Secondary Analyses from the HEALTH and FAITH Trials Guest Editors: Emil H

NOVEMBER 2020 VOLUME 34 NUMBER 11 SUPPLEMENT

Orthopaedic Trauma Association AOTrauma North America Belgian Orthopaedic Trauma Association Canadian Orthopaedic Trauma Society Foundation for Orthopedic Trauma International Society for Fracture Repair The Japanese Society for Fracture Repair

Outcomes After Hip Fractures: Secondary Analyses From the HEALTH and FAITH Trials Guest Editors: Emil H. Schemitsch, MD, FRSC(C) and Sheila Sprague, PhD

Publication provided by the Orthopaedic Trauma Association. Additional funding support provided by McMaster University.

The views and opinions expressed in this supplement are those of the authors and do not necessarily reﬂ ect the views of the editors of Journal of Orthopaedic Trauma. Volume 34 Number 11 Supplement November 2020

Supplemental digital content is Contents available for the article. Supplement Articles Si Introduction Emil H. Schemitsch and Sheila Sprague Sii Study Summaries FAITH Investigators and HEALTH Investigators* S1 The FAITH and HEALTH Trials: Are We Studying Different Hip Fracture Patient Populations? Michael Blankstein, Emil H. Schemitsch, Sofia Bzovsky, Daniel Axelrod, Rudolf W. Poolman, Frede Frihagen, Mohit Bhandari, Marc Swiontkowski, Sheila Sprague, and Patrick C. Schottel on behalf of the FAITH and HEALTH Investigators* S9 Arthroplasty Versus Internal Fixation for the Treatment of Undisplaced Femoral Neck Fractures: A Retrospective Cohort Study Shaikh Afaq, Nathan N. O'Hara, Emil H. Schemitsch, Sofia Bzovsky, Sheila Sprague, Rudolf W. Poolman, Frede Frihagen, Diane Heels-Ansdell, Mohit Bhandari, Marc Swiontkowski, and Gerard P. Slobogean on behalf of the FAITH and HEALTH Investigators* S15 Factors Associated With Mortality After Surgical Management of Femoral Neck Fractures Sofia Bzovsky, Marianne Comeau-Gauthier, Emil H. Schemitsch, Marc Swiontkowski, Diane Heels-Ansdell, Frede Frihagen, Mohit Bhandari, and Sheila Sprague on behalf of the FAITH and HEALTH Investigators* S22 Predictors of Loss to Follow-up in Hip Fracture Trials: A Secondary Analysis of the FAITH and HEALTH Trials Surabhi Sivaratnam, Marianne Comeau-Gauthier, Sheila Sprague, Emil H. Schemitsch, Rudolf W. Poolman, Frede Frihagen, Mohit Bhandari, Marc Swiontkowski, and Sofia Bzovsky on behalf of the FAITH and HEALTH Investigators* S29 What Predicts Health-Related Quality of Life for Patients With Displaced Femoral Neck Fractures Managed With Arthroplasty? A Secondary Analysis of the HEALTH Trial Daniel Axelrod, Marianne Comeau-Gauthier, Sofia Bzovsky, Emil H. Schemitsch, Rudolf W. Poolman, Frede Frihagen, Ernesto Guerra-Farfán, Diane Heels-Ansdell, Mohit Bhandari, and Sheila Sprague on behalf of the HEALTH Investigators*

(continued next page)

Journal of Orthopaedic Trauma (ISSN 0890-5339) is published monthly by Wolters Kluwer Health, Inc., at 14700 Citicorp Drive, Bldg 3, Hagerstown, MD 21742. Periodicals postage paid at Hagerstown, MD and at additional mailing offices. Postmaster: Send address changes to Journal of Orthopaedic Trauma, P.O. Box 1610, Hagerstown, MD 21740. Journal of Orthopaedic Trauma • Volume 34, Number 11 Supplement, November 2020

Contents (continued)

S37 Is Total Hip Arthroplasty a Cost-Effective Option for Management of Displaced Femoral Neck Fractures? A Trial-Based Analysis of the HEALTH Study Daniel Axelrod, Jean-Éric Tarride, Seper Ekhtiari, Gordon Blackhouse, Herman Johal, Sofia Bzovsky, Emil H. Schemitsch, Diane Heels-Ansdell, Mohit Bhandari, and Sheila Sprague on behalf of the HEALTH Investigators* S42 Predictors of Medical Serious Adverse Events in Hip Fracture Patients Treated With Arthroplasty David Neilly, David R. W. MacDonald, Sheila Sprague, Sofia Bzovsky, Daniel Axelrod, Rudolf W. Poolman, Frede Frihagen, Diane Heels-Ansdell, Mohit Bhandari, Emil H. Schemitsch and Iain M. Stevenson on behalf of the HEALTH Investigators S49 What Factors Increase Revision Surgery Risk When Treating Displaced Femoral Neck Fractures With Arthroplasty: A Secondary Analysis of the HEALTH Trial Michael Blankstein, Emil H. Schemitsch, Sofia Bzovsky, Rudolf W. Poolman, Frede Frihagen, Daniel Axelrod, Diane Heels-Ansdell, Mohit Bhandari, Sheila Sprague, and Patrick C. Schottel on behalf of the HEALTH Investigators* S55 Predictors of Long-Term Pain After Hip Arthroplasty in Patients With Femoral Neck Fractures: A Cohort Study Atefeh Noori, Sheila Sprague, Sofia Bzovsky, Emil H. Schemitsch, Rudolf W. Poolman, Frede Frihagen, Daniel Axelrod, Diane Heels-Ansdell, Mohit Bhandari, and Jason W. Busse on behalf of the HEALTH Investigators* S64 Who Did the Arthroplasty? Hip Fracture Surgery Reoperation Rates are Not Affected by Type of Training—An Analysis of the HEALTH Database Ryan D. DeAngelis, Gregory T. Minutillo, Matthew K. Stein, Emil H. Schemitsch, Sofia Bzovsky, Sheila Sprague, Mohit Bhandari, Derek J. Donegan, and Samir Mehta on behalf of the HEALTH Investigators* S70 Venous Thromboembolism in Hip Fracture Patients: A Subanalysis of the FAITH and HEALTH Trials David R. W. MacDonald, David Neilly, Prism S. Schneider, Sofia Bzovsky, Sheila Sprague, Daniel Axelrod, Rudolf W. Poolman, Frede Frihagen, Mohit Bhandari, Marc Swiontkowski, Emil H. Schemitsch, and Iain M. Stevenson on behalf of the FAITH and HEALTH Investigators* S76 Clockwise Torque of Sliding Hip Screws: Is There a Right Side? Franka S. Würdemann, Rudolf W. Poolman, Pieta Krijnen, Sofia Bzovsky, Sheila Sprague, Bart L. Kaptein, Johannes H. Hegeman, Emil H. Schemitsch, Mohit Bhandari, Marc Swiontkowski, and Inger B. Schipper on behalf of the Dutch Hip Fracture Audit Group and FAITH Investigators*

*See Contributors page for a complete listing of the HEALTH and FAITH Investigators. Journal of Orthopaedic Trauma • Volume 34, Number 11 Supplement, November 2020

Contents (continued)

Business and production offices are located at Two Commerce Square, 2001 Market St., Philadelphia, PA 19103. Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. The Journal of Orthopaedic Trauma was founded in partnership with the Orthopaedic Trauma Association (OTA) and is the official publication of the OTA. Annual Subscription Rates: United States—$822.00 Individual, $2,172.00 Institution, $335.00 In training. Rest of the World–$960.00 Individual, $2,454.00 Institution, $352.00 In training. Single copy rate $205. All prices include a handling charge. Subscriptions outside of North America must add $28.00 for airfreight delivery. United States residents of AL, CO, DC, FL, GA, HI, IA, ID, IN, KS, KY, LA, MD, MO, ND, NM, NV, PR, RI, SC, SD, UT, VT, WA, WV add state sales tax. The GST tax of 7% must be added to all orders shipped to Canada (Wolters Kluwer Health, Inc.’s GST Identification #895524239, Publications Mail Agreement #617687). Subscription prices outside the United States must be prepaid. Prices subject to change without notice. Visit us online at www.lww.com. Individual and in-training subscription rates include print and access to the online version. Institutional rates are for print only; online subscriptions are available via Ovid. Institutions can choose to purchase a print and online subscription together for a discounted rate. Institutions that wish to purchase a print subscription, please contact Wolters Kluwer Health, Inc. at 14700 Citicorp Drive, Bldg 3, Hagerstown, MD 21742; phone 800-638-3030 (outside the United States 301-223-2300); fax 301-223-2400. Institutions that wish to purchase an online subscription or online with print, please contact the Ovid Regional Sales Office near you or visit www.ovid.com/site/index.jsp and select Contact and Locations. Website: www.jorthotrauma.com

Listed in Index Medicus/MEDLINE, Current Contents/Clinical Medicine, Scisearch, Research Alert, Focus On: Sports Science and Medicine, Sociedad Iberoamerican de Información Científica.

Address for non-member subscription information, orders, or change of address (except Japan): 14700 Citicorp Drive, Bldg 3, Hagerstown, MD 21742; phone 800-638-3030 (outside the United States 301-223-2300); fax 301-223-2400. In Japan, contact Wolters Kluwer Health Japan Co., Ltd., Forecast Mita Building 5th floor, 1-3-31 Mita Minato-ku, Tokyo, Japan 108-0073. Telephone: +81 3 5427 1969. E-mail: [email protected]

Wolters Kluwer Health cannot be held responsible for errors or for any consequences arising from the use of the information contained in this journal. The appearance of scientific reports and/or workshops, or any other material in the Journal of Orthopaedic Trauma does not constitute an endorsement or approval by Wolters Kluwer Health of the findings, data, conclusions, recommendations, procedures, results, or any other aspect of the content of such articles. The appearance of advertising in the Journal of Orthopaedic Trauma does not constitute an endorsement or approval by Wolters Kluwer Health, Inc. of the quality or value of the products advertised or any of the claims, data, conclusions, recommendations, procedures, results or any other information included in the advertisements.

Permissions and photocopying: For permission and/or rights to use content for which the copyright holder is Wolters Kluwer or the society we have partnered with the Copyright Clearance Center to provide permissions for our products through their RightsLink service, please go to the journal’s website and after clicking on the relevant article, click on the “Get Content & Permissions” link under the “Article Tools” box that appears on the right side of the page. For questions about the Rightslink service, e-mail [email protected] or call 877-622-5543 (U.S. Only) or 978-777-9929. Permissions FAQs and information on author’s permission requests are available at https://shop.lww.com/journal-permission. For additional permission inquiries, please contact [email protected]. For translation rights requests, contact [email protected]. For license to republish and distribute requests, contact HealthLicensing@ wolterskluwer.com. For special projects and reprints (U.S./Canada), contact Alan Moore at [email protected] or [email protected]. For special projects and reprints (non-U.S./Canada), contact Avia Potashnik at [email protected] or [email protected].

Instructions for Authors may be accessed on the journal’s website at www.jorthotrauma.com. SUPPLEMENT ARTICLE

Contributors FAITH Investigators Steering Committee: Mohit Bhandari (Chair, McMaster University), Marc Swiontkowski (University of Minnesota), PJ Devereaux (McMaster University), Gordon Guyatt (McMaster University), Martin J. Heetveld (Spaarne Gasthuis, Haarlem), Kyle Jeray (Greenville Health System), Susan Liew (The Alfred), Martin Richardson (University of Melbourne), Emil H. Schemitsch (University of Western Ontario), Lehana Thabane (McMaster University), Paul Tornetta III (Boston University Medical Center), and Stephen D. Walter (McMaster University) Global Methods Centre: Mohit Bhandari (Principal Investigator); Sheila Sprague (Research Methodologist); Paula McKay (Manager); Taryn Scott, Alisha Garibaldi, Helena Viveiros, Marilyn Swinton, (Research Coordination); Mark Gichuru, Sofia Bzovsky (Adjudication Coordination); Diane Heels-Ansdell, Qi Zhou (Statistical Analysis); Lisa Buckingham, Aravin Duraikannan (Data Management); Deborah Maddock, Nicole Simunovic (Grants Management) (McMaster University) United States Methods Centre: Marc Swiontkowski (Principal Investigator); Julie Agel (Research Coordination) (University of Minnesota) Netherlands Method Centre: Martin J. Heetveld (Principal Investigator); Esther M.M. Van Lieshout (Research Coordination); Stephanie M. Zielinski (Trial Coordination) (Erasmus MC, University Medical Center Rotterdam) United Kingdom Method Centre: Amar Rangan (Principal Investigator), Birgit C. Hanusch, Lucksy Kottam, Rachel Clarkson (Research Coordination) (The James Cook University Hospital) Adjudication Committee: Gregory J. Della Rocca (Chair) (Duke University), Robert Haverlag (Onze Lieve Vrouwe Gasthuis), Susan Liew (The Alfred), Gerard P. Slobogean (University of Maryland, Baltimore), Kyle Jeray (Greenville Health System) Participating Clinical Sites: Canada Royal Columbian Hospital / Fraser Health Authority / University of British Columbia - Robert McCormack, Kelly Apostle, Dory Boyer, Farhad Moola, Bertrand Perey, Trevor Stone, Darius Viskontas, H. Michael Lemke, Mauri Zomar, Karyn Moon, Raely Moon, Amber Oatt Foothills Medical Centre - Richard E. Buckley, Paul Duffy, Robert Korley, Shannon Puloski, James Powell, Kelly Johnston, Kimberly Carcary, Melissa Lorenzo, Ross McKercher London Health Sciences Centre - David Sanders, Mark MacLeod, Abdel-Rahman Lawendy, Christina Tieszer Sunnybrook Health Sciences Centre - David Stephen, Hans Kreder, Richard Jenkinson, Markku Nousiainen, Terry Axelrod, John Murnaghan, Diane Nam, Veronica Wadey, Albert Yee, Katrine Milner, Monica Kunz, Wesley Ghent St. Michael’s Hospital - Emil H. Schemitsch, Michael D. McKee, Jeremy A. Hall, Aaron Nauth, Henry Ahn, Daniel B. Whelan, Milena R. Vicente, Lisa M. Wild, Ryan M. Khan, Jennifer T. Hidy Queen Elizabeth II Health Sciences Centre - Chad Coles, Ross Leighton, Michael Biddulph, David Johnston, Mark Glazebrook, David Alexander, Catherine Coady, Michael Dunbar, J. David Amirault, Michael Gross, William Oxner, Gerald Reardon, C. Glen Richardson, J. Andrew Trenholm, Ivan Wong, Kelly Trask, Shelley MacDonald, Gwendolyn Dobbin Human Mobility Research Centre, Queen’s University and Kingston General Hospital - Ryan Bicknell, Jeff Yach, Davide Bardana, Gavin Wood, Mark Harrison, David Yen, Sue Lambert, Fiona Howells, Angela Ward Oakville Trafalgar Memorial Hospital - Paul Zalzal, Heather Brien, V. Naumetz, Brad Weening, Nicole Simunovic The Ottawa Hospital - Eugene K. Wai, Steve Papp, Wade T. Gofton, Allen Liew, Stephen P. Kingwell, Garth Johnson, Joseph O’Neil, Darren M. Roffey, Vivian Borsella Juravinski Hospital and Cancer Centre - Victoria Avram United States Boone Hospital Center – Columbia Orthopaedic Group - Todd M. Oliver, Vicki Jones, Michelle Vogt Orthopaedic Associates of Michigan - Clifford B. Jones, James R. Ringler, Terrence J. Endres, Debra L. Sietsema, Jane E. Walker Greenville Health System - Kyle J. Jeray, J. Scott Broderick, David R. Goetz, Thomas B. Pace, Thomas M. Schaller, Scott E. Porter, Michael L. Beckish, John D. Adams, Benjamin B. Barden, Aaron T. Creek, Stephen H. Finley, Jonathan L. Foret, Garland K. Gudger Jr, Richard W. Gurich Jr, Austin D. Hill, Steven M. Hollenbeck, Lyle T. Jackson, Kevin K. Kruse, III, Wesley G. Lackey, Justin W. Langan, Julia Lee, Lauren C. Leffler, Timothy J. Miller, R. Lee Murphy, Jr., Lawrence K. O’Malley II, Melissa E. Peters, Dustin M. Price, John A. Tanksley, Jr., Erick T. Torres, Dylan J. Watson, Scott T. Watson, Stephanie L. Tanner, Rebecca G. Snider, Lauren A. Nastoff, Shea A. Bielby, Robert J. Teasdall Regions Hospital-University of Minnesota - Julie A. Switzer, Peter A. Cole, Sarah A. Anderson, Paul M. Lafferty, Mengnai Li, Thuan V. Ly, Scott B. Marston, Amy L. Foley, Sandy Vang, David M. Wright

J Orthop Trauma • Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com Contributors J Orthop Trauma • Volume 34, Number 11 Supplement, November 2020

Lahey Hospital & Medical Center - Andrew J. Marcantonio, Michael S.H. Kain, Richard Iorio, Lawrence M. Specht, John F. Tilzey, Margaret J. Lobo, John S. Garfi MetroHealth Medical Center - Heather A. Vallier, Andrea Dolenc, Mary Breslin Miami Valley Hospital - Michael J. Prayson, Richard Laughlin, L. Joseph Rubino, Jedediah May, Geoffrey Ryan Rieser, Liz Dulaney-Cripe, Chris Gayton St. Elizabeth Youngstown Hospital - James Shaer, Tyson Schrickel, Barbara Hileman University of Rochester Medical Center - John T. Gorczyca, Jonathan M. Gross, Catherine A. Humphrey, Stephen Kates, John P. Ketz, Krista Noble, Allison W. McIntyre, Kaili Pecorella Colorado Orthopedic Consultants - Craig A. Davis, Stuart Weinerman, Peter Weingarten, Philip Stull, Stephen Lindenbaum, Michael Hewitt, John Schwappach, Janell K. Baker, Tori Rutherford, Heike Newman, Shane Lieberman, Erin Finn, Kristin Robbins, Meghan Hurley, Lindsey Lyle, Khalis Mitchell, Kieran Browner, Erica Whatley, Krystal Payton, Christina Reeves St. Louis University Hospital - Lisa K. Cannada, David E. Karges, Sarah A. Dawson University of Pennsylvania - Samir Mehta, John Esterhai, Jaimo Ahn, Derek Donegan, Annamarie D. Horan, Patrick J. Hesketh, Evan R. Bannister Northwest Orthopaedic Specialists - Jonathan P. Keeve, Christopher G. Anderson, Michael D. McDonald, Jodi M. Hoffman University of Pittsburgh Medical Center - Ivan Tarkin, Peter Siska, Gary Gruen, Andrew Evans, Dana J. Farrell, James Irrgang, Arlene Luther Mayo Clinic - William W. Cross III, Joseph R. Cass, Stephen A. Sems, Michael E. Torchia, Tyson Scrabeck Texas Tech University Health Sciences Center – Lubbock - Mark Jenkins, Jules Dumais, Amanda W. Romero University Orthopaedic Associates, LLC - Carlos A. Sagebien, Mark S. Butler, James T. Monica, Patricia Seuffert US Army Institute of Surgical Research - Joseph R. Hsu, Daniel Stinner, James Ficke, Michael Charlton, Matthew Napierala, Mary Fan Boston University Medical Center - Paul Tornetta III, Chadi Tannoury, Hope Carlisle, Heather Silva UC Health/University of Cincinnati Medical Center - Michael Archdeacon, Ryan Finnan, Toan Le, John Wyrick, Shelley Hess Scott and White Memorial Hospital - Michael L. Brennan, Robert Probe, Evelyn Kile, Kelli Mills, Lydia Clipper, Michelle Yu, Katie Erwin Geisinger Medical Center - Daniel Horwitz, Kent Strohecker, Teresa K. Swenson Hennepin County Medical Center - Andrew H. Schmidt, Jerald R. Westberg Kaiser Permanente - Kamran Aurang, Gary Zohman, Brett Peterson, Roger B. Huff OrthoIndy Trauma St. Vincent Trauma Center - Joseph Baele, Timothy Weber, Matt Edison Santa Clara Valley Medical Center - Jessica Cooper McBeth Indiana University – Eskenazi Health Services - Karl Shively, Janos P. Ertl, Brian Mullis, J. Andrew Parr, Ripley Worman, Valda Frizzell, Molly M. Moore Mission Hospital Research Institute - Charles J. DePaolo, Rachel Alosky, Leslie E. Shell, Lynne Hampton, Stephanie Shepard, Tracy Nanney, Claudine Cuento Dartmouth-Hitchcock Medical Center - Robert V. Cantu, Eric R. Henderson, Linda S. Eickhoff Denver Health and Hospital Authority - E. Mark Hammerberg, Philip Stahel, David Hak, Cyril Mauffrey, Corey Henderson, Hannah Gissel, Douglas Gibula University of California Irvine Medical Center - David P. Zamorano, Martin C. Tynan, Deeba Pourmand, Deanna Lawson University of Missouri Health Care - Gregory J. Della Rocca, Brett D. Crist, Yvonne M. Murtha, Linda K. Anderson Covenant Healthcare of Saginaw - Colleen Linehan, Lindsey Pilling Hartford Hospital - Courtland G. Lewis, Stephanie Caminiti, Raymond J. Sullivan, Elizabeth Roper Vanderbilt University Medical Center - William Obremskey, Philip Kregor, Justin E. Richards, Kenya Stringfellow Western Slope Study Group - Michael P. Dohm, Abby Zellar www.jorthotrauma.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. J Orthop Trauma • Volume 34, Number 11 Supplement, November 2020 Contributors

The Netherlands St. Antonius Ziekenhuis - Michiel J.M. Segers, Jacco A.C. Zijl, Bart Verhoeven, Anke B. Smits, Jean Paul P.M. de Vries, Bram Fioole, Henk van der Hoeven, Evert B.M. Theunissen, Tammo S. de Vries Reilingh, Lonneke Govaert, Philippe Wittich, Maurits de Brauw, Jan Wille, Peter M.N.Y.M. Go, Ewan D. Ritchie, Ronald N. Wessel, Eric R. Hammacher Kennemer Gasthuis - Martin J. Heetveld, Gijs A. Visser, Heyn Stockmann, Rob Silvis, Jaap P. Snellen, Bram Rijbroek, Joris J.G. Scheepers, Erik G.J. Vermeulen, Michiel P.C. Siroen, Ronald Vuylsteke, Hans L.F. Brom, Herman Rijna IJsselland Ziekenhuis - Piet A.R. de Rijcke, Cees L. Koppert, Steven E. Buijk, Richard P.R. Groenendijk, Imro Dawson, Geert W.M. Tetteroo, Milko M.M. Bruijninckx, Pascal G. Doornebosch, Eelco J.R. de Graaf Reinier de Graaf Gasthuis - Maarten van der Elst, Carmen C. van der Pol, Martijne van ’t Riet, Tom M. Karsten, Mark R. de Vries, Laurents P.S. Stassen, Niels W.L. Schep, G. Ben Schmidt, W.H. Hoffman Onze Lieve Vrouwe Gasthuis - Rudolf W. Poolman, Maarten P. Simons, Frank H.W.M. van der Heijden, W. Jaap Willems, Frank R.A.J. de Meulemeester, Cor P. van der Hart, Kahn Turckan, Sebastiaan Festen, Frank de Nies, Robert Haverlag, Nico J.M. Out, Jan Bosma University Medical Center St. Radboud - Albert van Kampen, Jan Biert, Arie B. van Vugt, Michael J.R. Edwards, Taco J. Blokhuis, Jan Paul M. Frölke, Leo M.G. Geeraedts, Jean W.M. Gardeniers, Edward C.T.H. Tan, Lodewijk M.S.J. Poelhekke, Maarten C. de Waal Malefijt, Bart Schreurs Maasstad Ziekenhuis - Gert R Roukema, Hong A. Josaputra, Paul Keller, Peter D. de Rooij, Hans Kuiken, Han Boxma, Berry I. Cleffken, Ronald Liem Medisch Centrum Haaglanden - Steven J. Rhemrev, Coks H.R. Bosman, Alexander de Mol van Otterloo, Jochem Hoogendoorn, Alexander C. de Vries, Sven A.G. Meylaerts St. Elisabeth Ziekenhuis - Michiel H.J. Verhofstad, Joost Meijer, Teun van Egmond, Frank H.W.M. van der Heijden, Igor van der Brand Erasmus MC, University Medical Center Rotterdam - Peter Patka, Martin G. Eversdijk, Rolf Peters, Dennis Den Hartog, Oscar J.F. Van Waes, Pim Oprel Tergooi Ziekenhuizen - Harm M van der Vis, Martin Campo, Ronald Verhagen, G.H. Robert Albers, Arthur W. Zurcher University Medical Center Utrecht - Rogier K.J. Simmermacher, Jeroen van Mulken, Karlijn van Wessem, Taco J. Blokhuis, Steven M. van Gaalen, Luke P.H. Leenen Bronovo Ziekenhuis - Maarten W.G.A. Bronkhorst, Onno R. Guicherit Academic Medical Center - J. Carel Goslings, Robert Haverlag, Kees Jan Ponsen Germany University of Aachen Medical Center - Hans-Christoph Pape, Matthias Knobe, Roman Pfeifer Norway Oslo University Hospital - Frede Frihagen, John Clarke-Jenssen, Geir Hjorthaug, Torben Ianssen, Asgeir Amundsen, Jan Egil Brattgjerd, Tor Borch, Berthe Bøe, Bernhard Flatøy, Sondre Hasselund, Knut Jørgen Haug, Kim Hemlock, Tor Magne Hoseth, Geir Jomaas, Thomas Kibsgård, Tarjei Lona, Gilbert Moatshe, Oliver Müller, Marius Molund, Tor Nicolaisen, Fredrik Nilsen, Jonas Rydinge, Morten Smedsrud, Are Stødle, Axel Trommer, Stein Ugland, Anders Karlsten, Guri Ekås, Elise Berg Vesterhus, Anne Christine Brekke United Kingdom University Hospital of North Tees - Peter Hull, Sophie Lewis, Simone Evans (Cambridge University Hospitals, England); Rajesh Nanda, Rajanikanth Logishetty, Sanjeev Anand, Carol Bowler University Hospital of North Durham and Darlington Memorial Hospital - Andrew Jennings, Graham Chuter, Glynis Rose, Gillian Horner Wexham Park Hospital - Callum Clark, Kate Eke Northumbria Healthcare NHS Foundation Trust - Mike Reed, Dominic Inman, Chris Herriott, Christine Dobb Australia The Alfred - Susan Liew, Harvinder Bedi, Ashley Carr, Hamish Curry, Andrew Chia, Steve Csongvay, Craig Donohue, Stephen Doig, Elton Edwards, Greg Etherington, Max Esser, Andrew Gong, Arvind Jain, Doug Li, Russell Miller, Ash Moaveni, Matthias Russ, Lu Ton, Otis Wang, Adam Dowrick, Zoe Murdoch, Claire Sage Royal Brisbane and Women’s Hospital - Kevin Tetsworth, Geoff Donald, Patrick Weinrauch, Paul Pincus, Steven Yang, Brett Halliday, Trevor Gervais, Michael Holt, Annette Flynn Royal Melbourne Hospital - Marinis Pirpiris, David Love, Andrew Bucknill, Richard J Farrugia

India RLB Hospital and Research Centre - Mahesh Bhatia, Vinod Arora, Vivek Tyagi Nirmal Hospital - Ajay Gupta, Neeraj Jain, Farah Khan Satellite Orthopaedic Hospital and Research Centre - Ateet Sharma, Amir Sanghavi, Mittal Trivedi Highway Hospital - Anil Rai, Subash, Kamal Rai Popular Hospital - Vineet Yadav, Sanjay Singh, Kamal Rai Madhuraj Nursing Home - Amal Shankar Prasad, Vimlesh Mishra M.S. Rammaiah Medical College & Hospital - D.C. Sundaresh, Angshuman Khanna St John’s Medical College Hospital - Joe Joseph Cherian, Davy J Olakkengil, Gaurav Sharma Sunshine Hospital - Akhil Dadi, Naveen Palla, Utsav Ganguly Unity Health Complex - B. Sachidananda Rai, Janakiraman Rajakumar HEALTH Investigators Steering Committee: Mohit Bhandari (Chair, McMaster University), P.J. Devereaux (McMaster University), Thomas A. Einhorn (New York University Langone Medical Center), Frede Frihagen (Oslo University Hospital), Ernesto Guerra-Farfán (Hospital Vall d'Hebrón), Kenneth J. Koval (Orlando Regional Medical Centre), Rudolf W. Poolman (OLVG), Emil H. Schemitsch (University of Western Ontario), Kevin Tetsworth (University of Queensland), Lehana Thabane (McMaster University), Stephen D. Walter (McMaster University), Gordon H. Guyatt (McMaster University). Global Methods Centre: Mohit Bhandari (Principal Investigator); Sheila Sprague (Research Director); Paula McKay (Manager); Kim Madden, Sofia Bzovsky, Kerry Tai, Taryn Scott, Marilyn Swinton (Research Coordination); Naveen Khan (Adjudication Coordination); Diane Heels-Ansdell (Statistical Analysis); Lisa Buckingham, Aravin Duraikannan (Data Management); (McMaster University) United States Methods Centre: Thomas A. Einhorn (Principal Investigator); Heather Silva (Research Coordination) (Boston University Medical Center). Thomas A. Einhorn, David Novikov, Daniel P. Waren (New York University) Netherlands Methods Centre: Ydo V. Kleinlugtenbelt, Ellie B.M. Landman (Deventer Ziekenhuis; Definitive). Martin J. Heetveld, Rudolf W. Poolman, Esther M.M. Van Lieshout, Paul T.P.W. Burgers (Erasmus MC, University Medical Center Rotterdam, Rotterdam; Pilot) United Kingdom Methods Centre: Amar Rangan (The James Cook University Hospital, South Tees Hospitals NHS Foundation Trust, University of Oxford, University of York), Birgit Hanusch, Lucksy Kottam (The James Cook University Hospital, South Tees Hospitals NHS Foundation Trust) Central Adjudication Committee: Robert D. Zura (Chair, Louisiana State University Health Sciences Center), Victoria Avram (McMaster University), Ajay Manjoo (McMaster University) Data and Safety Monitoring Board (CIHR): John Antoniou (Chair), Tim Ramsay, Earl R. Bogoch, Andrew Trenholm Data and Safety Monitoring Board (NIH): Stephen Lyman (Chair), Madhu Mazumdar, Kevin J. Bozic, Mark Luborsky, Stuart Goodman, Susan Muray, Benjamin K. Potter Statistical Support: Diane Heels-Ansdell, Lehana Thabane, Nathan N. O’Hara, Gerard P. Slobogean, Quazi Ibrahim Participating Clinical Sites: Canada University of Calgary - Richard Buckley, Robert Korley, Paul Duffy, Shannon Puloski, Eldridge Batuyong, Christopher Martin, James MacKenzie, Jason Werle, Kelly Johnston, Neil White, Marcia Clark, Gregory Abelseth, Stephen Hunt, Justin LeBlanc, Prism Schneider, Kimberly Carcary, Ross McKercher, Lisa Murphy, Melissa Lorenzo, Tanja Harrison, Leah Schultz, Aftab Akbari, Stephanie Yee, Tina Samuel, Saboura Mahdavi University of British Columbia/Fraser Health Authority/Royal Columbian Hospital - Trevor B. Stone, Bertrand Perey, Farhad Moola, Darius Viskontas, Kelly Apostle, Dory Boyer, H. Michael Lemke, Robert McCormack, Mauri Zomar, Bindu Mohan, Karyn Moon, Raely Pritchard, Brenda Chen Fan, Kyrsten Payne St. Michael’s Hospital - Aaron Nauth, Emil H. Schemitsch, Michael D. McKee, Jeremy A. Hall, Daniel Whelan, Sarah Ward, David Walmsley, Earl R. Bogoch, James P Waddell, Henry Ahn, Timothy R. Daniels, Milena R. Vicente, Jennifer T. Hidy, Melanie T. MacNevin, Paril Suthar Sunnybrook Health Sciences Centre - Hans Kreder, David Stephen, Richard Jenkinson, Markku Nousiainen, Terry Axelrod, Veronica Wadey, Sebastian Tomescu, Monica Kunz, Katrine Milner, Melanie MacNevin, Ria De Gorter (Cagaanan), Wesley Ghent, Fathima Adamsahib, Araby Sivananthan, Aimee Theriault, Michelle Arakgi, Phumeena Balasuberamaniam, Ravi Tuazon www.jorthotrauma.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. J Orthop Trauma • Volume 34, Number 11 Supplement, November 2020 Contributors

