. Imamura M, Takahashi M, Sasaki I, et al. Effects of the pathway of bile 2. Jabłonska B, Lampe P. Iatrogenic bile duct injuries: etiology, diagnosis flow on the digestion of FAT and the release of gastrointestinal hor- and management. World J Gastroenterol 2009;15:4097-104. mones. Am J Gastroenterol 1988;83:386-92. 3. Jabłonska B, Olakowski M, Lampe P, et al. Quality-of-life assessment 24. Gazzaniga GM, Filauro M, Mori L. Surgical treatment of iatrogenic in the treatment of iatrogenic bile duct injuries: hepaticojejunostomy lesions of the proximal common bile duct. World J Surg 2001;25:1254-9. versus end-to-end biliary reconstructions. ANZ J Surg 2012;82:923-7. 25. Kohneh Shahri N, Lasnier C, Paineau J. Bile duct injuries at laparo- 4. Ishiko T, Egawa H, Kasahara M, et al. Duct-to-duct biliary recon- scopic cholecystectomy: early repair results. Ann Chir 2005;130:218-23. struction in living donor liver transplantation utilizing right lobe graft. 26. Strasberg SM, Helton WS. An analytical review of vasculobiliary injury Ann Surg 2002;236:235-40. in laparoscopic and open cholecystectomy. HPB (Oxford) 2011;13:1-14. 5. Kasahara M, Egawa H, Takada Y, et al. Biliary reconstruction in 27. Koffron A, Ferrario M, Parsons W, et al. Failed primary manage- right lobe living-donor liver transplantation: comparison of different ment of iatrogenic biliary injury: incidence and significance of con- techniques in 321 recipients. Ann Surg 2006;243:559-66. comitant hepatic arterial disruption. Surgery 2001;130:722-8. 6. Wadhawan M, Kumar A, Gupta S, et al. Post-transplant biliary 28. Schmidt SC, Langrehr JM, Hintze RE, et al. Long-term results and complications — an analysis from a predominantly living donor risk factors influencing outcome of major bile duct injuries follow- liver transplant centre. J Gastroenterol Hepatol 2013;28:1056-60. ing cholecystectomy. Br J Surg 2005;92:76-82. 7. Neuhaus P, Blumhardt G, Bechstein WO, et al. Technique and results 29. Madariaga JR, Dodson SF, Selby R, et al. Corrective treatment and of biliary reconstruction using side-to-side choledochocholedochos- anatomic considerations for laparoscopic cholecystectomy injuries. tomy in 300 orthotopic liver transplants. Ann Surg 1994;219:426-34. J Am Coll Surg 1994;179:321-5. 8. Wang SF, Huang ZY, Chen XP. Biliary complications after living 30. Sarno G, Al-Sarira AA, Ghaneh P, et al. Cholecystectomy-related donor liver transplantation. Liver Transpl 2011;17:1127-36. bile duct and vasculobiliary injuries. Br J Surg 2012;99:1129-36.

276 J can chir, Vol. 57, No 4, août 2014

end-jablon.indd 276 14-07-16 10:53 AM REVIEW

31. Alves A, Farges O, Nicolet J, et al. Incidence and consequence of an hepatic artery injury in patients with postcholecystectomy bile duct strictures. Ann Surg 2003;238:93-6. CJS’s top viewed articles* 32. Stewart L, Robinson TN, Lee CM, et al. Right hepatic artery injury associated with laparoscopic bile duct injury: incidence, mechanism, and consequences. J Gastrointest Surg 2004;8:523-30. 1. Research questions, hypotheses and objectives 33. Jabłonska B, Lampe P. Reconstructive biliary surgery in the treatment Farrugia et al. of iatrogenic bile duct injuries. In: Brzozowski T, editor. New advances in the basic and clinical gastroenterology. Rijeka (HR): InTech; 2012:477-494. Can J Surg 2010;53(4):278-81 34. Jabłonska B. The arterial blood supply of the extrahepatic biliary 2. Complications associated with laparoscopic tract — surgical aspects. Pol J Surg 2008;80:336-42. sleeve gastrectomy for morbid obesity: 35. Castaldo ET, Pinson CW, Feurer ID, et al. Continuous versus a surgeon’s guide interrupted suture for end-to-end biliary anastomosis during liver transplantation gives equal results. Liver Transpl 2007;13:234-8. Sarkhosh et al. 36. Wojcicki M, Milkiewicz P, Silva M. Biliary tract complications after Can J Surg 2013;56(5):347-52 liver transplantation: a review. Dig Surg 2008;25:245-57. 3. Tracheostomy: from insertion to decannulation 37. Scatton O, Meunier B, Cherqui D, et al. Randomized trial of chole Engels et al. dochocholedochostomy with or without a T tube in orthotopic liver Can J Surg 2009;52(5):427-33 transplantation. Ann Surg 2001;233:432-7. 38. López-Andújar R, Montalvá Orón E, Frangi Carregnato A, et al. 4. Treatment of an infected total hip replacement T-tube or no T-tube in cadaveric orthotopic liver transplantation: with the PROSTALAC system the eternal dilemma: results of a prospective and randomized clinical Scharfenberger et al. trial. Ann Surg 2013;258:21-9. Can J Surg 2007;50(1):24-8 39. Sotiropoulos GC, Sgourakis G, Radtke A, et al. Orthotopic liver transplantation: T-tube or not T-tube? Systematic review and 5. Managing complications associated with meta-analysis of results. Transplantation 2009;87:1672-80. laparoscopic Roux-en-Y gastric bypass for 40. Huang WD, Jiang JK, Lu YQ. Value of T-tube in biliary tract morbid obesity reconstruction during orthotopic liver transplantation: a meta- Sahle Griffith et al. analysis. J Zhejiang Univ Sci B 2011;12:357-64. Can J Surg 2012;55(5):329-36 41. Pekolj J, Alvarez FA, Palavecino M, et al. Intraoperative management and repair of bile duct injuries sustained during 10,123 laparoscopic 6. Hardware removal after tibial fracture has healed cholecystectomies in a high-volume referral center. J Am Coll Surg Sidky and Buckley 2013;216:894-901. Can J Surg 2008;51(4):263-8 42. Robinson TN, Stiegmann GV, Durham JD, et al. Management of major bile duct injury associated with laparoscopic cholecystectomy. 7. Complications associated with adjustable gastric Surg Endosc 2001;15:1381-5. banding for morbid obesity: a surgeon’s guide 43. Lillemoe KD, Melton GB, Cameron JL, et al. Postoperative bile duct Eid et al. strictures: management and outcome in the 1990s. Ann Surg 2000;232 Can J Surg 2011;54(1):61-6 :430-41. 44. Yoshiya S, Shirabe K, Matsumoto Y, et al. Rendezvous ductoplasty 8. Blinding: Who, what, when, why, how? for biliary anastomotic stricture after living-donor liver transplanta- Karanicolas et al. tion. Transplantation 2013;95:1278-83. Can J Surg 2010;53(5):345-8 45. Hsieh TH, Mekeel KL, Crowell MD, et al. Endoscopic treatment of 9. All superior pubic ramus fractures are not anastomotic biliary strictures after living donor liver transplantation: outcomes after maximal stent therapy. Gastrointest Endosc 2013;77: created equal 47-54. Steinitz et al. Can J Surg 2004;47(6):422-5 10. Preoperative bowel preparation for patients undergoing elective colorectal surgery: a clinical practice guideline endorsed by the Canadian Society of Colon and Rectal Surgeons Eskicioglu et al. Can J Surg 2010;53(6):385-95

*Based on page views on PubMed Central of research, reviews, commentaries and discussions in surgery. Updated July 14, 2014.

Can J Surg, Vol. 57, No. 4, August 2014 277

end-jablon.indd 277 14-07-16 10:53 AM REVIEW • REVUE

Systematic review on the inclusion of patients with cognitive impairment in hip fracture trials: a missed opportunity?

Simran Mundi, BHSc, MSc Background: More than 320 000 hip fractures occur annually in North America. An (Cand) estimated 30% of this population have cognitive impairment. We sought to determine the extent to which patients with cognitive impairment or dementia have been Harman Chaudhry, MD included in randomized controlled trials (RCTs) assessing hip fracture management. Mohit Bhandari, MD, PhD Methods: We conducted a systematic search of 3 electronic journal databases of articles published between January 2000 and June 2010. Studies were screened in duplicate From the Division of Orthopaedic Sur- to collect English-language RCTs assessing operative interventions for femoral head, gery, Department of Surgery, McMaster neck or intertrochanteric fractures. We systematically collected descriptive data and University, Hamilton, Ont. used the χ2 test for comparison between groups as appropriate.

Poster presented at the Canadian Ortho- Results: We screened 1201 abstracts, 72 of which were eligible for inclusion in our paedic Residents Association annual review. Femoral neck and intertrochanteric fractures were equally represented. Thirty- meeting, the Canadian Orthopaedic three (46%) studies did not report the inclusion or exclusion of patients with cognitive Association annual meeting and the impairment. Nineteen (26%) studies explicitly included cognitively impaired patients, Orthopaedic Trauma Association (OTA) whereas 20 (28%) excluded them. Only 2 trials (3%) reported outcomes specific to annual meeting. cognitively impaired patients. Fourteen trials (19.4%) reported the use of a validated cognitive assessment tool. None of the trials that reported inclusion of cognitively Accepted for publication impaired patients were from North American centres. Nov. 11, 2013 Conclusion: One in 3 patients with hip fractures have concomitant cognitive impairment, yet 8 of 10 hip fracture trials excluded or ignored this population. The ambigu- Correspondence to: ity or exclusion of these patients misses an opportunity to study outcomes and identify M. Bhandari factors associated with improved prognosis. 293 Wellington St. N, Suite 110 Hamilton ON L8L 8E7 [email protected] Contexte : On dénombre plus de 320 000 fractures de la hanche chaque année en Amérique du Nord et on estime que 30 % de ces personnes ont une atteinte cognitive. Nous avons voulu déterminer dans quelle mesure les patients qui souffrent d’une DOI: 10.1503/cjs.023413 atteinte cognitive ou de démence ont été inclus dans les essais randomisés et contrôlés (ERC) portant sur la prise en charge de la fracture de la hanche. Méthodes : Nous avons procédé à une interrogation systématique de 3 bases de don- nées de journaux électroniques pour recenser les articles publiés entre janvier 2000 et juin 2010. Les études ont été passées en revue en parallèle pour dégager les ERC de langue anglaise ayant évalué des interventions chirurgicales pour fractures de la tête ou du col du fémur ou fractures intertrochantériennes. Nous avons recueilli les don- nées descriptives de manière systématique et utilisé le test du χ2 pour comparer des groupes entre eux, selon le cas. Résultats : Nous avons passé en revue 1201 résumés, dont 72 répondaient à nos critères d’admissibilité. Les fractures du col du fémur et intertrochantériennes étaient représentées en proportions égales. Trente-trois études (46 %) ne faisaient aucune mention de l’inclusion ou de l’exclusion des patients souffrant d’atteinte cognitive. Dix-neuf (26 %) études incluaient expressément des patients souffrant d’atteinte cognitive, tandis que 20 (28 %) les excluaient. Seulement 2 essais (3 %) ont fait état de résultats spécifiques aux patients souffrant d’atteinte cognitive. Quatorze essais (19,4 %) ont déclaré utiliser un outil d’évaluation cognitive validé. Aucun des essais ayant mentionné l’inclusion de patients souffrant d’atteinte cognitive ne provenait de centres nord-américains. Conclusion : Un patient victime d’une fracture de la hanche sur 3 souffrait concomi- tamment d’une atteinte cognitive et pourtant, 8 essais sur 10 portant sur la fracture de la hanche ont exclus ou ignoré cette population. L’ambiguïté vis-à-vis de ces patients ou leur exclusion est une occasion manquée d’étudier les paramètres et de relever les facteurs associés à un pronostic plus favorable.