Vancouver General Hospital - Peter J. O’Brien, Piotr A. Blachut, Henry M. Broekhuyse, Pierre Guy, Kelly A. Lefaivre, Gerard P. Slobogean, Dean G. Malish, Raman Johal, Jessica Peattie, Abdullah Mamun, Benita Okocha, Irene Leung Hôpital de l’Enfant-Jésus – Étienne Belzile, Luc Bedard, Martin Bédard, Bernard Laliberté, Jean Lamontagne, Martin Lesieur, Luc Petitclerc, Marie-ève Roger, Stéphane Pelet, Luc Lemire, Hélène Côté Queen Elizabeth II Health Sciences Centre - Chad Coles, Ross Leighton, C. Glen Richardson, Michael Biddulph, Michael Gross, Michael Dunbar, J. David Amirault, David Alexander, Catherine Coady, Mark Glazebrook, David Johnston, William Oxner, Gerald Reardon, Ivan Wong, Kelly Trask, Shelley MacDonald Memorial University of Newfoundland - Andrew Furey, Craig Stone, Keegan Au, William Moores, Peter Rockwood, Carl Moores, Daniel Squire, Sarah Anthony, Minnie Parsons, Valisha Keough, Erin Baker Lakeridge Health – Samir Chhabra, Osama Gharsaa, Abdulmajid Ibrahim, Esmat Dessouki, Omar Dessouki, Aileen Manganaro, Kelly Fusco Hôpital Saint-François d'Assise – François Marquis, Annie Arteau, Marc Bouchard, Alexandre Leclerc, Michèle Angers, Sylvain Belzile, Sylvie Turmel Hôpital du Sacré-Coeur de Montréal – G-Yves Laflamme, Benoit Benoit, Pierre Ranger, Michel Malo, Julio Fernandes, Julie Fournier, Karine Tardif Ottawa Hospital Research Institute, The Ottawa Hospital – Robert J. Feibel, Hesham Abdelbary, Christopher Kennedy, Paul Kim, Paul Beaule, Joel Werier, Alberto Carli, Peter Lapner, Geoff Dervin, Katie McIlquham, Sara A. Ruggiero, Carly St. Germain, Heather Cosgrove, Johanna Dobransky, Cheryl Kreviazuk Humber River Hospital – Sebastian Rodriguez-Elizalde, Anna Rose Soliman, Sagar Desai, John Townley, Iriss Mariano, Theresa B. Pedutem, Mary June Francisco United States Lahey Hospital and Medical Center - Andrew Marcantonio, Michael Kain, Lawrence Specht, John Tilzey, John Garfi Mayo Clinic – Andrew Sems, Tab Mabry, Joseph Cass, Michael Torchia, William Cross, Michael Taunton, Joaquin Sanchez-Sotelo, Rafael Sierra, Barbara Foreman Marshall University - Franklin D. Shuler, James Day, Tigran Garabekyan, Felix Cheung, Ali Oliashirazi, Jonathon Salava, Linda Morgan, Timothy Wilson-Byrne, Grant Buchanan, Milad Modarresi, Mary Beth Cordle Indiana University – Brian Mullis, Karl Shively, Andrew Parr, Janos Ertl, Ripley Worman, Mark Webster, Judd Cummings, Valda Frizzell, Molly Moore New York University – Thomas A. Einhorn, Kenneth A. Egol, Nirmal C. Tejwani, Roy I. Davidovitch, Ran Schwarzkopf, Jonathan M. Vigdorchik, James D. Slover, Sanjit R. Konda, Toni M. McLaurin, David Novikov, Pankajkumar Patel, Daniel P. Waren Orthopaedic Associates of Michigan – Clifford B. Jones, James R. Ringler, Debra L. Sietsema University Orthopaedic Associates – Carlos Sagebien, David Harwood, Stephen Kayiaros, Patricia Seuffert The CORE Institute – Clifford B. Jones, Russell Meldrum, Sarim S. Ahmed, Debra L. Sietsema Texas Tech University – Enes Kanlic, Amr Abdelgawad, Juan Shunia University of Pennsylvania – Samir Mehta, John L. Esterhai Jr, Jaimo Ahn, Derek Donegan, Atul Kamath, Neil Sheth, Annamarie Horan, Patrick Hesketh, Thomas Rose Duke University Medical Center – Rachel Reilly, Robert D. Zura, Steven Olson, Kendra Marr, Maria Manson, Cameron Howes University of Mississippi Medical Center – Benjamin M. Stronach, Patrick F. Bergin, Jennifer S. Barr, Clay A. Spitler, Josie M. Hydrick, Leslie N. Johnson Boston University Medical Center – Thomas A. Einhorn, Paul Tornetta III, William R. Creevy, Heather Silva, Michelle J. Lespasio, Hope Carlisle Colorado Orthopedic Consultants – Craig Davis, Philip Stull, Stewart Weinerman, Peter Weingarten, Steven Lindenbaum, Michael Hewitt, Rebecca Danielwicz, Janell Baker Mission Hospital Research Institute – Charles DePaolo, Christina Riggsbee, Tracey Nanney Park Nicollet Institute – Gregg Strathy, Kathleen Peter, Paul Johnson, and Meaghan Morton Rubin Institute for Advanced Orthopaedics – Michael Mont, Donald E. Delanois, Bhaveen Kapadia, Kimona Issa, and Marylou Mullen

SUNY Buffalo – Mark Anders, Christopher Mutty, Matthew Phillips, Sridhar Rachala, Mary Bayers-Thering Allegheny – Timothy J. Sauber, Edward Westrick, Joseph Guth, Michael H. Maher University of Utah - Thomas F. Higgins, Erik N. Kubiak, Jeremy Gililland, David Rothberg, Christopher Peters, Christopher Pelt, Ami R. Stuart, Kirby Corbey, Ashley Neese, Mark T. Russell, Victor Solis, Andrea Gurule University of California Irvine - David Zamorano, John A. Scolaro, Douglas Kiester, Deeba Pourmand St. Elizabeth Youngstown Hospital - Tyson T. Schrickel, John V. Gentile, Barbara M. Hileman, Elisha A. Chance Rothman Institute – Javad Parvizi, Tiffany Morrison Emory University School of Medicine – James Roberson, Thomas Bradbury, Greg Erens, Kyle Webb. University of Arizona – Michael Dohm, Cindy Fastje, José Luis Camarena, Maria Gordon The Netherlands Deventer Ziekenhuis – Ydo V. Kleinlugtenbelt, Hans-Peter W. van Jonbergen, Willem H. Roerdink, Joost M. Reuver, Alexander F.W. Barnaart, Elvira R. Flikweert, Danielle D. Langeloo, Rinco C.T. Koorevaar, Hannie H.F. Elskamp-Meijerman, Ellie B.M. Landman Amphia Ziekenhuis - Leon H.G.J. Elmans, Joost A.A.M. van den Hout, Adrianus J.P. Joosten, Ad F.A. van Beurden, Stefan B.T. Bolder, Denise Eygendaal, Adrianus F.C.M. Moonen, Rutger C.I. van Geenen, Eric A. Hoebink, Robert Wagenmakers, Wouter van Helden Tergooiziekenhuizen - Harm M. van der Vis, Lijkele Beimers, Jasper de Vries, Arthur W. Zurcher, G.H. Rob Albers, Maarten Rademakers, Stefan Breugem, Ibo van der Haven, Pieter Jan Damen, Gythe H. Bulstra, Martin M. Campo, Mathijs P. Somford, Daniël Haverkamp OLVG - Rudolf W. Poolman, Bauke W. Kooistra, Willem Jan Kleyn Molekamp, Frank R.A.J. de Meulemeester, Arthur E.B. Kleipool, Robert Haverlag, Maarten P. Simons, Eduard L.A.R. Mutsaerts, Diederik H.R. Kempen, Mark C. Altena, Dirk Jan F. Moojen, John (S.J.) Ham, Arnard van der Zwan, Derek F.P. van Deurzen, Michel P.J. van den Bekerom, Vanessa A.B. Scholtes, Mariëlla E.M. Volkers, Amanda D. Klaassen, Loes W.A.H. van Beers, Nienke W. Willigenburg, Ihsan Amajjar, Joint Research Group Isala Clinic – Sven H. van Helden, R.G. Zuurmond, Adriaan Mostert, P. Houben, G. Solinge, W.R. Spanjersberg, R. Nijveldt, K. van Egmond, Lonneke Buitenhuis Gelre Ziekenhuizen - Hugo W. Bolhuis, Pieter H.J. Bullens, Mike Hogervorst, Karin E. de Kroon, Rob H. Jansen, Ferry Steenstra, Eric E.J. Raven Spaarne Ziekenhuis - Peter A. Nolte, Michel P.J. van den Bekerom, Tjitte de Jong, Arthur van Noort, Diederik A. Vergroesen, Bernard G. Schutte Diaconessenhuis Leiden - Rover Krips, J. Bernard Mullers, Hans Schüller Medisch Centrum Haaglanden - Stefan B. Keizer, Jan-Willem A. Swen, Peter H.C. den Hollander, Bregje J.W. Thomassen IJsselland Ziekenhuis - W. Peter J. Fontijne, Saskia C. Wiersma, Bastiaan Boetes, Edgar J.T. ten Holder Ruwaard van Putten Ziekenhuis - Rob Kooijman, Roelf R. Postema, René J.T.M. Bleker, Harald I.H. Lampe Slotervaartziekenhuis - Lein Schuman, John Cheung, Frank van Bommel, W. Paul C.A. Winia, Daniel Haverkamp, Harm van der Vis Leids Universitair Medisch Centrum - Huub J.L. van der Heide, Jochem Nagels, Enrike H.M.J. van der Linden-van der Zwaag Gelderse Vallei - Jan P. van Dijk, Wouter H. van Helden Flevoziekenhuis - Mark L.M. Falke, Frans J. Kurek, Adrianus C.H. Slingerland Spain Hospital Vall d'Hebrón - Ernesto Guerra Farfán, Jordi Teixidor Serra, Jordi Tomás Hernández, Vicente Molero García, Jordi Selga Marsà, Juan Antonio Porcel Vázquez, Joan Minguell Monyart, Maria Villar Casares, Jaume Mestre Torres, Sandra Ponce Ruiz, Teresa Sanclemente Boli, Judith Sánchez-Raya, José Andrés Peiró, Diego Collado Gastalver, Víctor Barro Ojeda, Iñaki Mimendia Sancho, Maria Jurado Ruiz, Carla Carbonell Rosell, M. Cristina García Martínez, Luken Zubizarreta Barrutia, Martí Plomer Sánchez, Felipe Moreira Borim, Jorge Nuñez Camarena and Antoni Fraguas Castany, Yaiza García Sánchez, Míriam Garrido Clua Hospital Universitari Mútua de Terrassa – Pablo Castillón, Martí Bernaus, Francesc Anglés, Agustí Bartra Hospital Dr. Josep Trueta - Miguel Angel Froufe Siota, Diana Noriego, Samer Al-dirra, Xavier Madirolas, Cristina Martinez Merchan Hospital Universitario Costa del Sol - Enrique Guerado, Encarnación Cruz Ocaña, Juan Ramon Cano Manuel Godino, Miguel Hirschfeld, Laura Ramos, Antonio Royo, Adolfo Galán, Paqui Fontalba, Pilar Gavira www.jorthotrauma.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. – Si | 2019; 2006;17: 2011;45:15 N Engl J Med. 1363. – Osteoporos Int. 289. – Indian J Orthop. c 166. , – b 1527. – 1997;157:1357 xation in the operative management of 1992;2:285 fi CMAJ. . Canada Year Book, 2011. Available at: https:// REFERENCES 1990;252:163 2017;389:1519 ACKNOWLEDGMENTS Osteoporos Int. Lancet. 2208. – 1733. – The authors thank Mohit Bhandari (Principal The data collected as part of the FAITH and HEALTH a Bzovsky for leading the coordination of this supplement, RTICLE fi 22. Statistics Canada. Seniors www150.statcan.gc.ca. Accessed June 23, 2020. ability associated with1726 osteoporotic fractures. States. Numbers, costs, andClin potential Orthop effects Relat of Res. postmenopausal estrogen. of hip fracture in Canada. wide projection. fracture: worldwide geographic variation. (FAITH) Investigators. Fracture hip fractures (FAITH):trolled trial. an international, multicentre, randomised con- hip arthroplasty or hemiarthroplasty for381:2199 hip fracture. and Sheila Sprague, PhD A 3. 4. Johnell O, Kanis JA. An estimate of the5. worldwide prevalence Cummings S, and Rubin dis- S, Black D. The future of hip fractures6. in Papadimitropoulos the E, United Coyte P, Josse R, et al. Current and projected rates Centres, and all participatingful clinical conduct sites of forsupport their the success- for FAITH the andSo secondary HEALTH analyses. trials The andas authors for also their well thank Marianne as Comeau-Gauthier forviewing Diane their each Heels-Ansdell, contributionsstatistical manuscript accuracy in Daniel included and intellectual re- in Axelrod, content. the and supplement for Investigator for theEinhorn (Principal HEALTH Investigator andMarc for FAITH the Swiontkowski HEALTH trials), (Principaltrial), trial), Thomas and Investigator the for members the of FAITH the trial committees, the Methods 1. Cooper C, Campion G, Melton LJ.2. Hip Dhanwal fractures in DK, the Dennison elderly: a EM, world- Harvey NC, et al. Epidemiology of hip 7. Fixation using Alternative Implants for the Treatment of Hip fractures trials provide a uniqueclinical opportunity questions to in answersupplement highly this relevant fracture is population.clinical dedicated The questions to following using datatrials. addressing from 12 the FAITH highly and HEALTH relevant 8. HEALTH Investigators, Bhandari M, Einhorn TA, Guyatt G, et al. Total a s ’ UPPLEMENT Division of c Introduction S 6 Demographic – years and that 3 1,4 1,2 -Aventis, personal fees from fi xation. We conducted 2 multicen- fi Emil H. Schemitsch, MD, FRCSC INTRODUCTION Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com 2020;34:Si) Please refer to the Study Summaries portion of Department of Health Research Methods, Evidence, and Impact, b 7,8 Department of Surgery, University of Western Ontario, London, ON, a Over half of hip fractures (54%) are femoral neck Hip fractures rank in the top 10 of all-cause disability Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Canada; McMaster University, Hamilton, ON, Canada; and Osteosynthesis andPentapharm, Trauma personal Care fees from Foundation, Sano personal fees from personal fees fromOrthopaedic ITS, editorial Trauma, or boardTrauma governing or board Association, committee for editorial memberTrauma the or for Association Journal the governing of International, Orthopaedic board board for or the committee Orthopaedic member for the from Amgencommittee Co, member for research thefees from Canadian DePuy, support Orthopaedic board or Association,or from committee personal committee member member for Biocomposites, the for Hip the board Society, board International or Society for Fracture Repair, Orthopaedic Surgery,Hamilton, Department ON, Canada. of Surgery, McMaster University, London, ON N6A 5A5, Canada (e-mail: [email protected]). Swemac, and personalS. fees from Sprague Zimmer, reportsHealth, outside employment editorial the from submitted or Globalment work. from Research governing McMaster Solutions University, board Inc., outside and the for employ- submitted BMS work. Women Saunders/Mosby-Elsevier, personal fees from Smithfees & Nephew, from personal Springer, personal fees from Stryker, personal fees from J Orthop Trauma projections by Statistics2036, 1 Canada in indicate 4 Canadians that, will by be older the than year 65 the annual number ofin hip Canada fractures and is 500,000 likely to in exceed the 88,000 United States. Hemiarthroplasty) trial compared thearthroplasty intervention with of total hemiarthroplasty hip inof 1441 age patients 50fractures. or years older with low-energy displaced femoral neck age or olderfemoral with neck aEvaluation fracture, with low-energy while Alternatives displaced the of or HEALTH Total undisplaced (Hip Hip Arthroplasty Fracture vs. options for(Fixation fractures using of AlternativeHip fractures) Implants the trial forscrews femoral compared with the the a neck. intervention Treatment sliding hip of The of screw cancellous in FAITH 1079 patients 50 years of treated with arthroplasty,and whereas treated one-third with are internal ter, undisplaced randomized controlled trials focusing on the management fractures, two-thirds of which are typically displaced and to more than 6 million per year by 2050. worldwide, and the number of hip fractures is expected to rise J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. ( the supplement for further details on these 2 trials. E.H. Schemitsch reports personal fees from Acumed, LLC, personal fees Reprints: Emil H. Schemitsch, MD, FRCSC, 339 Windermere Rd,Copyright Box © 5339 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001938 Health, Inc. All rights reserved.

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 = ii S P | as well xation as 1 3.44; fi – measured with the dence interval [CI] fi ered on ClinicalTrials. 0.05). Avascular necro- . = 0.79). Hip instability or Unplanned revision surgery P Unplanned revision surgery 1 g, relieve pain, treat infection, dverse events between groups P er with low energy displaced fracture healing, and mortality iarthroplasty) trial was a 1441- 3 -related quality of life, function, nctional outcomes, health state ality, and other adverse events, s were enrolled from 80 clinical cant difference was found between 2 fi 1.40; 4.09). Function, as – – domized controlled trial comparing the Research Ethics Boards/Institutional Review = 0.18). Health-related quality of life, functional HEALTH STUDY OVERVIEW ’ The FAITH trial was regist P 2 The HEALTH (hip fracture evaluations with alternatives 1.09; – RTICLE =0.82). A P as the primary study results. gov (NCT00761813) and approvedResearch by Ethics the Board Hamilton (#06-402)clinical Integrated as well sites asBoards. by all The participating protocol has been previously published outcomes, health statealso utilities, did not differ betweensis groups was ( more commonpared to in the the cancellous sliding0.0319). screws However, hip group no screw (HR signi groupthe 1.91, as number 1.06 of com- medically( related a within 24 months didfollows: not 107 differ of bygroup type versus 542 117 of of patients surgical 537group (20%) patients (22%) (hazard in in the ratio the0.63 cancellous screws [HR] sliding 0.83, hip 95% screw con of total hip arthroplastypatient vs. multicenter hem ran intervention of totalpatients hip 50 arthroplasty yearsfemoral with of neck hemiarthroplasty fractures. age in Patient sites or in old theFinland, the United Netherlands, States, Norway,and Spain, South Canada, Africa. the The Australia, United primarysurgerywithin24monthsofthefemoralneckfracture. Kingdom, outcome New was unplanned Zealand, revision Secondary outcomeship-related included complications, health death, seriousand adverse overall events, health end points. within 24 monthsfollows: did 57 not of differassigned 718 between to treatment patients total(8.3%) groups (7.9%) who hip as who had arthroplasty(HR had been versus 0.95, been randomly 60 95% randomly assigneddislocation of occurred CI to in 723 hemiarthroplasty 0.64 34arthroplasty patients and patients 17 (4.7%) patients assigned (2.4%)(HR, assigned to 2.00; to 99% total hemiarthroplasty CI, hip 0.97 postsurgery. Theunplanned primary revision outcome surgeryfracture of within to 24 promote the months fractureor FAITH healin of improve function. femoral trial The neck secondarythe was outcomes impact were to of determine health-related a sliding quality hip ofutilities, screw life, versus fracture fu cancellous healing,such screws mort on asbreakage or avascular failure, and necrosis, infection. nonunion, malunion, implant ” UPPLEMENT S ndings. cial views of the fi fi Study Summaries FAITH Investigators and HEALTH Investigators tervention of cancellous screws lia, the Netherlands, Norway, Outcomes After Hip Fractures: “ xation in the operative manage- Siii) Services Incorporated. Dr. Bhandari was fi – lled from 81 clinical sites in the ’ ict of interest. fl Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com 2020;34:Sii FAITH STUDY OVERVIEW SUPPLEMENT STUDY SUMMARIES The FAITH (fracture This supplement titled, Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. London, ON N6A 5A5 (e-mail: [email protected]). for HealthEngland. Research The funding Comprehensivestudy; sources the Clinical had collection, no management, Researchor analysis, role the or in Network interpretation preparation, design of review, the in or or data; approval conduct of of the the manuscript. authors and doesNational Institutes not of necessarilyalso Health. supported Research represent reported by theLocal in The Research this County of Network publication Durham which was & operates as Tees part Valley of Comprehensive the National Institute (McMaster University, Hamilton, ON, Canada).supported The FAITH by trial the was also NationalSkin Institute Diseases of of Arthritis theR01AR055267-01A1. and National Musculoskeletal Institutes The and of content Health under is Award solely Number the responsibility of the Netherlands Organisation for82310-97-11032), Physicians Health Researchalso and funded, in Development part,provided (80- through funding the for the EarlyChair present Research in study Award Musculoskeletal as Trauma Program well which which as is by unrelated to a the Canada present Research study or the preparation, review, orwas approval of supported the by manuscript. research TheResearch FAITH grants trial (MOP-106630 from and the MCT-87771), Canadian(1R01AR055267-01A1), National Institutes Institutes of Stichting of Health Health NutsOhra (SNO-T-0602-43), the sche wetensehappen (ZonMw) (17088.2503), Sophiesfor Minde Orthopaedic Foundation Research,Orthopaedics. The McMaster funding sources Surgical hadstudy; Associates, no role the and in collection, design management, Stryker or analysis, conduct of or the interpretation of the data; Institutes of Health Research (CIHR)Health (MCT-90168), National Institutes (NIH) of (1UM1AR063386-01), ZorgOnderzoek Nederland-medi- J Orthop Trauma below for a summary of each trial and their includes the manuscripts forses 12 preplanned that secondarycontrolled analy- use trials data of from older hip 2 fracture large patients. multicenter Please randomized see Secondary Analyses From the HEALTH and FAITH Trials Germany, the Unitedclinically Kingdom, at and 1 India and 10 and weeks were and assessed 6, 9, 12, 18, and 24 months with a sliding hip screwwith in 1079 a patients 50fracture. years low of Patients age energy or were older United displaced enro States, or Canada, undisplaced Austra femoral neck ment ofcontrolled hip trial comparing fractures) the in trial was a multicenter randomized J Orthop Trauma Reprints: Emil H. Schemitsch, MD, FRCSC, 339 Windermere Rd, BoxCopyright 5339, © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001927 Health, Inc. All rights reserved. Accepted for publication AugustThe 10, authors 2020. report noThe con HEALTH trial was supported by research grants from the Canadian (

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. , e e t, T- ed lor A1. S1 sonal | n, VT c s from rom the Department d f , cantly more c fi nancial or material support fi a Bzovsky, MSc, fi So Department of Surgery, University of Division of Orthopaedic Surgery, Oslo b e b Services Incorporated. M. Bhandari was ’ Sheila Sprague, PhD, ict of interest. fl g tests were performed to compare differences 2 x Frede Frihagen, MD, PhD, FRCSC, cial views of the National Institutes of Health. Research d fi nancial or material support for the Journal of Bone and Joint fi 0.001). HEALTH trial patients were signi cantly lower than that of HEALTH patients (66 vs. 79 The mean age of the 1079 FAITH trial patients was 72 fi , P RTICLE A patients with displaced fractures werepatients. then T-tests compared and with HEALTH was signi years; versus 79 yearspatients for were older, the mostlymore 1441 White, comorbidities HEALTH used than more FAITH trialpatients, medication, patients. patients. 32% and Of (346/1079) HEALTH had the had 1079 FAITH displaced trial fractures. Their mean age Results: for sex, age, osteoporosis status, and ASA class. Marc Swiontkowski, MD, f , Emil H. Schemitsch, MD, FRCSC, c a xation and UPPLEMENT fi -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees from Smith & Nephew, fi S on behalf of the FAITH and HEALTH Investigators a Rudolf W. Poolman, MD, PhD, c nancial or material support from Saunders/Mosby-Elsevier, publishing royalties, fi Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; c Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada; and f Lippincott Williams & Wilkins, outside the submitted work. S. Sprague reports editorial or governing board for BMS Hip Fracture Patient Populations? — Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com 50 years. We evaluated whether patient popula- $ Over the past decade, 2 randomized controlled trials

American, publishing royalties, Patient demographics, medical comorbidities, and frac- and Patrick C. Schottel, MD s Health, employment from Global Research Solutions Inc, and employment from McMaster University, outside the submitted work. P. C. Schottel Mohit Bhandari, MD, PhD, FRCSC, ’ — Department of Orthopaedics and Rehabilitation, University of Vermont Medical Center, South Burlington, VT;

-Aventis, outside the submitted work. M. Swiontkowski reports board or committee member for the American Orthopaedic Association, consultant to a ﬁ Michael Blankstein, MD, MSc, FRCSC, Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN. Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Surgery Synthes, and personal feesAphria, research from support Zimmer, from outside FerringSano Pharmaceuticals, the research submitted support work. andthe personal Minnesota M. Board fees Bhandari of from Medical reports Pendopharma, Practice, and research editorial research or support governing support board from and and personal Acumed, publishing fee royalties, LLC, research support from personal fees frompersonal Pentopharm, fees personal from fees Springer,Poolman from personal Sano reports fees from boardOrthopaedics, Stryker, or outside personal fees the committee from submitted member Swemac, work. and for F. personal the Frihagen fees reports Dutch from personal Zimmer, Orthopaedic outside fees Association, the from submitted research Amgen work. Co, support R. personal W. from fees Lima, from and Smith & research Nephew, support personal from fees Link from fees from Amgen Co,DePuy, board research or support committee fromITS, member Biocomposites, editorial for or board the governing or Hipgoverning board Society, committee for board board the member for or Journal for the committee of Orthopaedic the member Orthopaedic Trauma Canadian Trauma, for Association board Orthopaedic the International, or International Association, board committee Society personal or member for fees for committee Fracture the from member Repair, Orthopaedic for personal Trauma the fees Association, Osteosynthesis from editoria and Trauma Care Foundation The content is solelyreported in the this responsibility publication was ofNational also the supported Institute by authors for The Health and Countystudy; Research Durham does & the Comprehensive not Tees collection, Clinical Valley necessarily Research management, Comprehensive represent Local Network analysis, Research the in or Network, England. of interpretation which The operates of as funding the part sources of data; had the or no the role preparation, in design review, or or conduct approval of of th the manuscript. 0602-43), the Netherlands Organisationalso for Health funded, Research in andMusculoskeletal Development Trauma, part, (80-82310-97-11032), which through Physicians isby unrelated the the to National Early Institute the of present Research Arthritis study and Award (McMaster Musculoskeletal and University, Program Skin Hamilton, that Diseases Ontario, of Canada). provided the The National funding Institutes FAITH of trial for Health was under the also Award Number support present R01AR055267-01 study and by a Canada Research Chair in (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. interpretation Research, The of fundingCanadian the sources Institutes data; of had or Health the no Research preparation, role (MOP-106630 review, and in MCT-87771), or design National approval or Institutes of of conduct the Health of manuscript. (1R01AR055267-01A1), the The Stichting NutsOhra study; FAITH (SNO- trial the was collection, supported managemen by research grants f g 05405 (e-mail: [email protected]). from Wolters KluwerWomen Health reports paid consultancy from Synthes, outside the submitted work. The remaining authors report no con University Hospital, Oslo, Norway; Western Ontario, London, ON, Canada;

of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden University Medical Center, Leiden, the Netherlands; The FAITH and HEALTH Trials: Are We Studying Different Reprints: Michael Blankstein, MD, MSc, FRCSC, Department of Orthopaedics and Rehabilitation, Robert T Stafford Hall, 95 Carrigan Drive, Burlingto M. Blankstein reports stock or stock options from 7D Surgical, outside the reported work. E. H. Schemitsch reports personal fees from Acumed, LLC, per The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health ture characteristics from both trials were compared. FAITH trial J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Methods: Background: Daniel Axelrod, MD, MSc (Cand), cohort from the FAITH trial to HEALTH trial patients. arthroplasty) for the treatmentin of low-energy patients femoral aged necktions fractures in both thecharacteristics FAITH and and HEALTH compared trials had the different displaced baseline femoral neck fracture were performed to evaluate 2 surgical strategies (internal Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001930 Health, Inc. All rights reserved.

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Blankstein et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 likely to be female, have ASA classification Class III/IV/V, and carry TABLE 1. a diagnosis of osteoporosis, as compared with the subgroup of FAITH and HEALTH Trials Patient Enrollment FAITH patients with displaced femoral neck fractures (P , 0.001). Eligibility Criteria FAITH Inclusion Criteria HEALTH Inclusion Criteria fi Conclusion: This study demonstrates signi cant differences Adult men or women aged 50 years Adult men or women aged 50 between patients enrolled in the 2 trials. Although both studies and older (with no upper age limit) years and older (with no upper focused on femoral neck fractures with similar enrollment criteria, age limit) patient populations differed. This sheds light on a noteworthy Fracture of the femoral neck Fracture of the femoral neck limitation of discordant patient enrollment into randomized trials, confirmed with either AP or lateral confirmed with either AP or lateral despite similar eligibility criteria. hip radiographs, CT, MRI hip radiographs, CT, MRI Operative treatment of displaced Displaced fracture that is not, in the Key Words: baseline characteristics, femoral neck fracture fractures within 4 days of presenting judgment of the attending surgeon, to the emergency room optimally managed by reduction and Level of Evidence: Therapeutic Level II. See Instructions for internal fixation Authors for a complete description of levels of evidence. Operative treatment of undisplaced Operative treatment within 3 days of fractures within 7 days of presenting the patient being medically cleared (J Orthop Trauma 2020;34:S1–S8) to the ER for surgery Patient was ambulatory before Patient was ambulatory before INTRODUCTION fracture, although they may have fracture, although they may have Significant controversies exist regarding the ideal used an aid such as a cane or a used an aid such as a cane or a treatment of undisplaced and displaced femoral neck fractures walker walker in both the young and geriatric populations. In young Anticipated medical optimization for Anticipated medical optimization for operative fixation of the hip arthroplasty of the hip patients, internal fixation is typically employed to preserve Provision of informed consent by Provision of informed consent by a healthy functional hip joint. Although arthroplasty out- patient or legal guardian patient or proxy comes are quite predictable and successful, revision surgery is Low-energy fracture, in the Low-energy fracture (defined as a occasionally needed due to issues such as component judgment of the attending surgeon fall from standing height) loosening, instability, adverse tissue reaction, periprosthetic No other major trauma (defined as No other major trauma (defined as fractures, and infections.1 When hip arthroplasty is performed an injury severity score .16) an injury severity score ,17) in young patients, the lifetime risk of revision is higher, and Assurance that surgeons with 2 expertise in both total hip the poor success rates of those revisions are problematic. In arthroplasty and hemiarthroplasty the elderly population, arthroplasty is the mainstay of treat- were available to perform surgery ment, especially in patients with displaced fractures. Intriguingly, recent studies suggest that in elderly patients, FAITH Exclusion Criteria HEALTH Exclusion Criteria even undisplaced and minimally displaced fractures can have Patients not suitable for internal Patient not suitable for fi fl improved clinical outcomes with arthroplasty as compared xation (ie, severe osteoarthritis, hemiarthroplasty (eg, in ammatory rheumatoid arthritis, or pathologic fi 3–5 arthritis, RA, pathologic fracture, or with internal xation. fracture) severe hip OA) fi When considering both internal xation and replace- Associated major injuries of the Associated major injuries of the ment, the ideal implant choice remains controversial. The lower extremity (ie, ipsilateral or lower extremity (ie, ipsilateral or FAITH trial randomized patients with both undisplaced and contralateral fractures of the foot, contralateral fractures of the foot, displaced femoral neck fractures to receive either cancellous ankle, tibia, fibula, knee, or femur; ankle, tibia, fibula, knee, or femur; 6 dislocations of the ankle, knee, or dislocations of the ankle, knee, or screws or a sliding hip screw. The HEALTH trial random- hip; or femoral head defects or hip; or femoral head defects or ized patients with displaced femoral neck fractures to either a fracture) fracture) 7 hemiarthroplasty or a total hip replacement. Both studies Retained implant around the affected Retained implant around the affected enrolled patients aged 50 years and older with a low-energy hip hip fracture mechanism. Patient enrollment eligibility criteria of Infection around the hip (ie, soft Infection around the hip (soft tissue both studies are displayed in Table 1. However, the extent of tissue/bone) or bone) similarity of patients who were actually enrolled in both stud- Patients with disorders of known Patients with a disorder of bone ies is not known. The aim of this study was to evaluate bone metabolism except metabolism other than osteoporosis osteoporosis (ie, Paget disease, renal (ie, Paget disease, renal whether the patient populations enrolled into both multicenter osteodystrophy, osteomalacia) osteodystrophy, osteomalacia) trials were different from one another in respect to demo- Patients with a history of frank Patients with a previous history of graphics and medical comorbidities. Furthermore, we specif- dementia that would interfere with frank dementia that would interfere ically compared the displaced femoral neck fractures that assessment of the primary outcome with assessment of the primary were treated with internal fixation in the FAITH trial with (ie, reoperation at 2 years) outcome (ie, secondary procedures the HEALTH trial patients. We hypothesized that the 2 study at 2 years) Likely problems, in the judgment of Likely problems, in the judgment of populations would be similar. the investigators, with maintaining the investigators, with maintaining follow-up follow-up Patients whose fracture occurred as a METHODS result of an act of violence Baseline demographics and fracture characteristics of AP, anteroposterior; CT, computerized tomography; MRI, magnetic resonance all enrolled patients from both the FAITH and HEALTH trials imaging; OA, osteoarthritis; RA, rheumatoid arthritis. were compared. Factors included were age, sex, ethnicity,

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Characteristics of FAITH and HEALTH Participants body mass index (BMI), prefracture living status and medication, and had more comorbidities than the FAITH functional status, medications, and major comorbidities. All patients. When specifically comparing the patients with dis- patients enrolled in the HEALTH trial had a displaced placed fractures from the FAITH trial with those in the femoral neck fracture (Garden III and IV).8 The FAITH trial HEALTH study, we found that HEALTH study patients were patients were divided using the Garden classification between significantly older, more often female, were ASA Class III/IV/ undisplaced (Garden I and II) and displaced (Garden III and V, and had osteoporosis. This demonstrates that there was IV) fractures. The FAITH trial patients with displaced frac- unintentional discordant patient enrollment despite nearly tures were then separately compared with the HEALTH identical inclusion and exclusion criteria. It is conceivable that patients. To assess severity of fracture displacement in each surgeons were less likely to enroll unhealthy older patients with trial, the numbers of patients with Garden III and IV fractures any fracture displacement in an internal fixation trial. in each trial were also compared. The importance of scrutinizing a trial’s inclusion and exclusion criteria, as well as the subsequently enrolled study Statistical Analysis cohort cannot be overstated. Not appreciating the sometimes Statistical analysis using t-tests and x2 tests were then used subtle exclusion of a specific patient population from a study to compare the differences for the following factors using the a can lead to findings that result in potentially unintended priori principle: sex, age, osteoporosis status, and American changes in surgical practice. For instance, a 2015 landmark Society of Anesthesiologists (ASA) class. To avoid multiple study, Proximal Fracture of the Humerus Evaluation by comparisons, given the large numbers of variables collected in Randomization (PROFHER), prospectively randomized dis- each study, the authors elected to statistically analyze only the 4 placed proximal humerus fracture patients aged 16 years and fundamental factors above. All tests were 2 tailed with alpha = older between operative and nonoperative management.9 0.05. Continuous data were presented with means and SDs, and They found that operative management provided no patient- categorical data were presented as frequencies and proportions. reported clinical outcome benefit compared with nonoperative treatment. This has understandably led to greater interest in nonoperative management of displaced proximal humerus RESULTS fractures. However, it is important to note that of the 1250 The mean age of the 1079 patients in the FAITH trial patients assessed for study eligibility, there were 1000 was 72 (SD, 12) years, as opposed to 79 (SD, 8) years for the patients who were excluded using rationale such as “clear 1441 patients in the HEALTH trial. The HEALTH patients indication for surgery” and “other reasons.” Having broad were older, more often White and less ethnically diverse, used inclusion criteria, as well as exclusion criteria that give a more medication, and had more comorbidities. Table 2 sum- surgeon the ability to not enroll patients based on their per- marizes all the patient demographic comparisons between the sonal discretion can have unintentional consequences, such as FAITH and HEALTH trial participants. often overlooked selection bias. In the FAITH trial, it was The fracture characteristics of FAITH and HEALTH trial reported that 21% of patients required a revision within 24 patients are summarized in Table 3. Of the 1079 patients in the months of their index surgery. Although a 21% revision rate FAITH trial, 32% (346/1079) of the patients had displaced is relatively acceptable for internal fixation in a geriatric fem- fractures. Of these displaced fracture patients, 72% were clas- oral neck fracture cohort, it is easy to overlook that the study sified as Garden 3, as opposed to 44% in the HEALTH trial. enrolled a disproportionate number of undisplaced fractures Tables 4 and 5 summarize the demographic and fracture char- (68%; 733/1079). Overlooking that important point may acteristics of the displaced fracture patients from the FAITH result in misinterpreting the findings and concluding that trial as compared with the HEALTH trial patients. internal fixation of all geriatric femoral neck fractures is suc- Comparison of selected characteristics of displaced cessful in approximately 80% of patients. If a more represen- fracture patients in the FAITH and HEALTH trials found tational geriatric femoral neck fracture population had been that the mean age of the HEALTH study patients was enrolled, it is likely that the revision rate would have been significantly higher than that of the patients in the FAITH higher. One proposal would be to obtain a general consensus study (79 vs. 66 years; P , 0.001). The HEALTH trial for standard study inclusion and exclusion criteria. Leaving patients were significantly more likely to be female, have enrollment criteria up to the investigator or surgeon enrolling ASA classification Class III/IV/V, and have a known diagno- the patient can oftentimes introduce unintended selection sis of osteoporosis, as compared with the subgroup of FAITH bias. Greater focus on this aspect of study design is needed. patients with displaced femoral neck fractures (P , 0.001). Patients’ age was highlighted to be a significant differ- Table 6 summarizes the subgroup comparisons. ence between the patients enrolled in both the HEALTH and FAITH studies. In a recent systematic review and meta- analysis of internal fixation versus arthroplasty for the treat- DISCUSSION ment of nondisplaced femoral neck fractures in the elderly, Although the inclusion criteria for both the FAITH and patients’ average age was approximately 80 years. The review HEALTH studies were very similar and included patients older concluded that when treating nondisplaced and minimally than 50 years who sustained low-energy femoral neck fractures, displaced femoral neck fractures in the elderly, hemiarthro- this study critically evaluated the patients who were actually plasty may reduce the relative risk of reoperation by 70% enrolled in both studies. Overall, the HEALTH patients were when compared with internal fixation. Nevertheless, the older, more often White and less ethnically diverse, used more authors proposed that their results are most applicable to

TABLE 2. Patient Demographics in FAITH and HEALTH Trials Characteristic CS, n = 537 SHS, n = 542 THA, n = 718 HA, n = 723 Age, mean (SD) n = 537 n = 542 n = 718 n = 722 72.0 (12.3) 72.2 (12.0) 79.1 (8.3) 78.6 (8.6) Age, n (%) n = 535 n = 535 n = 718 n = 722 50–70 y 238 (44.5) 245 (45.8) 136 (18.9) 149 (20.6) 71–80 y 147 (27.5) 138 (25.8) 249 (34.7) 247 (34.2) $81 y 150 (28.0) 152 (28.4) 333 (46.4) 326 (45.1) Sex, n (%) n = 535 n = 535 n = 718 n = 722 Male 210 (39.3) 212 (39.6) 208 (29.0) 223 (30.9) Female 325 (60.7) 323 (60.4) 510 (71.0) 499 (69.1) Ethnicity, n (%) n = 535 n = 533 n = 716 n = 721 Indigenous 3 (0.6) 1 (0.2) 2 (0.3) 1 (0.1) South Asian 65 (12.1) 65 (12.2) 3 (0.4) 6 (0.8) East Asian 4 (0.7) 6 (1.1) 7 (1.0) 7 (1.0) Hispanic/Latino 1 (0.2) 3 (0.6) 7 (1.0) 6 (0.8) White 444 (83.0) 436 (81.8) 683 (95.4) 684 (94.9) Black 18 (3.4) 22 (4.1) 12 (1.7) 15 (2.1) Middle Eastern 0 (0.0) 0 (0.0) 2 (0.3) 2 (0.3) BMI (kg/m2), n (%) n = 528 n = 530 n = 697 n = 705 Underweight ,18.5 33 (6.3) 37 (7.0) 35 (5.0) 38 (5.4) Normal weight 18.5–24.9 300 (56.8) 276 (52.1) 357 (51.2) 336 (47.7) Overweight 25–29.9 148 (28.0) 159 (30.0) 217 (31.1) 243 (34.5) Obese 30–39.9 47 (8.9) 58 (10.9) 77 (11.1) 83 (11.8) Morbidly obese $40 0 (0.0) 0 (0.0) 11 (1.6) 5 (0.7) Prefracture living status, n (%) n = 537 n = 542 n = 718 n = 723 Institutionalized 31 (5.8) 30 (5.5) 30 (4.2) 27 (3.7) Not institutionalized 506 (94.2) 512 (94.5) 688 (95.8) 696 (96.3) Pre-fracture functional status, n (%) n = 537 n = 542 n = 718 n = 723 Use of Aid 110 (20.5) 121 (22.3) 187 (26.0) 182 (25.2) Independent ambulator 427 (79.5) 421 (77.7) 531 (74.0) 541 (74.8) Current medications, n (%) n = 534 n = 535 n = 715 n = 722 None 179 (33.5) 170 (31.8) 114 (15.9) 127 (17.6) NSAIDs 64 (12.0) 86 (16.1) 91 (12.7) 90 (12.5) Analgesics: opioid 69 (12.9) 58 (10.8) 63 (8.8) 55 (7.6) Glucocorticoids Not collected Not collected 26 (3.6) 23 (3.2) Anabolic steroid therapy Not collected Not collected 3 (0.4) 1 (0.1) Hormone replacement therapy Not collected Not collected 30 (4.2) 34 (4.7) Bisphosphonates Not collected Not collected 50 (7.0) 47 (6.5) Other osteoporosis medications 73 (13.6) 67 (12.5) 28 (3.9) 15 (2.1) Anti-hypertension medications 252 (47.1) 244 (45.6) 407 (56.9) 402 (55.7) Pulmonary (respiratory system) 69 (12.9) 58 (10.8) 81 (11.3) 87 (12.1) Medications General cardiac medications 167 (31.2) 167 (31.2) 296 (41.4) 278 (38.5) Calcium Not collected Not collected 140 (19.6) 139 (19.3) Calcitonin (mialcalcin) Not collected Not collected 2 (0.3) 1 (0.1) Vitamin D Not collected Not collected 165 (23.1) 160 (22.2) Prior surgery to affected hip, n (%) N = 535 N = 535 N = 714 N = 722 0 (0) 3 (0.6) 2 (0.3) 1 (0.1) Major comorbidities, n (%) N = 537 N = 542 N = 715 N = 722 Osteopenia Not collected Not collected 28 (3.9) 30 (4.2) Osteoporosis 19 (3.5) 14 (2.6) 114 (15.9) 110 (15.2) Lung disease 103 (19.2) 87 (16.1) 127 (17.8) 122 (16.9) Diabetes 79 (14.7) 82 (15.1) 135 (18.9) 145 (20.1) Ulcers or stomach disease 81 (15.1) 68 (12.5) 49 (6.9) 67 (9.3) Kidney disease 58 (10.8) 34 (6.3) 71 (9.9) 67 (9.3)

TABLE 2. (Continued) Patient Demographics in FAITH and HEALTH Trials Characteristic CS, n = 537 SHS, n = 542 THA, n = 718 HA, n = 723 Anemia or other blood disease 55 (10.2) 40 (7.4) 48 (6.7) 55 (7.6) Depression 92 (17.1) 76 (14.0) 70 (9.8) 84 (11.6) Cancer 63 (11.7) 59 (10.9) 65 (9.1) 80 (11.1) Osteoarthritis, degenerative 121 (22.5) 160 (29.5) 111 (15.5) 91 (12.6) arthritis Back pain 104 (19.4) 106 (19.6) 64 (9.0) 71 (9.9) Rheumatoid arthritis 26 (4.8) 6 (1.1) 13 (1.8) 21 (2.9) Heart disease 157 (29.2) 156 (28.8) 247 (34.6) 249 (34.5) High blood pressure 276 (51.4) 281 (51.8) 434 (60.7) 443 (61.4)

BMI, body mass index; CS, cancellous screws; HA, hemiarthroplasty; NSAIDS, nonsteroidal anti-inﬂammatory drugs; SHS, sliding hip screw; THA, total hip arthroplasty.

patients in their late 70s and 80s and may not be generalizable from 2008 to 2011 highlighted this trend.10 Few orthopaedic to younger patients.5 Likewise, our study reveals that the randomized controlled trials performed in the United States overall results of both the FAITH and HEALTH trials might reported information on patients’ race or ethnicity. Among be applicable to dissimilar age groups. The signiﬁcantly lower trials that reported demographic race/ethnicity data, the inclu- age in the FAITH group would suggest that surgeons are sion of minority patients was considerably lower than would be more likely to consider or attempt internal ﬁxation of dis- anticipated on the basis of census demographics. Failure to placed femoral neck fractures in younger healthy male represent racial diversity may result in decreased generalizabil- subjects. ity of study conclusions across different populations. Bias The patients enrolled in the HEALTH trial were more against patients with dementia was also recently recognized often White and less ethnically diverse than the FAITH patient as a concern because patients with dementia and hip fractures population. The lack of diversity in patient enrollment for often get excluded from clinical trials. In the FAITH study, a orthopaedic studies should not go unnoticed. A systematic total of 1690 patients were excluded, 22.2% (375/1690) due to review of orthopaedic randomized controlled trials published dementia or cognitive impairment. In the HEALTH study, 36%

TABLE 3. Fracture Characteristics in FAITH and HEALTH Trials Characteristic CS, n = 537 SHS, n = 542 THA, n = 718 HA, n = 723 Fractured hip, n (%) n = 535 n = 535 n = 715 n = 722 Left 281 (52.5) 280 (52.3) 386 (54.0) 386 (53.5) Right 254 (47.5) 255 (47.7) 329 (46.0) 336 (46.5) Level of the fracture line, n (%) n = 536 n = 535 n = 715 n = 722 Subcapital 351 (65.5) 331 (61.9) 434 (60.7) 456 (63.2) Midcervical 154 (28.7) 159 (29.7) 251 (35.1) 230 (31.9) Basal 31 (5.8) 45 (8.4) 30 (4.2) 36 (5.0) Garden classiﬁcation, n (%) n = 537 n = 542 n = 715 n = 722 Garden I (undisplaced) 277 (51.7) 257 (48.0) N/A N/A Garden II (undisplaced) 92 (17.2) 99 (18.5) N/A N/A Garden III (displaced) 128 (23.9) 121 (22.6) 311 (43.5) 320 (44.3) Garden IV (displaced) 39 (7.3) 58 (10.8) 404 (56.5) 402 (55.7) Pauwels’ classiﬁcation, n (%) n = 536 n = 535 n = 714 n = 721 Type I 59 (11.0) 59 (11.0) 70 (9.8) 47 (6.5) Type II 394 (73.5) 398 (74.4) 404 (56.6) 367 (50.9) Type III 83 (15.5) 78 (14.6) 240 (33.6) 307 (42.6) Mechanism of injury, n (%) n = 534 n = 533 n = 715 n = 722 Fall from standing 521 (97.6) 515 (96.7) 696 (97.3) 700 (97.0) Spontaneous fracture 6 (1.1) 13 (2.4) 16 (2.2) 14 (1.9) Fall from small height Not collected Not collected 3 (0.4) 8 (1.1) Other low energy trauma 7 (1.3) 5 (0.9) Not collected Not collected Additional fractures or injuries, n (%) n = 535 n = 535 n = 714 n = 722 72 (13.5) 67 (12.5) 61 (8.5) 60 (8.3)

CS, cancellous screws; HA, Hemi-arthroplasty; SHS, sliding hip screw; THA, total hip arthroplasty.

TABLE 4. Demographics of Displaced Fracture Patients in FAITH and HEALTH Trials Characteristic CS, n = 167 SHS, n = 179 THA, n = 718 HA, n = 723 Age, mean (SD) n = 166 n = 179 n = 718 n = 722 66.1 (11.5) 67.7 (11.1) 79.1 (8.3) 78.6 (8.6) Age, n (%) n = 166 n = 179 n = 718 n = 722 50–70 y 107 (64.1) 111 (62.0) 136 (18.9) 149 (20.6) 71–80 y 41 (24.6) 42 (23.5) 249 (34.7) 247 (34.2) $81 y 18 (10.8) 26 (14.5) 333 (46.4) 326 (45.1) Sex, n (%) n = 166 n = 179 n = 718 n = 722 Male 92 (55.1) 98 (54.7) 208 (29.0) 223 (30.9) Female 74 (44.3) 81 (45.3) 510 (71.0) 499 (69.1) Ethnicity, n (%) n = 166 n = 179 n = 716 n = 721 Indigenous 2 (1.2) 1 (0.6) 2 (0.3) 1 (0.1) South Asian 40 (24.0) 41 (22.9) 3 (0.4) 6 (0.8) East Asian 1 (0.6) 1 (0.6) 7 (1.0) 7 (1.0) Hispanic/Latino 1 (0.6) 0 (0.0) 7 (1.0) 6 (0.8) White 120 (71.9) 133 (74.3) 683 (95.4) 684 (94.9) Black 2 (1.2) 3 (1.7) 12 (1.7) 15 (2.1) Middle Eastern 0 (0.0) 0 (0.0) 2 (0.3) 2 (0.3) BMI (kg/m2), n (%) n = 164 n = 174 n = 697 n = 705 Underweight ,18.5 7 (4.2) 8 (4.5) 35 (5.0) 38 (5.4) Normal weight 18.5–24.9 85 (50.9) 90 (50.3) 357 (51.2) 336 (47.7) Overweight 25–29.9 54 (32.3) 60 (33.5) 217 (31.1) 243 (34.5) Obese 30–39.9 16 (9.6) 14 (7.8) 77 (11.1) 83 (11.8) Morbidly obese $40 2 (1.2) 2 (1.1) 11 (1.6) 5 (0.7) Pre-fracture living status, n (%) n = 167 n = 179 n = 718 n = 723 Institutionalized 5 (3.0) 12 (6.7) 30 (4.2) 27 (3.7) Not institutionalized 162 (97.0) 167 (93.3) 688 (95.8) 696 (96.3) Pre-fracture functional status, n (%) n = 167 n = 179 n = 718 n = 723 Use of Aid 18 (10.8) 30 (16.8) 187 (26.0) 182 (25.2) Independent Ambulator 149 (89.2) 149 (83.2) 531 (74.0) 541 (74.8) Current medications, n (%) n = 166 n = 179 n = 715 n = 722 None 87 (52.1) 78 (43.6) 114 (15.9) 127 (17.6) NSAIDs 17 (10.2) 37 (20.7) 91 (12.7) 90 (12.5) Analgesics: Opioid 6 (3.6) 9 (5.0) 63 (8.8) 55 (7.6) Glucocorticoids Not collected Not collected 26 (3.6) 23 (3.2) Anabolic steroid therapy Not collected Not collected 3 (0.4) 1 (0.1) Hormone replacement therapy Not collected Not collected 30 (4.2) 34 (4.7) Bisphosphonates Not collected Not collected 50 (7.0) 47 (6.5) Other osteoporosis medications 12 (7.2) 15 (8.4) 28 (3.9) 15 (2.1) Anti-hypertension medications 65 (38.9) 70 (39.1) 407 (56.9) 402 (55.7) Pulmonary (respiratory system) 16 (9.6) 16 (8.9) 81 (11.3) 87 (12.1) Medications General cardiac medications 29 (17.4) 40 (22.3) 296 (41.4) 278 (38.5) Calcium Not collected Not collected 140 (19.6) 139 (19.3) Calcitonin (mialcalcin) Not collected Not collected 2 (0.3) 1 (0.1) Vitamin D Not collected Not collected 165 (23.1) 160 (22.2) Prior surgery to Affected hip, n (%) n = 166 n = 179 n = 714 n = 722 0 (0.0) 1 (0.6) 2 (0.3) 1 (0.1) Major comorbidities, n (%) n = 166 n = 179 n = 715 n = 722 Osteopenia Not collected Not collected 28 (3.9) 30 (4.2) Osteoporosis 5 (3.0) 10 (5.6) 114 (15.9) 110 (15.2) Lung disease 24 (14.4) 15 (8.4) 127 (17.8) 122 (16.9) Diabetes 24 (14.4) 17 (9.5) 135 (18.9) 145 (20.1) Ulcers or stomach disease 9 (5.4) 13 (7.3) 49 (6.9) 67 (9.3) Kidney disease 7 (4.2) 6 (3.4) 71 (9.9) 67 (9.3)

TABLE 4. (Continued) Demographics of Displaced Fracture Patients in FAITH and HEALTH Trials Characteristic CS, n = 167 SHS, n = 179 THA, n = 718 HA, n = 723 Anemia or other blood disease 4 (2.4) 6 (3.4) 48 (6.7) 55 (7.6) Depression 12 (7.2) 11 (6.1) 70 (9.8) 84 (11.6) Cancer 7 (4.2) 7 (3.9) 65 (9.1) 80 (11.1) Osteoarthritis, degenerative 6 (3.6) 11 (6.1) 111 (15.5) 91 (12.6) arthritis Back pain 9 (5.4) 8 (4.5) 64 (9.0) 71 (9.9) Rheumatoid arthritis 5 (3.0) 1 (0.6) 13 (1.8) 21 (2.9) Heart disease 28 (16.8) 31 (17.3) 247 (34.6) 249 (34.5) High blood pressure 68 (40.7) 72 (40.2) 434 (60.7) 443 (61.4)

BMI, body mass index; CS, cancellous screws; HA, hemiarthroplasty; NSAIDS, nonsteroidal anti-inﬂammatory drugs; SHS, sliding hip screw; THA, total hip arthroplasty.

(207/575) of the excluded patients were not enrolled due to In conclusion, our study highlights the discrepancy in dementia/cognitive impairment.11 patient enrollment of 2 large multicenter prospective random- A limitation of this study is that not all demographic-, ized control trials evaluating the surgical treatment of low- comorbidity-, and injury-related data between studies could be energy femoral neck fractures, despite using similar inclusion compared to avoid performing multiple comparisons. and exclusion criteria. The patients in the FAITH trial with Therefore, there are likely other patient characteristics that displaced femoral neck fractures that were treated with may be significantly different between the studies that may not internal fixation were younger, healthier male subjects with have been identified. Furthermore, the FAITH and HEALTH lower rates of osteoporosis compared with the HEALTH trial trials were never designed for intertrial comparisons. There patients treated with arthroplasty. These limitations must be were some differences in the eligibility criteria for each trial, considered when selecting internal fixation or arthroplasty for and although similar data were collected, the 2 trials were not low-energy hip fracture patients aged 50 years and older. This meant to be compared. Statistical differences between variables study highlights an opportunity to form greater investigator could therefore be misleading. Finally, we are unsure how agreement on universal study inclusion and exclusion criteria. prevalent discordant patient enrollment is between randomized Leaving this at the discretion of each individual surgeon or trials evaluating the same injury in orthopaedics. investigator may introduce subtle enrollment bias, and greater

TABLE 5. Fracture Characteristics of Displaced Fracture Patients in FAITH and HEALTH Trials Characteristic CS, n = 167 SHS, n = 179 THA, n = 718 HA, n = 723 Fractured hip, n (%) n = 167 n = 179 n = 715 n = 722 Left 91 (54.5) 103 (57.5) 386 (54.0) 386 (53.5) Right 75 (44.9) 76 (42.5) 329 (46.0) 336 (46.5) Level of the fracture line, n (%) n = 167 n = 179 n = 715 n = 722 Subcapital 61 (36.5) 74 (41.3) 434 (60.7) 456 (63.2) Midcervical 86 (51.5) 87 (48.6) 251 (35.1) 230 (31.9) Basal 19 (11.4) 18 (10.1) 30 (4.2) 36 (5.0) Garden classiﬁcation, n (%) n = 167 n = 179 n = 715 n = 722 Garden III (displaced) 128 (76.6) 121 (67.6) 311 (43.5) 320 (44.3) Garden IV (displaced) 39 (23.4) 58 (32.4) 404 (56.5) 402 (55.7) Pauwels’ classiﬁcation, n (%) n = 167 n = 179 n = 714 n = 721 Type I 10 (6.0) 10 (5.6) 70 (9.8) 47 (6.5) Type II 91 (54.5) 101 (56.4) 404 (56.6) 367 (50.9) Type III 65 (38.9) 68 (38.0) 240 (33.6) 307 (42.6) Mechanism of injury, n (%) n = 167 n = 179 n = 715 n = 722 Fall from standing 162 (97.0) 170 (95.0) 696 (97.3) 700 (97.0) Spontaneous fracture 1 (0.6) 6 (3.4) 16 (2.2) 14 (1.9) Fall from small height Not collected Not collected 3 (0.4) 8 (1.1) Other low energy trauma 3 (1.8) 3 (1.7) Not collected Not collected Additional fractures or injuries, n (%) n = 167 n = 179 n = 714 n = 722 12 (7.2) 9 (5.0) 61 (8.5) 60 (8.3)

CS, cancellous screws; HA, hemiarthroplasty; SHS, sliding hip screw; THA, total hip arthroplasty.

REFERENCES TABLE 6. Comparison of Baseline Characteristics in FAITH and 1. Ledford CK, Perry KI, Hanssen AD, et al. What are the contemporary HEALTH Trial Displaced Fracture Patients etiologies for revision surgery and revision after primary, noncemented – FAITH, HEALTH, total hip arthroplasty? J Am Acad Orthop Surg. 2019;27:933 938. 2. Kuijpers MFL, Hannink G, van Steenbergen LN, et al. Outcome of Characteristic n = 346 n = 1441 P* revision hip arthroplasty in patients younger than 55 years: an analysis , Age, mean (SD) n = 346 n = 1440 0.001 of 1,037 revisions in the Dutch Arthroplasty Register. Acta Orthop. 66.9 (11.3) 78.8 (8.4) 2020;91:165–170. Sex, n (%) n = 345 n = 1440 ,0.001 3. Chen JY, She GR, Luo SM, et al. Hemiarthroplasty compared with fi Male 190 (54.9) 431 (29.9) internal xation for treatment of nondisplaced femoral neck fractures in elderly patients: a retrospective study. Injury. 2020;51:1021–1024. Female 155 (44.8) 1009 (70.1) 4. Dolatowski FC, Frihagen F, Bartels S, et al. Screw fixation versus hemi- ASA classification, n (%) n = 346 n = 1441 ,0.001 arthroplasty for nondisplaced femoral neck fractures in elderly patients: a Class I/II 274 (79.2) 652 (45.2) multicenter randomized controlled trial. J Bone Joint Surg Am. 2019;101: Class III/IV/V 72 (20.8) 789 (54.8) 136–144. ʼ fi Osteoporosis, n (%) n = 346 n = 1441 ,0.001 5. Richards JT, Overmann AL, O Hara NN, et al. Internal xation versus arthroplasty for the treatment of nondisplaced femoral neck fractures in 15 (4.3) 224 (15.6) the elderly: a systematic review and meta-analysis. J Orthop Trauma. – *P-values calculated by t-tests and x2 tests. 2020;34:42 48. ASA, American Society of Anesthesiologists. 6. Fixation using Alternative Implants for the Treatment of Hip fractures, I, Fracture fixation in the operative management of hip fractures (FAITH): an international, multicentre, randomised controlled trial. Lancet. 2017; focus on this aspect of study design is needed. A possible 389:1519–1527. concern, however, would be that the narrower the study 7. Investigators H, Einhorn TA, Guyatt G, Schemitsch EH, et al. Total hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019; inclusion criteria, the less generalizable the results would be, – fi 381:2199 2208. especially when attempting to apply the clinical ndings 8. Kazley JM, Banerjee S, Abousayed MM, et al. Classifications in brief: worldwide. Future work should be directed at the prevention garden classification of femoral neck fractures. Clin Orthop Relat Res. of study enrollment selection bias by limiting the ability to 2018;476:441–445. exclude patients based on surgeon opinion, especially when 9. Rangan A, Handoll H, Brealey S, et al. Surgical vs nonsurgical treatment of adults with displaced fractures of the proximal humerus: the treatment equipoise exists. PROFHER randomized clinical trial. JAMA. 2015;313:1037–1347. 10. Somerson JS, Bhandari M, Vaughan CT, et al. Lack of diversity in orthopaedic trials conducted in the United States. J Bone Joint Surg ACKNOWLEDGMENTS Am. 2014;96:e56. 11. Hebert-Davies J, Laflamme GY, Rouleau D, et al. Bias towards demen- The authors thank the HEALTH and FAITH Investigators tia: are hip fracture trials excluding too many patients? A systematic (http://links.lww.com/JOT/B247). review. Injury. 2012;43:1978–1984.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. s c er m ch ith 2), on, US S9 orial ague | Health f nuscript. g Department of b nancial or material fi a Bzovsky, MSc, fi cant reduction in the odds ict of interest. Department of Orthopedic and fi fl iding hip screw or multiple e So The primary outcome was b cial views of the National Institutes of fi Marc Swiontkowski, MD, Division of Orthopaedic Surgery, Oslo University d f , c Arthroplasty (n = 1441), including total hip arthro- Frede Frihagen, MD, PhD, FRCSC, The 24-month mortality rate was 15.0% (n = 327). nancial or material support from Saunders/Mosby-Elsevier, and xation (n = 734), including sl e fi fi Lippincott Williams & Wilkins, outside the submitted work. G. P. RTICLE — A Results: Intervention: Main Outcome Measurement: plasty and hemiarthroplasty,internal performed forcancellous a screws, displaced performed for FNF an versus undisplaced FNF. Arthroplasty was associated with a signi reoperation and health-related quality of life. mortality within 24 months of injury. Secondary outcomes included Emil H. Schemitsch, MD, FRCSC, a on behalf of the FAITH and HEALTH Investigators Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, c UPPLEMENT a -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees from Smith & Nephew, fi S Cohort Study s Health, employment from Global Research Solutions Inc., and employment from McMaster ’ American, publishing royalties, — Mohit Bhandari, MD, PhD, FRCSC, Hara, MHA, ’ d Rudolf W. Poolman, MD, PhD, -Aventis, outside the submitted work. M. Swiontkowski reports board or committee member for the American d fi , c Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN. g xation for undisplaced femoral neck fi Nathan N. O a nancial or material support from Wolters Kluwer Health fi Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com Services Incorporated. M. Bhandari was also funded, in part, through the Early Research Award Program that provided funding for this study ’ Patients aged 50 years or older with a FNF.