ore than 320 000 hip fractures occur annually in Search strategy North America.1,2 As hip fracture is a condition M most common among elderly individuals, its We performed a systematic search of the medical litera- societal burden is expected to grow as the North American ture to identify all relevant RCTs published between population continues to age. By 2040, the number of indi- January 2000 and June 2010. Two investigators searched viduals older than 65 is forecasted to increase from 3 electronic medical databases (Medline, Embase, 34.8 million to 77.2 million, resulting in an annual hip PubMed) using the following search terms: “hip frac- fracture incidence of greater than 580 000.1 Dementia — a ture*” OR “femoral neck fracture*” OR “femoral head chronic form of cognitive impairment — is prevalent in fracture*” OR “intertrochanteric fracture*” OR “subcap- the elderly population as well, and co-occurrence of this ital fracture*” alongside appropriate database subject condition with hip fracture is not infrequent. By some esti- headings (i.e., MeSH, Emtree). Given the eligibility cri- mates, 30% of patients who sustain a hip fracture also have teria outlined, we placed the following limits on the cognitive impairment or dementia.3–5 searches: publication in the English language, RCT and There is early evidence to suggest that patients with publication in January 2000 or later. We included a sys- dementia typically experience poorer functional outcomes tematic PubMed search as a supplementary query to and increased morbidity and mortality following a hip ensure no pertinent trials were overlooked. This search fracture.6–15 Identifying strategies to optimize outcomes was done with limits to predetermined journals that were in hip fracture patients with dementia is therefore critic judged to be high-yield: The Journal of Bone & Joint Sur- ally important; however, the extent to which this issue is gery (American and British volumes), Clinical Orthopaedic addressed in orthopedic surgery randomized controlled and Related Research, Acta Orthopaedica and The Journal of trials (RCTs) has not been well-elucidated. Exclusion of Orthopaedic Trauma. patients with dementia from surgical RCTs could potentially undermine the applicability of trial results to this Data extraction sizeable subgroup. We conducted a systematic review to analyze the inclu- Standardized data extraction forms were developed a sion of patients with cognitive impairment and dementia priori. For each study, characteristics of the trial, in hip fracture RCTs conducted over the course of the including geographical location, sample size, number past decade. This information will provide an important of centres, mean patient population age and sex ratios, consideration to both clinicians managing hip fracture were recorded. Furthermore, we documented the frac- patients with dementia and researchers designing future ture type and operative interventions assessed, as well hip fracture RCTs. as the significance of the results of the primary outcome measure. Methods For each trial, we evaluated whether patients with dementia or other forms of cognitive impairment were We performed a systematic review of RCTs involving hip explicitly included or explicitly excluded. We also fracture operative treatments to determine the extent to recorded the cognitive assessment tool used to make a which patients with cognitive impairment or dementia diagnosis for inclusion or exclusion. If inclusion status were included. We used applicable components of the could not be ascertained based on the published manu- PRISMA 2009 checklist as a framework for this review. script or if no mention was made regarding the strategy for cognitive assessment, then we considered it “not Eligibility criteria and study selection reported.” Finally, for studies including patients with dementia, we determined whether a subgroup analysis Criteria for inclusion in this review were established a was performed and assessed the results of such analyses. priori, and all studies satisfied the following parameters: For studies excluding patients with dementia, we RCT study design; assessment of an operative interven- recorded if a rationale for exclusion was provided, and if tion for femoral head, femoral neck, or intertrochan- so, we noted the reason provided. teric fractures; publication in English; original publica- All data extraction was done in duplicate, and any dis- tion; and publication date between January 2000 and crepancies were resolved by consensus among the June 2010. reviewers. We used a 2-step review process to screen and select eligible trials. The first step entailed a review of all Statistical analysis titles and abstracts yielded by our search strategy. Studies meeting the inclusion criteria and those with We systematically collected descriptive data and used the equivocal eligibility were retrieved for full-text review χ2 statistical test for comparison between groups as appro- and data retrieval. priate. Our primary variable of interest was the number of

E142 J can chir, Vol. 57, No 4, août 2014 REVIEW studies that included patients with dementia compared American centres. There were no significant differences with those that excluded patients with dementia. between RCTs that included or excluded these patients in terms of patient age, number of centres, or operative pro- Results cedures compared. Of the 19 studies that included this patient population, 10 specified dementia as the form of Our search identified a total of 1201 studies published cognitive impairment, whereas the remaining 9 did not. between January 2000 and June 2010 for screening of titles Thirty-three studies (46%) failed to report the inclusion or and abstracts. Of the 1201 studies, 92 trials were deemed exclusion of patients with cognitive impairment from their potentially eligible and retrieved for full text review. Of trials (Fig. 2 and Table 2). these studies, 72 were included for final review (Fig. 1). Fourteen of 72 trials (19.4%) reported the use of a validated cognitive assessment tool. This included formal Study characteristics tests, such as the Mini-Mental State Exam (MMSE). A single additional trial used a cognitive assessment tool that The majority of studies were conducted in Europe (79%) was not validated. and involved a single centre (65%). Sample sizes ranged Of the 19 studies including patients with cognitive from 19 to 569 patients. An equivalent number of studies impairment, only 2 studies highlighted outcomes of this assessed the management of femoral neck fractures (n = population. The first study tested surgical interventions 36) and intertrochanteric fractures (n = 36). More than in cognitively impaired patients only, while the second half (51%) of the studies compared methods of internal conducted a subgroup analysis on this population. With fixation, 22% compared methods of arthroplasty, and respect to the 20 studies excluding patients with cognitive 15% compared arthroplasty to internal fixation. Studies reported significant findings 18% of the time (Table 1). Table 1. Characteristics of RCTs

Characteristic Studies, no. (%) Inclusion of patients with dementia Geographical location Europe 57 (79) Among the 72 RCTs included in this review, 19 studies North America 5 (7) included both cognitively intact and impaired patients, East Asia 5 (7) and 1 of these studies reported dementia or cognitive South Asia 3 (4) impairment as the focus of the paper. Other 2 (3) Nineteen studies (26%) explicitly included patients with No. of centres cognitive impairment and 20 studies (28%) explicitly Single centre 47 (65) excluded such patients, as stated in their methodology or as Multicentre 16 (22) evident in the paper. None of the RCTs that reported Not reported 9 (13) inclusion of cognitively impaired patients were from North Sample size < 50 10 (14) 50–100 21 (29) 100–150 19 (26) > 150 22 (31) Medline, Embase, PubMed Titles and abstracts screened Type of fracture n = 1201 Femoral head 0 Femoral neck 36 (50) Intertrochanteric 36 (50) Type of treatment Full manuscript review, n = 92 IF v. IF 37 (51) Arthroplasty 16 (22) HA v. HA 11 (15) THA v. HA 3 (4) Included, n = 72 Excluded, n = 20 THA v. THA 2 (3) • Duplicate publication, n = 2 Arthroplasty v. IF 11 (15) • Long-term follow-up of Other 8 (11) previously published RCT, n = 5 Significance of results • Not RCT or published in English, Significant 13 (18) n = 5 Not significant 14 (19) • Unable to access, n = 8 Mixed 45 (63)

HA = hip arthroplasty; IF = internal fixation; RCT = randomized controlled trial; THA = total hip Fig. 1. Systematic search strategy for article inclusion. RCT = arthroplasty. randomized controlled trial.

Can J Surg, Vol. 57, No. 4, August 2014 E143 REVUE impairment, only 6 (30%) attempted to provide a ration ture management. Furthermore, validated screening ale within the published manuscript for the exclusion of tools for cognitive impairment are rarely used in those such patients. Reasons were the patients’ inability to pro- studies that purport to explicitly include or exclude vide informed consent (1 study) and the aim of evaluating these patients. Finally, we were able to identify only outcomes in an active or a mentally competent subpopu- 2 trials that evaluated interventions specifically for lation (5 studies). patients with cognitive impairment. One of these studies included only patients with cognitive impair- Discussion ment, while another conducted a subgroup analysis for this patient population.10,16 Our systematic review evaluated 72 RCTs in an Previous literature has provided contrasting attempt to delineate the degree to which patients with results. In a systematic review of 17 RCTs conducted cognitive impairment and dementia are being incor over a period of 20 years, Herbert-Davies17 and col- porated into orthopedic trials on hip fracture manage- leagues found that 13 (76%) RCTs explicitly included ment. Our results indicate that patients with cognitive patients with dementia, while 4 (24%) explicitly impairment are seldom included (26%) and are rarely excluded this patient population. A possible explana- the focus (1%) of RCTs evaluating operative hip frac- tion for this discrepancy is that the authors evaluated only RCTs that reported sufficient data on the number of patients with dementia. Studies offering a quantitative description of patients with dementia are 26% certainly more likely to include such patients. Our review was more comprehensive to the extent that our analysis was based solely upon reporting of inclusion Dementia patients included 46% and exclusion status, irrespective of further quantita- Dementia patients excluded Inclusion not reported tive reporting. An assumption that outcomes are similar in patients with and without cognitive impairment is not supported by the evidence. For instance, Panula and col- 28% leagues11 reviewed the charts of 428 hip fracture patients in a Finnish hospital registry and correlated Fig. 2. Proportion of studies including, excluding, and not these to the official cause of death statistics in Finland. reporting patients with cognitive impairment or dementia. The investigators found that patients with dementia

Table 2. Inclusion and exclusion by study characteristics

Group; no.

Dementia Dementia Characteristic patients included patients excluded Not reported p value Geographical location Europe 19 15 23 0.020 North America 0 4 1 0.040 East Asia 0 1 4 0.32 South Asia 0 0 3 > 0.99 Other 0 0 2 > 0.99 Type of fracture Neck 10 14 12 0.27 Intertrochanteric 9 6 21 0.27 Operative intervention IF v. IF 11 6 20 0.08 Arthroplasty v. arthroplasty 2 6 8 0.13 Arthroplasty v. IF 5 5 1 0.93 Other 1 3 4 0.32 Results Significant 5 4 4 0.64 Not significant 4 3 7 0.62 Mixed 10 13 22 0.43

IF = internal fixation.