To compare the 24-month risk of mortality between Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada; d Secondary data analysis of 2 multinational randomized Retrospective cohort study. Department of Orthopaedics, R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, Baltimore, MD;

a Hara reports stock or stock options for Arbutus Medical Inc, outside the submitted work. E. H. Schemitsch reports personal fees from Acumed, LLC, ’ Diane Heels-Ansdell, MSc, Sheila Sprague, PhD, Undisplaced Femoral Neck Fractures: A Retrospective and Gerard P. Slobogean, MD, MPH Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. LLC, research support fromsupport Aphria, and research personal supportOrthopaedic from fees Association, Ferring consultant from to Pharmaceuticals, Sano thesupport research Minnesota support for Board and of the Medical personal Journal Practice, fees editorial of from or Pendopharma, governing Bone board and and and resear publishing Joint royalties, Surgery reports editorial orUniversity, governing outside the board submittedLima, for work. and R. BMS research W. support Women from Poolman& Link reports Orthopaedics, Nephew, board outside personal or the fees submitted committee from work. member F. Synthes, for Frihagen and the reports personal personal Dutch fees fees Orthopaedic from from Association, Amgen Zimmer, research Co, personal outside support fees the fro from submitted Sm work. M. Bhandari reports research support from Acumed, from DePuy, board orfrom committee ITS, member editorial or for governing theor board Hip governing for the board Society, Journal for board of thepersonal or Orthopaedic Orthopaedic committee Trauma, fees Trauma board member Association from or for International, committeepersonal Pentopharm, board the member fees or for personal International from committee the Springer, Society fees Orthopaedic member personal for Trauma for from fees Association, Fracture the from Sano edit Repair, Osteosynthesis Stryker, and personal personal fees Trauma fees from Care Swemac, Foundati and personal fees from Zimmer, outside the submitted work. S. Spr personal fees from Amgen Co, research support from Biocomposites, board or committee member for the Canadian Orthopaedic Association, personal fee publishing royalties, operates as part ofdesign the or conduct National of Institute the for study; Health the Research collection, management, Comprehensive analysis, Clinical or Research interpretation Network of the in data; England. or The the funding preparation, sources review, or had approval no of role the in ma and by a Canada Researchwas Chair also in supported Musculoskeletal by TraumaR01AR055267-01A1. the that The National content is Institute is unrelated solely to ofHealth. the this Arthritis Research responsibility study and of reported (McMaster Musculoskeletal the University, in and authors Hamilton, this and Skin ON, does publication Diseases Canada). not was The of necessarily FAITH the also represent trial National the supported of Institutes by of The Health County under Durham Award and Numb Tees Valley Comprehensive Local Research Network that (1R01AR055267-01A1), Stichting NutsOhra (SNO-T-0602-43), theand Netherlands Physicians Organization for Health Research and Development (80-82310-97-1103 Canada; 22 South Greene St, Baltimore, MD 21201 (e-mail: [email protected]). Slobogean reports editorialAssociation, research or support governing fromDepartment the of board Patient-Centered Defense, for and Outcomes paid Research the consultant Institute, Journal for paid Zimmer, consultant of outside for the Orthopaedic Smith submitted Trauma, & work. Nephew, The board research remaining or support authors report for committee no the member con for the Orthopaedic Trauma Hospital, Oslo, Norway; and Trauma Surgery, OLVG, Amsterdam and Leiden University Medical Center, Leiden, the Netherlands;

Surgery, University of Western Ontario, London, ON, Canada; Arthroplasty Versus Internal Fixation for the Treatment of Shaikh Afaq, MD, N. N. O arthroplasty andfractures internal (FNFs). controlled trials. J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Participants: Design: Setting: Objective: The FAITH trial was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of Reprints: Gerard P. Slobogean, MD, MPH, Department of Orthopaedics,Copyright R © Adams 2020 Cowley Wolters ShockDOI: Trauma Kluwer 10.1097/BOT.0000000000001940 Center, Health, University Inc. of All Maryland rights School reserved. of Medicine,

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Afaq et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 of mortality [adjusted odds ratio (aOR): 0.56, 95% confidence 80 centers. HEALTH participants were randomized to either interval (CI): 0.44–0.72, P , 0.01] compared with treatment with total hip arthroplasty or hemiarthroplasty to treat a displaced internal fixation. 11.4% (n = 248) of the study patients required FNF. The FAITH trial was also an international multicenter reoperation within 24 months of injury. The odds of reoperation randomized clinical trial performed at 81 centers from 2008 were 59% lower with arthroplasty treatment than with internal through 2014. In FAITH, 1108 participants were randomized fixation (aOR: 0.41, 95% CI: 0.32–0.55, P , 0.01). The 24- to either multiple cancellous screws or sliding hip screw fix- month SF-12 physical component scores were 2.7 points higher ation for the treatment of a low-energy FNF. Both trials were in arthroplasty patients compared with internal fixation patients coordinated by McMaster University and were approved by (95% CI: 1.6–3.8, P , 0.01). the Ethics Committee at McMaster University and all participating centers. Conclusions: Our findings suggest arthroplasty for a FNF may reduce the risk of mortality and reoperation compared with internal Eligibility Criteria fi fi xation of undisplaced fractures. This nding is counter to many We included participants aged 50 years or older with a current surgical practices but consistent with a mounting body of low-energy fracture of the femoral neck. Eligible participants evidence. Before widespread adoption of arthroplasty for undis- fi fi must have been able to ambulate either without assistance or placed fractures, these results should be con rmed in a de nitive assistance from an aid before the injury. Patients with rheuma- comparative trial. toid arthritis, pathological fractures or severe osteoarthritis of the Key Words: arthroplasty, internal fixation, undisplaced femoral hip, other major injuries of the lower extremities, retained neck fracture implant around the affected hip, infection around the hip, disorder of bone metabolism, and previous history of dementia Level of Evidence: Therapeutic Level III. See Instructions for were excluded. The FAITH trial included all FNFs amenable to Authors for a complete description of levels of evidence. surgical fixation but primarily included minimally displaced (J Orthop Trauma 2020;34:S9–S14) fractures, whereas HEALTH only included displaced fractures being treated with arthroplasty. All fractures were classified by BACKGROUND each trial’s Central Adjudication Committee using the Garden fi Two decades ago, the optimal treatment of displaced classi cation, and Garden I and II fractures were considered 11,12 femoral neck fractures (FNFs) was controversial.1 A undisplaced. mounting body of evidence has since shifted clinical prac- Because the target population for this study was tice in favor of arthroplasty for displaced FNF manage- undisplaced fractures, we excluded the displaced fracture ment.2,3 However, for undisplaced FNFs, internal fixation patients from the FAITH trial. However, we included all remains the preferred approach as a less invasive procedure patients from the HEALTH trial because the degree of with less expensive implants. A recent randomized trial fracture displacement has no effect on healing since the comparing hemiarthroplasty with internal fixation in non- femoral head and neck is resected and should have no displaced FNFs contested the current practice,4 suggesting impact on the outcome after the arthroplasty procedure. improved mobility and fewer major reoperations with hemiarthroplasty. This finding was consistent with another Study Treatments small trial and 2 recent retrospective cohort studies in min- Participants in the arthroplasty treatment group received imally displaced FNF patients.5–7 either total hip arthroplasty or hemiarthroplasty as per the Given the limited comparative data for the optimal randomized allocation within the HEALTH trial. Treating fi treatment of undisplaced FNFs, we aimed to determine if surgeons had to meet a prede ned threshold for surgical fi arthroplasty in patients aged 50 years or older with displaced expertise in the procedures. Similarly, the internal xation FNFs was associated with reduced mortality, fewer reoperations, group comprised patients that were randomly assigned a and higher health-related quality of life compared with internal sliding hip screw or multiple cancellous screws. The sliding fixationforundisplacedFNFsinpatientsaged50yearsorolder. hip screw was a single large-diameter, partly threaded screw fi We hypothesized that arthroplasty would decrease mortality, af xed to the proximal femur with a side plate. Treatment reduce reoperations, and increase health-related quality of life with cancellous screws required a minimum of 2 threaded scores.8 We also assessed the variation in treatment effect across screws with a 6.5-mm diameter. Surgeons treating partici- fi various prefracture risk profiles. We hypothesized that the ben- pants with internal xation were required to have performed fi efits of arthroplasty treatment would be greater in high-risk FNF at least 25 hip fracture xation procedures in their career, patients. including at least 5 procedures in the year before the study treatment.

METHODS Study Outcomes The primary end point was mortality within 24 months of Study Design and Procedures injury. The secondary outcomes included reoperation within 24 This retrospective cohort study combined data from the months and health-related quality of life. Reoperation was HEALTH and FAITH clinical trials.9,10 The HEALTH trial deﬁned as any operation subsequent to the initial procedure to was an international, expertise-based clinical trial conducted promote fracture healing, relieve pain, treat infection, or improve between 2013 and 2016 that enrolled 1495 participants from function. Health-related quality of life was measured with the

12-item Short-Form General Health Survey (SF-12), which quality of life scores was estimated using linear regression. reports physical component summary (PCS) and mental com- Estimates for physical health quality of life included ASA ponent summary (MCS) scores on a scale of 0–100.13 SF-12 classification, additional injuries, and preinjury back pain as scores were normalized to a population mean of 50, with higher confounders. Mental health quality of life estimates were con- scores implying greater health. ditioned on sex and preinjury depression as confounders. To assess variations in the treatment effect on mortality Statistical Analysis across risk strata, we used the risk modeling approach 14 fl Patient characteristics were described using counts with described by Kent et al. Brie y, we developed a risk pre- proportions and means with SDs, depending on the distribution diction model for mortality omitting the treatment variable. of the data. Categorial data were compared between the The data-informed risk model included age, sex, ASA classifi treatment arms using x2 tests. Continuous variables were com- cation, prefracture functional status, cancer, and anemia or pared using t tests. To measure for potential confounding, we other blood diseases with a C-statistic of 0.76. Based on the assessed for the distribution of covariates between the 2 surgi- risk model, the probability of mortality was assigned to each cal exposure groups and outcomes. If the covariates differed patient. We then binned the study sample into risk quartiles significantly between the exposure and outcome or resulted in a based on the probability of death. In each risk quartile, we change in the effect measure of 10% or greater when added to calculated the association between treatment and mortality on the unadjusted model, they were included in the final adjusted a relative (odds ratio) and absolute scale (risk difference). analysis. Logistic regression models were used to determine Statistical analyses were performed using SAS 9.4 (SAS the association of treatment with mortality and reoperation. All Institute, Cary, NC) and R version 4.0.0 (Vienna, Austria). 15 covariates listed under Table 1 were assessed for potential Missing covariate data were imputed using multiple imputations. confounding. Using the above criteria, the final adjusted model for mortality included age and the American Society of Anesthesiologists (ASA) classification. ASA classification RESULTS was the only confounder included in the reoperation model. Of the 2175 patients included in the study, 1441 were The association between treatment and 24-month health-related treated with arthroplasty, and 734 were treated with internal

TABLE 1. Patient Characteristics Characteristic Arthroplasty (n = 1441) Internal Fixation (n = 734) P Age, y, mean (SD) 78.8 (8.4) 74.2 (11.9) ,0.01 Female, n (%) 1009 (70.1%) 488 (66.5%) 0.09 BMI, kg/m2, mean (SD) 25.1 (4.8) 24.3 (4.5) ,0.01 Mechanism of injury, n (%) ,0.01 Fall from standing 1396 (97.2%) 711 (97.3%) Spontaneous fractures 30 (2.1%) 15 (2.1%) Fall from small height 11 (0.8%) 0 (0.0%) Others 0 (0.0%) 5 (0.7%) ASA classiﬁcation, n (%) ,0.01 Class I–II 652 (45.3%) 414 (56.4%) Class III–IV 789 (54.8%) 320 (43.6%) Prefracture functional status, n (%) 0.95 Independent ambulator 1072 (74.4%) 547 (74.5%) Use of aid 369 (25.6%) 187 (25.5%) Prefracture living status, n (%) 0.22 Institutionalized 65 (4.5%) 42 (5.7%) Not institutionalized 1376 (95.5%) 692 (94.3%) Additional injuries, n (%) 61 (4.3%) 110 (15.0%) ,0.01 Hypertension, n (%) 877 (61.0%) 412 (56.4%) 0.04 Diabetes, n (%) 280 (19.5%) 116 (15.9%) 0.04 Back pain, n (%) 135 (9.4%) 163 (22.4%) ,0.01 Cancer, n (%) 145 (10.1%) 102 (14.0%) ,0.01 Anemia or other blood diseases, 103 (7.2%) 82 (11.3%) ,0.01 n (%) Ulcer, n (%) 116 (8.1%) 119 (16.3%) ,0.01 Depression, n (%) 154 (10.7%) 130 (17.8%) ,0.01

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

TABLE 2. Study Outcomes Arthroplasty (n = 1441) Internal Fixation (n = 734) Crude OR (95% CI) P Adjusted OR (95% CI) P Mortality, n (%) 198 (13.7%) 129 (17.6%) 0.75 (0.59–0.95) 0.02 0.56 (0.44–0.73) ,0.01 Reoperation, n (%) 117 (8.1%) 131 (17.9%) 0.41 (0.31–0.53) ,0.01 0.41 (0.32–0.55) ,0.01

Arthroplasty Internal Fixation Crude Difference (n = 1006) (n = 490) (95% CI) P Adjusted Difference (95% CI) P 24-mo SF-12 PCS, mean (SD) 38.8 (9.9) 36.1 (9.9) 2.7 (1.7–3.8) ,0.01 2.7 (1.6–3.8) ,0.01 24-mo SF-12 MCS, mean (SD) 52.3 (10.6) 51.2 (14.5) 1.1 (20.1–2.3) 0.07 0.9 (20.3–2.0) 0.14

Only a subset of arthroplasty and internal ﬁxation patients were administered the SF-12 survey as per the study protocols. PCS, physical component summary; MCS, mental component summary.

fixation (Table 1). Patients in the arthroplasty group were [95% confidence interval (CI): 0.37–0.78] and a risk reduc- older (78.8 vs. 74.2 years, P , 0.01) and more likely to have tion of 14.1% (95% CI: 5.7%–22.6%). an ASA classification of 3–5 (54.8% vs. 43.6%, P , 0.01). We did not observe a significant difference in the proportion of patients requiring an aid before injury (25.6% vs. 25.5%, P DISCUSSION = 0.95) or institutionalized before the fracture (4.5% vs. 5.7%, The findings of our study suggest that arthroplasty P = 0.22). The 24-month mortality rate was 15.0% (n = 327). reduced the risk of mortality and reoperation within 24 Treatment with arthroplasty significantly reduced the odds of months after injury compared with internal fixation in mortality [adjusted odds ratio (aOR): 0.56, 95% CI: 0.44– undisplaced FNFs. Arthroplasty was also associated with a 0.72, P , 0.01] compared with treatment with internal fixa- mild, yet statistically significant, improvement in overall tion (Table 2). 11.4% (n = 248) of the study patients required physical health. The study treatment was not associated with reoperation within 24 months of injury. The odds of reopera- a difference in overall mental health. We observed that tion were 59% lower with arthroplasty treatment than those arthroplasty provided the strongest protective effects against with treatment by internal fixation (aOR: 0.41, 95% CI: 0.32– mortality in our high-risk strata patients. 0.55, P , 0.01). The 24-month SF-12 physical component Previous studies have suggested that arthroplasty is scores were 2.7 points higher in arthroplasty patients than associated with lower reoperations when compared with those of internal fixation patients (95% CI: 1.6–3.8, P , internal fixation of minimally displaced fractures.4–8 0.01). No difference in 24-month SF-12 mental component Consistent with the 2 previous trials and 2 previous observa- scores was observed between the 2 treatment groups (adjusted tional analyses in this population,4–7 we observed a similar difference: 0.9, 95% CI: 20.3—2.0, P = 0.14). magnitude of reduced odds of reoperation associated with The characteristics of the patients based on their arthroplasty. However, the survival benefits observed in our mortality risk quartile are available in Table 3. The mean risk study contradict the 2 previous observational studies,5,6 which of mortality was 3.7% in the lowest risk quartile, 7.4% in the suggested internal fixation was protective against mortality. In low-mid risk quartile, 16.0% in the mid-high risk quartile, and these studies, the effects of selection bias cannot be deter- 33.1% in the high-risk quartile (Fig. 1). The point estimates mined, and therefore, it is important to note that the prospec- for ORs and risk differences favored arthroplasty across the 4 tive randomized trial comparing the arthroplasty and internal risk quartiles. However, only significant differences were fixation treatments for mortality is consistent with our study observed in the highest risk quartile with an OR of 0.54 results.4 In the trial by Dolatowski et al,4 mortality was

TABLE 3. Characteristic of Mortality Risk Quartiles Low Risk (1) (N = 544) Low-Mid Risk (2) (N = 544) Mid-High Risk (3) (N = 544) High Risk (4) (N = 543) Age, yrs, mean (SD) 67.2 (8.9) 77.1 (7.8) 80.9 (7.3) 83.8 (7.0) Female, n (%) 449 (82.5%) 382 (70.2%) 362 (66.5%) 305 (56.2%) ASA classiﬁcation, n (%) Class I–II 505 (92.8%) 345 (63.4%) 177 (32.5%) 39 (7.2%) Class III–IV 39 (7.2%) 199 (36.6%) 367 (67.5%) 504 (92.8%) Prefracture functional status, n (%) Independent ambulator 538 (98.9%) 522 (96.0%) 407 (74.8%) 152 (28.0%) Use of aid 6 (1.1%) 22 (4.0%) 137 (25.2%) 391 (72.0%) Anemia or other blood diseases, 1 (0.2%) 11 (2.0%) 37 (6.8%) 137 (25.2%) n (%) Cancer, n (%) 9 (1.7%) 32 (5.9%) 62 (11.4%) 146 (26.9%)

ASA, American Society of Anesthesiologists.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Arthroplasty Versus Internal Fixation reported as a secondary outcome, and although the benefits of hemiarthroplasty were not statistically significant in their study (P = 0.11), the point estimate suggested a 60% relative risk reduction. Similarly, consistent with our SF-12 PCS results, the trial by Dolatowski et al observed a 10% benefit in health-related quality of life with hemiarthroplasty compared with internal fixation.4 There are several possible mechanisms for the observed benefits of arthroplasty in this patient population. Despite the more invasive procedure, a hip arthroplasty immediately creates a stable weight-bearing construct. This may contribute to less pain and earlier clinically significant mobilization, particularly because delayed mobilization has been previously demonstrated to lead to reduced function and lower survival in hip fracture patients.16 Similarly, arthroplasty is associated with less reoperations than internal fixation, and this likely contributes to reduced mortality and improved clinical outcomes. In our study, this reduction in mortality was observed despite HEALTH participants remaining at a greater risk for mortality. These risk factors included older age and higher ASA prevalence. Therefore, although both clinical trials may have recruitment biases, such as HEALTH potentially recruit- ing a healthier population of displaced FNF patients suitable for randomization to total hip arthroplasty, the benefits of arthroplasty seem to still outweigh the difference in mortality risk factors. To better understand potential participant recruitment biases between the trials and the substantial variation in risk of mortality in the pooled study population, we used a novel risk modeling approach to assess the heterogeneity of treatment effect (HTE).14 HTE describes variation in the magnitude of treatment effect basedonclinicallyrelevant patient attributes (subgroups).17 The risk modeling technique allowed us to account for an imbalance in baseline covariates between the treatment groups, assess the prognostic value of the observed variables, and estimate differences in treatment effects across clinically unique strata. Although all HTE treatment estimates favored arthroplasty, the most substantial mortality benefits were observed in the oldest, sickest patients. With more than 2000 patients, our study is 3 times larger than the 4 previous studies combined.4–7 We used data from 2 recent, high-quality, multinational, randomized trials.9,10 The risk modeling technique for analyzing HTE represents an improvement over classic subgroup analysis, which is prone to low statistical power, multiplicity, and weak previous theory on relative effect modifiers and, therefore, sus- ceptible to false-negative and false-positive findings. As patients have many attributes that simultaneously affect the study outcome, modeling their combined effects produced a more patient centered and clinically actionable estimate of treatment effects. Despite the strengths of the study, there were some limitations. With regards to osteoporosis or smoking status, an important prognostic factor for both mortality and failure, FIGURE 1. Association between treatment and mortality by only HEALTH had data for osteoporosis, whereas smoking risk quartile. The figure presented the mean risk of mortality status was collected only under FAITH. For this reason, we with ranges by risk quartile (top), and the association between were unable to include these variables in our analysis. treatment and mortality by quartile based on observed odds Furthermore, although the treatment decision for internal ratios (middle) and risk differences (bottom).

fixation has traditionally been based on fracture displacement multicenter randomized controlled trial. J Bone Joint Surg Am. 2019;101: defined within the Garden classification (type I and II), more 136–144. recent studies have highlighted the importance of also 5. Sikand M, Wenn R, Moran CG. Mortality following surgery for undisplaced intracapsular hip fractures. Injury. 2004;35:1015–1019. assessing the fracture alignment on the lateral hip radiograph. 6. Kang JS, Jeon YS, Ahn CH, et al. Osteosynthesis versus endoprosthesis In certain fractures that seem minimally displaced on ante- for the treatment of femoral neck fracture in Asian elderly patients. BMC roposterior radiographs, large amounts of posterior angulation Musculoskelet Disord. 2016;17:264. on the lateral view are highly predictive of internal fixation 7. Lu Q, Tang G, Zhao X, et al. Hemiarthroplasty versus internal fixation in failure.18 These previous findings suggest another layer of super-aged patients with undisplaced femoral neck fractures: a 5-year treatment decision complexity that favors widespread adop- follow-up of randomized controlled trial. Arch Orthop Trauma Surg. – tion of arthroplasty in this patient population. 2017;137:27 35. 8. Richards JT, Overmann AL, O’Hara NN, et al. Internal fixation versus Currently, the conventional treatment for an undis- fi arthroplasty for the treatment of nondisplaced femoral neck fractures in placed or minimally displaced FNF is internal xation. the elderly: a systematic review and meta-analysis. J Orthop Trauma. Arthroplasty has been typically disregarded in this fracture 2020;34:42–48. population given the invasive nature of the procedure, the cost 9. Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or of the implant, and the presumed good clinical results from hemiarthroplasty for hip fracture. N Engl J Med. 2019;381:2199–2208. internal fixation. As newer data continue to emerge, it 10. Fracture fixation in the operative management of hip fractures (FAITH): becomes apparent that the results of internal fixation for an international, multicentre, randomised controlled trial. Lancet. 2017; 389:1519–1527. undisplaced fractures are often poor and that hip arthroplasty 11. Garden RS. Low-angle fixation in fractures of the femoral neck. J Bone reduces reoperations and potentially reduces mortality. Our Joint Surg Br. 1961;43:647–663. results continue to challenge the prevailing practice of 12. Kazley JM, Banerjee S, Abousayed MM, et al. Classifications in brief: internal fixation for these fractures; however, definitive Garden classification of femoral neck fractures. Clin Orthop Relat Res. evidence from a large, appropriately powered trial remains 2018;476:441–445. warranted before a widespread practice change. 13. Shah CH, Brown JD. Reliability and validity of the Short-Form 12 Item Version 2 (SF-12v2) health-related quality of life survey and disutilities associated with relevant conditions in the U.S. Older adult population. J Clin Med. 2020;9:661. ACKNOWLEDGMENTS 14. Kent DM, Paulus JK, van Klaveren D, et al. The predictive approaches to The authors thank the HEALTH and FAITH treatment effect heterogeneity (PATH) statement. Ann Intern Med. 2020; Investigators (http://links.lww.com/JOT/B245). 172:35–45. 15. Royston P, White IR. Multiple imputation by chained equations (MICE): implementation in Stata. J Stat Softw. 2011;45:1–20. REFERENCES 16. Siu AL, Penrod JD, Boockvar KS, et al. Early ambulation after hip 1. Bhandari M, Devereaux PJ, Tornetta P III, et al. Operative management fracture: effects on function and mortality. Arch Intern Med. 2006;166: of displaced femoral neck fractures in elderly patients. An international 766–771. survey. J Bone Joint Surg Am. 2005;87:2122–2130. 17. Varadhan R, Segal JB, Boyd CM, et al. A framework for the analysis of 2. Bhandari M, Swiontkowski M. Management of acute hip fracture. N Engl J Med. 2017;377:2053–2062. heterogeneity of treatment effect in patient-centered outcomes research. J – 3. Bhandari M, Devereaux PJ, Swiontkowski MF, et al. Internal fixation Clin Epidemiol. 2013;66:818 825. compared with arthroplasty for displaced fractures of the femoral neck. A 18. Okike K, Udogwu UN, Isaac M, et al. Not all Garden-I and II femoral meta-analysis. J Bone Joint Surg Am. 2003;85:1673–1681. neck fractures in the elderly should Be fixed: effect of posterior tilt on 4. Dolatowski FC, Frihagen F, Bartels S, et al. Screw fixation versus hemi- rates of subsequent arthroplasty. J Bone Joint Surg Am. 2019;101:1852– arthroplasty for nondisplaced femoral neck fractures in elderly patients: a 1859.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. e d t, T- he tee ittee nd a S15 (NIH) b s from fi | rom the nancial or fi 0.001), and = 0.02), use P , e P 0.001) were associ- , xation vs. joint arthro- fi P -Aventis, personal fees from fi Services Incorporated. M. Bhandari ’ Department of Health Research Methods, d nancial or material support for the Journal of Department of Surgery, University of Western fi b = 0.04) were associated with a higher risk of P = 0.03), lower body mass index ( P on behalf of the FAITH and HEALTH Emil H. Schemitsch, MD, FRCSC, Our analysis found that factors that are indicative of Frede Frihagen, MD, PhD, d , 0.001), and kidney disease ( a a d , = 0.004), use of an ambulatory aid before femoral neck cial views of the National Institutes of Health. Research reported in P fi P RTICLE = 0.002), American Society of Anesthesiologists (ASA) class III/ A P ( IV/V ( fracture ( ated with a higher risk offracture. mortality Older within age 24 ( months ofof femoral an neck ambulatory aid beforehaving a femoral comorbidity neck ( fracture ( Conclusions: mortality within 90 days of femoral neck fracture. a poorer health status werewithin associated with 24 a higher monthsdifference risk of of in mortality treatment femoralplasty) methods neck on (internal the fracture. risk of We mortality. did not Division of Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway. e -Aventis, outside the submitted work. S. Sprague reports editorial or governing board for fi Investigators UPPLEMENT nancial or material support from Saunders/Mosby-Elsevier, and publishing royalties, S fi and Sheila Sprague, PhD d Diane Heels-Ansdell, MSc, , a c Lippincott Williams & Wilkins, outside the submitted work. F. Frihagen reports personal fees from Amgen — 0.001), lower body mass index , days and 24 months of hip fracture. Marianne Comeau-Gauthier, MD, MSc, ict of interest. P fl Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN; c a American, publishing royalties, — Management of Femoral Neck Fractures Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com Hip fractures are recognized as one of the most s Health, employment from Global Research Solutions Inc, and employment from McMaster University, outside the submitted work. The ’ 50 years enrolled in the FAITH and HEALTH trials. Marc Swiontkowski, MD, Factors Associated With Mortality After Surgical $ Two multivariable Cox proportional hazards regressions a Bzovsky, MSc, Division of Orthopaedic Surgery, Department of Surgery, McMaster University, 293 Wellington St. N, Suite 110, Hamilton, ON fi Ninety-one (4.1%) and 304 (13.5%) of 2247 participants Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; a Bzovsky, MSc, a fi So Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. consultant to the Minnesota BoardBone of and Medical Joint Practice, Surgery editorial or governing board and publishing royalties, member for the Orthopaedicmember Trauma for Association, editorial theSaunders/Mosby-Elsevier, or Osteosynthesis personal governing fees and board from for Traumapersonal Smith the fees Care & Orthopaedic from Nephew, Foundation, Trauma Zimmer, personal outside Association personal fees the International, from fees submitted board Springer, work. from or personal M. commit Pentopharm, fees Swiontkowski from personal reports Stryker, board fees personal or fees from committee from member Sano Swemac, for an the American Orthopaedic Association, Institute for Health Research Comprehensivecollection, Clinical management, Research analysis, Network or in interpretation England. The of funding the sourcesfor data; had the or no Canadian the role Orthopaedic preparation, in Association,International design review, personal Society or or fees for conduct from approval Fracture of DePuy, of Repair, the board the personal study; or t manuscript. fees committee from member ITS, for editorial the Hip or Society, governing board board or for committee the member Journal for of the Orthopaedic Trauma, board or comm Musculoskeletal Trauma, whichNational is Institute unrelated of to Arthritiscontent and this is Musculoskeletal solely study and the responsibility (McMaster Skinthis of Diseases University, the publication of authors was Hamilton, and the also does ON, National supported not Institutes necessarily by Canada). represent of The The the County Health of FAITH Durham under & Award trial Tees Number was Valley R01AR055267-01A1. Comprehensive also Th Local supported Research by Network which the operates as part of the National analysis, or interpretation ofCanadian the Institutes data; of or Health the Research0602-43), preparation, (MOP-106630 review, the and MCT-87771), or Netherlands National approval Organisation Instituteswas of for of also the Health Health funded, manuscript. (1R01AR055267-01A1), in Research The Stichting part, NutsOhra and FAITH through (SNO- trial Development the was (80-82310-97-11032), Early supported Research and by Award Physicians Program research which grants provided f funding for the present study as well as by a Canada Research Chair in (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundation for Stryker Orthopaedic Orthopaedics. Research, The funding sources had no role in design or conduct of the study; the collection, managemen Ontario, London, ON, Canada; Evidence, and Impact, McMaster University, Hamilton, ON, Canada; and material support from Wolters Kluwer Health research support from Acumed, LLC,Pendopharma, research and support research from support Aphria,BMS and research support personal Women from fees Ferring from Pharmaceuticals,remaining Sano research authors support report and no personal fee con Co, personal fees from Smith & Nephew, personal fees from Synthes, and personal fees from Zimmer, outside the submitted work. M. Bhandari reports L8L 8E7, Canada (e-mail: [email protected]). were usedassociated to with mortality investigate within 90 potential prognostic factors that may be J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Results: Methods: Background: Mohit Bhandari, MD, PhD, FRCSC, died within 90fracture, days respectively. Older and age ( 24 months of suffering a femoral neck that follow. Mortalitywithin rates one year postsurgery of canincreased fracture. risk We range of aimed mortality from to withinpatients identify 14% 24 aged factors months to of associated a with 58% femoral neck fracture in devastating injuries impacting older adults because of the complications E. H. Schemitsch reports personal fees from Acumed, LLC, personal fees from Amgen Co, research support from Biocomposites, board or committee member The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health Reprints: So Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001937 Health, Inc. All rights reserved.

Key Words: femoral neck fracture, mortality, internal fixation, ar- orthopaedic surgeons, potential prognostic factors were throplasty selected a priori from the study case report forms. Study (FAITH vs. HEALTH) was included as one of the potential Level of Evidence: Therapeutic Level II. See Instructions for prognostic factors to account for difference in intervention Authors for a complete description of levels of evidence. status (surgical method of fracture management) between (J Orthop Trauma 2020;34:S15–S21) the 2 studies. Because similar eligibility criteria were used to enroll participants into the FAITH and HEALTH trials INTRODUCTION and given that the survival experience was similar in both Hip fractures have become a major public health trials, it was deemed appropriate to combine data from both problem and are recognized as one of the most devastating trials to evaluate our research questions. injuries impacting older adults because of the complications that follow, which include chronic pain, diminished physical Statistical Analyses health, quality of life, and premature death.1 As the global Complete-case analyses using multivariable Cox pro- population continues to age, the yearly incidence of hip portional hazards regressions were used to investigate the fractures has been projected to reach 6.26 million by 2050, association between the selected baseline, fracture character- 1,2 as compared to 1.66 million in 1990. Despite fractures of istic, and surgical prognostic factors and the risk of mortality the femoral neck typically being managed surgically, with within 90 days and 24 months of hip fracture in separate fi either internal xation or arthroplasty methods, morbidity models. Mortality of the FAITH and HEALTH trial partic- 3–5 and mortality rates remain high postsurgery. Asystematic ipants within 90 days and 24 months of hip fracture was review of 70 trials, published between 1981 and 2012, found chosen as the dependent variable in each model, with that mortality rates for femoral neck fracture patients were continent entered into the model as a stratification variable. ; similarovera31-yearperiod( 20%), whereas another Participants who did not experience the event of mortality review reported that mortality rates can range from 14% to were censored at 24 months (730 days) or time of last visit. 3,6 58% within one year of fracture. In addition, the Fixation The analyses included continent to account for potential using Alternative Implants for the Treatment of Hip frac- differences in treatment practices between the geographic tures (FAITH) and Hip fracture Evaluation with locations that participants were recruited from. Selected ALternatives of Total hip arthroplasty versus independent variables were entered into the multivariable Hemiarthroplasty (HEALTH) trials have found similar high Cox proportional hazards regression models simultaneously. mortality rates of 14.5% and 13.7%, respectively, within 24 Results were reported as adjusted hazard ratios (HR), 95% 7,8 months of hip fracture. confidence intervals (CIs), and P-values. All tests were 2- Although the high mortality rate of hip fracture patients tailed with alpha = 0.05.10 All analyses were performed with is well documented, factors associated with mortality after hip SPSS version 25. fracture have not been thoroughly examined in a large, global patient population. Identifying which factors are associated with mortality may assist surgeons in their treatment decisions and, ultimately, enhance the care of hip fracture RESULTS patients. The objectives of this analysis were to identify Baseline Participant Characteristics factors associated with an increased risk of mortality within 90 days and 24 months of a femoral neck fracture in patients One thousand six and 1241 participants enrolled in the aged 50 years or older enrolled in the FAITH and HEALTH FAITH and HEALTH trials, respectively, totaling 2247 trials. participants, had complete data for the independent variables included in the mortality models. The majority of participants were female (57.2%), the mean age was 75.3 years (SD 10.8), METHODS and the mean body mass index (BMI) was 24.8 (SD 4.7). Before injury, the majority of participants were independent Selection of Factors ambulators (77.0%), lived in a noninstitutionalized setting With a binary outcome, having fewer than 10 events for (95.5%), did not use osteoporosis medications (92.6%), and each predictor variable can result in overfitted, unstable did not have any comorbidities, with the exception of high models.9 Given that mortality was experienced within 24 blood pressure (56.7%). The typical femoral neck fracture months of femoral neck fracture in a total of 304 participants was displaced (69.7%), with a fracture line at the subcapital from the FAITH (150 participants) and HEALTH (154 par- level (59.5%), and assigned a type II Pauwels classification ticipants) trials, the number of predictor variables included in (57.0%).11,12 Furthermore, the majority of participants had the first model was limited to 30 parameters. Because mor- either a Class I or Class II American Society of tality was experienced within 90 days of femoral neck frac- Anesthesiologists (ASA) classification at baseline (54.4%). ture in a total of 91 participants from the FAITH (40 The event of mortality was experienced in 304 [13.5% total; participants) and HEALTH (51 participants) trials, the num- 150 (49.3%) from FAITH and 154 (50.7%) from HEALTH] ber of predictor variables included in the second model was and 91 [4.1% total; 40 (44.0%) from FAITH and 51 (56.0%) limited to 10 parameters. Based on biologic rationale, pre- from HEALTH] participants within 24 months and 90 days of vious reports in the literature, and discussion with 2 femoral neck fracture, respectively (Table 1).

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Mortality in Femoral Neck Fracture Patients

TABLE 1. Baseline Participant Characteristics Variable Death within 2 Years, N = 304 No Death Within 2 Years, N = 1943 Total, N = 2247 Age, mean (SD) y 80.6 (8.3) 74.5 (10.9) 75.3 (10.8) Sex, n (%) Female 155 (51.0) 1130 (58.2) 1285 (57.2) Male 149 (49.0) 813 (41.8) 962 (42.8) BMI, mean (SD) kg/m2 24.0 (4.9) 25.0 (4.6) 24.8 (4.7) ASA classification, n (%) Class I/II 94 (30.9) 1128 (58.1) 1222 (54.4) Class III/IV/V 210 (69.1) 815 (41.9) 1025 (45.6) Prefracture living setting, n (%) Not institutionalized 276 (90.8) 1871 (96.3) 2147 (95.5) Institutionalized 28 (9.2) 72 (3.7) 100 (4.5) Prefracture functional status, n (%) Independent ambulator 158 (52.0) 1572 (80.9) 1730 (77.0) Using ambulatory aid 146 (48.0) 371 (19.1) 517 (23.0) Study, n (%) FAITH (internal fixation) 150 (49.3) 856 (44.1) 1006 (44.8) HEALTH (joint replacement) 154 (50.7) 1087 (55.9) 1241 (55.2) Baseline osteoporosis medication use, n (%) Yes 31 (10.2) 136 (7.0) 167 (7.4) No 273 (89.8) 1807 (93.0) 2080 (92.6) Heart disease, n (%) Yes 149 (49.0) 574 (29.5) 723 (32.2) No 155 (51.0) 1369 (70.5) 1524 (67.8) High blood pressure, n (%) Yes 209 (68.8) 1066 (54.9) 1275 (56.7) No 95 (31.3) 877 (45.1) 972 (43.3) Lung disease, n (%) Yes 77 (25.3) 315 (16.2) 392 (17.4) No 227 (74.7) 1628 (83.8) 1855 (82.6) Diabetes, n (%) Yes 72 (23.7) 323 (16.6) 395 (17.6) No 232 (76.3) 1620 (83.4) 1852 (82.4) Kidney disease, n (%) Yes 64 (21.1) 140 (7.2) 204 (9.1) No 240 (78.9) 1803 (92.8) 2043 (90.9) Depression, n (%) Yes 47 (15.5) 242 (12.5) 289 (12.9) No 257 (84.5) 1701 (87.5) 1958 (87.1) Fracture displacement, n (%) Undisplaced 124 (40.8) 556 (28.6) 680 (30.3) Displaced 180 (59.2) 1387 (71.4) 1567 (69.7) Level of the fracture line, n (%) Subcapital 212 (69.7) 1126 (58.0) 1338 (59.5) Midcervical 73 (24.0) 704 (36.2) 777 (34.6) Basal 19 (6.3) 113 (5.8) 132 (5.9) Pauwels classification, n (%) Type I 50 (16.4) 237 (12.2) 287 (12.8) Type II 184 (60.5) 1097 (56.5) 1281 (57.0) Type III 70 (23.0) 609 (31.3) 679 (30.2) Time from injury to surgery, mean 58.1 (69.7) 50.8 (74.2) 51.8 (73.7) (SD) hours Length of surgery, mean (SD) 71.1 (40.4) 75.3 (37.8) 74.7 (38.2) minutes

(continued on next page)

TABLE 1. (Continued) Baseline Participant Characteristics Variable Death within 2 Years, N = 304 No Death Within 2 Years, N = 1943 Total, N = 2247 Type of anesthesia, n (%) General 139 (45.7) 1042 (53.6) 1181 (52.6) Regional 165 (54.3) 901 (46.4) 1066 (47.4) Intraoperative blood loss, mean 203.0 (200.5) 207.4 (180.4) 206.8 (183.2) (SD), mL Postoperative thromboprophylaxis, n (%) Yes 301 (99.0) 1932 (99.4) 2233 (99.4) No 3 (1.0) 11 (0.6) 14 (0.6) Majority of procedure performed by, n (%) Surgeon 187 (61.5) 1188 (61.1) 1375 (61.2) Resident 99 (32.6) 641 (33.0) 740 (32.9) Fellow 18 (5.9) 114 (5.9) 132 (5.9)

ASA, American Society of Anesthesiologists; BMI, body mass index; SD, standard devation.

Predictors of 24-Month Mortality before femoral neck fracture, and having a comorbidity were Older age (HR 1.42 for every 10-year increase, 95% CI factors found to be associated with a higher risk of mortality 1.22–1.65; P , 0.001), lower BMI (HR 1.23 for every 5- within 90 days of femoral neck fracture in this analysis. point decrease, 95% CI 1.08–1.39; P = 0.002), American Overall, the majority of the findings from our analysis support Society of Anesthesiologists (ASA) class III/IV/V (HR 1.53 the existing literature concerning predictive variables of – vs. class I/II, 95% CI 1.14–2.04; P = 0.004), use of an ambu- mortality after hip fracture.6,13 18 latory aid before femoral neck fracture (HR 2.10 vs. ambulat- Increasing age was a factor found to be associated with ing independently, 95% CI 1.63–2.71; P , 0.001), and a higher risk of 1-, 2-, 3-, 5-, and 10-year mortality across – kidney disease (HR 2.14, 95% CI 1.60–2.86; P , 0.001) several previously completed studies.6,13 22 Other studies were associated with a higher risk of mortality within 24 have identified kidney disease, including renal failure, which months of femoral neck fracture (Table 2). We did not find is the final stage of chronic kidney disease, to be an essential a difference in treatment methods (internal fixation vs. joint predictor of mortality following hip fracture.14,20,23,24 arthroplasty) in the risk of mortality. In addition, no other Although we did control for ASA grade in our analysis, factors were significantly associated with mortality (P . patients with chronic kidney disease usually have more co- 0.05). morbidities and more postoperative complications, which are highly likely to contribute to mortality within 90 days of hip Predictors of 90-Day Mortality fracture surgery.23,24 Ninety-nine (4.1%) of 2247 participants died within 90 Lower BMI being predictive of mortality within 24 days of femoral neck fracture. Older age (HR 1.34 for every 10- months of hip fracture was another finding from our analysis year increase, 95% CI 1.04–1.73; P = 0.03), lower BMI (HR that supports the current published literature.18,25,26 In partic- 1.33 for every 5-point decrease, 95% CI 1.04–1.70; P =0.02), ular, one study reported that patients with a BMI .26 had a use of an ambulatory aid before femoral neck fracture (HR 4.39 lower mortality rate than those with BMI ,22 (odds ratio 2.6 vs. ambulating independently, 95% CI 2.74–7.02; P , 0.001), for the 1-year survival, 95% CI 1.20–5.50; P = 0.006).25 and having a comorbidity (HR 2.27, 95% CI 1.02–5.04; P = Similarly, another study found that compared to participants 0.04) were associated with a higher risk of mortality within 90 of normal weight, mortality in overweight (HR 0.74, 95% CI days of femoral neck fracture (Table 3). No other factors were 0.62–0.88; P = 0.001) and obese (HR 0.74, 95% CI 0.60– significantly associated with mortality (P . 0.05). 0.91; P = 0.004) participants was significantly lower after sustaining a hip fracture.26 Although the observation of higher mortality in individuals with a lower BMI is consistent DISCUSSSION among the published literature, further research is still Using data from the FAITH and HEALTH trials, we required to understand the reasons for reduced mortality in investigated which factors are associated with mortality obese and overweight participants when compared to those of within 90 days and 24 months of femoral neck fracture. normal weight.26 There is evidence suggesting that poor Older age, lower BMI, ASA class III/IV/V, use of an nutritional status is linked to increased mortality in hip frac- ambulatory aid before femoral neck fracture, and kidney ture patients, which may explain why underweight partici- disease were factors found to be associated with a higher risk pants with a lower BMI are at a higher risk of mortality.22,27 of mortality within 24 months of femoral neck fracture in this The impact of time from injury to hip surgery has been analysis. Older age, lower BMI, use of an ambulatory aid a topic of discussion among orthopaedic surgeons. In our

TABLE 2. Factors Associated With 24-Month Mortality TABLE 3. Factors Associated With 90-Day Mortality (n = 2247; (n = 2247; 304 Events) 91 Events) Independent Variable HR (95% CI) P Independent Variable HR (95% CI) P Age (10-y increase) 1.42 (1.22–1.65) ,0.001 Age (10-y increase) 1.34 (1.04–1.73) 0.03 BMI (5-point decrease) 1.23 (1.08–1.39) 0.002 BMI (5-point decrease) 1.33 (1.04–1.70) 0.02 ASA classification Prefracture functional status Class III/IV/V vs. Class I/II 1.53 (1.14–2.04) 0.004 Using ambulatory aid vs. 4.39 (2.74–7.02) ,0.001 Prefracture functional status independent ambulator Using ambulatory aid vs. 2.10 (1.63–2.71) ,0.001 Comorbidities independent ambulator Yes vs. No 2.27 (1.02–5.04) 0.04 Kidney disease ASA classification Yes vs. no 2.14 (1.60–2.86) ,0.001 Class III/IV/V vs. Class I/II 1.44 (0.87–2.38) 0.15 Sex Study Male vs. female 1.03 (0.80–1.32) 0.84 FAITH (internal fixation) vs. 0.88 (0.53–1.46) 0.62 Prefracture living setting HEALTH (joint replacement) Institutionalized vs. not 1.40 (0.93–2.11) 0.11 Time from injury to surgery (hours) 1.00 (0.998–1.003) 0.69 institutionalized Length of surgery (minutes) 0.99 (0.992–1.004) 0.51 Study Type of anesthesia FAITH (internal fixation) vs. 0.79 (0.48–1.31) 0.36 General vs. Regional 0.88 (0.55–1.41) 0.58 HEALTH (joint replacement) Baseline osteoporosis medication use ASA, American Society of Anesthesiologists; BMI, body mass index. Yes vs. no 1.02 (0.69–1.52) 0.91 Heart disease Yes vs. no 1.14 (0.88–1.47) 0.33 HIP ATTACK (The Hip Fracture Accelerated Surgical High blood pressure Treatment and Care Track) trial aimed to definitively answer Yes vs. no 1.02 (0.79–1.33) 0.86 this particular question by examining whether accelerated Lung disease care (surgery within 6 hours of hip fracture diagnosis) versus Yes vs. no 1.26 (0.96–1.66) 0.10 standard care (surgery within 24 hours) could reduce mortal- Diabetes ity and major complications in hip fracture patients aged 45 Yes vs. no 1.22 (0.92–1.62) 0.17 years or older.28 The trial did not find a reduction in mortality Depression at 30 and 90 days when surgery was conducted within 6 Yes vs. no 0.98 (0.70–1.36) 0.89 hours, as well as found no difference between the surgery Fracture displacement groups on major complications occurring within 90 days.28 Displaced vs. undisplaced 0.75 (0.48–1.18) 0.21 However, it has been proposed that earlier surgery may Level of the fracture line shorten the period of immobility, which may improve func- Midcervical vs. subcapital 0.85 (0.64–1.12) tional outcomes, lessen costs and, ultimately, reduce morbid- Basal vs. subcapital 1.48 (0.90–2.41) Overall: 0.10 ity and mortality in the long run. Pauwels classification Several studies found that a high ASA classification as Type I vs. type III 1.15 (0.77–1.72) well as ambulating with an assistive device before hip fracture Type II vs. type III 1.08 (0.80–1.46) Overall: 0.79 was predictive of mortality.13,18,21,22,27 It has been hypothe- Time from injury to surgery (h) 1.001 (0.99–1.002) 0.33 sized that the use of an ambulatory aid may lead to higher Length of surgery (mins) 1.001 (0.99–1.004) 0.67 mortality rates because it delays the patient’s ability to imme- Type of anesthesia diately bear weight after hip fracture surgery or may be a General vs. Regional 1.04 (0.79–1.36) 0.79 surrogate for frailty.29 Intraoperative blood loss (mL) 1.00 (1.00–1.001) 0.21 Despite our findings indicating otherwise, there have Postoperative thromboprophylaxis been studies published reporting that being institutionalized Yes vs. no 0.58 (0.14–2.40) 0.45 before femoral neck fracture is associated with a higher risk Majority of procedure performed by of mortality.20,30 In a retrospective cohort study of 3992 Resident vs. surgeon 0.98 (0.76–1.27) patients older than 45 years having an osteoporotic hip frac- Fellow vs. surgeon 1.05 (0.64–1.72) Overall: 0.97 ture, institutionalization before injury occurrence (incidence – , ASA, American Society of Anesthesiologists; BMI, body mass index. rate ratio 1.48, 95% CI 1.36 1.60; P 0.05) was associated with a higher risk of mortality within 5 years of injury.20 In our analysis, we found that treatment method analysis, the mean time from injury to surgery was 57.9 hours (internal fixation in the FAITH trial vs. joint arthroplasty in (SD 215.4 hours) and not found to be significantly associated the HEALTH trial) was not significantly associated with with mortality within 24 months (HR 1.00, 95% CI 1.00– mortality (P = 0.36) overall within 24 months nor within 90 1.00; P = 0.86) nor within 90 days (HR 1.00, 95% CI days (P = 0.62) of fracture. Numerous published studies sup- 0.998–1.003; P = 0.69) of fracture. The recently completed port our finding that there is no overall difference in mortality

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Bzovsky et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 between patients treated with arthroplasty versus internal fix- ACKNOWLEDGMENTS ation.27,31–34 However, some published literature has sug- The authors thank the HEALTH and FAITH gested trends of internal fixation favoring mortality Investigators (http://links.lww.com/JOT/B226). reduction.27,34,35 A prospective multicenter study of 697 patients aged over 80 years, presenting with a fracture of the upper femur, treated with either internal fixation or arthro- REFERENCES plasty found that in their univariate analysis, mortality was 1. Dhanwal DK, Dennison EM, Harvey NC, et al. Epidemiology of hip higher with arthroplasty (25%) than with internal fixation fracture: worldwide geographic variation. Indian J Orthop. 2011;45: (17%; P = 0.002), but that the effect no longer existed in their 15–22. multivariate analysis (P . 0.05).27 Bhandari et al35 conducted 2. Cooper C, Campion G, Melton LJ. Hip fractures in the elderly: a world- a meta-analysis to determine the effect of arthroplasty (hemi- wide projection. Osteoporos Int. 1992;2:285–289. 3. Mundi S, Pindiprolu B, Simunovic N, et al. Similar mortality rates in hip arthroplasty, bipolar arthroplasty, and total hip arthroplasty) fracture patients over the past 31 Years: a systematic review of RCTs. versus internal fixation (with a fixed-angle screw and plate or Acta Orthop. 2014;85:54–59. with multiple screws) on rate of mortality in patients aged 65 4. Leibson C, Tosteson A, Gabriel S, et al. Mortality, disability, and nursing years or older with a displaced femoral neck fracture. The home use for persons with and without hip fracture: a population-based authors found a trend toward an increase in the relative risk study. J Am Geriatr Soc. 2002;50:1644–1650. fi 5. Carpintero P, Caeiro JR, Carpintero R, et al. Complications of hip frac- of death in the rst 4 months and at one year after arthroplasty tures: a review. World J Orthop. 2014;5:402–411. compared with internal fixation. Another meta-analysis of 6. Schnell S, Friedman SM, Mendelson DA, et al. The 1-year mortality of elderly patients with displaced femoral neck fractures com- patients treated in a hip fracture Program for elders. Geriatr Orthop Surg pared the effect of arthroplasty versus internal fixation on Rehabil. 2010;1:6–14. various clinical outcomes and found that, overall, there was 7. Fixation using Alternative Implants for the Treatment of Hip fractures (FAITH) Investigators. Fracture fixation in the operative management of increased postoperative risk for mortality with arthroplasty as hip fractures (FAITH): an international, multicentre, randomised con- compared to internal fixation, but that there was no statisti- trolled trial. Lancet. 2017;389:1519–1527. cally significant difference between the 2 groups at the dif- 8. HEALTH Investigators, Bhandari M, Einhorn TA, Jay RM, et al. Total ferent follow-up times.34 hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019; The primary strength of this analysis is the preplanned 381:2199–2208. 9. Peduzzi P, Concato J, Feinstein AR, et al. Importance of events per analysis of a large number of well characterized patients from independent variable in proportional hazards regression analysis II. over 130 clinical sites across 12 countries that were included Accuracy and precision of regression estimates. J Clin Epidemiol. when examining the research questions. Such a large and diverse 1995;48:1503–1510. sample size increases the external validity and generalizability of 10. Sprague S, Schemitsch EH, Swiontkowski M, et al. Factors associated the research findings from this analysis. Another strength is that with revision surgery after internal fixation of hip fractures. J Orthop Trauma. 2018;32:223–230. we focused on femoral neck fractures, whereas most published 11. Pauwels F. Der Schenkelhalsbruch: Ein Mechanisches Problem. studies on this topic have combined all hip fracture types Stuttgart, Germany: Ferdinand Enke; 1935. together. However, not being able to include all participants 12. Bartonícek J. Pauwels’ classification of femoral neck fractures: correct from the FAITH and HEALTH trials in this analysis as a result interpretation of the original. J Orthop Trauma. 2001;15:358–360. of missing data was a limitation. In addition, because only those 13. Paksima N, Koval KJ, Aharanoff G, et al. Predictors of mortality after hip fracture: a 10-year prospective study. Bull NYU Hosp Jt Dis. 2008;66: variables that were collected as part of the FAITH and HEALTH 111–117. trials could be used in the analysis, it may be possible that not all 14. Endo A, Baer HJ, Nagao M, et al. Prediction model of in-hospital mor- factors associated with mortality were captured. Both trials also tality after hip fracture surgery. J Orthop Trauma. 2018;32:34–38. had strict exclusion criteria that may have resulted in healthier 15. Khan MA, Hossain FS, Ahmed I, et al. Predictors of early mortality after patients being enrolled. For example, patients must have been hip fracture surgery. Int Orthop. 2013;37:2119–2124. 16. Karadsheh MS, Rodriguez EK, Harris MB, et al. Mortality and revision ambulatory before the fracture (although they may have used an surgery are increased in patients with Parkinson’s disease and fractures of aidsuchasacaneorwalker)andhadnohistoryoffrank the femoral neck. Clin Orthop Relat Res. 2015;473:3272–3279. dementia to be included. Given the possibility that healthier 17. Lin JCF, Liang WM. Mortality, readmission, and reoperation after hip patients were enrolled in both trials, this may explain, in part, a fracture in nonagenarians. BMC Musculoskelet Disord. 2017;18:144. lower 24-month mortality rate of 13.5%, as compared to higher 18. Juliebo V, Krogseth M, Skovlund E, et al. Medical treatment predicts mortality after hip fracture. J Gerontol Ser A Biol Sci Med Sci. 2010;65A:442–449. rates in the literature. 19. Mukka S, Knutsson B, Krupic F, et al. The influence of cognitive status on In conclusion, given the major public health problem outcome and walking ability after hemiarthroplasty for femoral neck fracture: a surrounding femoral neck fractures in older adults, it is prospective cohort study. Eur J Orthop Surg Traumatol. 2017;27:653–658. necessary to gain a further understanding of which patients 20. Guzon-Illescas O, Perez Fernandez E, Crespí Villarias N, et al. Mortality are at risk for mortality in order for surgeons to be able to after osteoporotic hip fracture: incidence, trends, and associated factors. J Orthop Surg Res. 2019;14:203–211. provide their patients with an accurate prognosis following 21. Åhman R, Siverhall PF, Snygg J, et al. Determinants of mortality after injury and to develop intervention strategies. Our analysis hip fracture surgery in Sweden: a registry-based retrospective cohort found that factors that are indicative of a poorer health status, study. Sci Rep. 2018;8:15695. such as older age, lower BMI, worse ASA class, use of an 22. Folbert E, Hegeman J, Vermeer M, et al. Improved 1-year mortality in ambulatory aid, and kidney disease, were associated with a elderly patients with a hip fracture following integrated orthogeriatric fi treatment. Osteoporos Int. 2017;28:269–277. higher risk of mortality. We did not nd treatment methods 23. Hsiue PP, Seo LJ, Sanaiha Y, et al. Effect of kidney disease on hemi- (internal fixation vs. joint arthroplasty) to be overall signif- arthroplasty outcomes after femoral neck fractures. J Orthop Trauma. icantly associated with mortality. 2019;33:583–589.

24. Puvanesarajah V, Amin R, Qureshi R, et al. Outcomes following surgical 30. Chang W, Lv H, Feng C, et al. Preventable risk factors of mortality after management of femoral neck fractures in elderly dialysis-dependent hip fracture surgery: systematic review and meta-analysis. Int J Surg. patients. Arch Orthop Trauma Surg. 2018;138:757–764. 2018;52:320–328. 25. Flodin L, Laurin A, Lökk J, et al. Increased 1-year survival and discharge 31. Ma HH, Chou TFA, Tsai SW, et al. Outcomes of internal fixation versus to independent living in overweight hip fracture patients: a prospective hemiarthroplasty for elderly patients with an undisplaced femoral neck frac- study of 843 patients. Acta Orthop. 2016;87:146–151. ture: a systematic review and meta-analysis. J Orthop Surg Res. 2019;14:320. 26. Prieto-Alhambra D, Premaor MO, Avilés FF, et al. Relationship between 32. Richards JT, Overmann AL, O’Hara NN, et al. Internal fixation versus arthro- mortality and BMI after fracture: a population-based study of men and plasty for the treatment of nondisplaced femoral neck fractures in the elderly: a women aged $40 years. J Bone Miner Res. 2014;29:1737–1744. systematic review and meta-analysis. J Orthop Trauma. 2020;34:42–48. 27. Duriez P, Devaux T, Chantelot C, et al. Is arthroplasty preferable to 33. Ye CY, Liu A, Xu MY, et al. Arthroplasty versus internal fixation for internal fixation for the treatment of extracapsular fracture of the upper displaced intracapsular femoral neck fracture in the elderly: systematic femur in the elderly? Orthop Traumatol Surg Res. 2016;102:689–694. review and meta-analysis of short- and long-term effectiveness. Chin 28. HIP ATTACK Investigators. Accelerated surgery versus standard care in Med J Engl. 2016;129:2630–2638. hip fracture (HIP ATTACK): an international, randomised, controlled 34. Wang J, Jiang B, Marshall RJ, et al. Arthroplasty or internal fixation for trial. Lancet. 2020;395:698–708. displaced femoral neck fractures: which is the optimal alternative for 29. Malik AT, Quatman-Yates C, Phieffer LS, et al. Factors associated with elderly patients? A meta-analysis. Int Orthop. 2009;33:1179–1187. inability to bear weight following hip fracture surgery: an analysis of the 35. Bhandari M, Devereaux PJ, Swiontkowski MF, et al. Internal fixation ACS-NSQIP hip fracture procedure targeted database. Geriatr Orthop compared with arthroplasty for displaced fractures of the femoral neck. A Surg Rehabil. 2019;10:2151459319837481. meta-analysis. J Bone Joint Surg Am. 2003;85-A:1673–1681.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. e d h t, P f T- he tee ittee ports (NIH) erning rom the d Division of e -values. All P a nancial or material fi = 0.003), enrolled from a -Aventis, personal fees from P fi Lippincott Williams & Wilkins, Division of Orthopaedic Surgery, s Health, employment from Global b — ’ Services Incorporated. M. Bhandari ’ Marc Swiontkowski, MD, b , 0.001), and treated with arthroplasty ( a , 0.001) were associated with an increased P dence intervals, and associated American, publishing royalties, , fi — P -Aventis, outside the submitted work. M. Swiontkowski = 0.02), from a hospital in Oceania, India, or South fi Rudolf W. Poolman, MD, PhD, P c 0.001) as compared to a European hospital, and treated We were unable to contact 123 of 2520 participants (4.9%) , cial views of the National Institutes of Health. Research reported in fi P Volume 34, Number 11 Supplement, November 2020 RTICLE 0.001) were associated with an increased odds of not completing A tests were 2-tailed with alpha = 0.05. odds ratios, 95% con Results: United States ( Africa ( with arthroplasty ( odds of consent withdrawal. for their 24-month follow-upconsent from visits the and trial. 124 Beingnon-European non-White (4.9%) hospital ( withdrew ( their , the 24-month follow-up visit. Being enrolled from a hospital in the Marianne Comeau-Gauthier, MD, MSc, Department of Surgery, University of Western Ontario, London, ON, Canada; a c Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN. ict of interest. f fl nancial or material support from Wolters Kluwer Health J Orthop Trauma fi UPPLEMENT S Mohit Bhandari, MD, PhD, FRCSC, on behalf of the FAITH and HEALTH Investigators e b Emil H. Schemitsch, MD, FRCSC, b , a nancial or material support for the Journal of Bone and Joint Surgery fi a Bzovsky, MSc fi culties locating and communicating with fi

Hip fracture trials often suffer substantial loss to and So Surabhi Sivaratnam, BHSc, MD (Cand),

We conducted 2 multivariable logistic regression a Bzovsky, MSc, Division of Orthopaedic Surgery, Department of Surgery, McMaster University, 293 Wellington St. N, Suite 110, Hamilton, ON ﬁ Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada; Secondary Analysis of the FAITH and HEALTH Trials a www.jorthotrauma.com Sheila Sprague, PhD, | Predictors of Loss to Follow-up in Hip Fracture Trials: A Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Department of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden University Medical Center, Leiden, the Netherlands; d member for the Osteosynthesis and Trauma Care Foundation, personal fees from Pentopharm, personal fees from Sano collection, management, analysis, or interpretation of thefor data; the or Canadian the Orthopaedic preparation, Association,International review, personal Society or fees for from approval Fracture DePuy,member of Repair, board for the personal or the manuscript. fees committee Orthopaedic from member Trauma ITS, for editorial Association, the or Hip editorial Society, governing or board board governing or for board committee the for member Journal the for of the Orthopaedic Orthopaedic Trauma Trauma, Association board International, or board comm or commit National Institute of Arthritiscontent and is Musculoskeletal solely the and responsibility Skinthis of Diseases the publication of authors was and the also does National supportedInstitute not by Institutes for necessarily represent The Health of the County Research Health of Durham Comprehensive under & Clinical Award Tees Research Number Valley Network R01AR055267-01A1. Comprehensive in Th Local England. Research The Network, funding which sources operates had as no part role of in the design National or conduct of the study; t Canadian Institutes of Health Research0602-43), (MOP-106630 the and MCT-87771), Netherlands National Organisation Instituteswas of for also Health Health (1R01AR055267-01A1), funded, Research Stichting in NutsOhraMusculoskeletal and (SNO- part, Development Trauma, (80-82310-97-11032), through which and the is Physicians Early unrelated Research to Award this Program, study which provided (McMaster funding University, for Hamilton, this ON, study Canada). as The well FAITH as trial by was a also Canada Research supported Chair by in the (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. interpretation Research, The of funding the sources data; had or the no preparation, role review, in or design approval or of conduct the of manuscript. the The study; FAITH trial the was collection, supported managemen by research grants f Department of Surgery, McMaster University, Hamilton, ON, Canada; L8L 8E7, Canada (e-mail: [email protected]). reports board or committee memberboard for and the publishing American Orthopaedic royalties, Association, consultant to the Minnesota Boardoutside of the Medical Practice, submitted editorial work. or The gov remaining authors report no con Dutch Orthopaedic Association, research supportpersonal from fees Lima, from and Amgen researchwork. support Co, M. from personal Bhandari Link fees Orthopaedics, reportssupport outside from research and the Smith personal support submitted & fees from work. Nephew, F. from Acumed, Frihagen personal Pendopharma, LLC, re fees and research research from support support Synthes, from and and Aphria, personal personal research fees fees support from from Sano from Zimmer, Ferring outside Pharmaceuticals, the researc submitted Saunders/Mosby-Elsevier, personal fees frompersonal Smith fees & from Zimmer, Nephew, outside personalResearch the fees Solutions submitted Inc, from work. and S. Springer, employment Sprague personal from reports fees McMaster editorial University, or from outside governing Stryker, the board personal submitted for work. fees BMS R. Women from W. Swemac, Poolman an reports board or committee member for the Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway; and support from Saunders/Mosby-Elsevier, publishing royalties, Frede Frihagen, MD, PhD, FRCSC, S22 Accepted for publication AugustFrom 11, the 2020. Methods: Background: to contact participants atconsent 24 within months 24 postfracture months and of their withdrawal fracture. of Results were reported as analyses to determine which factors were predictive of being unable associated with their withdrawal of consent. participants or when participants, ortheir their consent. family We members, withdraw aimed towith determine being which unable factors to weretheir contact associated 24-month FAITH follow-up and and HEALTH to participants also for determine which factors were follow-up due to dif E. H. Schemitsch reports personal fees from Acumed, LLC, personal fees from Amgen Co, research support from Biocomposites, board or committee member The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health Reprints: So Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001928 Health, Inc. All rights reserved.