E144 J can chir, Vol. 57, No 4, août 2014 REVIEW

Competing interests: None declared. who sustained hip fractures had a more than 3-fold increased risk of death than those with dementia in the Contributors: All authors designed the study. S. Mundi and H. Chaudhry acquired the data, which all authors analyzed, S. Mundi general population. Similarly, in a chart review of and H. Chaudhry wrote the article, which M. Bhandari reviewed. All 495 hip fracture patients in the United States, Bentler authors approved the final version for publication. and colleagues18 demonstrated that patients with dementia were 45% more likely to die postinjury than References patients without dementia. Some early evidence indicates that patients with cogni- . 1 Schemitsch E, Bhandari M. Femoral neck fractures: controversies and evidence. J Orthop Trauma 2009;23:385. tive impairment may actually have different intervention- 2. Cooper C, Campion G, Melton LJ. Hip fractures in the elderly: a specific outcomes as well. An RCT performed by world-wide projection. Osteoporos Int 1992;2:285-9. Johannson and colleagues10 comparing total hip arthro- 3. Stenvall M, Berggren M, Lundstrom M, et al. A multidisciplinary plasty to internal fixation for hip fracture demonstrated intervention program improved the outcome after hip fracture for an inversion of outcomes among hip fracture patients people with dementia — subgroup analyses of a randomized controlled trial. Arch Gerontol Geriatr 2012;54:e284-9. with cognitive impairment. Specifically, the investigators 4. Lundström M, Olofsson B, Stenvall M, et al. Postoperative delirium found a 5% reoperation rate with internal fixation and a in old patients with femoral neck fracture: a randomized interven- 32% dislocation rate after arthroplasty in patients with tion study. Aging Clin Exp Res 2007;19:178-86. cognitive impairment. This pattern was reversed in cog- 5. Juliebø V, Krogseth M, Skovlund E, et al. Delirium is not associ- nitively intact patients, who experienced a 60% reopera- ated with mortality in elderly hip fracture patients. Dement Geriatr Cogn Disord 2010;30:112-20. tion rate after internal fixation and a 12% dislocation rate . 6 Samuelsson B, Hedstrom MI, Ponzer S, et al. Gender difference 10 after arthroplasty. Purposly studying patients with cog- and cognitive aspects on functional outcome after hip fracture — a nitive impairment would help identify such differences, 2 years’ follow up of 2,134 patients. Age Ageing 2009;38:686-92. thereby better informing orthopedic practice. . 7 Givens JL, Sanft TB, Marcantonio ER. Functional recovery after hip fracture: the combined effects of depressive symptoms, cognitive impairment, and delirium. J Am Geriatr Soc 2008;56:1075-9. Limitations . 8 Rogmark C, Johnell O. Primary arthroplasty is better than internal fixation of displaced femoral neck fractures: a meta-analysis of 14 Our study has several strengths. As mentioned, we used a randomized studies with 2,289 patients. Acta Orthop 2006;77:359-67. systematic search strategy to identify eligible studies and 9. Clayer MT, Bauze RJ. Morbidity and mortality following fractures applied this search across 3 medical databases to collect a of the femoral neck and trochanteric region: analysis of risk factors. J Trauma 1989;29:1673-8. large sample of 72 RCTs. Two reviewers extracted all 10. Johansson T, Jacobsson SA, Ivarsson I, et al. Internal fixation versus data independently and in duplicate. We were able to total hip arthroplasty in the treatment of displaced femoral neck capture a broad range of studies with respect to geo- fractures. Acta Orthop Scand 2000;71:597-602. graphic location, type of hip fracture and operative inter- 11. Panula J, Pihlajamaki H, Mattila VM, et al. Mortality and cause of vention. Unfortunately, our study did have the limitation death in hip fracture patients ages 65 or older — a population- based study. BMC Musculoskelet Disord 2011;12:105. of excluding 8 potentially relevant articles owing to inac- 12. Huusko TM, Karppi P, Avikainen V, et al. Randomised, clinically cessibility. Given our large sample size and the findings controlled trial of intensive geriatric rehabilitation in patients with of our study, it is unlikely that the inclusion of such stud- hip fracture: subgroup analysis of patients with dementia. BMJ ies would have substantially altered our results. 2000;321:1107-11. 13. Kyo T, Takaoka K, Ono K. Femoral neck fracture. Factors related to ambulation and prognosis. Clin Orthop Relat Res 1993; (292):215-22. onclusion C 14. Lieberman D, Fried V, Castel H, et al. Factors related to successful rehabilitation after hip fracture: a casecontrol study. Disabil Rehabil The ambiguity and outright exclusion of patients with 1996;18:224-30. cognitive impairment in RCTs challenges the apparent 15. Nightingale S, Holmes J, Mason J, et al. Psychiatric illness and external validity of these trials. The selection of certain mortality after hip fracture. Lancet 2001;357:1264-5. 16. Blomfeldt R, Tornkvist H, Ponzer S, et al. Internal fixation versus primary outcomes, such as patient-reported question- hemiarthroplasty for displaced fractures of the femoral neck in naires, may necessarily preclude inclusion of patients elderly patients with severe cognitive impairment. J Bone Joint Surg with cognitive impairment in an RCT. However, given Br 2005;87:523-9. the size of this subpopulation, we believe that exclusion 17. Hebert-Davies J, Laflamme GY, Rouleau D, et al. Bias towards without explanation is no longer acceptable. We pro- dementia: Are hip fracture trials excluding to many patients? A systematic review. Injury 2012;43:1978-84. pose a “call for inclusion” of patients with cognitive dys- 18. Bentler SE, Liu L, Obrizan M, et al. The aftermath of hip fracture: function to identify interventions that improve survival discharge placement, functional status change, and mortality. Am J and function in this patient population. Epidemiol 2009;170:1290-9.

Can J Surg, Vol. 57, No. 4, August 2014 E145 DISCUSSIONS IN SURGERY DISCUSSIONS EN CHIRURGIE

Technique to achieve the symmetry of the new inframammary fold

Marcello Pozzi, MD* Summary Giovanni Zoccali, MD*† The literature outlines several surgical techniques to restore inframmam- * mary fold definition, but symmetry of the fold is often left to irreproducible Ernesto Maria Buccheri, MD procedures. We report our personal technique to restore the symmetry of Roy de Vita, MD* the inframmammary fold during multistep breast reconstruction.

From the *Department of Plastic and Reconstructive Surgery, IFO - “Regina Elena” National Cancer Institute, Rome, n multistep breast reconstruction, the inframammary fold is often dis- Italy, and †L’Aquila University, Depart- torted as a result of imperfect tissue expansion.1,2 The literature outlines ment of Life, Health & Enviromental several surgical techniques to restore inframmammary fold definition, but Sciences, Plastic Reconstructive and I 3–5 Aesthetic Surgery Section. L’Aquila, symmetry of the fold is often left to irreproducible procedures. We report Italy our personal technique to restore the symmetry of the inframmammary fold during multistep breast reconstruction. Accepted for publication The surgeon begins by drawing the midline with the patient in an Nov. 7, 2013 upright position. The physiologic inframammary fold is pointed out- ward, and a perpendicular line is drawn from its lower point (point A) to Correspondence to: G. Zoccali the midline (point B). From point B, a second line is drawn until it Department of Plastic and Reconstructive reaches the lower point of the contralateral inframammary fold (point Surgery, IFO C). From point C, a perpendicular line to the midline is drawn. The sur- Regina Elena National Cancer Institute geon moves the new inframammary fold upward the exact distance Via Elio Chianesi 53, 00133 Rome, Italy between point B and the projection of point C on the midline (Fig. 1, [email protected] panel 1, x-distance). During this surgical procedure, an appropriate dermo-adipose flap must DOI: 10.1503/cjs.026913 be raised to extend the dissection beyond the future inframammary fold. Moving up the new inframammary fold along the x-distance, the CB segment becomes perpendicular to the midline, reaching the same height as the contralateral fold (Fig. 1, panel 2). In order to obtain adequate symmetry of the medial portion of the inframammary fold, the surgeon first draws the BA segment followed by a bisector line drawn from the inner corner (point B) toward the inframammary fold (i.e., the x-distance from point B to the inframammary fold). A new line, equal to the x-distance, is drawn perpendicular to the bisector line (Fig. 1, panel 3). During surgery, the new inframmammary fold has to be fixed at the lateral extreme of this segment, ensuring the same position as the contralateral fold (Fig. 1, panel 4). To ensure symmetry of the lateral part of inframmammary fold, the surgeon draws a line from point A to the anter ior axillary line (point D), creating the y-distance (Fig. 1, panel 5). A bisector line is drawn from point D toward the inframammary fold, creating the x-distance. On the contralateral side, a new segment is drawn from point C to the anterior axillary line (point D, y-distance), followed by a bisector line (x-distance; Fig. 1, panel 5). During surgery, the new inframammary fold lateral extreme must be fixed at the apex of the bisector line (Fig. 1, panel 6). The restoration of a well-defined fold during reconstructive or cosmetic surgery is a fundamental step toward an excellent result. It is

personal-zoccali.indd 278 14-07-16 12:08 PM DISCUSSIONS IN SURGERY

common opinion that the optimum conformation of an B inframmammary fold is an angle of 90°; a variation of x A this angle results in less definition.3,4 Of equal impor 1 C tance to definition is the symmetry between inframam mary folds. Symmetry is often left to “at a glance” pro cedures without scientific basis. C B x A Our method is simple, quick to perform, reliable and 2 reproducible, allowing the surgeon to perform an infra mammary fold to restore and ensure symmetry to the contralateral breast. To our knowledge, no previous reports on this issue x x B have been published in literature. 3 Competing interests: None declared.

References x x B 1. Bogetti P, Cravero L, Spagnoli G, et al. Aesthetic role of the surgic 4 ally rebuilt inframammary fold for implant-based breast reconstruc X X tion after mastectomy. J Plast Reconstr Aesthet Surg 2007;60:1225-32. C 2. Handel N, Jensen JA. An improved technique for creation of the A D D inframammary fold in silicone implant breast reconstruction. Plast Reconstr Surg 1992;89:558-62. Y Y 5 3. Agha-Mohammadi S, Hurwitz DJ. Management of upper abdominal X X laxity after massive weight loss: reverse abdominoplasty and infra mammary fold reconstruction. Aesthetic Plast Surg 2010;34:226-31. A D D C 4. Nava M, Quattrone P, Riggio E. Focus on the breast fascial system: a new approach for inframammary fold reconstruction. Plast Reconstr 6 Y Y Surg 1998;102:1034-45. Fig. 1. Technique to achieve symmetry of the new inframam- 5. Versaci AD. A method of reconstructing a pendulous breast utiliz mary fold. ing the tissue expander. Plast Reconstr Surg 1987;80:387-95.