Discussion: Certain factors may be predictive of loss to follow-up implemented extensive retention strategies, not all partici- in hip fracture trials. We suggest that the identification of such pants were compliant with follow-up. Using the data from factors may be used to inform and improve retention strategies in these trials to identify factors that may be predictive of future orthopaedic hip fracture trials. being unable to contact participants may allow future researchers to develop more targeted strategies that miti- Key Words: loss to follow-up, withdrawal of consent, hip fractures gate the risk of LTFU and ensure that the patients retained Level of Evidence: Prognostic Level II. See Instructions for represent the sample that was recruited. Authors for a complete description of levels of evidence. (J Orthop Trauma 2020;34:S22–S28) METHODS INTRODUCTION Definition of Loss to Follow-up in HEALTH and FAITH Trials Adifficulty frequently faced in hip fracture trials is loss to follow-up (LTFU), which includes participants not Participants in the HEALTH and FAITH trials were deemed LTFU after their failure to complete the trial 24- returning for their follow-up appointments or being fi unable to reach participants by telephone for data collec- month visit due to dif culties locating participants or with tion, as well as participants, or their family members, participants, or their family members, withdrawing their withdrawing their consent to participate in the trial.1–3 consent. Exhaustive measures had been taken by the trial Even the most rigorously designed orthopaedic hip frac- personnel to contact patients before marking them as lost to ture randomized controlled trials (RCTs) incur the prob- follow-up. These measures included contacting the patient, lem of participant noncompliance with follow-up.1–4 High their primary care physician, as well as 3 alternative LTFU rates may lower the statistical power of a study’s contacts. Strategies used in the trials to minimize LTFU sample size and can substantially bias the study’sresults included aligning the follow-up schedule with that of their if the outcomes of participants remaining in the study normal surgical fracture clinic visits, numerous attempts to 3,5 contact participants or individuals affiliated with partici- differ from those who were LTFU. Moreover, high fi LTFU may result in a study’s findings no longer reflecting pants, routine veri cation of contact, and the selection of the sample of patients who were recruited. Therefore, the outcomes for which data could be collected through minimization of LTFU is an important methodological telephone rather than an in-person visit. 4 consideration. Statistical Methods The LTFU rates within trials can be quite high, with previous orthopaedic trauma RCTs reporting up to 28% Demographics LTFU.1 Although hip fracture trials in older adults seem to We examined the demographic and injury characteris- have lower rates of LTFU, ranging typically from 2% to tics of participants who study personnel were unable to – 12%,6 8 including the FAITH (6.9%) and HEALTH contact at 24 months, those who withdrew their consent from (12.0%) trials,9,10 it has been reported that approximately the trials, and those who completed their 24-month follow-up 20% of orthopaedic RCTs fail to report their LTFU.3 visit. Mean values and SDs were presented for continuous Therefore, these rates may actually be higher and remain variables, whereas frequencies and percentages were pre- an important source of potential bias.3 In other areas of sented for categorical variables. orthopaedics, LTFU rates exceed the commonly accepted 20% LTFU threshold.11,12 Recent research suggests that Analyses this 20% threshold may not be strict enough to mitigate We conducted 2 multivariable logistic regression the risk of bias incurred due to LTFU in orthopaedic analyses to investigate the association between selected trials.13 Zelle et al13 made this suggestion based on their factors and not completing the 24-month follow-up and findings that numerous statistically significant orthopaedic withdrawing consent, respectively. In the first analysis, the trials’ results became nonsignificant when simulated with dependent variable was not completing their 24-month an LTFU rate of 20%. Therefore, Zelle et al13 instead rec- follow-up visit for the reason that they could not be ommend that researchers aim for a follow-up rate that contacted in the HEALTH and FAITH studies during the exceeds 80%. 24-month follow-up period. It should be noted that Although the HEALTH and FAITH trials both participants who died during the 24-month follow-up achieved greater than 80% follow-up, the trial data can period were not included as being LTFU. The number of offer insights for both clinicians and researchers on independent variables and corresponding levels included in participant characteristics that may be associated with the 2 analyses were based on the recommendations of being unable to contact participants. The objective of this Peduzzietal.14 Notably, a low number of events per vari- preplanned analysis was to investigate and identify which able (EPV) (ie, EPV values less than 10) can lead to major baseline characteristics were predictive of being unable to problems involving model bias, precision, and significance. contact participants within 24 months of femoral neck In this analysis, the number of events was the number of fracture in patients aged 50 years or older participating in patients who were lost to follow-up, which was 123. To the HEALTH and FAITH trials.9,10 Although both of these avoid overfitting and ensure that our EPV was greater than large, international studies were rigorously designed and 10, we included 10 variables, with a total of 12 levels. The

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Sivaratnam et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 selection of these independent variables was based on pre- Factors Associated with LTFU vious literature and expert opinions. The factors included Participants of Indigenous, South Asian, East Asian, age, sex, ethnicity [White vs. other (Indigenous, South Black, or Hispanic/Latino ethnicity (OR vs. White 2.17, 95% Asian, East Asian, Black, or Hispanic/Latino)], body mass CI: 1.30–3.61; P = 0.003) were at an increased odds of not index (underweight [body mass index ,18.5] versus other completing the 24 month follow-up in the FAITH and (body mass index $18.5), prefracture living setting (insti- HEALTH trials. Participants who were enrolled from a hos- tutionalized vs. not institutionalized), prefracture func- pital in Canada (OR vs. Europe 2.51, 95% CI: 1.42–4.44; P , tional status (use of an aid vs. independent ambulation), 0.001), from a hospital in the United States (OR vs. Europe American Society for Anesthesiologists (ASA) physical 4.72, 95% CI: 2.73–8.17; P , 0.001), from a hospital in status (Class I/II vs. Class III/IV/V), hospital location Oceania, India, or South Africa (OR vs. Europe 3.91, 95% [Canada vs. the United States vs. Europe vs. Other CI: 1.96–7.82; P , 0.001), and treated with arthroplasty (OR (Oceania, India, and South Africa)], comorbidities (present vs. internal fixation 2.67, 95% CI: 1.71–4.17; P , 0.001) vs. not present), and type of surgical treatment (arthro- were also at an increased odds. Age, sex, body mass index, plasty vs. internal fixation). We also conducted a multivari- prefracture living setting, prefracture functional status, ASA able logistic regression analysis to investigate the physical status, and presence of a comorbidity were not found association between the same selected factors and with- to be associated with not completing the 24-month visit drawal of consent. The dependent variable was withdrawal (Table 2). of consent in the HEALTH and FAITH studies during the Similar to the first analysis, participants who were 24-month follow-up period. The statistical output of these enrolled from a hospital in the United States (OR vs. Europe analyses was reported as odds ratios (ORs) with corre- 1.77, 95% CI: 1.10–2.86; P = 0.02), from a hospital in sponding 95% confidence intervals (CIs) and associated Oceania, India, or South Africa (OR vs. Europe 2.75, 95% P-values. All analyses were conducted using R software CI: 1.53–4.95; P , 0.001), and treated with arthroplasty (OR (version 3.6.1, R Foundation for Statistical Computing, vs. internal fixation 3.10, 95% CI: 1.92–5.01; P , 0.001) Vienna, Austria). were at an increased odds of withdrawing consent in the FAITH and HEALTH trials. In contrast to the prior analysis, we found that not having a comorbidity approached statistical significance (OR 1.79, 95% CI 0.99–3.23; P = 0.053) for RESULTS being associated with an increased odds of withdrawing con- Participant Characteristics sent. Age, ethnicity, body mass index, prefracture living setting, prefracture functional status, and ASA physical status There were 247 of 2520 (9.8%) participants in were also not found to be associated with withdrawing con- HEALTH and FAITH who were LTFU. Specifically, 123 sent (Table 2). of 2520 (4.9%) participants from the HEALTH (77/1441 participants, 5.3%) and FAITH (46/1079 participants, 4.3%) trials did not complete their 24-month visit. In addition, 124 DISCUSSION of 2520 (4.9%) participants withdrew their consent to The results of this analysis show that certain participant participate in the HEALTH (96/1441, 6.7%) and FAITH characteristics and clinical variables, including ethnicity, (28/1079 participants, 2.6%) trials. country of the participating hospital, and type of surgical Table 1 presents the demographic and injury charac- treatment, have an impact on LTFU in femoral neck fracture teristics of participants who did not complete their patients aged 50 years or older. Specifically, participants with 24-month follow-up visit, withdrew consent, and those an ethnicity other than White, enrolled from a hospital not who were not LTFU. For those 123 participants who did located in Europe, and treated with arthroplasty were at an not complete the 24-month visit (excluding those who increased odds of being lost to follow-up. Surprisingly, other withdrew consent), the mean age was 74.1 (SD 12.0), most factors, including sex, age, body mass index, prefracture were female (57.7%), White (75.2%), enrolled from a hos- living setting, prefracture functional status, and ASA class, pital in North America (39.0%), and were not living in an were found to not be associated with LTFU in this population. institutionalized setting prefracture (95.1%). For those 124 Our results as well as other orthopaedic studies on predicting participants who withdrew their consent from the trial, the factors of LTFU have highlighted significant health care mean age was 78.4 (SD 9.8), most were female (71.0%), disparities that need to be addressed in terms of ethnicity, White (92.7%), enrolled from a hospital in North America disability, unemployment, and social support, to name a (40.3%), and were not living in an institutionalized setting few.1,15–18 Identifying predictors of LTFU will help future prefracture (91.9%). Among those 2397 participants who trials in targeting their “retention efforts” in the participants were not lost to follow-up, the mean age was 75.9 years associated with higher risk of LTFU. (SD 10.7), most were female (66.4%), White (90.4%), Previous literature supports our finding that the country enrolled from a hospital in Europe (43.7%), and were not of the participating hospital is associated with LTFU. A living in an institutionalized setting prefracture (95.0%). It systematic review investigating LTFU rates in 559 orthopae- should be noted that in general, the majority of participants dic studies of 131,836 participants identified that treatment in enrolled in both the FAITH and HEALTH trials were the United States as compared to other countries was a White (82.4% and 95.1%, respectively). predictive factor of LTFU (P = 0.01).3 Somerson et al3

TABLE 1. Patient Demographics Did Not Complete the 24-Month Variable Not LTFU*, N = 2273 Visit, N = 123 Withdrew Consent, N = 124 Age, mean (SD) (y) 75.7 (10.7) 74.1 (12.0) 78.4 (9.8) Sex, n (%) Male 770 (33.9) 52 (42.3) 36 (29.0) Female 1503 (66.1) 71 (57.7) 88 (71.0) Ethnic origin, n (%) Indigenous 5 (0.2) 2 (1.6) 0 (0.0) South Asian 125 (5.5) 12 (9.8) 2 (1.6) East Asian 20 (0.9) 4 (3.3) 1 (0.8) Black 57 (2.5) 9 (7.3) 5 (4.0) Hispanic or Latino 14 (0.6) 3 (2.4) 0 (0.0) White 2041 (89.8) 91 (74.0) 115 (92.7) Current drugs, n (%)† None 532 (23.4) 34 (27.6) 24 (19.4) NSAIDs 299 (13.2) 20 (16.3) 12 (9.7) General cardiac 824 (36.3) 39 (31.7) 45 (36.3) Opioid analgesics 188 (8.3) 15 (12.2) 14 (11.3) Pulmonary drugs 271 (11.9) 12 (9.8) 12 (9.7) Antihypertension drugs 1189 (52.3) 56 (45.5) 60 (48.4) Osteoporosis drugs 253 (11.1) 10 (8.1) 14 (11.3) BMI, n (%) N = 2232 N = 112 N = 115 Underweight (,18.5) 123 (5.4) 10 (8.1) 10 (8.1) Normal Weight (18.5–24.9) 1152 (50.7) 62 (50.4) 50 (40.3) Overweight (25.0–29.9) 701 (30.8) 33 (26.8) 37 (29.8) Obese (.30.0) 256 (11.3) 7 (5.7) 18 (14.5) Prefracture living setting, n (%) Institutionalized 110 (4.8) 6 (4.9) 10 (8.1) Not institutionalized 2163 (95.2) 117 (95.1) 114 (91.9) Prefracture functional status, n (%) Use of aid 539 (23.7) 28 (22.8) 33 (26.6) Independent ambulator 1734 (76.3) 95 (77.2) 91 (73.4) ASA physical status, n (%) Class I/II 1213 (53.4) 67 (54.5) 56 (45.2) Class III/IV/V 1060 (46.6) 56 (45.5) 68 (54.8) Centre, n (%) Canada 518 (22.8) 28 (22.8) 15 (12.1) United States 546 (24.0) 48 (39.0) 35 (28.2) Europe 994 (43.7) 23 (18.7) 54 (43.5) Oceania 101 (4.4) 8 (6.5) 19 (15.3) Africa 5 (0.2) 1 (0.8) 0 (0.0) India 109 (4.8) 15 (12.2) 1 (0.8) Comorbidity, n (%) Yes 2007 (88.3) 102 (82.9) 107 (86.3) No 266 (11.7) 21 (17.1) 17 (13.7) Surgical treatment, n (%) Arthroplasty 1268 (55.8) 77 (62.6) 96 (77.4) Internal ﬁxation 1005 (44.2) 46 (37.4) 28 (22.6)

*Participants who died during the 24-month follow-up period were captured under the “Not LTFU” group. †More than one drug could be selected. ASA, American Society of Anesthesiologists; BMI, body mass index; LTFU, lost to follow-up; SD, standard deviation. indicated that the reasons for this difference remain unclear privately funded health care system) were more likely to be and are likely multifactorial. Similarly, in an RCT of 2381 LTFU than those who received treatment in Canada, Europe, adults with an open fracture, participants who received treat- or Australia (which have publicly funded health care systems) ment in the United States (which has a predominantly (OR 3.56, 95% CI 2.46–5.17; P , 0.001).4 Although we also

TABLE 2. Prognostic Variables Associated With LTFU (n = 2520) Did Not Complete the Withdrawal of Consent 24-Month Visit (123 Events) (124 Events) Variable OR (95% CI) P OR (95% CI) P Ethnicity Other vs. White 2.17 (1.30–3.61) 0.003 1.79 (0.83–3.89) 0.14 Hospital location Overall: ,0.001 Overall: ,0.001 Canada vs. Europe 2.51 (1.42–4.44) ,0.001 0.58 (0.32–1.05) 0.07 United States vs. Europe 4.72 (2.73–8.17) ,0.001 1.77 (1.10–2.86) 0.02 Other vs. Europe 3.91 (1.96–7.82) ,0.001 2.75 (1.53–4.95) ,0.001 Surgical treatment arthroplasty vs. internal ﬁxation 2.67 (1.71–4.17) ,0.001 3.10 (1.92–5.01) ,0.001 Age (10-y increase) 0.93 (0.77–1.13) 0.51 1.14 (0.92–1.41) 0.22 Sex Male vs. female 1.34 (0.91–1.98) 0.20 0.85 (0.56–1.28) 0.45 BMI Underweight (BMI ,18.5) vs. 1.58 (0.79–3.17) 0.18 1.51 (0.76–3.02) 0.24 other (BMI $18.5) Prefracture living setting Institutionalized vs. not 0.87 (0.36–2.07) 0.63 1.71 (0.85–3.45) 0.13 institutionalized Prefracture functional status Use of aid vs. independent 1.08 (0.67–1.75) 0.57 1.00 (0.64–1.57) 0.99 ambulator ASA physical status Class III/IV/V vs. Class I/II 0.77 (0.50–1.18) 0.40 1.12 (0.74–1.69) 0.59 Comorbidity No vs. Yes 1.27 (0.71–2.22) 0.78 1.79 (0.99–3.23) 0.053

ASA, American Society of Anesthesiologists; BMI, body mass index; LTFU, loss to follow-up; OR, odds ratio. identified that participants enrolled in hospitals located in Similar to our findings, a cohort study of 3202 Canada and other counties (Oceania, India, and South patients undergoing anterior cruciate ligament reconstruc- Africa) were at a higher risk of LTFU as compared to those tion found that those of non-White ethnicity (black: OR enrolled in Europe, we are unaware of any studies that have 3.64; other non-White: OR 1.81) were at a high risk for made comparisons between these locations. Similar to our 2-year LTFU.15 Ramkumar et al15 suggested that the most primary analysis, our secondary analysis demonstrated that pragmatic denominator among the described patients at risk participants enrolled in hospitals located in the United for LTFU was possibly socioeconomic status, but they were States and other countries (Oceania, India, and South unable to directly address it in their study other than con- Africa) were at an increased odds of withdrawing consent trolling for level of education and marital status in their as compared to those enrolled in Europe, but to the best of analyses, which still showed an increased LTFU associated our knowledge, there have not been previously conducted with non-White ethnicities. Other orthopaedic studies have studies evaluating predictors of consent withdrawal in ortho- shown that when controlling for factors, including age, sex, paedics. Our results suggest that there may be geographic insurance, income, and education, those of non-White differences that may have enhanced follow-up and compli- ethnicity were less likely to receive operative fixation for a ance. In Norway and several other European countries, where clavicle fracture,21 and benefits from total knee arthroplasty nearly all hospitals are government-run, patients “belong” to a or total hip arthroplasty.22 In addition, other studies have hospital, which is determined by their proximity to the hos- suggested better compliance with follow-up when patients pital. This European model of “belonging” to a certain hos- and their physician share the same ethnic identity.23,24 pital may prove to be more successful in participant retention Similar to Ramkumar et al, we were unable to address because for most health issues, patients will have to go to that socioeconomic status in our analyses. At this time, there is particular hospital. In addition, these patients may also feel limited research on what factors contribute to multicultural more responsible toward their assigned hospital because they participants’ recruitment and retention in studies. It is know there is a long-term relationship. To expand on this highly complex, and as the studies above suggested, there further, there have been some studies that have found are other reasons why those of non-White ethnicity access decreased participant retention with increased living distance health care differently and this probably has an impact on from their hospital.16,19,20 follow-up visit compliance. Collection of information on

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Follow-up in Hip Fracture Trials level of education, working status, marital status, social odds of a patient being LTFU.4 Another study found that each support, and income would allow for researchers to assess additional year in age reduced the odds of a participant chang- the effect of socioeconomic status independently. An ing some of their contact information by 2.2% (OR 0.98, 95% improved understanding of these factors would help ensure CI: 0.97–0.99), which would allow for study personnel to more that patients recruited and retained in studies would be easily contact study participants who are older. representative of the diverse populations that they are Overall, the literature investigating LTFU in orthopae- drawn from. Ensuring diversity would also assist in making dic trials remains scarce and has not reached consensus on study findings more generalizable to the whole population. which factors are predictive of LTFU. Previous literature has Our findings indicated that participants whose femoral also identified strategies that may be implemented to prevent neck fracture was treated with arthroplasty (OR vs. internal LTFU in orthopaedic settings.4 For example, a Cochrane fixation 2.67, 95% CI: 1.71–4.17; P , 0.001) were associated review suggested that “incentives, communication strategies, with an increased odds of LTFU, as well as an increased odds new questionnaire strategies, behavioural or motivational of withdrawal of consent (OR vs. internal fixation 3.10, 95% strategies, case management, and methodological strategies” CI: 1.92–5.01; P , 0.001) within 24 months of femoral neck could be used to improve retention rates.30 Some of these fracture. To the best of our knowledge, no other orthopaedic strategies were implemented in the FAITH and HEALTH trauma studies have evaluated the impact of surgical treat- studies. For example, telephone follow-ups were conducted ment for femoral neck fracture patients on LTFU, nor on when in-clinic visits were not possible.5,13 However, these withdrawal of consent. However, a systematic review and strategies may have been more effective if they were focused meta-analysis of 8 RCTs comparing arthroplasty to internal on participants who were at higher risk of being lost to fixation in 2206 elderly patients with a displaced intracapsular follow-up. We hypothesized that the results of the present femoral neck fracture found that those in the arthroplasty analysis would allow research personnel to be more cognizant group reported significantly lower complications (risk ratio: of which participants are more likely to be lost to follow-up, 0.56, 95% CI: 0.38–0.80; P , 0.01), reoperations (risk ratio: allowing for targeted implementation of retention strategies. 0.17, 95% CI: 0.13–0.22; P , 0.00001), revision rates (risk However, given our findings, a future area of study may entail ratio: 0.11, 95% CI: 0.08–0.16; P , 0.00001), and better evaluating these targeted strategies to help increase retention. function compared to those treated with internal fixation, This analysis has both strengths and limitations. The and they were less likely to suffer postoperative pain.25 strengths of this analysis include the use of data from 2 RCTs These findings lead us to suggest that those experiencing that were rigorously designed by the same principal investi- more complications postsurgery may be more likely to return gators, followed similar follow-up regimens, and had large to clinic for a follow-up visit and, therefore, not be lost to sample sizes. In addition, these RCTs enrolled patients from follow-up. In addition, the increased odds of LTFU in arthro- numerous countries, and thus both of these studies had high plasty patients may be related to arthroplasty being a more external validity. To the best of our knowledge, determining definitive treatment, whereas internal fixation patients require predictors of withdrawal of consent has not previously been monitoring to ensure fracture healing, which takes time and evaluated in the field of orthopaedics and therefore, adds may translate to a greater attendance at follow-up visits. novel findings to the literature. Some of the limitations of the Another study investigating hip fractures among elderly present analysis are as follows. Given that our data were patients found that LTFU was highest among patients living retrospectively obtained from 2 completed RCTs, we were in institutionalized settings, such as nursing homes, as well as only able to include previously collected variables in the patients who were not independent ambulators, which model. Socioeconomic factors such as alcohol consumption, differed from our findings of no statistically significant smoking habits, and insurance status may have been impor- differences between these groups of patients.26 tant predictors of LTFU, but such data were not collected in It should also be noted that there is conflicting evidence the HEALTH and FAITH studies and, therefore, we were regarding the association of certain factors with LTFU.1,15,27,28 unable to analyze these factors. Because the FAITH and Although not focusing on the elderly hip fracture population, HEALTH trials included older patients, most individuals for example, previous orthopaedic trauma studies of younger were already retired and for that reason, we were unable to adults investigating LTFU found that male sex is predictive of include employment status before injury as a factor in the LTFU.1,15 These findings were supported by other literature model. This factor may have been predictive of LTFU in a indicating that women are more inclined to visit their doctors younger population, as has been suggested in previous than men, which in turn might imply that they would be more literature.1,18,31 Moreover, given that 123 patients were lost likely to attend follow-up appointments.27 By contrast, a study to follow-up, we were limited in the number of factors and investigating hip fractures among elderly people found that sex corresponding levels that could be included in our model. was not associated with LTFU.28 Agemaynotbeanimportant Overall, we conclude that certain characteristics and factor in predicting LTFU in the elderly population for the clinical variables may be predictive of LTFU in orthopaedic reason that older participants are less transient and may be trials; however, further research must be done. In our study easier to be contacted by study personnel, but other studies of elderly hip fracture patients with a 9.8% LTFU rate, have demonstrated that younger adults may be at a higher risk participants with an ethnicity other than White, enrolled of LTFU compared to older adults.4,29 Specifically, in an RCT from a hospital not located in Europe, and treated with of 2381 adults with an open fracture, an age of ,30 years (OR arthroplasty were at an increased odds of being lost to 2.16, 95% CI: 1.19–3.95; P = 0.012) significantly increased the follow-up. Predicting factors for LTFU have previously

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Sivaratnam et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 been mainly described for the trauma population so it is 14. Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number highly likely that the elderly hip fracture population does not of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379. share the same predictors. Additional research is necessary, 15. Ramkumar PN, Tariq MB, Amendola A, et al. Risk factors for loss to given that current literature on this frequently encountered follow-up in 3202 patients at 2 years after anterior cruciate ligament problem remains scarce, especially in the older hip fracture reconstruction: implications for identifying health disparities in the population. We suggest that the identification of such factors MOON prospective cohort study. Am J Sports Med. 2019;47:3173–3180. may be used to inform and improve retention strategies in 16. Patterson JT, Albright PD, Jackson JH, et al. Travel barriers, unemployment, and external fixation predict loss to follow-up after surgical man- future orthopaedic hip fracture trials. agement of lower extremity fractures in Dar es Salaam, Tanzania. OTA Int. 2020;3:e061–e067. 17. Sleat GKJ, Lefaivre KA, Broekhuyse HM, et al. Predicting completion of ACKNOWLEDGMENTS follow-up in prospective orthopaedic trauma research. OTA Int. 2019;2: e047–e052. The authors thank the HEALTH Investigators and the 18. ten Berg PWL, Ring D. Patients lost to follow-up after metacarpal frac- FAITH Investigators (http://links.lww.com/JOT/B250). tures. J Hand Surg Am. 2012;37:42–46. 19. Wharton MG, Shultz CL, Packard BD, et al. Patient compliance with REFERENCES follow-up after open reduction and internal fixation for treating malleolar – 1. Zelle BA, Buttacavoli FA, Shroff JB, et al. Loss of followup in ortho- ankle fractures: a retrospective review. UNM Orthop Res J. 2019;8:82 84. paedic trauma: who is getting lost to follow-up? J Orthop Trauma. 2015; 20. Badenhorst D, Van der Westhuizen C, Britz E, et al. Lost to follow-up: 29:510–515. challenges to conducting orthopaedic research in South Africa. South Afr – 2. Sprague S, Leece P, Bhandari M, et al. Limiting loss to follow-up in a Med J. 2018;108:917 921. fi multicenter randomized trial in orthopedic surgery. Control Clin Trials. 21. Schairer W, Nwachukwu B, Warren R, et al. Operative xation for 2003;24:719–725. clavicle fractures-socioeconomic differences persist despite overall 3. Somerson JS, Bartush KC, Shroff JB, et al. Loss to follow-up in orthopae- population increases in utilization. J Orthop Trauma. 2017;31: – dic clinical trials: a systematic review. Int Orthop. 2016;40:2213–2219. e167 e172. fi 4. Madden K, Scott T, McKay P, et al. Predicting and preventing loss to 22. Irgit K, Nelson C. De ning racial and ethnic disparities in THA and – follow-up of adult trauma patients in randomized controlled trials: an TKA. Clin Orthop Relat Res. 2011;469:1817 1823. example from the FLOW trial. J Bone Joint Surg. 2017;99:1086–1092. 23. Saha S, Arbelaez J, Cooper L. Patient-physician relationships and racial 5. Murray DW, Britton AR, Bulstrode CJ. Loss to follow-up matters. J disparities in the quality of health care. Am J Public Health. 2003;93: – Bone Joint Surg. 1997;79:254–257. 1713 1719. 6. Frihagen F, Nordsletten L, Madsen JE. Hemiarthroplasty or internal 24. Laveist T, Nuru-Jeter A. Is Doctor-patient race concordance associated fixation for intracapsular displaced femoral neck fractures: randomised with greater satisfaction with care? J Heal Soc Behav. 2002;43:296– controlled trial. BMJ. 2007;335:1251–1254. 306. 7. Parker MJ, Pryor G, Gurusamy K. Hemiarthroplasty versus internal fix- 25. Ye CY, Liu A, Xu MY, et al. Arthroplasty versus internal fixation for ation for displaced intracapsular hip fractures: a long-term follow-up of a displaced intracapsular femoral neck fracture in the elderly: systematic randomised trial. Injury. 2010;41:370–373. review and meta-analysis of short- and long-term effectiveness. Chin 8. Chammout GK, Mukka SS, Carlsson T, et al. Total hip replacement Med J Engl. 2016;129:2630–2638. versus open reduction and internal fixation of displaced femoral neck 26. Nygard H, Matre K, Fevang JM. Evaluation of timed up and go test as a fractures: a randomized long-term follow-up study. J Bone Joint Surg. tool to measure postoperative function and prediction of one year walk- 2012;94:1921–1928. ing ability for patients with hip fracture. Clin Rehabil. 2016;30:472–480. 9. HEALTH Investigators, Bhandari M, Einhorn TA, Guyatt G, et al. Total 27. Bertakis K, Azari R, Helms L, et al. Gender differences in the utilization hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019; of health care services. J Fam Pr. 2000;49:147–152. 381:2199–2208. 28. Wilson RT, Chase GA, Chrischilles EA, et al. Hip fracture risk among 10. Collaborators. Fixation using alternative implants for the treatment of hip community-dwelling elderly people in the United States: a prospective fractures (FAITH) investigators. fracture fixation in the operative man- study of physical, cognitive, and socioeconomic indicators. Am J Public agement of hip fractures (FAITH): an international, multicentre, rando- Health. 2006;96:1210–1218. mised controlled trial. Lancet. 2017;389:1519–1527. 29. London D, Stepan J, Goldfarb C, et al. The (In)stability of 21st 11. Sackett DL, Straus SE, Richardson WS, et al. Evidence-based Medicine century orthopedic patient contact information and its implications —How to Practice and Teach EBM. New York, NY: Churchill on clinical research: a cross-sectional study. Clin Trials. 2017;14: Livingstone; 1997. 187–191. 12. Kristman V, Manno M, Côté P. Loss to follow-up in cohort studies: how 30. Brueton VC, Tierney J, Stenning S, et al. Strategies to improve retention much is too much? Eur J Epidemiol. 2003;19:751–760. in randomised trials. Cochrane Database Syst Rev. 2013;12:MR000032. 13. Zelle BA, Bhandari M, Sanchez AI, et al. Loss of follow-up in ortho- 31. Norquist BM, Goldberg BA, Matsen FA. Challenges in evaluating paedic trauma: is 80% follow-up still acceptable? J Orthop Trauma. patients lost to follow-up in clinical studies of rotator cuff tears. J 2013;27:177–181. Bone Joint Surg Am. 2000;82:838–842.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. , h c t, , fi tee 110, P S29 a (NIH) f , d | a , board or committee = 0.01) and preoper- P -Aventis, personal fees from a Bzovsky, MSc, fi 0.01) were associated with fi , So P a S36) – Department of Surgery, University of Western b cation (AMD 1.99, Frede Frihagen, MD, PhD, fi Prognostic Level II. c 0.01), whereas higher American Society of Journal of Orthopaedic Trauma 2020;34:S29 , P Patients receiving THA are likely to receive small and health-related quality of life, function, femoral neck Mohit Bhandari, MD, PhD, FRCSC, 2.40, a Department of Traumatology, Orthopaedic Surgery and Emergency, e 2 RTICLE A J Orthop Trauma 0.01). sion was associated with lower functional scores (AMD 7.73, lower WOMAC function scores. Preoperative treatment for depres- Anesthesiologists classi ative use of an aid (AMD 5.39, ment of PCS scoresscores over over time time. compared with THA(AMD monopolar improved HA (but WOMAC not function bipolar HA) characteristics seem toimprovement play among a femoral neck larger fracture role patients. Key in Words: predicting functional fractures Conclusion: Level of Evidence: clinically unimportant improvementscompared in with health those utilityment receiving and monopolar function compared HA with and those little receiving improve- bipolar HA. Patient-speci ( -Aventis, outside the submitted work. S. Sprague reports editorial or governing board fi on behalf of the HEALTH Investigators f , UPPLEMENT a S Rudolf W. Poolman, MD, PhD, Marianne Comeau-Gauthier, MD, MSc, b a Diane Heels-Ansdell, MSc, Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada. 0.01), preoperative use f e , P d no clinically important 2.64, ict of interest. = 0.04) demonstrated less improve- Department of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden University Medical Center, Leiden, the ith HRQoL and function in this fl c 2 P 0.01), and partial weight-bearing status , 1.38, P 2 = 0.02), whereas higher American Society of and Sheila Sprague, PhD Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com P s Health, employment from Global Research Solutions Inc, and employment from McMaster University, outside the submitted work. The 2.66, ’ Hebrón, Barcelona, Spain; and ’ 2 Total hip arthroplasty (THA) has been argued to Division of Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway;

d Using repeated measures regression, we estimated the THA as compared to monopolar HA, but not bipolar HA,

Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; a Emil H. Schemitsch, MD, FRCSC, Ernesto Guerra-Farfán, MD, With Displaced Femoral Neck Fractures Managed With Arthroplasty? A Secondary Analysis of the HEALTH Trial Daniel Axelrod, MD, MSc (Cand),

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. What Predicts Health-Related Quality of Life for Patients Netherlands; for the Canadian OrthopaedicInternational Association, personal Society fees for from Fracture DePuy,member Repair, board for personal or fees the committee from Orthopaedic membermember ITS, Trauma for editorial for Association, the or Hip editorial the governing Society, or board Osteosynthesis board governing for or and board the committee for Trauma member the for Care the Orthopaedic Foundation, Trauma Association personal International, fees board from or commit Pentopharm, personal fees from Sano (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. Research, interpretation The of funding the sources data; had or no the role preparation, in design review, or or conduct approval of of the the study; manuscript. the collection, managemen Hospital Vall D Ontario, London, ON, Canada; Hamilton, ON L8L 8E7 (e-mail: [email protected]). support from Acumed,Pendopharma, LLC, and research research supportfor support BMS and from Women personal Aphria, feesremaining research from authors report Sano support no from con Ferring Pharmaceuticals, research support and personal fees from Saunders/Mosby-Elsevier, personal feesand from personal Smith fees & fromresearch Nephew, Zimmer, support outside personal from the Lima, feespersonal submitted and fees from work. research from support Springer, R. Smith from W. & personal Link Nephew, Poolman fees Orthopaedics, personal reports fees outside from board from the Stryker, or Synthes, submitted and committee personal work. personal member F. fees fees for Frihagen from from the reports Zimmer, outside Swemac, personal Dutch the fees Orthopaedic submitted from Association, work. Amgen M. Co Bhandari reports researc Anesthesiologists score (AMD postoperatively (AMD was more likely to[AMD] 1.88 improve points, PCS scores (adjusted mean difference of an aid (AMD J Orthop Trauma From the Accepted for publication August 11, 2020. Results: Methods: Background: functional advantages ofexplores THA factors over associatedpopulation. HA. w The current analysis improve health-related quality offemoral life neck (HRQoL) fracture and(HA). patients function compared in The with hemiarthroplasty HEALTH trial showe Universities Osteoarthritis Index (WOMAC)variables. function score] and 23 association between HRQoL andphysical component function score [Short (PCS) and Form-12 Western Ontario (SF-12) and McMaster The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National InstitutesE. of H. Health Schemitsch reports personal fees from Acumed, LLC, personal fees from Amgen Co, research support from Biocomposites, board or committee member Reprints: Daniel Axelrod, MD, MSc (Cand), Division ofCopyright Orthopaedic Surgery, © Department 2020 of WoltersDOI: Surgery, Kluwer McMaster 10.1097/BOT.0000000000001933 Health, University, Inc. 293 Wellington All St. rights N, reserved. Suite

INTRODUCTION and physical function in patients with hip or knee osteoarthri- Arthroplasty is the gold standard for displaced femoral tis and has been validated in patients who sustained a femoral neck fractures in the elderly.1–4 It is justified by predictable neck fracture.17 Hip function was assessed using the physical good outcomes, satisfactory joint function, early full weight- function subcomponent score from the WOMAC question- bearing, and rapid recovery,5,6 which are believed to be the naire, which ranges from 0 to 68 with higher scores indicating foundations for successful rehabilitation in the elderly. more functional limitations. The minimally important differ- Although a partial joint replacement [hemiarthroplasty ence (MID) of 4 for the SF-12 and 7 for the WOMAC was (HA)] has historically been the treatment of choice, there determined a priori in the HEALTH trial.14,15 All question- has been an increase in use of total joint replacements [total naires were administered by research personnel at baseline hip arthroplasty (THA)] for treatment of femoral neck frac- (recollection of prefracture status), 1 week, 10 weeks, and tures.7 Nevertheless, there is a lack of consensus among 6, 9, 12, 18, and 24 months after surgery. orthopaedic surgeons in regards to technical considerations and whether a partial joint replacement or a total joint replace- Selection of Baseline Factors ment is the most appropriate treatment for patients with dis- We selected baseline factors a priori based on biologic placed femoral neck fractures.8–13 rationale and previous reports in the literature. For each The most recent evidence suggests shorter operation potential factor, we proposed a priori a hypothesized effect for time10–12 and lower dislocation rate in favor of HA for the each dependent variable (ie, SF-12 PCS and WOMAC first 4 postoperative years,8,10–13 at the expense of late ace- function score). To avoid an overfitted or unstable model, tabular erosion.10,12 Both implants demonstrate similar length we used a rule of thumb that there should be at least 10 times of hospitalization,9,10,12 infection rate,8,10,11 patient survivor- the number of observations as there are factors in a regression ship,8,10–12 and overall complication rate.8–11 HA is associ- model. Given that there were more than 900 participants with ated with a lower revision rate in the 5 years after surgery,12 multiple observations included in each model, we were not at but is expected to surpass the revision rate of THA beyond 5 high risk of overfitting. We classified all baseline factors into years.10,12 Although it is unclear whether THA brings any 1 of 3 groups (participant characteristics, perioperative additional functional benefits,8–13 some argue that the lower characteristics, and postoperative characteristics). Participant reoperation rates with THA in patients expected to live more characteristics included age, sex, body mass index (BMI), than 5 years will lead to meaningful improvements in the American Society of Anesthesiologists (ASA) classification, patient’s health-related quality of life (HRQoL) and function. prefracture living setting, prefracture functional status, and Using data prospectively collected as part of the Hip the following comorbidities requiring treatment: diabetes, Fracture Evaluation with Alternatives of Total Hip respiratory disease, rheumatoid arthritis, cardiac disease, Arthroplasty versus Hemiarthroplasty (HEALTH) trial data depression, renal disease, hematologic disease (incl. anemia), (ClinicalTrials.gov NCT00556842),14,15 we aimed to investi- and cancer. Perioperative characteristics included type of gate the effect of THA, monopolar HA, and bipolar HA, anesthetic, surgical approach, implant type, time from injury along with other demographic and perioperative factors, on to surgery, and use of cement in either acetabulum or femur. the patients’ HRQoL and functional outcomes. Although sig- Postoperative characteristics initially included postoperative nificant research has been performed to investigate the differ- weight-bearing status, postoperative living status, and post- ence between HA and THA, we believe that factors other than operative ambulatory status. implant choice play a larger role in predicting HRQoL and physical function in patients aged 50 years and older presenting with a low-energy, isolated, displaced, femoral neck Definition of Baseline Factors fracture. Participant Characteristics We analyzed age and BMI as continuous variables. We METHODS analyzed all other baseline participant characteristic variables categorically, ie, sex (male vs. female), ASA classification (ie, Health-Related Quality of Life class I/II vs. class III/IV/V), prefracture living setting (ie, The HEALTH trial prospectively collected HRQoL and institutionalized vs. not institutionalized), prefracture func- hip function assessment as secondary outcomes for partici- tional status (using ambulatory aid vs. independent ambula- pants. HRQoL was measured using the Short Form-12 Health tor), and all medical comorbidities as either requiring Survey, from which the physical component summary scores treatment or not (those without disease also were counted (SF-12 PCS) were obtained. The SF-12 measures self- as not requiring treatment). reported HRQoL through an eight-domain profile of functional health and well-being and physical and mental health Perioperative Characteristics summary measures.16 Each domain was scored separately We analyzed most perioperative characteristic variables from 0 (lowest level of health) to 100 (highest level of health) categorically [eg, type of anesthetic (ie, general anesthetic vs. using standardized scoring methods to calculate a norm-based spinal/other), surgical approach, treatment group, and use of physical component score (PCS). The Western Ontario and cement]. Surgical approach was analyzed as direct anterior McMaster Universities Osteoarthritis Index (WOMAC) ques- approach (DAA), anterolateral/lateral, or posterolateral/ tionnaire is a 24-item instrument used to assess pain, stiffness, posterior. Treatment group was monopolar HA, bipolar HA,

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Femoral Neck Fractures Managed With Arthroplasty or THA. Time from injury to surgery was analyzed as a treated with an HA and 453 with a THA. As shown in continuous variable. Table 1, the mean age of included participants was 78.2 years (SD = 10.14) with the majority being women (73.5%, n = Postoperative Characteristics 681). Most patients were independently mobile (78.5%, n = We analyzed all postoperative characteristics as cate- 727) and living independently (97.4%, n = 903/927). An gorical variables (eg, postoperative weight-bearing status). approximately equivalent number of relatively healthy Postoperative weight-bearing status was analyzed as full patients (ASA classes I and II) and patients with 1 or more weight-bearing, partial weight-bearing, or non weight- moderate to severe diseases (ASA classes III, IV, and V) were bearing. included in the analysis. Instructions of partial weight-bearing or non–weight-bearing were given to 43.5% of the partici- Statistical Analysis pants (n = 403). The median time from injury to surgery was Our statistical analysis plan was determined a priori. 32.2 hours (range: 4.4–1268 hours). The incidence of We included HEALTH trial participants with complete data depression before injury in this population was 9.9% (n = 89). for all factors and respective baseline HRQoL and hip The majority of orthopaedic surgeons used the anterolateral/ function measures in each model. We used descriptive lateral approach (60%, n = 555), followed by posterior/ statistics to summarize all factors (frequencies and percent- posterolateral approach (37.5%, n = 347) and finally the ages for categorical variables and mean values, medians, and DAA (2.5%, n = 23). In this secondary analysis, 73 patients ranges for continuous variables). Before performing the (8%) underwent a second operation and 281 (30.4%) suffered multivariable analysis, we evaluated each pairwise associa- serious adverse events. tion or correlation between the independent variables. We decided a priori that if any variables were highly correlated Factors Associated with Postfracture SF-12 (ie, 0.7 or higher), only 1 variable would be included in the PCS model. A total of 756 participants met the inclusion criteria for We conducted 2 repeated measures models with the SF-12 PCS model. After adjusting for baseline SF-12 PCS participant variable as the random effect. Analyses were scores, the use of THA was associated with increased self- conducted to account for clustering within participants across perceived global physical health as measured by the SF-12 multiple time points. We used SF-12 PCS and WOMAC PCS (AMD 1.88 compared with those who received monop- function scores as the dependent variables (1 for each model). olar HA, 95% CI: 0.25–3.51, P = 0.024) (Table 2). There was fi We included all factors speci ed above as independent no measured difference in SF-12 PCS scores between those fi variables in xed effects, as well as time of HRQoL or who received THA or bipolar HA (P = 0.07). The following function assessment (6, 12, and 24 months after surgery). factors were associated with significantly lower mean post- Moreover, we used an autoregressive correlation structure to fracture SF-12 PCS (indicating worse HRQoL): ASA class inform the model that each patient observation was expected III/IV/V (AMD 22.64 compared with class I/II, 95% CI: to be correlated with their previous observation. 23.89 to 21.38, P , 0.001), preoperative use of an aid We decided a priori that the SF-12 PCS and WOMAC (AMD 22.66 compared with those who did not use an aid, function would be parallel primary outcomes. We anticipated 95% CI: 24.26 to 21.06, P , 0.001), and restricted weight- that the results would be similar across the SF-12 PCS and bearing status postoperatively (AMD 21.38 compared with WOMAC analyses as they measure similar attributes. those who were fully weight-bearing, 95% CI: 22.71 to Overall, we considered factors that were associated with 20.04, P = 0.04). None of the statistically significant outcome in both models as being more plausible (and those AMDs reached the MID for the SF-12 PCS, and no other which were inconsistently associated to be less plausible). factors were found to be significantly associated with post- Results were presented as adjusted mean differences (AMDs), fracture SF-12 PCS (age, sex, BMI, living status preopera- fi 95% con dence intervals (CIs), and P values. All tests were tively, diabetic disease, respiratory disease, rheumatoid 2-tailed with alpha = 0.05. A P value less than 0.05 was disease, cardiac disease, renal disease, hematologic disease, fi considered statistically signi cant. Generalized linear models depression, and cancer). Consequently, although a statisti- were created with Gaussian distributions. In addition, a gen- cally significant difference was observed among some of eralized estimated equation model was undertaken as a sen- the hypothesized predictors of HRQoL, none could be con- fi sitivity analysis. Goodness of t of the model was evaluated sidered to produce clinically important differences to patient through assessment of deviance. We used R (v3.6.1 open care. access online) for statistical analyses. Factors Associated With Postfracture RESULTS WOMAC Functional Scores A total of 707 participants met the inclusion criteria for Baseline Characteristics the WOMAC function model. After adjusting for baseline From the 1441 patients randomized to THA (n = 718) WOMAC physical function scores, the use of THA reduced or HA (n = 723) included in the final analysis as part of the postoperative WOMAC function scores (representing better HEALTH trial, 927 patients met the inclusion criteria for at function), but not to a clinically important difference (AMD least 1 model in the current analysis from which 471 were 22.40 compared with those who received monopolar HA,

TABLE 1. TABLE 1. Characteristics of all Patients Included in the HRQoL (Continued) Characteristics of all Patients Included in Analysis (n = 927) the HRQoL Analysis (n = 927) Incidence of Factors Incidence of Factors Participant Baseline Characteristics Cement for the femur, n (%) Mean age, yrs (SD) 78.25 (0.14) Yes 576 (62.2) Sex, n (%) No 350 (37.8) Male 246 (26.5) Postop weight-bearing status, n (%) Female 681 (73.5) Full weight-bearing 524 (56.5) Mean BMI, kg/m2 (SD) 25.31 (4.7) – fi Partial weight-bearing/non 403 (43.5) ASA classi cation, n (%) weight-bearing Class I/II 461 (49.8) Postoperative characteristics Class III/IV/V 466 (50.2) Reoperation, n (%) Prefracture functional status, n (%) No 854 (92.0) Any ambulatory aid 200 (21.5) Yes 73 (8.0) Independent ambulation 727 (78.5) Serious adverse events, n (%) Prefracture living, n (%) No 646 (69.6) Not institutionalized 903 (97.4) Yes 281 (30.4) Institutionalized 24 (2.6) Readmission (any cause), n (%) Comorbidities, diabetes, n (%) No 240 (88.6) Yes 129 (14.8) Yes 31 (11.4) No 740 (85.2) Respiratory disease Missing data were found in all categories except sex, ASA classification, prefracture functional status, prefracture living situation, postop weight-bearing status, reoperation, Yes 94 (10.7) serious adverse event, and readmission (any cause). No 778 (89.3) ASA, American Society of Anesthesiologists; BMI, body mass index; DAA, direct Rheumatoid arthritis anterior approach; GA, general anesthesia; HA, hemiarthroplasty; SD, standard deviation; THA, total hip arthroplasty. Yes 20 (2.1) No 902 (97.9) Cardiac disease Yes 207 (25.0) 95% CI: 24.49 to 20.31, P = 0.024) (Table 3). By contrast, No 621 (75.0) use of THA did not improve WOMAC functional scores over Kidney disease bipolar HA (P = 0.273). The following factors were associ- Yes 33 (3.8) ated with significantly higher mean postfracture WOMAC No 845 (96.2) functional scores (indicating worse function): ASA class III/ Hematologic disease IV/V (AMD 1.99 compared with class I/II, 95% CI: 0.41– Yes 33 (3.7) 3.58, P = 0.014), preoperative use of an aid (AMD 5.39 No 861 (96.3) compared with those who did not use an aid, 95% CI: Depression 3.29–7.49, P , 0.001), and receiving treatment for depres- Yes 89 (9.9) sion (AMD 7.73 compared with those not diagnosed or not No 815 (90.1) receiving treatment, 95% CI: 2.12–13.34, P = 0.007). No Perioperative characteristics other factors were found to be significantly associated with Type of anesthetic, n (%) postfracture WOMAC function scores (age, sex, BMI, living GA 367 (40.0) status preoperatively, diabetic disease, respiratory disease, Spinal/other 555 (60.0) rheumatoid disease, cardiac disease, renal disease, hemato- Surgical approach, n (%) logic disease, and cancer). Of the statistically significant DAA 23 (2.5) AMDs, only patients receiving treatment for depression Anterolateral/lateral 555 (60.0) reached the MID for the WOMAC function scores. Posterior/posterolateral 347 (37.5) Treatment group, n (%) DISCUSSION Monopolar HA 211 (22.8) Bipolar HA 260 (28.1) The HEALTH trial is one of the largest multicenter THA 453 (49.1) randomized controlled trials comparing the use of THA Time from injury to surgery, hours 57.4 (4.35 to 1268) versus HA in the treatment of isolated, displaced femoral (range) neck fractures in patients aged 50 years and older.14,15 This Cement for the acetabulum, n (%) secondary analysis reports prospectively collected data inves- Yes 120 (26.5) tigating predictors of HRQoL and functional outcomes in this No 333 (73.5) population when treated with arthroplasty. Prognostic factors of functional independence have been only scarcely

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Femoral Neck Fractures Managed With Arthroplasty determined after arthroplasty treatment of femoral neck frac- report a decline in function and global physical health com- tures. The use of THA compared with monopolar HA, but not pared with preinjury independent ambulators. As defined by bipolar HA, was associated with a statistically significant the MID, the deterioration in function over the postoperative improvement over time in SF-12 PCS and WOMAC physical period of 24 months was not sufficient to affect patients in a function component scores. Functional outcomes were nega- clinically meaningful way. Undeniably, patients with limited tively influenced by severe systemic disease (ASA classifica- mobility before injury are likely to experience less functional tion III, IV, and V), preoperative use of walking aids, improvement from their baseline limitation over the course of depression, and the surgeons’ postoperative partial weight- rehabilitation as compared to fully ambulatory patients. bearing orders. Importantly, none of these associations ex- Age may give a general indication of expected recovery ceeded the MID, except for presence of depression preoper- after a hip fracture,36,41,43,45 but only to a certain extent.39,40 atively. In other words, only the presence of depression had Similarly, BMI,43,46 sex,39,40,45 and time from injury to sur- enough influence to lead to an observable and important dete- gery36,43 have limited predictive value in regard to medium- rioration in function that would be considered clinically and long-term functional outcomes. Our study did not reveal meaningful from the patient’s perspective. an association with functional outcomes and any of these Presence of depressive symptoms preoperatively had a characteristics after treatment with arthroplasty. We believe strong influence on the WOMAC physical function outcome this population of older patients with hip fractures to be score, although it did not influence the SF-12 PCS. The highly complex. Functional outcomes are most likely the WOMAC physical function score does not specifically assess results of a combination of the individual’s characteristics, mental health; yet, it is inherently influenced by the patient’s but also of other social health determinants, such as environ- psychological status,16–18 whereas the SF-12 PCS correlates ment, education, community and more, which are not consid- poorly with mental health status.16,19 The WOMAC also has ered here. greater power than the short-form questionnaire to detect Here, we report a statistically significant, but clinically minimal changes in the context of hip or knee joint patholo- nonmeaningful, increase in global physical health and gies.20 Depression strongly affects one’s perception of func- physical function after THA for a displaced femoral neck tional status,21 and patients with depressive symptoms often fracture, as compared to monopolar HA, but not bipolar HA, describe themselves as more impaired than what is reflected at least for the first 2 years postoperatively. In general, THA from objective performance measures.22 However, objec- may be considered to be equivalent to HA from the patient’s tively, they are less likely to engage in rehabilitation,23,24 perspective in this patient population, but this is a highly have poorer functional independence, and demonstrate slower controversial topic, and several RCTs in the past 20 years recovery.25–28 Furthermore, the incidence of depression in have been conducted to determine the influence of implant this study was found to be twice the prevalence found in choice on functional outcomes.15,37,38,47–60 The functional the general population.29 These patients may benefit from improvement seen with THA in these studies is debatable, early psychological intervention to achieve potential func- and careful examination reveals a difference in mean function tional gains. Some interventions have been proven to be of less than 5 points between HA and THA at 2–3 years effective such as comprehensive and interdisciplinary care postoperatively. Ultimately, our results align with previous efforts for hip fracture patients which have been associated research and suggest that THA will offer a small yet unim- with a lesser risk of depression,30–33 while 2 randomized portant benefit to patients receiving treatment for displaced controlled trials have shown that simple psychological coun- femoral neck fractures, when compared with either bipolar or seling significantly improved recovery and physical function monopolar HA. at the 1-month and 6-month follow-ups.34,35 Other surgical parameters included the surgical The presence of each individually considered comor- approach and the use of cement. Both factors were not bidity did not influence functional outcomes. However, associated with any effect on functional outcomes in our patients with ASA classification III to V, which refers to study. Similar to the literature for primary THA, benefits severity rather than presence or absence of comorbidities, observed with some surgical approaches, if any, disappear by were more likely to report worsening physical function and 6 months postoperatively.61 Barenius et al57 reported global physical health over time after a hip fracture. Reuling improved functional outcomes with cemented femoral com- et al36 found that the presence of comorbidities negatively ponents, regardless of the type of arthroplasty. This study affected functional outcomes, but their results did not show randomized THA and HA as the surgical treatment for fem- an association with ASA classification. Previous reports have oral neck fracture, regardless of use of cement, which may identified high severity ASA to be an independent risk of explain the absence of any observed effect on HRQoL mea- mortality,37,38 although its effect on HRQoL and function surements. Surgeons were free to decide which patients after arthroplasty has not been consistent.36,39–41 It is likely would require additional fixation with cement, and they likely a complex interplay between disease severity and the additive chose whom would benefit most from it. effect of multimorbidity that adversely influences functional Early, unrestricted weight-bearing is the gold standard outcomes. after joint replacement for a hip fracture.62 It is simple and Preinjury functional level,41 cognitive function,40–42 safe,63 promotes better and earlier recovery,42,43,64 and seems and preinjury ambulatory status43,44 are the most established to be one of the only factors under the control of the ortho- predictors of functional recovery. In this study, participants paedic surgeon. Delayed weight-bearing leads to a higher with prefracture mobility impairment were more likely to incidence of any complications,65 is associated with poor

TABLE 2. Determinants of Global Physical Health Using TABLE 3. Determinants of Physical Function Using Repeated Repeated Measures Multilevel Model Regression With the SF- Measures Multilevel Model Regression With the WOMAC 12 PCS as the Dependent Variable for 756 Patients, Adjusted Physical Function as the Dependent Variable for 707 Patients, for Baseline SF-12 PCS Adjusted for Baseline WOMAC Physical Function Score Independent Variable AMD (99% CI)* P Independent Variable AMD (99% CI)* P Baseline characteristics Baseline characteristics Age (10-y increments) 20.62 (21.43 to 0.18) 0.128 Age (10-y increments) 20.41 (21.44 to 0.62) 0.432 Female 21.33 (22.70 to 0.03) 0.56 Female 0.65 (21.11 to 2.42) 0.467 BMI (5-point increments) 20.14 (20.76 to 0.47) 0.643 BMI (5-point increments) 0.19 (20.60 to 0.98) 0.644 ASA classiﬁcation III–V (ref. 22.64 (23.89 to 21.38) ,0.001 ASA class III–V (ref. class I and 1.99 (0.41 to 3.58) 0.014 class I and II) II) Dependent ambulator 22.66 (24.26 to 21.06) ,0.001 Dependent ambulator 5.39 (3.29 to 7.49) ,0.001 Institutionalized 22.16 (26.22 to 1.91) 0.299 Institutionalized 0.54 (26.05 to 4.97) 0.848 Comorbidities Comorbidities Diabetic 1.98 (22.14 to 6.10) 0.346 Diabetic 1.02 (24.19 to 6.23) 0.701 Respiratory disease 0.52 (23.48 to 4.53) 0.797 Respiratory disease 20.25 (25.48 to 4.98) 0.924 Rheumatoid disease 22.37 (213.90 to 9.07) 0.685 Rheumatoid disease 5.63 (28.56 to 19.81) 0.436 Cardiac disease 23.49 (28.19 to 1.20) 0.144 Cardiac disease 4.74 (21.38 to 10.87) 0.129 Renal disease 22.22 (25.18 to 0.73) 0.140 Renal disease 3.39 (20.41 to 7.19) 0.08 Hematologic disease 20.59 (24.30 to 3.12) 0.755 Hematologic disease 3.88 (21.06 to 8.81) 0.123 Depression 22.38 (26.56 to 1.79) 0.262 Depression 7.73 (2.12 to 13.34) 0.007 Cancer 0.97 (21.56 to 3.50) 0.452 Cancer 22.01 (25.32 to 1.29) 0.232 Perioperative characteristics Perioperative characteristics Surgical approach (reference: Surgical approach (reference: DAA) DAA) Anterolateral/lateral 23.17 (27.37 to 1.02) 0.138 Anterolateral/lateral 3.26 (1.82 to 8.34) 0.208 Posterior 22.17 (26.38 to 2.03) 0.311 Posterior 4.16 (20.93 to 9.26) 0.109 Implant (reference: THA) Implant (reference: THA) Bipolar HA 21.25 (22.6 to 0.11) 0.07 Bipolar HA 1.12 (20.63 to 2.87) 0.209 Monopolar HA 21.88 (0.25 to 3.51) 0.024 Monopolar HA 2.40 (24.49 to 20.31) 0.024 Time from injury to surgery (hours) 20.00 (20.01 to 0.00) 0.204 Time from injury to surgery (hours) 0.00 (20.01 to 0.01) 0.673 Cemented femoral component 20.43 (21.74 to 0.87) 0.518 Cemented femoral component 20.04 (21.71 to 1.64) 0.966 Postoperative characteristics Postoperative characteristics Partial weight-bearing (ref. 21.38 (22.71 to 20.04) 0.043 Full weight-bearing post- 20.37 (22.08 to 1.34) 0.672 weight-bearing as tolerated) operatively

*MID was set at 4 points for the 12-item Short-Form Health Survey PCS (SF-12 *MID was set at 7 points for the Western Ontario and McMaster Universities PCS). Osteoarthritis Index (WOMAC). Significance = p values , 0.05. Significance = p values , 0.05. AMD, adjusted mean difference; ASA, American Society of Anesthesiologists; AMD, adjusted mean difference; ASA, American Society of Anesthesiologists; BMI, body mass index; CI, confidence interval; DAA, direct anterior approach; HA, BMI, body mass index; CI, confidence interval; DAA, direct anterior approach; HA, hemiarthroplasty; SD, standard deviation; THA, total hip arthroplasty. hemiarthroplasty; SD, standard deviation; THA, total hip arthroplasty. compliance,66,67 and was associated with statistically signifi- unable to predict outcomes beyond 2 years and are, therefore, cant worse global physical health in our study, although it did unable to determine whether the rate of acetabular erosion or not reach the MID. Although it is unknown why close to 45% conversion of HA into THA is significant. Six RCTs reported of participants in this study were given instructions for partial long-term follow-up and found no difference in functional weight-bearing or non–weight-bearing, we recommend outcomes between HA and THA at 3 years,53,55 4 years,50 5 reconsideration of any instructions around restricted weight- years,49 8 years,68 and 12 years.59 Two reported worse func- bearing in this patient population. tional outcomes with HA at 3 years69 or 5 years,58 but again, The influence of complications, readmissions, reopera- they reported a difference in mean score of less than 5 points tions, or dislocations on functional outcomes could not be between the 2 groups. Furthermore, despite best possible assessed. It is likely that any reoperation or complication will practices, missed follow-ups may have weakened the strength have an impact on functional outcomes. In the HEALTH trial, of correlation between time points and may explain some of the rate of overall reoperations was 7.9% in 718 THA patients the nonstatistically significant findings. and 8.3% in 723 HA patients and was not significantly Elderly patients who suffer a displaced femoral neck different. Consequently, although the rate of reoperation may fracture can be successfully treated with either HA or THA. negatively influence functional outcomes, the effect on each Although we agree that the surgeon may use his clinical group may be similar. From the results of this study, we are judgment in the choice of implant, patients can expect

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Femoral Neck Fractures Managed With Arthroplasty similar functional outcomes at 2 years with either treatment 14. Bhandari M, Devereaux PJ, Einhorn TA, et al. Hip fracture evaluation strategy. A hip fracture is a life-changing event, and the with alternatives of total hip arthroplasty versus hemiarthroplasty fi (HEALTH): protocol for a multicentre randomised trial. BMJ Open. absence of de nitive and strong predictors affecting func- 2015;5:e006263. tional outcomes in this study indicates that health determi- 15. Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or nants other than surgical parameters are of greater hemiarthroplasty for hip fracture. N Engl J Med. 2019;381:2199–2208. consequence. The critical event here is the hip fracture itself 16. Ostendorf M, van Stel HF, Buskens E, et al. Patient-reported outcome in total hip replacement. A comparison of five instruments of health status. J and all the circumstances leading to the injury, not the – fi Bone Joint Surg Br. 2004;86:801 808. surgery. Further re nement of surgical techniques may not 17. Burgers PT, Poolman RW, Van Bakel TM, et al. Reliability, validity, and lead to improvement in patients’ outcomes without address- responsiveness of the Western Ontario and McMaster Universities ing the highly complex issue of hip fractures with interdis- Osteoarthritis Index for elderly patients with a femoral neck fracture. J ciplinary and comprehensive care teams focusing on Bone Joint Surg Am. 2015;97:751–757. 18. Hirschmann MT, Testa E, Amsler F, et al. The unhappy total knee recovery and global care. arthroplasty (TKA) patient: higher WOMAC and lower KSS in depressed patients prior and after TKA. Knee Surg Sports Traumatol Arthrosc. 2013;21:2405–2411. ACKNOWLEDGMENTS 19. Clement ND, MacDonald D, Burnett R. Primary total knee replacement in patients with mental disability improves their mental health and knee The authors thank the HEALTH Investigators (http:// function: a prospective study. Bone Joint J. 2013;95-b:360–366. links.lww.com/JOT/B248). 20. Davies GM, Watson DJ, Bellamy N. Comparison of the responsiveness and relative effect size of the Western Ontario and McMaster Universities Osteoarthritis Index and the short-form Medical Outcomes Study Survey REFERENCES in a randomized, clinical trial of osteoarthritis patients. Arthritis Care 1. Richards JT, Overmann AL, O’Hara NN, et al. Internal fixation versus Res. 1999;12:172–179. arthroplasty for the treatment of nondisplaced femoral neck fractures in 21. Grigsby J, Kaye K, Kowalsky J, et al. Relationship between functional the elderly: a systematic review and meta-analysis. J Orthop Trauma. status and the capacity to regulate behavior among elderly persons fol- 2020;34:42–48. lowing hip fracture. Rehabil Psychol. 2002;47:291. 2. Parker MJ, Gurusamy K. Internal fixation versus arthroplasty for intra- 22. Kempen GIJM, Steverink N, Ormel J, et al. The assessment of ADL capsular proximal femoral fractures in adults. Cochrane Database Syst among frail elderly in an interview survey: self-report versus Rev. 2006:Cd001708. performance-based tests and determinants of discrepancies. J Gerontol 3. Frihagen F, Nordsletten L, Madsen JE. Hemiarthroplasty or internal B Psychol Sci Soc Sci. 1996;51:P254–P260. fixation for intracapsular displaced femoral neck fractures: randomised 23. Lenze EJ, Munin MC, Dew MA, et al. Adverse effects of depression and controlled trial. BMJ. 2007;335:1251–1254. cognitive impairment on rehabilitation participation and recovery from 4. Rogmark C, Carlsson A, Johnell O, et al. A prospective randomised trial hip fracture. Int J Geriatr Psychiatry. 2004;19:472–478. of internal fixation versus arthroplasty for displaced fractures of the neck 24. Resnick B, Orwig D, Hawkes W, et al. The relationship between psy- of the femur. Functional outcome for 450 patients at two years. J Bone chosocial state and exercise behavior of older women 2 months after hip Joint Surg Br. 2002;84:183–188. fracture. Rehabil Nurs. 2007;32:139–149. 5. Lu Q, Tang G, Zhao X, et al. Hemiarthroplasty versus internal fixation in 25. Morghen S, Bellelli G, Manuele S, et al. Moderate to severe depressive super-aged patients with undisplaced femoral neck fractures: a 5-year symptoms and rehabilitation outcome in older adults with hip fracture. follow-up of randomized controlled trial. Arch Orthop Trauma Surg. Int J Geriatr Psychiatry. 2011;26:1136–1143. 2017;137:27–35. 26. Buecking B, Struewer J, Waldermann A, et al. What determines health- 6. Dolatowski FC, Frihagen F, Bartels S, et al. Screw fixation versus hemi- related quality of life in hip fracture patients at the end of acute care?—a arthroplasty for nondisplaced femoral neck fractures in elderly patients: a prospective observational study. Osteoporos Int. 2014;25:475–484. multicenter randomized controlled trial. J Bone Joint Surg Am. 2019;101: 27. Feng L, Scherer SC, Tan BY, et al. Comorbid cognitive impairment and 136–144. depression is a significant predictor of poor outcomes in hip fracture 7. Boniello AJ, Lieber AM, Denehy K, et al. National trends in total hip rehabilitation. Int Psychogeriatr. 2010;22:246–253. arthroplasty for traumatic hip fractures: an analysis of a nationwide all- 28. Shyu YI, Chen MC, Cheng HS, et al. Severity of depression risk predicts payer database. World J Orthop. 2020;11:18–26. health outcomes and recovery following surgery for hip-fractured elders. 8. Lewis DP, Wæver D, Thorninger R, et al. Hemiarthroplasty vs total hip Osteoporos Int. 2008;19:1541–1547. arthroplasty for the management of displaced neck of femur fractures: a 29. Salk RH, Hyde JS, Abramson LY. Gender differences in depression in systematic review and meta-analysis. J Arthroplasty. 2019;34:1837– representative national samples: meta-analyses of diagnoses and symp- 1843. toms. Psychol Bull. 2017;143:783–822. 9. Metcalfe D, Judge A, Perry DC, et al. Total hip arthroplasty versus 30. Tseng MY, Shyu YL, Liang J, et al. Interdisciplinary intervention hemiarthroplasty for independently mobile older adults with intracapsular reduced the risk of being persistently depressive among older patients hip fractures. BMC Musculoskelet Disord. 2019;20:226. with hip fracture. Geriatr Gerontol Int. 2016;16:1145–1152. 10. Wang F, Zhang H, Zhang Z, et al. Comparison of bipolar hemiarthro- 31. Zhang M Effect of HBM rehabilitation exercises on depression, anxiety plasty and total hip arthroplasty for displaced femoral neck fractures in and health belief in elderly patients with osteoporotic fracture. Psychiatr the healthy elderly: a meta-analysis. BMC Musculoskelet Disord. 2015; Danub. 2017;29:466–472. 16:229. 32. Wang S, Qiu Y. Application of predictive nursing care in elderly patients 11. Xu D, Li X, Bi F, et al. Hemiarthroplasty compared with total hip with fractures that underwent total hip arthroplasty procedures. Int J Clin arthroplasty for displaced fractures of femoral neck in the elderly: a Exp Med. 2019;12:5888–5894. systematic review and meta-analysis of fourteen randomized clinical 33. Shyu YIL, Liang J, Tseng MY, et al. Comprehensive care improves trials. Int J Clin Exp Med. 2018;11:5430–5443. health outcomes among elderly Taiwanese patients with hip fracture. J 12. Migliorini F, Trivellas A, Driessen A, et al. Hemiarthroplasty versus total Gerontol A Biol Sci Med Sci. 2013;68:188–197. arthroplasty for displaced femoral neck fractures in the elderly: meta- 34. Gambatesa M, D’Ambrosio A, D’Antini D, et al. Counseling, quality of analysis of randomized clinical trials. Archives Orthop Trauma Surg. life, and acute postoperative pain in elderly patients with hip fracture. J 2020:1–10. Multidiscip Healthc. 2013;6:335–346. 13. Yoo JI, Cha YH, Kim JT, et al. Clinical outcomes of bipolar hemiar- 35. Allegrante JP, Peterson MGE, Cornell CN, et al. Methodological chal- throplasty versus total hip arthroplasty: assessing the potential impact of lenges of multiple-component intervention: lessons learned from a ran- cement use and pre-injury activity levels in elderly patients with femoral domized controlled trial of functional recovery after hip fracture. HSS J. neck fractures. Hip Pelvis. 2019;31:63–74. 2007;3:63–70.