Can J Surg, Vol. 57, No. 4, August 2014 279

personal-zoccali.indd 279 14-07-16 12:08 PM DISCUSSIONS IN SURGERY DISCUSSIONS EN CHIRURGIE

Users’ guide to the surgical literature: how to evaluate clinical practice guidelines

Christopher J. Coroneos, MD*† Clinical scenario Sophocles H. Voineskos, MD*† Sylvie D. Cornacchi, MSc*‡ You are a young general surgeon in a community practice. A new consultation is scheduled in your clinic: a 60-year-old woman presenting with chronic Charlie H. Goldsmith, PhD†‡§ venous ulcers on both legs. She has no other notable medical history. Over *‡ Teegan A. Ignacy, BSc the previous 5 years, she had a number of debridements and skin grafts per- Achilleas Thoma, MD, MSc*†‡ formed by a surgeon who recently retired. She is very frustrated by her odor- ous oozing ulcers and is embarrassed to visit her family. You review the From the *Division of Plastic and Recon- records at your hospital but cannot find much useful information. You do not structive Surgery, Department of Surgery, McMaster University, Hamilton, Ont., know why her previous surgeries and nonsurgical treatments have failed. You †Department of Clinical Epidemiology know that venous ulcers have different etiologies, such as persisting edema, and Biostatistics, McMaster University, superinfections or concomitant arterial insufficiency. You decide to review Hamilton, Ont., ‡Surgical Outcomes the literature guidelines that give specific recommendations on the manage- Research Centre (SOURCE), McMaster University, Hamilton, Ont., and the ment of venous ulcers to ensure an option on the treatment algorithm has not §Faculty of Health Sciences, Simon Fraser been overlooked. University, Burnaby, British Columbia, Canada and Arthritis Research Centre of Literature search Canada, Richmond, BC

Accepted for publication As described in a previous article in the “users’ guide to the surgical litera- Oct. 21, 2013 ture” series,1 you begin with a Medline search. The terms “venous ulcer” and “guideline” are entered separately. Based on medical subject heading Correspondence to: (MESH) terms, Medline prompts inclusion of the terms “leg ulcer/” or “vari- A. Thoma cose ulcer/” and “guideline” or “practice guideline,” respectively. These 101-206 James St. S Hamilton ON L8P 3A9 terms are combined and results limited to the English language, yielding [email protected] 10 articles.2–11 Three of them do not relate to venous ulcers,2,3,6 and 3 were published before the year 2000.8–10 Two are nonspecific for ulcers,5,7 and 1 DOI: 10.1503/cjs.029612 focuses on prevention.4 You select the article entitled, “Guidelines for the treatment of venous ulcers,”11 which appears to address your question. You print the guideline and review it before your next visit with the patient.

Introduction

Clinical practice guidelines (CPGs) are defined in the literature as “system atically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances.”12–15 They distill a large body of literature on a topic into a format that is high-yield and easy for physicians to use. Worldwide, surgeons perform 200 million procedures annually.16 There is constant effort to optimize this complex and expensive health care facet.16 Sur- geons are faced with difficult management decisions while balancing evidence- based recommendations.17 When trial evidence exists, it often cannot be per- fectly applied to specific patient presentations. It is difficult to independently condense primary research for each patient. Moreover, health care providers and insurers are increasingly concerned with quality improvement and cost effectiveness. Guidelines aim to balance these factors,12,15 and direct consistent

user-coroneos.indd 280 14-07-16 10:57 AM DISCUSSIONS IN SURGERY

and reliable care.17 The number of surgical guidelines available Robson and colleagues11 summarize the management of in the literature is increasing. However, CPGs vary in their venous ulcers in 8 categories: diagnosis, compression, quality and sometimes deviate from high methodologic al infection control, wound bed preparation, dressings, sur- rigor.16 It is necessary for surgeons to be able to appraise gery, adjuvant agents and long-term maintenance. How- CPGs before deciding to adopt their recommendations. ever, the guideline does not include specific PIPOH cri Since 1990, CPGs have been an increasingly popular teria. For example, the authors need to be more specific in tool influencing physician practice.18,19 More than 20 tools Recommendation #6.3: “Less extensive surgery on the to interpret and appraise CPGs have been published;14 the venous system, such as superficial venous ablation, endo latest is the AGREE-II instrument (appraisal of guidelines venous laser ablation, or valvuloplasty, especially when for research and evaluation).13 It was originally released in combined with compression therapy, can be useful in 2003 to address guideline development, reporting and decreasing the recurrence of venous ulcers (Level I).” The evaluation. Two further studies,20,21 have refined the reader must carefully consider patient population and instrument, now recognized as the methodological stan- health care setting in this recommendation. Venous ulcers dard in guideline evaluation.14 are associated with comorbidity. If our hypothetical patient In this article, we discuss a practical approach to the had diabetes or an inflammatory disorder, interventions appraisal of a CPG; Box 1 contains the key items readers would differ. Further, procedures such as endovenous laser should consider when using a CPG in surgery. As in previ- ablation may not be available in every health care setting. ous users’ guide to the surgical literature articles,22 we use a condensed framework to approach a guideline from a sur- Who was involved in guideline development? gical perspective. This will provide surgeons with a prac (ie. authors, reviewers, patients, readers) tical approach to interpreting and applying recommend No guideline is developed in isolation, free from potential ations in a CPG, using the guideline by Robson and bias.26,27 These biases may be subconscious and difficult to colleagues11 as an example. detect.28 Surgeons must critically consider how and why the guideline has been created. What inherent biases may Are the recommendations valid? the authors have? Organizing committees and professional organizations beyond those listed in the authorship Is there a clear statement of a clinical problem? may have reviewed the guideline. While reviewing evi- Like other publications, CPGs address a defined problem dence and providing recommendations, each group will in a specific group of patients. Surgeons must always con- have their own influence.29 Who are the people in these sider whether the CPG recommendations can be applied roles? Who are they representing? What is their exper- to their own patients.15,23 The PIPOH items (patient pop- tise? Guideline panels and authorship are often sponsored ulation, intevention(s), professionals/patients, outcomes to by the pharmaceutical industry in some capacity.28 be considered, health care setting) are suggested in the The authorship should be analyzed for surgical input. ADAPTE process (www.adapte.org) to frame the content Review by a surgical association or publishing in a surgical and clinical question in a guideline.24 Readers should use journal demonstrate evaluation with surgical familiarity. these categories to decide if the recommendations pre- Even when presented with the same research evidence, sented are representative of their patient and treatment professional groups can differ in their recommendations.30 goals. Surgeons are cautioned in applying CPGs not Shaneyfelt and colleagues15 identify examples in breast and designed for their patient populations.25 Subtle differences prostate cancer: a cancer interest group may support adoption in any category can alter the CPG’s applicability. of new, costly population screening interventions despite limited effectiveness, whereas public health groups may not Box 1. Users’ guides for an article on clinical practice view the intervention as a cost-effective strategy in the guidelines general population. I. Are the recommendations valid? Robson and colleagues’ work11 is developed by the 1. Is there a clear statement of a clinical problem? Wound Healing Society, with grant support from its 2. Who was involved in guideline development (i.e., authors, reviewers, educational/charitable arm, the Wound Healing Founda- patients, readers)? tion.31 The guideline is published in Wound Repair and 3. How is the guideline reviewed? Regeneration by the Wound Healing Society. The guideline 4. What literature are recommendations based on? II. What recommendations are made? has been developed and published by the same association, 5. Are useful recommendations presented? indicating a potential conflict of interest. Further, CPGs 6. How do authors move from evidence to recommendations? authored by research/care societies may be of lower quality 16 III. Will the results help me in caring for my patients? than those published by guideline societies. The CPG by 7. Were all outcomes considered (surgical outcomes versus natural Robson and colleagues listis its authors and their affiliations course of disease)? and positions. The group is composed of academicians, pri- 8. Will I be able to implement these recommendations? vate practice physicians, podiatrists, nurse clinicians,

Can J Surg, Vol. 57, No. 4, August 2014 281

user-coroneos.indd 281 14-07-16 10:57 AM DISCUSSIONS EN CHIRURGIE

research nurses, industrial scientists and an epidemiologist. What is the evidence base? The diverse author group reflects the multidisciplinary Authors should use appropriate methodology to support approach to chronic wounds and helps to reduce profes- their recommendations. A transparent and structured sional bias.32 However, little detail is given to the roles of methodology reflects rigorous development. This is evalu- each professional, and there is no mention of surgeons. ated stepwise, beginning with the search strategy, apprais- ing evidence and grading recommendations. Similar to the How is the guideline reviewed? rigor of a systematic review, a good CPG will reflect a body Surgeons should scrutinize the review and revision proof high-quality research with coherent results.17 Quality of cess of CPGs. Like any other publication, CPGs are sub- referenced studies should be clear. Issues with blinding, ject to peer review. Beyond the guideline’s sponsoring allocation concealment and equal expertise among groups association and authors, surgeons should be sure that are unique challenges in surgical RCTs.37 Unfortunately, independent experts are involved.33 This includes experts CPGs in surgery can rarely depend solely on high-level evi- in medical and research methodology and possibly dence (systematic reviews, randomized controlled trials patient groups. The process should be transparent. Com- [RCTs]).38 In some surgical areas, observational studies and mentary and editing from the review panel should be case reports may be the only evidence available,39 and these included or available in a supplement. Prior to dissemina- must be analyzed for confounding and bias. Often the avail- tion, guidelines may be pilot tested on small patient sam- able evidence is not of high quality.38 However, guidelines ples to ensure applicability. Authors should describe the addressing questions without available high-quality evi- process of reviewing and updating the guideline on an dence are still important in guiding physician decision mak- ongoing basis.26 Some groups establish a team monitoring ing.40 Moreover, these complex situations require a trans- for new evidence, whereas others provide a predeter- parent and rigorous methodology.40 mined schedule of updates to their guidelines.32 Guide- Guidelines should include, either in the text or support- lines with extensive readership and consistent new ing documentation, a statement detailing the development research findings (e.g., ACCP34 and ACLS35 guidelines) process. The availability of this process is a good predictor often schedule new releases. of the CPG’s overall rigor.41 Search strategies should In Robson and colleagues’ work,11 specific revision meth- incorporate multiple databases and a search of grey litera- odology is lacking. There is no indication the Wound Heal- ture (unpublished sources, such as conferences and thesis ing Society has reviewed the guideline, despite their spon- work). Explicit inclusion and exclusion criteria should be sorship. Details of expert review, review scales, specialties/ defined, and the assessment of the validity of the evidence disciplines of reviewers and edits suggested during the should be reproducible and consistent among studies.36 A review process would all be pertinent to the surgeon. lag time exists between guideline development and publi- Without insight into who approved the CPG, it is difficult cation. The CPG by Robson and colleagues11 was devel- to discern the potential biases that would impact surgical oped in October 2005 and published in the November/ decision making.33 No procedure is defined for revision. December 2006 issue of Wound Repair and Regeneration. In Given that leaders in the ever changing field of wound man- some rapidly changing specialties, it is possible that new agement provided this CPG, surgeons should expect a evidence becomes available within this lag time. schedule of updates. The CPG by Robson and colleagues11 includes a meth- Editorial independence and funding should be declared ods section. While databases are listed, no search terms are with all forms of research. Surgeons should be critical in specified to ensure a reproducible search methodology for assessing the interests of governing bodies or pharmaceutical/ the references cited. Without search terms, transparency is equipment sponsors. The translation of primary literature difficult to establish. Robson and colleagues specify that to clinical recommendations requires judgment.15 Surgeons their methodology differs from that of previous publica- must ensure this judgment is not biased.36 For example, tions, including laboratory/animal studies and findings does a company marketing dressings have any stake in these extrapolating from treatment of other ulcers. Beyond this, recommendations? Do involved professional organizations their process for selecting evidence is vague. have a monetary or public interest? The interests of Robson and colleagues11 succeed in defining the level of authors, Wound Repair and Regeneration and the Wound evidence for each recommendation. For example, they cite Healing Society are not discussed in the CPG by Robson the following literature for Recommendation #2.1: and colleagues. It is difficult to interpret the biases authors “Cullum N, Nelson EA, Fletcher AW, Sheldon TA. Com- may impose on the CPG outside of credentials listed. No pression for venous leg ulcers. The Cochrane Database of information is available for financial or research support of Systematic Reviews. (2001 Issue 2) The Cochrane Collab members. Robson and colleagues succeed in not emphasiz- oration. John Wiley & Sons Ltd. [STAT, 23 RCT].” This ing the use of brand name products. Instead it indicates the illustrates that the recommendation is based on a meta- evidence-based properties of a dressing that improve analysis of 23 RCTs and provides readers with a reference wound care. to the original data. Recommendations are followed by

282 J can chir, Vol. 57, No 4, août 2014

user-coroneos.indd 282 14-07-16 10:57 AM DISCUSSIONS IN SURGERY

contributing references, each marked with 1 of 8 levels of primary literature. The guideline succeeds in presenting evidence: STAT (Statistical analysis, meta-analysis, con- surgical, nonsurgical and preventative surgical options for sensus statement by commissioned panel of experts), RCT, venous ulcers where applicable. LIT REV (literature review), CLIN S (clinical case series), RETRO S (retrospective series review), EXP (experimen- How do authors move from evidence to tal laboratory or animal study), TECH (technique or recommendations? methodology description) or PATH S (pathological series Arriving at a guideline recommendation is complex, com- review). Of the 41 grouped recommendations made, 5 do bining best evidence, clinical decision making and patient not reference RCT or higher levels of evidence: 1.3, 1.4, preferences.42 Good CPGs will provide simple, straightfor- 4.3, 5.4 and 6.4. While RCTs, systematic reviews and ward care recommendations despite the complexities behind meta-analyses represent the top of the level of evidence them. When authors use a systematic method to arrive at a hierarchy, there is no discussion of the merits of each ref- judgment, recommendations are more clear and accurate in erence. Preferably, the RCTs should each be evaluated for guiding practice.43 Using this methodology, CPG authors individual methodological quality, especially given the should provide a strength or grade for each recommenda- unique issues in surgical trials. tion.43 This provides surgeons an indication of the confidence authors have in the literature, level of evidence and What recommendations are made? real-world effectiveness behind each of their recommendations.17 While CPG authors use a variety of methods to Are useful recommendations presented? grade recommendations,17 use of a consistent and transpar- Surgeons use guidelines for specific and practical evidence- ent methodology allows CPGs to be compared across differ- based advice to direct patient care. An RCT measuring ent fields and specialties.43 The GRADE39 methodology is physician practice finds specific recommendations leading used widely, including the Cochrane Collaboration and to more appropriate and fewer inappropriate clinical tests UpToDate.39 The GRADE methodology uses a simple sys- when compared with unspecific recommendations.32 For tem to categorize the quality of evidence into 4 levels (high, surgical CPGs and decision making, choices for patient moderate, low and very low) and strength of recommenda- care can often be reduced to a decision tree (e.g., nonoper- tions (strong or weak). Authors interpret methodology, het- ative v. procedure X v. procedure Y). From a surgical erogeneity, directness, precision and publication bias of each standpoint, attention to this paradigm is critical. Given a primary paper.44 For example, the Society for Vascular Sur- patient presentation, readers will turn to CPGs to illustrate gery adopts the GRADE framework and has a transparent both when a procedure should be performed and which methodology in forming their rigorous, patient-important procedure should be performed if different options exist. guideline recommendations.45 In Robson and colleagues’ work,11 recommendations are Robson and colleagues11 do not describe the strengths specific in most cases. For example, their Recommendation and limitations in the body of evidence for each recommen- #1.4 states, “Apparent venous ulcers that have been open dation. There is no formal tool used to illustrate the quality continuously without signs of healing for 3 months or that of each paper cited. A classification is used to indicate the do not demonstrate any response to treatment after strength of each recommendation. This helps illustrate the 6 weeks should be biopsied for histological diagnosis judgment process for each recommendation. However, the (Level III),” and Recommendation #1.1 states, “Gross authors do not include patient values in their judgment: arterial disease should be ruled out by establishing that • “Level I: Meta-analysis of multiple RCTs or at least pedal pulses are present on physical examination and/or 2 RCTs support the intervention of the guideline. that the ankle: brachial index (ABI) is > 0.8. (Any ABI < 1.0 Another route would be multiple laboratory or animal suggests a degree of vascular disease and compression ther- experiments with at least 2 clinical series supporting the apy is usually considered to be contraindicated with an ABI laboratory results.” < 0.7) […] (Level I).” These 2 examples reflect objective • “Level II: Less than Level I, but at least 1 RCT and at recommendations for venous ulcers. least 2 significant clinical series or expert opinion papers The clarity of other recommendations could be with literature reviews support the intervention. Experi- improved. For example, Recommendation #4.1 states, mental evidence that is quite convincing, but not yet “Examination of the patient as a whole is important to supported by adequate human experience, is included.” evaluate and correct causes of tissue damage. This includes • “Level III: Suggestive data of proof of principle, but factors such as (A) systemic diseases and medications, (B) lacking sufficient data, such as meta-analysis, RCT or nutrition, and (C) tissue perfusion and oxygenation (Level multiple clinical series.” II).” What specific diseases and medications are most The suggestion in the guideline can be positive or nega- important for venous ulcers? What components of a nutri- tive at the proposed level (e.g., meta-analysis and 2 RCTs tion workup are relevant? What typically needs to be sup- stating intervention is not of use in treating venous ulcers).” plemented? This information should be provided from the A high level of evidence may not lead to a strong

Can J Surg, Vol. 57, No. 4, August 2014 283

user-coroneos.indd 283 14-07-16 10:57 AM DISCUSSIONS EN CHIRURGIE

recommendation. For example, Recommendation #7b.4 ommendation #6.3 states, “Less extensive surgery on the states, “Negative pressure wound therapy may be useful venous system, such as superficial venous ablation, endo prior to a skin graft/flap by helping promote the develop- venous laser ablation, or valvuloplasty, especially when ment of granulation tissue in the wound base, or postopera- combined with compression therapy, can be useful in tively by preventing shearing and removing exudates. How- decreasing the recurrence of venous ulcers (Level I).” This ever, its reported experience in venous ulcers is limited recommendation should be scrutinized because authors can (Level II).” Despite high-level evidence there has not been expand on the specific indication of each procedure com- an illustration of clinical effectiveness, and the impact of pared with traditional deep ligation of multiple perforating therapy may outweigh its potential benefits to patients. veins and previously mentioned subfascial endoscopic per- A classification of each recommendation’s strength is forator surgery. This approach to surgical decision making missing. Grading recommendations based on this system would be helpful to readers. would allow for comparisons among recommendations in this guideline. No consensus methodology (e.g., Delphi Will I be able to implement these recommendations? method46) is included. Without explicit methodology, it is Moving from primary evidence to CPGs, authors consider difficult to ascertain how the CPG committee arrived at the potential barriers in offering these procedures to their recommendations. Insight on how decisions were patients.32 When using a guideline, surgeons interpret rec- made is necessary for surgeons to apply findings in their ommendations in their own setting. An academic tertiary own decision making. The aforementioned GRADE meth- care centre and community hospital have different patient odology provides structure to the review process and limits populations, resources and support personnel.36 Applicability the bias of “expert opinion” where evidence is unclear. and assessment of barriers is often overlooked,47 especially in surgery. Guidelines are expected to illustrate how recom- Will the results help me care for my patients? mendations can be applied in the settings the authors intended.41 Are all outcomes considered? Robson and colleagues11 describe the necessary com The process used to select the relevant outcomes and ponents for proper management of venous ulcers. While importance of these outcomes must be explicit and sensible. the CPG touches on the multidisciplinary care required in The importance of a certain outcome is directly related to the preoperative workup, operative/postoperative man- what a patient cares about most. Therefore, CPG authors agement and follow-up, the barriers and difficulties in this need to describe the methods with which the outcomes process are not specifically discussed. The CPG focuses were chosen and a description of the process used to decide solely on interventions. Practically, surgeons are most on the importance of each outcome. Information on who often limited by the resources available to them. Using a was involved in outcome choice as well as how values were CPG, surgeons must consider if their own resources assigned to outcomes should be apparent in the guideline. would support recommendations. Are new, expensive Surgical decision making, like other recommendations, dressings more effective in treatment? Questions of cost- can often be reduced to analysis of benefit versus risk and effectiveness and economic analysis are increasingly harm.44 Guidelines should identify not only the interven- important to answer.18 tions of interest, but also sensible alternatives. Surgeons must consider whether the benefits of the treatment dis- Resolution cussed outweigh not only the side effects and risks of treatment, but also the implications of another treatment or no Although Robson and colleagues’ work11 is not specific to a treatment. For example, under what circumstances does the particular population or any comorbid conditions, you benefit of diagnostic laparoscopy outweigh the risk? In a have no reason to believe that it is not applicable to your CPG for basal cell carcinoma,43 authors weigh surgical exci- patient. You consider the recommendations in this guide- sion against curettage and desiccation, cryotherapy, radia- line in a stepwise manner. A biopsy of the ulcer first rules tion, chemotherapy and carbon dioxide laser. Consider- out malignancy. A quantitative biopsy rules out clincially ations include the clinical situation, availability of equipment important bacterial contamination. You proceed to debride and patient values/risk profiles among other variables. the ulcer in your clinic to minimize the bacterial medium. Robson and colleagues11 provide a thorough approach to Home care services are used for daily moist dressing in workup and treatment of venous ulcers. Surgical CPGs are addition to compression to minimize edema. Two weeks sometimes guilty of focusing on the surgical aspects of care later, the ulcer has a clean base, and you perform a split while ignoring other aspects of patient management. The thickness skin graft. With weekly outpatient follow-up, to nonsurgical multidisciplinary approach is well defined for the patient’s surprise the ulcer proceeds to heal for the first workup, allowing readers a guide to the workup and preop- time in 5 years. This is not the end of the story though. erative preparation of a venous ulcer. Operative interven- You recommend that the patient should continue applying tions should be compared more directly. For example, Rec- the compression dressings for life to avoid recurrence.

284 J can chir, Vol. 57, No 4, août 2014

user-coroneos.indd 284 14-07-16 10:57 AM DISCUSSIONS IN SURGERY

Competing interests: None declared. 19. Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines.BMJ 1999; Contributors: C. Coroneos, S. Voineskos, T. Ignacy and A. Thoma designed the study. All authors acquired and interpreted the data. 318:527-30. C. Coroneos, T. Ignacy and A. Thoma wrote the article, which all 20. Brouwers MC, Kho ME, Browman GP, et al. Development of the authors reviewed and approved for publication. AGREE II, part 1: performance, usefulness and areas for improvement. CMAJ 2010;182:1045-52. References 21. Brouwers MC, Kho ME, Browman GP, et al. Development of the AGREE II, part 2: assessment of validity of items and tools to support 1. Birch DW, Eady A, Robertson D, et al. Users’ guide to the surgical application. CMAJ 2010;182:E472-8. literature: how to perform a literature search. Can J Surg 2003; 46:136-41. 22. Thoma A, Farrokhyar F, Bhandari M, et al. Users’ guide to the sur gical literature. How to assess a randomized controlled trial in sur- 2. Wienert V, Gerlach H, Gallenkemper G, et al. Medical compres- gery. Can J Surg 2004;47:200-8. sion stocking (MCS). J Dtsch Dermatol Ges 2008;6:410-5. 23. Grol R, Dalhuijsen J, Thomas S, et al. Attributes of clinical guidelines 3. Hopf HW, Ueno C, Aslam R, et al. Guidelines for the prevention of that influence use of guidelines in general practice: observational lower extremity arterial ulcers. Wound Repair Regen 2008;16:175-88. study. BMJ 1998;317:858-61. 4. Robson MC, Cooper DM, Aslam R, et al. Guidelines for the preven- 24. Graham ID, Harrison MB, Brouwers M, et al. Facilitating the use of tion of venous ulcers. Wound Repair Regen 2008;16:147-50. evidence in practice: evaluating and adapting clinical practice guidelines 5. Nicolaides AN, Allegra C, Bergan J, et al. Management of chronic for local use by health care organizations. J Obstet Gynecol Neonatal Nurs venous disorders of the lower limbs: guidelines according to scientific 2002;31:599-611. evidence. Int Angiol 2008;27:1-59. 25. Hayward RS, Wilson MC, Tunis SR, et al. Users’ guides to the 6. Hopf HW, Ueno C, Aslam R, et al. Guidelines for the treatment of medical literature. VIII. How to use clinical practice guidelines. A. arterial insufficiency ulcers.Wound Repair Regen 2006;14:693-710. Are the recommendations valid? The Evidence-Based Medicine Working Group. JAMA 1995;274:570-4. 7. Agus GB, Allegra C, Antignani PL, et al. Guidelines for the diagnosis and therapy of the vein and lymphatic disorders. Int Angiol 2005; 26. Thomson R, Lavender M, Madhok R. How to ensure that guide- 24:107-68. lines are effective. BMJ 1995;311:237-42. 8. McInnes E, Cullum N, Nelson A, et al. RCN guideline on the 27. McNicol M, Layton A, Morgan G. Team working: the key to management of leg ulcers. Nurs Stand 1998;13:61-3. implementing guidelines? Qual Health Care 1993;2:215-6. 9. Update: venous leg ulcer guideline. University of Pennsylvania. 28. Shaneyfelt TM, Centor RM. Reassessment of clinical practice Ostomy Wound Manage 1997;43:80-2. guidelines: go gently into that good night. JAMA 2009;301:868-9. 10. Douglas WS, Simpson NB. Guidelines for the management of 29. Thomas S. Standard setting in the Netherlands: impact of the human chronic venous leg ulceration. Report of a multidisciplinary work- factor on guideline development. Br J Gen Pract 1994;44:242-3. shop. British Association of Dermatologists and the Research Unit 30. Coulter I, Adams A, Shekelle P. Impact of varying panel membership of the Royal College of Physicians. Br J Dermatol 1995;132:446-52. on ratings of appropriateness in consensus panels: a comparison of a 11. Robson MC, Cooper DM, Aslam R, et al. Guidelines for the treat- multi- and single disciplinary panel. Health Serv Res 1995;30:577-91. ment of venous ulcers. Wound Repair Regen 2006;14:649-62. 31. Wound Healing Foundation. About the Wound Healing Founda- 12. Field M, Lohr K, eds. Committee to Advise the Public Health Service on tion [website]. Available: www.woundhealfoundation.net/about Clinical Practice Guidelines: clinical practice guidelines directions for a new /index.htm (accessed 2012 Apr. 15). program. Washington (DC): National Academy Press; 1990. 32. Shekelle PG, Woolf SH, Eccles M, et al. Clinical guidelines: devel- 13. Brouwers MC, Kho ME, Browman GP, et al. AGREE II: advanc- oping guidelines. BMJ 1999;318:593-6. ing guideline development, reporting and evaluation in health care. 33. Murphy MK, Black NA, Lamping DL, et al. Consensus develop- CMAJ 2010;182:E839-42. ment methods, and their use in clinical guideline development. 14. Vlayen J, Aertgeerts B, Hannes K, et al. A systematic review of Health Technol Assess 1998;2:i-iv, 1-88. appraisal tools for clinical practice guidelines: multiple similarities 34. Holbrook A, Schulman S, Witt DM, et al. Evidence-based manage- and one common deficit.Int J Qual Health Care 2005;17:235-42. ment of anticoagulant therapy: antithrombotic therapy and preven- 15. Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are guidelines fol- tion of thrombosis, 9th ed: American College of Chest Physicians evi- lowing guidelines? The methodological quality of clinical practice dence-based clinical practice guidelines. Chest 2012;141(Suppl): guidelines in the peer-reviewed medical literature. JAMA 1999; e152S-84S. 281:1900-5. 35. Neumar RW, Otto CW, Link MS, et al. Part 8: adult advanced car- 16. Barajas-Nava L, Solà I, Delgado-Noguera M, et al. Quality assess- diovascular life support: 2010 American Heart Association guidelines ment of clinical practice guidelines in perioperative care: a systematic for cardiopulmonary resuscitation and emergency cardiovascular care. appraisal. Qual Saf Health Care 2010;19:e50. Circulation 2010;122(Suppl 3):S729-67. 17. Dahm P, Yeung LL, Gallucci M, et al. How to use a clinical practice 36. Burgers JS, Grol RPTM, Zaat JOM, et al. Mokkink HG a. Charac- guideline. J Urol 2009;181:472-9. teristics of effective clinical guidelines for general practice. Br J Gen Pract 2003;53:15-9. 18. Ray-Coquard I, Philip T, De Laroche G, et al. A controlled “before- after” study: impact of a clinical guidelines programme and regional 37. Poolman RW, Struijs PAA, Krips R, et al. Reporting of outcomes cancer network organization on medical practice. Br J Cancer in orthopaedic randomized trials: Does blinding of outcome asses- 2002;86:313-21. sors matter? J Bone Joint Surg Am 2007;89:550-8.

Can J Surg, Vol. 57, No. 4, August 2014 285

user-coroneos.indd 285 14-07-16 10:57 AM DISCUSSIONS EN CHIRURGIE

38. Cook DJ, Greengold NL, Ellrodt AG, et al. The relation between 43. Atkins D, Best D, Briss PA, et al. Grading quality of evidence and systematic reviews and practice guidelines. Ann Intern Med 1997; strength of recommendations. BMJ 2004;328:1490. 127:210-6. 44. Schünemann HJ, Vist GE, Jaeschke R, et al. Grading recommenda- 39. Guyatt GH, Prasad K, Schünemann HJ, et al. How to use a patient tions. In: Shanahan JF, Pancotti R, editors. Users’ guides to the medical management recommendation. In: Shanahan JF, Pancotti R, editors. literature — a manual for evidence based clinical practice. 2nd ed. New Users’ guides to the medical literature — a manual for evidence based clinical York (NY): McGraw-Hill Professional; 2008:679–701. practice. 2nd ed. New York (NY): McGraw-Hill Professional; 2008: 45. Murad MH, Montori VM, Sidawy AN, et al. Guideline methodol- 597–618. ogy of the Society for Vascular Surgery including the experience 40. Loblaw DA, Prestrud AA, Somerfield MR, et al. American Society of with the GRADE framework. J Vasc Surg 2011;53:1375-80. Clinical Oncology clinical practice guidelines: formal systematic 46. Vonk Noordegraaf A, Huirne JAF, Brölmann HAM, et al. Multi- review-based consensus methodology. J Clin Oncol 2012;30:3136-40. disciplinary convalescence recommendations after gynaecological 41. Fervers B, Burgers JS, Haugh MC, et al. Predictors of high quality surgery: a modified Delphi method among experts.BJOG 2011;118: clinical practice guidelines: examples in oncology. Int J Qual Health 1557-67. Care 2005;17:123-32. 47. Browman GP. Improving clinical practice guidelines for the 21st cen- 42. Guyatt G, Vist G, Falck-Ytter Y, et al. An emerging consensus on tury. Attitudinal barriers and not technology are the main challenges. grading recommendations? ACP J Club 2006;144:A8-9. Int J Technol Assess Health Care 2000;16:959-68.

Comment vous pouvez vous impliquer dans l’AMC ! L’AMC est vouée à jouer un rôle de chef de file auprès des médecins et à promouvoir les normes les plus élevées de santé et de soins de santé pour les Canadiens. Afin de renforcer l’Association et pour qu’elle représente véritablement tous les médecins du Canada, l’AMC a besoin de membres intéressés à occuper des charges élues et à siéger à des comités et des groupes consultatifs. La structure de l’AMC se compose d’organes de régie et d’entités consultatives élus par le Conseil général ou nommés par le Conseil d'admini - stration. Le Conseil d’administration, dont les membres sont élus par le Conseil général et représentent les associations médicales provinciales et territoriales, les résidents et les étudiants en médecine, est chargé de l’administration générale de l’AMC. Il rend compte des questions de régie au Conseil général. Les comités de l’AMC jouent le rôle de conseillers auprès du Conseil d’administration et présentent des recommandations au sujet de questions particulières intéressant les médecins et la population. Cinq comités principaux sont constitués principalement de représentants des régions, des résidents et des étudiants, tandis que les autres comités statutaires et spéciaux et les groupes de travail réunissent des personnes qui s’intéressent à des sujets précis et possèdent des compétences spécialisées. Des postes pourront devenir vacants dans un ou plusieurs de ces comités en cours d'année. Pour obtenir plus d’information au sujet des façons de participer, veuillez consulter http://www.cma.ca/centredesmembres/comment- vous-impliquer-dans-lamc ou communiquer avec Cherise Araujo Corporate and Governance Services Canadian Medical Association 1867 Alta Vista Drive, Ottawa ON K1G 5W8 Fax 613 526-7570, Tel 800 663-7336 x1949 [email protected] Votre participation peut faire la différence. Nous espérons avoir de vos nouvelles !

286 J can chir, Vol. 57, No 4, août 2014

user-coroneos.indd 286 14-07-16 10:57 AM DISCUSSIONS IN SURGERY DISCUSSIONS EN CHIRURGIE

Use of a novel energy technology for arresting ongoing liver surface and laceration hemorrhage

Chad G. Ball, MD, MSc Summary Persistent hemorrhage from liver capsular injuries has remained a technical From the Regional Trauma Program, challenge without an optimal solution. This report discusses an easy to use University of Calgary and the Foothills device that is commonly used within elective hepatic surgery and can be suc- Medical Centre, Calgary, Alta. cessful in arresting ongoing surface and laceration bleeding in patients with solid organ injuries. Accepted for publication Nov. 6, 2013

Competing interests: None declared. epatic hemorrhage is often life-threatening and difficult to arrest. While the algorithm for the treatment of ongoing liver bleeding is well Correspondence to: 1 C.G. Ball H described, technical considerations for stopping persistent hepatic Foothills Medical Centre surface/capsule bleeding are not. Traditional techniques include coagulation by 1403–29 St. NW high voltage cautery using a Bovie, topical hemostatic application and/or the Calgary AB T2N 4W4 delivery of ignited argon gas.2 This form of bleeding is particularly troublesome [email protected] when a surgeon unpacks a previously damage-controlled liver injury that had not been wrapped with a nonstick, plastic barrier. DOI: 10.1503/cjs.027113 Several patients (11) at the Foothills Medical Centre (FMC) have now been treated with a novel device that is commonly used for elective liver transection among hepatobiliary surgeons (Aquamantys; Medtronic). This instrument uti- lizes bipolar radiofrequency energy, which acts to ignite/boil dripping saline from a small, easy to manipulate handpiece instrument. This device is also excellent at sealing small to medium-sized bile ducts, thereby preventing subsequent bile leaks and collections. The 11 patients treated at FMC had multiple injuries (mean injury severity score 28) and required initial damage control packing with a return to the oper ating room within 24–72 hours once their physiology and biochemistry was stabil ized. All livers had been initially packed with standard laparotomy sponges and subsequently oozed substantial volumes of blood from the liver capsule and/or laceration itself when unpacked. Coagulation and cauterization with the Aqua- mantys device stopped the hemorrhage in all 11 patients immediately. Anecdo tally, the local effect of the Aquamantys device on the injured liver and associated hemorrhage was very similar to its effect during elective surgery on cirrhotic patients. In all cases, the 6.0 tip (round and blunt), as opposed to the 9.5 tip (sharp), was used. Interestingly, surgeons at FMC have also successfully used its deep tissue sealing ability for a splenic laceration in 1 patient with multiple injuries. The Aquamantys instrument has the potential to arrest ongoing hepatic surface/capsular bleeding as well as moderate hemorrhage associated with the liver laceration itself. Despite an impressive history and near ubiquitous use within the elective hepatic surgery arena, to my knowledge, this represents the first discussion of a potential use for this technology in injured patients. Fur- ther experience and study is required.

References

1. Kozar RA, Feliciano DV, Moore EE, et al. Western Trauma Association/critical decisions in trauma: operative management of adult blunt hepatic trauma. J Trauma 2011;71:1-5. 2. Pachter HL, Feliciano DV. Complex hepatic injuries. Surg Clin North Am 1996;76:763-82.

What would I want for to have a high-functioning team per- presentation. In this letter we provide my surgery? forming your surgery? The answer is this historic surgical background. obvious. The checklist can help oper- In the last century, breast surgery Team-oriented. Communicative. ating teams work better together. We has undergone dramatic changes in Transparent. These are words that hope our paper (Available at http dogma; it serves as a prime example of we want all patients to use when :// bcpsqc.ca/clinical-improvement how surgeons have made progress by describing their operating teams. /surgical-checklist/what-would-you challenging the limits of contempo- Teams that embody these character- -want-for-your-surgery/) offers con- rary doctrine. The origins of breast istics likely work well together and structive ideas on how Canada’s surgery for cancer can be traced back make few mistakes. But creating a surgical community can move forward to the 16th century.2 It was not until standardized surgical culture that as it aims to improve care for patients. 1894, however, that an American sur- encourages these qualities has proven Marlies van Dijk, RN, MSc geon, William Stewart Halsted challenging. Director Clinical Improvement, B.C. Patient (1852–1922), published his surgical Implementing the surgical safety Safety & Quality Council technique for breast cancer surgery in Vancouver, BC checklist can grow this culture by the form of the Halsted (radical) mas- framing how an operating team com- DOI: 10.1503/cjs.006414 tectomy.3 This technique involved municates; in turn, this can minimize excising the breast, lymph nodes and avoidable risks (like infections and References pectoralis major and minor, leaving allergic reactions) that endanger only skin covering the ribs. This patients. It makes sense. A recent 1. Urbach DR, Govindarajan A, Saskin R, et extensive en bloc tissue resection study,1 however, found no correlation al. Introduction of surgical safety check- resulted in considerable disfigurement lists in Ontario, Canada. N Engl J Med between the surgical checklist and 2014;370:1029-38. and morbidity from the resultant patient mortality. Does this mean that 2. LeBlanc J, Donnon T, Hutchison C, et al. weakened arm function and disabling there is no value in its application? Development of an orthopedic surgery lymphedema. Despite this, the vogue In British Columbia, a variety of trauma patient handover checklist. Can J toward more extensive surgery con- stakeholders from the surgical com- Surg 2014;57:8-14. tinued into the midtwentieth century munity have responded to this study with a “bigger is better” approach. In with the hope of highlighting the A century of breast surgery: fact, an American surgeon, Owen H. value of good teamwork and com from radical to minimal Wangensteen, was quoted as saying munication in the operating room. “Today, it should be said, I believe, There is tremendous value to the We read the recent article “What is the Halsted operation for cancer of checklist beyond its statistical signifi- the effect of screening mammog the breast is outmoded: it is not rad cance. It ensures that common objec- raphy on breast cancer incidence”1 ical enough”4. These newer opera- tives are being effectively communi- with great interest. The introduction tions involved extending the radical cated. It empowers all health of breast screening programs has surgery dissection into the neck and professionals to speak up if they opened many new uncertainties on mediastinum by supraradical mastec- notice a potential error. It gives the ideal management of women with tomy. This made no difference on patients a voice in determining their early breast cancer. This is especially patient outcomes; subsequent survival own care. true of in situ disease, as it is clear rates did not change in light of more Since there is very little education that a proportion of these women aggressive surgery. on teamwork for health professionals, may be overtreated. We feel your In 1948, Patey and Dyson (Lon- past studies2 promoting the benefits readers may be interested in the hist don) advocated for taking a step back of the checklist should not be dis- ory of breast cancer treatment and with pectoralis major–sparing surgery counted. Our group also suggests that the changes in surgical techniques as by using modified radical mastectomy hospitals invest resources and exper- a background for considering breast for breast cancer. The modified radical tise to provide teams with coaching cancer incidence over time. This hist mastectomy became popular and and training. This investment will ory is especially pertinent in light of slowly replaced more extensive surgery undoubtedly foster the use of tools only marginal reductions in the rate by the 1980s.5 Patey argued that the like the checklist. of late-stage cancer presentation,1 excision of pectoralis major did not The next time that you are in the suggesting the screening program is add any significant benefit but did con- operating room, imagine that it is you unlikely to eliminate the need for tribute to poor cosmetic outcomes and laying on the table. Would you want more extensive surgery for later-stage intraoperative blood loss. In this post-

Angeline Bhalerao, MB BS, BSc World War II era and particularly the involved part of the breast Foundation Doctor, Academic Medicine from the 1970s, advances in adjuvant (quadrantectomy). This was a radical Norfolk and Norwich University Hospital therapy (such as hormonal, chemo- idea at the time. An RCT of Norwich, United Kingdom therapy and radiotherapy) have been 701 women recruited from 1973 to DOI: 10.1503/cjs.005814 combined with less radical surgery to 1980 and followed up for 20 years achieve similar survival rates compared showed that the long-term survival with early more radical surgery. rate among women who undergo References The 1970s marked the age of breast-conserving surgery (with large-scale randomized controlled radiotherapy) was similar to that 1. Latosinsky S, Boileau JF, Bryant HE, et al. trials (RCTs) assessing the extent of among women who undergo radical CAGS and ACS evidence based reviews in surgery. 48. What is the effect of screen- surgery with objective outcomes. The mastectomy. The evolution and mini- ing mammography on breast cancer inci- Alabama Breast Cancer Project, Man- mization of breast surgery is being dence? Can J Surg 2014;57:67-9. chester Trial, the Cardiff–St. Mary’s echoed in the field of surgery to the 2. Ekmektzoglou KA, Xanthos T, German trial and several others compared axilla. A number of contemporary V, et al. Breast cancer: from the earliest more versus less extensive surgery. In studies have focused on the role of times through to the end of the 20th cen- 1971, the National Surgical Adjuvant management of the sentinel lymph tury. Eur J Obstet Gynecol Reprod Biol 2009; Breast and Bowel Project B-04 was node biopsy (Z0011, Amaros, 145:3-8. initiated and was to be the largest Supremo), stepping away from con- 3. Osborne MP. William Stewart Halsted: RCT on the subject. The trial ventional axillary clearance. his life and contributions to surgery. included 1079 women with clinically The future of breast surgery is Lancet Oncol 2007;8:256-65. negative axillary nodes who under- bright. The next decade will be marked 4. Tuttle TM, Owen H. Wangensteen, and went radical mastectomy, total mas- by shifts in paradigm toward less (with Jerome A Urban. The pursuit of extraaxil- tectomy without axillary dissection narrower acceptable margins) but more lary lymph node metastases from breast but with postoperative irradiation, or focused surgery to both the breast and cancer. J Am Coll Surg 2004;199:636-43. total mastectomy plus axillary dissec- axilla in light of advances in newer 5. Wilson RE, Donegan WL, Mettlin C, et tion only if their nodes became posi- radiotherapy and chemotherapy regi- al. The 1982 national survey of carcinoma tive. There is now 25-year follow-up mens. This is especially true with of the breast in the United States by the data from this study, which validated earlier stage cancer detection, which American College of Surgeons. Surg Gynecol Obstet 1984;159:309-18. other studies showing no advantage this data demonstrates. 1 from radical mastectomy.6 6. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up of a random- The true era of breast-conserving Gurdeep Singh Mannu, MB BS, BSc, ized trial comparing radical mastectomy, surgery is accredited to Umberto MRCSEd Academic Clinical Fellow, General Surgery total mastectomy, and total mastectomy Veronesi, an Italian oncologist who Oxford University Hospital NHS Trust followed by irradiation. N Engl J Med progressed the idea of removing only Oxford, United Kingdom 2002;347:567-75.

The Canadian Journal of Surgery is pleased to accept Le Journal canadien de chirurgie accepte volontiers les annonces career/classified advertisements. The deadline is 1 month before sur les carrières et annonces classées. Celles-ci doivent être issue date. reçues au JCC au plus tard 1 mois avant la date de parution.

Tarifs: Rates: Grand format: 1 page 1200 $; 2/3 page 900 $; 1/2 page vert/ Display ads: 1 page $1200; 2/3 page $900; 1/2 page vert/horiz $800; horiz 800 $; 1/3 page 650 $; 1/4 page 500 $; 1/6 page 400 $. 1/3 page $650; 1/4 page $500; 1/6 page $400. Word ads: $120 for Mot des annonces: 120 $ jusqu’à 40 mots et 1.20 $ par mot the first 40 words or less, additional words $1.20 each (additional supplémentaire (25 $ pour encadrement au trait). Encadré spécial $25 for frame). Special Display under 100 words, 55 × 55 mm, jusqu’à 100 mots, 55 × 55 mm, 205 $. L’emploi d’une boîte-réponse $205. $20 charge (first insertion only) for CJS confidential reply confidentielle au JCC donne lieu à la perception d’un supplément box numbers. de 20 $ (pour la première insertion seulement). VISA, MASTERCARD AND AMERICAN EXPRESS ACCEPTED. VISA, MASTERCARD ET AMERICAN EXPRESS ACCEPTÉS.

Advertisements should be sent to: Journal Advertising, Canadian Le texte des annonces doit être adressé à : Annonces publicitaires, Journal of Surgery, 1867 Alta Vista Dr., Ottawa ON K1G 5W8; Journal canadien de chirurgie, 1867, prom. Alta Vista, Ottawa tel 800 663-7336 or 613 731-8610 x2107/2041; fax 613 565-7488; (Ontario) K1G 5W8; tél 800 663-7336 ou 613 731-8610 x2107/2041; email [email protected] fax 613 565-7488; courriel [email protected] Veuillez faire parvenir les réponses au numéros de boîtes à Send all box number replies to: Box . . . , Canadian Journal of l’adresse suivante : Boîte . . . , Journal canadien de chirurgie, Surgery, 1867 Alta Vista Dr., Ottawa ON K1G 5W8. 1867, prom. Alta Vista, Ottawa, (Ontario) K1G 5W8. The Ontario Human Rights Code prohibits discriminatory Le Code des droits de la personne de l’Ontario interdit la employment advertising. discrimination dans la publicité relative à l’emploi.

CHUM IS RECRUITING A DIRECTOR FOR ITS DEPARTMENT OF SURGERY

The Centre hospitalier de l’Université de Montréal offers first and foremost Summary of the role and responsibilities highly specialized care to regional and supraregional adult clients in all medical The Director of the Department of Surgery will have the unique opportunity to specialties. Through its integrated activities in care, teaching, research, health participate and influence the creation of the new CHUM. They will be able to plan a technology assessment and health intervention methods, as well as quality service offering in the most modern infrastructures. They will have to promote the and patient experience, it has become the main development and knowledge academic mission and vision of excellence of their department. transfer pole. The ideal candidate must have an academic profile and a good understanding of the As of 2016, the CHUM will offer a renewed hospital experience in its new facility work performed in a university medical centre. They must have leadership qualities, located in downtown Montréal. management experience and good knowledge of the issues related to organizational The CHUM is an active member of the Réseau universitaire intégré de transition and move to the new CHUM. santé (RUIS) of the Université de Montréal. The institution gathers over Those interested in applying must have at least 5 years academic experience. They 10,000 employees; close to 1,000 doctors, dentists and pharmacists; must be eligible to obtain licensure to practise from the Collège des médecins du approximately 1,400 researchers, investigators and other members of the Québec. Knowledge in change management and clinical administration experience Centre de recherche du CHUM (CRCHUM); as well as 650 volunteers. With are assets. its major contribution in training doctors and health professionals in Quebec, the CHUM welcomes close to 6,000 students and interns each year. The CHUM’s annual budget ranges around $830 million. Its main areas of expertise The people interested are asked to send, under confidential seal, their résumé along are cancer, neurosciences, cardiovascular and metabolic, transplantation, with a letter of interest, before 4pm on Tuesday, September 2nd, 2014, by email, functional musculoskeletal, imagery, leading medical technologies, genetics at: [email protected] or by mail, at Madame Denyse Béchard, and biomarkers, as well as immunology and infectiology. ECMDP Office at: 1058, rue Saint Denis, porte 2312, Montréal (Québec) H2X 3J4 S-758

classified-aug14.indd 287 14-07-16 11:00 AM CHAIRE DE RECHERCHE FRANCIS-GLORIEUX SUR LES MALADIES MUSCULO-SQUELETTIQUES PÉDIATRIQUES Département de chirurgie pédiatrique Hôpitaux Shriners pour enfants, Université McGill Nous sommes à la recherche d’un scientifique hautement qualifié dans le domaine des maladies musculo-squelettiques pédiatriques (M.D. ou Ph.D. ou M.D., Ph.D.) pour être nommé titulaire de la Chaire de recherche Francis-Glorieux sur les maladies musculo-squelettiques créée par l’Hôpital Shriners pour enfants de Montréal et la Faculté de médecine de l’Université McGill. Le titulaire sera un scientifique de renommée internationale appelé à diriger une équipe de recherche très dynamique à l’Hôpital Shriners pour enfants du Canada. Le nouvel Hôpital Shriners pour enfants et son centre de recherche ouvriront leurs portes en 2015. Les installations abriteront un laboratoire de travaux pratiques de 2322 mètres carrés, un centre de recherche clinique de 185 mètres carrés, une petite installation réservée aux animaux, un laboratoire d’analyse du mouvement, une vaste salle de conférence et un laboratoire d’enseignement de techniques chirurgicales. Le nouvel hôpital Shriners sera adjacent à l’Hôpital de Montréal pour enfants, au Centre universitaire de santé McGill et à l’Institut de recherche avec lesquels il travaillera en étroite association. Le candidat devra jouir d’une réputation internationale dans le domaine de la recherche, bénéficier d’un financement soutenu par des organismes soumis à l’examen des pairs et avoir publié des travaux de recherche dans le domaine de la régénération osseuse ou des cellules souches. Le candidat devra posséder le leadership, la vision et l’expertise nécessaires pour faire avancer la recherche sur les maladies musculo-squelettiques et l’enseignement au Département de chirurgie pédiatrique. Le candidat devra : • attester d’une carrière universitaire dans le domaine de la médecine régénérative et de la recherche sur les cellules souches, ce domaine touchant à la recherche sur les maladies musculo-squelettiques; • posséder une expérience de premier plan dans le domaine de la médecine régénérative et notamment en recherche sur les cellules souches; • avoir une vision claire de l’évolution constante de la recherche sur les maladies musculo-squelettiques et avoir une capacité démontrée de partager sa vision, d’encourager les autres à prendre part à des activités de recherche au centre de recherche Shriners et de favoriser une approche interdisciplinaire en matière de recherche à l’Hôpital Shriners pour enfants, à l’Hôpital de Montréal pour enfants du Centre universitaire de santé McGill et à l’Université McGill; • posséder une expérience de la gestion d’un centre de recherche serait un atout. Le candidat doit être professeur agrégé ou professeur titulaire. L’Université McGill souscrit à la diversité et à l’équité en matière d’emploi. Elle accueille favorablement les demandes d’emploi des femmes, des peuples autochtones, des minorités visibles, des minorités ethniques, des personnes handicapées, des personnes de toutes orientations et identités sexuelles et d’autres personnes qui pourraient contribuer à une plus grande diversité. On encourage tous les candidats qualifiés à postuler, la priorité sera toutefois accordée aux Canadiens ainsi qu’aux résidents permanents. Date limite de présentation des candidatures : le 30 septembre 2014. Faire parvenir une lettre d’intérêt signée, un curriculum vitae et les noms de trois répondants à la personne suivante : Dr Jean-Pierre Farmer, président du Comité de recherche Chef du département de chirurgie pédiatrique, Université McGill a/s [email protected] ______FRANCIS GLORIEUX CHAIR IN PEDIATRIC MUSCULOSKELETAL RESEARCH Department of Pediatric Surgery Shriners Hospital, McGill University We are seeking a highly qualified scholar in pediatric musculoskeletal research (MD or Ph.D. or MD, Ph.D.) to assume the Francis Glorieux Chair in Musculoskeletal research at Shriners Hospital, McGill University’s Faculty of Medicine. It is anticipated that the prospective chair holder will be a world class scholar who will lead a vigorous research team at the Shriners Hospitals for Children - Canada. The new Shriners Hospital and its research centre will open its doors in 2015 and will include a 25,000 sq. ft. wet laboratory, a 2000 sq. ft. clinical investigation unit, a small animal facility as well as a motion analysis laboratory, a large conference room and a surgical skills laboratory. The new Shriners facility will be adjacent to, and partner with, the Montreal Children’s Hospital and the McGill University Health Centre and its Research Institute. It is expected that the candidate will have an international research reputation, including a track record of sustained funding from peer-review agencies, evidence of prior research publications in the field of bone regeneration/stem cell research and is expected to provide leadership, vision and expertise that will advance musculoskeletal research and teaching in the Department. Candidates must have: • Demonstrated academic accomplishments in regenerative medicine and stem cell research as it applies to musculoskeletal research; • Leadership experience in bone regeneration, and specifically stem cell research; • A clear vision of the evolving nature of musculoskeletal research and a demonstrated ability to share their vision and to encourage others to partake in research activities in the Shriners research centre and to foster an interdisciplinary approach to research with the Shriners Hospital for Children, the Montreal Children’s Hospital, McGill University Health Centre and McGill University; • Experience in the administrative aspects of a research centre would be an asset. Candidates should hold the academic rank of Associate or Full Professor. McGill University is committed to equity in employment and diversity. It welcomes applications from Aboriginal persons, persons with disabilities, ethnic minorities, persons of minority sexual orientation or gender identity, visible minorities, women and others who may contribute to diversification. All qualified applicants are encouraged to apply; however, Canadians and permanent residents will be given priority. Application deadline date: September 30, 2014. Please forward signed letter of interest, C.V. and names of three referees to: Dr. Jean-Pierre Farmer, Chairman, Search Committee Chair, Department of Pediatric Surgery, McGill University c/o [email protected] S-757

288 J can chir, Vol. 57, No 4, août 2014

classified-aug14.indd 288 14-07-16 11:00 AM GET THE BEST GOLF CLUBS THIS SEASON!

GLEN ABBEY GOLF CLUB GLENCAIRN GOLF CLUB

WYNDANCE GOLF CLUB ROCKY CREST GOLF CLUB

SPECIAL CLUBLINK MEMBERSHIP PRICING FOR CMA MEMBERS!

CMA MEMBERS A ClubLink membership gives you reciprocal access to more than RECEIVE 55 championship courses in Ontario, Quebec and Florida, PLUS · Reciprocal access to all Member Clubs. Clubs are categorized as Silver, Lowest available membership fee pricing Gold, Platinum and Prestige to suit all budgets and preferences. · Preferred tee times at your Home Club Long-term interest-free ﬁ nancing · No assessment fees Or .... Pay your membership fee in full and · Guests are welcome receive a 20-per-cent discount! · Optional access to more than 230 golf courses and clubs worldwide through ClubCorp Golf 365 days a year! · Discounts at ClubLink golf shops · Intra- and inter-club leagues and competitions, tournaments, pro-ams and special events · Organized trips, special offers, and discounts at resorts in Muskoka and Florida · Banquets, weddings, meeting room and corporate function discounts

TO TAKE ADVANTAGE OF THIS SPECIAL CMA MEMBERSHIP OPPORTUNITY, PLEASE CALL A CLUBLINK MEMBERSHIP CONSULTANT AT 1-800-661-1818 OR EMAIL [email protected].

TO LEARN MORE ABOUT CLUBLINK, VISIT WWW.CLUBLINK.CA.

CMACMA ADAD 2014.indd2014.indd 1 2014-02-142014-02-14 33:59:59:59 PMPM covercjs-aug14.indd 6 14-07-16 10:04 AM CLINICAL CONGRESS 2014 October 26–30, 2014 | Moscone Center | San Francisco

The Best Surgical Education All in One Place. Clinical Congress 2014. Where you want to be.

The Surgeon of the Future

covercjs-aug14.indd 2 14-07-16 10:04 AM