36. Reuling EM, Sierevelt IN, van den Bekerom MP, et al. Predictors of 53. Baker RP, Squires B, Gargan MF, et al. Total hip arthroplasty and hemi- functional outcome following femoral neck fractures treated with an arthroplasty in mobile, independent patients with a displaced intracapsu- arthroplasty: limitations of the Harris hip score. Arch Orthop Trauma lar fracture of the femoral neck. A randomized, controlled trial. J Bone Surg. 2012;132:249–256. Joint Surg Am. 2006;88:2583–2589. 37. Macaulay W, Nellans KW, Garvin KL, et al. Prospective randomized 54. Ravikumar KJ, Marsh G. Internal fixation versus hemiarthroplasty versus clinical trial comparing hemiarthroplasty to total hip arthroplasty in the total hip arthroplasty for displaced subcapital fractures of femur—13 year treatment of displaced femoral neck fractures: winner of the Dorr Award. results of a prospective randomised study. Injury. 2000;31:793–797. J Arthroplasty. 2008;23:2–8. 55. Ukaj S, Zhuri O, Ukaj F, et al. Dual mobility acetabular cup versus 38. Barishan FC, Akesen B, Atici T, et al. Comparison of hemiarthroplasty hemiarthroplasty in treatment of displaced femoral neck fractures in and total hip arthroplasty in elderly patients with displaced femoral neck elderly patients: comparative study and results at minimum 3-year fol- – fractures. J Int Med Res. 2018;46:2717 2730. low-up. Geriatr Orthop Surg Rehabil. 2019;10:2151459319848610. 39. Enocson A, Pettersson H, Ponzer S, et al. Quality of life after dislocation 56. Sonaje JC, Meena PK, Bansiwal RC, et al. Comparison of functional of hip arthroplasty: a prospective cohort study on 319 patients with outcome of bipolar hip arthroplasty and total hip replacement in dis- femoral neck fractures with a one-year follow-up. Qual Life Res. 2009; placed femoral neck fractures in elderly in a developing country: a 2- 18:1177–1184. year prospective study. Eur J Orthop Surg Traumatol. 2018;28:493–498. 40. Mukka S, Knutsson B, Krupic F, et al. The influence of cognitive status 57. Barenius B, Inngul C, Alagic Z, et al. A randomized controlled trial of on outcome and walking ability after hemiarthroplasty for femoral neck fracture: a prospective cohort study. Eur J Orthop Surg Traumatol. 2017; cemented versus cementless arthroplasty in patients with a displaced 27:653–658. femoral neck fracture: a four-year follow-up. Bone Joint J. 2018;100- – 41. Sanz-Reig J, Lizaur-Utrilla A, Serna-Berna R. Outcomes in nonage- B:1087 1093. narians after hemiarthroplasty for femoral neck fracture. A prospective 58. Xu F, Ke R, Gu Y, et al. Bipolar hemiarthroplasty vs. total hip replace- – matched cohort study. Hip Int. 2012;22:113–118. ment in elderly. Int J Clin Exp Med. 2017;10:7911 7920. 42. Siebens HC, Sharkey P, Aronow HU, et al. Outcomes and weight- 59. Tol M, van den Bekerom MPJ, Sierevelt IN, et al. Hemiarthroplasty or bearing status during rehabilitation after arthroplasty for hip fractures. total hip arthroplasty for the treatment of a displaced intracapsular frac- PM R. 2012;4:548–555. ture in active elderly patients: 12-year follow-up of randomised trial. 43. Mariconda M, Costa GG, Cerbasi S, et al. Factors predicting mobility Bone Joint J. 2017;99-B:250–254. and the change in activities of daily living after hip fracture: a 1-year 60. Sharma V, Awasthi B, Kumar K, et al. Outcome analysis of hemiarthro- prospective cohort study. J Orthop Trauma. 2016;30:71–77. plasty vs. total hip replacement in displaced femoral neck fractures in the 44. Kim JL, Jung JS, Kim SJ. Prediction of ambulatory status after hip elderly. J Clin Diagn Res. 2016;10:RC11–RC13. fracture surgery in patients over 60 years old. Ann Rehabil Med. 2016; 61. van der Sijp MPL, van Delft D, Krijnen P, et al. Surgical approaches and 40:666–674. hemiarthroplasty outcomes for femoral neck fractures: a meta-analysis. J 45. Haentjens P, Autier P, Barette M, et al. Predictors of functional outcome Arthroplasty. 2018;33:1617–1627. following intracapsular hip fracture in elderly women: a one-year pro- 62. Hoskins W, Webb D, Bingham R, et al. Evidence based management of spective cohort study. Injury. 2005;36:842–850. intracapsular neck of femur fractures. Hip Int. 2017;27:415–424. 46. Reider L, Hawkes W, Hebel JR, et al. The association between body 63. Tian P, Li ZJ, Xu GJ, et al. Partial versus early full weight bearing after mass index, weight loss and physical function in the year following a hip uncemented total hip arthroplasty: a meta-analysis. J Orthop Surg Res. fracture. J Nutr Health Aging. 2013;17:91–95. 2017;12:31. 47. Healy WL, Iorio R. Total hip arthroplasty: optimal treatment for dis- 64. Oldmeadow LB, Edwards ER, Kimmel LA, et al. No rest for the placed femoral neck fractures in elderly patients. Clin Orthop Relat wounded: early ambulation after hip surgery accelerates recovery. ANZ – Res. 2004;429:43 48. J Surg. 2006;76:607–611. 48. Cadossi M, Chiarello E, Savarino L, et al. A comparison of hemiarthro- 65. Warren J, Sundaram K, Anis H, et al. The association between weight- – plasty with a novel polycarbonate urethane acetabular component for bearing status and early complications in hip fractures. Eur J Orthop displaced intracapsular fractures of the femoral neck: a randomised – – Surg Traumatol. 2019;29:1419 1427. controlled trial in elderly patients. Bone Joint J. 2013;95-B:609 615. 66. Schaefer A, Hotfiel T, Pauser J, et al. Incompliance of total hip arthro- 49. Van den Bekerom MPJ, Hilverdink EF, Sierevelt IN, et al. A comparison plasty (THA) patients to limited weight bearing. Arch Orthop Trauma of hemiarthroplasty with total hip replacement for displaced intracapsular Surg. 2015;135:265–269. fracture of the femoral neck: a randomised controlled multicentre trial in 67. Seo H, Lee GJ, Shon HC, et al. Factors affecting compliance with patients aged 70 years and over. J Bone Joint Surg Br. 2010;92:1422– weight-bearing restriction and the amount of weight-bearing in the 1428. – 50. Mouzopoulos G, Stamatakos M, Arabatzi H, et al. The four-year func- elderly with femur or pelvic fractures. Ann Rehabil Med. 2020;44:109 tional result after a displaced subcapital hip fracture treated with three 116. different surgical options. Int Orthop. 2008;32:367–373. 68. Avery PP, Baker RP, Walton MJ, et al. Total hip replacement and hemi- 51. Blomfeldt R, Törnkvist H, Eriksson K, et al. A randomised controlled arthroplasty in mobile, independent patients with a displaced intracapsu- trial comparing bipolar hemiarthroplasty with total hip replacement for lar fracture of the femoral neck: a seven-to ten-year follow-up report of a displaced intracapsular fractures of the femoral neck in elderly patients. J prospective randomised controlled trial. J Bone Joint Surg Br. 2011;93: Bone Joint Surg Br. 2007;89:160–165. 1045–1048. 52. Keating JF, Grant A, Masson M, et al. Randomized comparison of 69. Hedbeck CJ, Enocson A, Lapidus G, et al. Comparison of bipolar hemi- reduction and fixation, bipolar hemiarthroplasty, and total hip arthro- arthroplasty with total hip arthroplasty for displaced femoral neck frac- plasty: treatment of displaced intracapsular hip fractures in healthy older tures: a concise four-year follow-up of a randomized trial. J Bone Joint patients. J Bone Joint Surg Am. 2006;88:249–260. Surg Am. 2011;93:445–450.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. t, s, b , and rom a S37 (NIH) e 110, | ns of this a a a Bzovsky, MSc, fi s Healthcare Hamilton, Hamilton, ON, ’ S41) So – Department of Health Research Methods, rst quartile), THA was both more b a fi Seper Ekhtiari, MD, Economic Level II. See Instructions for d -Aventis, outside the submitted work. S. Sprague , fi c , b Department of Surgery, University of Western Ontario, 2020;34:S37 e Our results suggest that for most patients, THA is Mohit Bhandari, MD, PhD, FRCSC, economic evaluation, displaced femoral neck fracture, s Hamilton, St. Joseph ’ a -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees fi RTICLE A J Orthop Trauma ict of interest. $100,000 to gaineffective one QALY, was the 12.8%patients probability and of younger THA 32.8%, thaneffective being respectively. and cost- 73 less In costly. ( yielded a Otherwise, marginal THA subgroup HRQoL was gains. of more expensive and not a cost-effectivemanagement treatment versus for HA. displaced However,younger THA femoral may patients. neck beimprovements fracture cost These in effective for quality patientsof of shorter experience life hospital stay with more and less fewer associated postoperative meaningful Key complications. cost Words: because total hip arthroplasty, hemiarthroplasty Level of Evidence: Conclusions: Authors for a complete description of levels( of evidence. fl Center for Health Economics and Policy Analysis (CHEPA), McMaster University, and c cant fi Jean-Éric Tarride, PhD, on behalf of the HEALTH Investigators a b , UPPLEMENT a Herman Johal, MD, MPH, FRCSC, S s Health, employment from Global Research Solutions Inc, and employment from McMaster ’ c , b Diane Heels-Ansdell, MSc, e s Web site (www.jorthotrauma.com). and Sheila Sprague, PhD ’ Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com Trial-Based Analysis of the HEALTH Study Displaced femoral neck fractures are a signi Daniel Axelrod, MD, MSc (Cand), McMaster Chair in Health Technology Management Hamilton, Hamilton, ON, Canada; and Health care resource utilization (HRU) and health-related d When compared with HA, THA was not cost-effective for Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; a Is Total Hip Arthroplasty a Cost-Effective Option for Gordon Blackhouse, MBA, MSc, Management of Displaced Femoral Neck Fractures? A Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Canada; Programs for Assessment of Technology in Health (PATH), The Research Institute of St. Joe Evidence, and Impact, McMaster University, Hamilton, ON, Canada; Hamilton, ON L8L 8E7, Canada (e-mail: [email protected]). reports editorial orUniversity, governing outside the board submitted for work. The BMS remaining Women article authors on report the no journal con Trauma Care Foundation, personalfrom fees Smith from & Pentopharm, Nephew,submitted personal personal work. fees fees M. from from Bhandari Springer, Sano research reports personal support research and fees support personal from from fees Stryker, from Acumed, personal Pendopharma, LLC, fees and research from research support Swemac, support from and and Aphria, personal personal fees research fees from support from Sano from Zimmer, Ferring outside Pharmaceutical the Acumed, LLC, personal fees frompersonal Amgen fees Co, research from support DePuy, frompersonal board Biocomposites, fees or board or committee from committee memberAssociation, member ITS, for for editorial editorial the the or Canadian or Hip Orthopaedic governing Society, governing Association, board board board or for for committee the member the Orthopaedic for Journal Trauma the of Association International Orthopaedic International, Society for board Trauma, Fracture or board Repair, committee or member committee for member the for Osteosynthesis the Orthopaedic Trauma (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. Research, interpretation The of funding the sources data; had or no the role preparation, in design review, or or conduct approval of of the the study; manuscript. the collection, managemen London, ON, Canada. Emil H. Schemitsch, MD, FRCSC, all patients with displacedgained). femoral neck If fractures ($150,000/QALY decision makers were willing to spend $50,000 or J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Results: Methods: Background: Dimensions (EQ-5D) scores, we(QALYs). derived A quality adjusted 1.5%applied. life discount Age years analyses rate perreported to age in group both 2019 were Canadian costs conducted. dollars. and All costs QALYs are was quality of lifecosted (HRQoL) data using were publicly collected postoperatively available and databases. Using EuroQol-5 improved functional outcomes whencost compared to effectiveness HA.economic To of analysis evaluate of THA the HEALTH study compared was undertaken. with HA, a trial-based source of morbidityhemiarthroplasty and (HA) mortality or and totalof hip can THA arthroplasty be (THA). have treated Proponents with argued either THA offers lower risk of revision, with Reprints: Daniel Axelrod, MD, MSc (Cand.), Division of Orthopaedic Surgery, Department of Surgery, McMaster University, 293 Wellington St. N, Suit Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001932 Health, Inc. All rights reserved. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versio The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National InstitutesH. of Johal Health reports board or committee member for the Orthopaedic Research Society, outside the submitted work. E. H. Schemitsch reports personal fees f

INTRODUCTION not captured for physiotherapy, we assumed that patients Hip fractures are the most common injury suffered by would use 7 physiotherapy visits. the elderly and contribute substantial morbidity to these The unit costs of secondary procedures were derived patients, who often have multiple pre-existing comorbidities from a patient cost estimator from the Canadian Institute of and are frail.1 Moreover, the mortality associated with hip Health Information.8 The patient cost estimator is an initiative fractures approaches 25% for all patients.2 The burden of which is informed by costing data submitted from all prov- these injuries—in both human and economic terms—is sub- inces across Canada, with the exception of Quebec. In addi- stantial. Up to 10 percent of patients with these injuries will tion, costs can be specified by province and by patient age. require a secondary operation, thus multiplying the economic For the purpose of this analysis, all cost estimates were taken impact of these injuries.3 These revision surgeries range from from the perspective of the Ontario health care system. Unit 2 to 3 and half times more costly than the primary procedure.3 costs of implants were taken from costing inventory at For displaced femoral neck fractures, both partial joint Hamilton General Hospital in Hamilton, Ontario. replacement of the hip (hemiarthroplasty, HA) and total joint We assumed the hourly rate for provincially covered replacement [total hip arthroplasty (THA)], are considered physiotherapy was $120.9 Unit costs used in this economic standard of care in the elderly.4 Those who have argued for evaluation are presented in Supplemental Digital Content 1 HA suggest that it represents a less morbid, less expensive (see Appendix 1, http://links.lww.com/JOT/B211). All costs and noninferior treatment option to THA. Those who argue are stated in 2019 Canadian dollars. for THA believe it yields fewer reoperations, and affords better quality of life, similar to patients who receive THA Health-Related Quality of Life and QALYs for osteoarthritis.5 Although both the Short Form-12 (SF-12) and the EQ- A recent meta-analysis suggested that THA for femoral 5D-5L questionnaires were collected in the HEALTH study, neck fractures did meaningfully improve quality of life for the economic results were derived from the EQ-5D, as this patients while reducing unplanned reoperations.4 However, instrument most easily allows the calculation of a health the meta-analysis was dependent on small, underpowered utility score, necessary to calculate quality-adjusted life years clinical trials with non-standardized adjudication of out- (QALYs) and is validated for use in this population.10 comes. Moreover, issues of confounding due to expertise bias QALYs combine quantity of life with health-related quality were pervasive in all studies. Ultimately, even after multiple of life and are used in cost-utility analyses to compare out- studies and meta-analyses,3,4,6 there remains uncertainty comes between interventions. regarding the best treatment for displaced femoral neck frac- The EQ-5D measures health-related quality of life in 5 tures in the elderly. dimensions (mobility, self-care, usual activities, pain/ To address this issue, a multi-national, multi-center discomfort, and anxiety/depression) with 3 levels of scoring expertise-based randomized controlled trial termed the Hip per attribute. Using a scoring algorithm, the individual Fracture Evaluation With Alternatives of Total Hip responses to each of the domains are transformed into a Arthroplasty versus Hemi-Arthroplasty (HEALTH) trial7 health utility score where 0 represents death and 1 the best was performed. The HEALTH study did not find any clini- imaginable health state. The minimal important difference for cally meaningful differences in either rate of unplanned reop- the EQ-5D was set at 0.05.11 QALYs were calculated for each erations or health utility when comparing THA to HA. At the intervention by weighting the utility scores by time spent in same time that the main trial was undertaken, relevant health health states using an area under the curve approach. care utilization data was also collected to facilitate an economic evaluation to inform health care policy leaders and key Cost-Effectiveness Analyses stakeholders in this domain. In this study, we report on the Utilizing a trial-based cost-effectiveness approach fol- results of these economic findings through a trial based eco- lowing good practice guidelines,12,13 use of THA versus HA nomic analysis. was assessed in incremental cost per QALY gained. Overall costs and outcomes were first evaluated to determine if either strategy was dominant (both less costly and better quality of METHODS life). If neither strategy was found to be dominant, incremental cost-effectiveness ratios (ICERs) were estimated by calcu- Study Design and Participants lating the difference in cost between THA and HA divided by The study design and clinical results of the HEALTH the difference in their effect (ie, QALYs). Uncertainty regard- trial are described elsewhere.7 Health care resource utilization ing costs and effects due to sampling variability associated and health-related quality of life [EuroQol Five-Dimension with the trial was measured using nonparametric bootstrap- (EQ-5D)] information were collected during follow-up visits ping techniques.14 at 6, 12, 24, 38, and 52 weeks, and at 18 and 24 months. The bootstrapping technique entails drawing a random sample from the original dataset (with replacement) and then Health Care Resource Utilization calculating the mean costs and effects associated with each A case report form was completed for each follow-up treatment group (ie, THA and HA). The sampling process visit to collect information on health-related quality of life, was repeated 2000 times to create a sampling distribution and secondary procedures, serious adverse events, physiotherapy generate means, SEs, and 95% confidence intervals (CIs) for visits and hospital readmissions. As the frequency of care was the incremental costs, incremental QALYs, and ICERs. Cost-

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Cost-Effective Option for Management of Displaced

TABLE 1. Average Cost per Patient (Can $ 2019) THA HA Difference (N = 718) (N = 723) Mean (SE) Mean (SE) Mean (95% CI) P Device $1789 $925 Initial procedure cost $16,756 ($1220) $15,732 ($840) $1024 (2$1880 to $3928) 0.10 Physiotherapy $592 ($10.3) $596 ($10.1) $4 (2$23 to $33) 0.74 Costs from adverse events $9469 ($1100) $7933 ($919) $1536 (2$1274 to $4346) 0.28 Costs from second procedures $4231 ($1083) $2541 ($526) $1690 (2$666 to $4047) 0.16 Payer perspective $32,851 ($2368) $27,358 ($1513) $5493 (2$14 to $11,000) 0.06

This represents the average overall costs attributed to each patient by each cost category, though most patients did not suffer from adverse events or undergo second procedures. effectiveness acceptability curves (CEACs) were used to with displaced femoral neck fractures was calculated at present the probability of THA to be cost effective at different $151,640/QALY gained (Fig. 1). Table 2 presents these willingness-to-pay thresholds (ie, $50,000/QALY gained or results. $100,000/QALY gained—2 commonly cited cost- Our sensitivity analyses produced similar results. When effectiveness thresholds) while taking into account the uncer- either a 0% or 3% discount rate was applied to both costs and tainty associated with the trial data.15 Analyses were con- QALYs at one year, the ICERs did not meaningfully change. ducted from a public payer and societal perspective over a Finally, when comparing imputed to not imputed EQ-5D 2-year time horizon. Both costs and QALYs collected after values, EQ-5D scores each decreased by 0.01 points for both one year were discounted at 1.5% per year, as per Canadian THA and HA, yielding no change in the difference in EQ-5D Agency for Drug and Technologies in Health (CADTH) scores. In addition, neither decreasing nor increasing the cost guidelines, to indicate that either a single QALY or amount of THA by 30% yielded meaningful differences in ICERs, of cost was more valuable now than in one year’s time.16 because implant costs only represented 10% of total costs per patient. Sensitivity and Subgroup Analyses Finally, a subgroup analysis by age quartile was ﬁ Multiple sensitivity analyses were planned. First, we performed. For patients younger than 73 ( rst quartile), planned to assess the impact of change in discount rate, THA was both less expensive (-$3693) and provided more varying from 0% to 1.5% (base case) to 3%. In addition, we health utility (0.10 QALYs). In contrast, for all other age performed a sensitivity analysis evaluating the difference in quartiles, HA provided similar utility at a reduced cost ICERs using a multiple imputation model (base case) and list- (Table 3). case deletion. Moreover, a sensitivity analysis was planned to The CEACs curves presented in Fig. 2 indicate that modify the cost of the technology (THA) by 30% in either when compared with HA, the probability of THA being direction, as recommended by CADTH.16 Finally, we per- cost-effective for treating operatively managed displaced fem- formed a sensitivity analysis with changing mortality rates oral neck fractures was 12.8% and 32.8%, if decision makers per group if mortality rates were found to be meaningfully different between THA and HA. We also performed a post- hoc subgroup analysis by age quartile, hypothesizing that THA would be most cost effective for younger quartiles. This analysis mirrors a similar age stratiﬁed analysis performed for the main HEALTH trial.

RESULTS Similar to the results of the clinical trial, there were no clinically meaningful differences between the 2 treatment groups in unplanned reoperation or health utility (Table 1). The costs from the single payer perspective were estimated to be $32,851 per patient (SE = $2368) for THA and $27,358 (SE = $2781) for HA. The total number of QALYs estimated using the EQ-5D score over the two-year time horizon was 1.40 (SE = 0.02) for THA and 1.36 (SE = 0.019) for HA. The difference of 0.04 QALYs (95% CI: 20.01 to 0.11) was not signiﬁcantly different. The resulting FIGURE 1. Cost-effectiveness plane. Editor’s Note: A color ICER of THA compared with HA for the treatment of patients image accompanies the online version of this article.

TABLE 2. Cost Effectiveness Analysis

Cost Mean (95% CI) QALYs Mean (95% CI) ICER D $/QALY THA HA Cost THA HA D QALYs Gained All patients $32,851 ($5275–$194,033) $27,358 ($5211–$144,953) $5493 1.40 (0–1.97) 1.36 (0–1.97) 0.04 (20.01 to 0.09) $151,640 (N = 1441)

Note that bootstrap sampling data is different than raw trial data and produces marginally different estimates to account for uncertainty. were willing to spend $50,000 or $100,000 to gain one recent meta-analysis including the HEALTH trial, the QALY, respectively. HEALTH trial contributed more than 50% of all patients to ever be evaluated in a randomized controlled trial comparing THA to HA.5 The dominance of this trial, in both sample size DISCUSSION and methodologic rigor, necessitates that it be evaluated as a To the authors knowledge, this is the most up-to-date separate entity and through a single trial-based analysis. contemporary trial-based economic evaluation comparing THA Our study has a number of strengths and some versus HA in elderly patients with displaced femoral neck limitations. As described above, the primary strength is that fractures. Previous studies have concluded, similar to our study, our evaluation is based on a multi-national, rigorously- that THA is likely not a cost effective option in the majority of conducted prospective randomized controlled trial which is patients who suffer femoral neck fractures.17,18 Consistent with the largest ever in its field. As is true of all trial-based the results of the HEALTH trial and other systematic reviews, analyses, using measured outcomes from patients rather than we did not observe any clinically significant differences in either using modeling practices reduces the number of assumptions reoperation or health-related quality of life by treatment group. on both costs and utility. In addition, through using a standard Moreover, through a methodologically rigorous trial-based eval- discount rate, performing bootstrapping techniques for sam- uation, we were able to assess whether or not THA was a cost- pling uncertainty, and conducting multiple sensitivity analy- effective intervention and under what circumstances the treat- ses determined a priori, we have adhered to good practice ment may be cost acceptable. guidelines for trial based economic evaluations.13,16 In addi- At both $50,000 and $100,000 willingness to pay tion, the use of CEACs easily illustrate the likelihood of THA thresholds, our cost effectiveness analysis revealed that to be cost-effective at given willingness to pay thresholds. compared with HA, THA is not likely to be a cost-effective A notable limitation of our economic evaluation is that intervention for all patients with displaced femoral neck cost and health-related quality of life data were not available fractures. Moreover, in no sensitivity analysis did THA for all patients. Even though EQ-5D scores were absent for represent a cost-effective intervention. However, in subgroup only approximately 8% of observations, relevant cost data analysis, for patients younger than 73 (the youngest quartile), such as medication use and return to work parameters were THA was a cost-effective intervention, because they received not collected throughout the trial. For missing health-related more QALYs for less estimated cost. In contrast, for all other quality of life data, multiple imputation techniques were used age quartiles, THA was more expensive and did not provide to generate complete data sets, and subsequent sensitivity more quality of life. Younger patients may be able to mobilize analyses were performed to generate confidence in these quicker with a THA as compared to the eldest cohort, estimates. Regarding return to work, the mean age of patients reducing the comparative additional cost to HA. Moreover, enrolled in this trial was 79 years (68 years) and thus patients younger patients may notice and appreciate the increased would be unlikely to be employed before their injury and to mobility offered from a THA, leading to higher health-related return to work, regardless of intervention.20 Moreover, cost- quality of life. ing data, in particular for orthopedic secondary procedures, Although previous economic evaluations have been was often not sufficiently specific (eg, all revisions without conducted comparing THA to HA, these were either infections were costed the same). However, the major cost performed using model approaches with significant assump- driver for these events was length of stay during hospital tions or trial-based approaches using small trials.18,19 In a readmission and, therefore, the severity of the event was more

TABLE 3. Subgroup Analysis of Cost Effectiveness by Age Quartile Cost Mean QALYs Mean THA HA D Cost THA HA D QALYs ICER $/QALY Gained Age ,73 $24,156 $27,849 -$3693 1.54 1.44 0.10 -$38,175 (dominated) Age 73–79 $42,441 $26,500 $15,941 1.48 1.45 0.03 $574,220 Age 80–84 $32,709 $21,947 $10,762 1.31 1.32 20.01 -$789,066 (dominated) Age 85 and older $31,918 $31,978 $-60 1.28 1.22 0.06 -$1072 (dominated)

2. Veronese N, Maggi S. Epidemiology and social costs of hip fracture. Injury. 2018;49:1458–1460. 3. Reeves RA, Schairer WW, Jevsevar DS. The national burden of periprosthetic hip fractures in the US: costs and risk factors for hospital readmission. Hip Int. 2019;29:550–557. 4. Lewis DP, Wæver D, Thorninger R, et al. Hemiarthroplasty vs total hip arthroplasty for the management of displaced neck of femur fractures: a systematic review and meta-analysis. J Arthroplasty. 2019;34:1837–1843.e2. 5. Ekthiari S, Gormly J, Axelrod D, et al. Total hip arthroplasty versus hemiarthroplasty for displaced femoral neck fracture: a systematic review and meta-analysis of randomized controlled trials. J Bone Joint Surg Am.2020 6. Carroll C, Stevenson M, Scope A, et al. Hemiarthroplasty and total hip arthroplasty for treating primary intracapsular fracture of the hip: a systematic review and cost-effectiveness analysis. In: NIHR Health Technology Assessment Programme: Executive Summaries. NIHR Journals Library; 2011;102(18),1638–1645. 7. HEALTH Investigators, Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019; 381:2199–2208. FIGURE 2. Cost-effectiveness acceptability curves. Editor’s 8. Canadian Institute of Health Intitatives. CIHI Patient Cost Estimator Note: A color image accompanies the online version of this [Internet]. 2020. Available at: https://www.cihi.ca/en/patient-cost- article. estimator. Accessed March 20, 2020. 9. Province of Ontario. Physiotherapy Reimbursement [Internet]. 2019. Available at: http://www.health.gov.on.ca/en/pro/programs/physio/ accurately captured in costing analysis. Another limitation of physio_pro_doctors_nurses.aspx. Accessed March 20, 2020. this analysis is the 2-year duration of the HEALTH study. It is 10. Parsons N, Griffin XL, Achten J, et al. Outcome assessment after hip possible that with 2 years of follow-up, all possible costs and fracture: is EQ-5D the answer? Bone Joint Res. 2014;3:69–75. outcomes following surgery may not have been accounted 11. McClure NS, Al Sayah F, Xie F, et al. Instrument-defined estimates of the minimally important difference for EQ-5D-5L index scores. Value for. In particular, it may take over 2 years for acetabular Heal. 2017;20:644–650. deterioration from a hemiarthroplasty to occur and to require 12. Ramsey SD, Willke RJ, Glick H, et al. Cost-effectiveness analysis along- revision. However, the best available data suggest that health side clinical trials II—an ISPOR good research practices task force outcomes following hip fractures take 2 years to stabilize report. Value Heal. 2015;18:161–172. (including rate of revision operations and functional 13. Husereau D, Drummond M, Petrou S, et al. Consolidated health eco- 21 nomic evaluation reporting standards (CHEERS) statement. BMJ. 2013; outcomes). 346:f1049–1055. Our analysis provides further evidence as to the cost 14. Hunink MGM, Bult JR, De Vries J, et al. Uncertainty in decision models effectiveness of THA in the treatment of patients with analyzing cost-effectiveness: the joint distribution of incremental costs displaced femoral neck fractures. We found no meaningful and effectiveness evaluated with a nonparametric bootstrap method. Med differences in health-related quality of life between THA and Decis Mak. 1998;18:337–346. 15. Woods B, Revill P, Sculpher M, et al. Country-level cost-effectiveness HA, whereas THA cost over $5000 more per patient. thresholds: initial estimates and the need for further research. Value Heal. However, for the youngest cohort of patients, THA was 2016;19:929–935. found to provide meaningful clinical benefit and at less cost. 16. Canadian Agency for Drugs and Technologies in Health. Guidelines for The cost effectiveness of THA for femoral neck fractures the Economic Evaluation of Health Technologies: Canada. 3rd ed. should be further evaluated prospectively and may justify Ottawa, ON; 2006. 17. Burgers PT, Hoogendoorn M, Van Woensel EAC, et al. Total medical reimbursement of the technology for this patient group. The costs of treating femoral neck fracture patients with hemi- or total hip results of this trial, although most applicable to the single arthroplasty: a cost analysis of a multicenter prospective study. payer health care system in Ontario, Canada, may be useful in Osteoporos Int. 2016;27:1999–2008. informing health policy decisions internationally. 18. Slover J, Hoffman MV, Malchau H, et al. A cost-effectiveness analysis of the arthroplasty options for displaced femoral neck fractures in the active, healthy, elderly population. J Arthroplasty. 2009;24:854–860. 19. Healy WL, Iorio R. Total hip arthroplasty: optimal treatment for dis- ACKNOWLEDGMENTS placed femoral neck fractures in elderly patients. Clin Orthop Relat The authors thank the HEALTH Investigators (http:// Res. 2004;429:43–48. 20. Statistics Canada. Labour Force Characteristics by Sex and Detailed Age Group, links.lww.com/JOT/B235). Annual, Inactive (X 1,000) [Internet]. 2019. Available at: https://www150.statcan. gc.ca/t1/tbl1/en/tv.action?pid=1410001701. Accessed March 15, 2020. REFERENCES 21. Ravikumar KJ, Marsh G. Internal fixation versus hemiarthroplasty versus 1. Rapp K, Büchele G, Dreinhöfer K, et al. Epidemiology of hip fractures. Z total hip arthroplasty for displaced subcapital fractures of femur–13 year Gerontol Geriatr. 2019;52:10–16. results of a prospective randomised study. Injury. 2000;31:793–797.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. , e h t, tee f (NIH) c , b = 0.59), and Department of f P , board or committee -Aventis, personal fees from fi Frede Frihagen, MD, d Division of Orthopaedic Surgery, b S48) – = 0.37) were not associated with a rmary, Aberdeen, United Kingdom AB25 Sheila Sprague, PhD, fi P a Prognostic Level II. See Instructions for INTRODUCTION Journal of Orthopaedic Trauma 2020;34:S42 Emil H. Schemitsch, MD, FRCSC, = 0.09), use of femoral cement ( c , P Total hip arthroplasty is associated with more SAEs b serious adverse events, displaced femoral neck fracture, Volume 34, Number 11 Supplement, November 2020 Elderly patients with hip fractures are often frail with RTICLE Department of Health Research Methods, Evidence, and Impact, c rmary, Aberdeen, United Kingdom; A J Orthop Trauma fi multiple medical comorbiditiesfrom and medical are serious at adverse risk events of (SAEs). suffering Although the cement and/or pressurization. Key Words: total hip arthroplasty, hemiarthroplasty in the immediate postoperativeselecting period, patients and care forplasty. should this A be higher treatment taken BMI may in compared beage protective alone with in does hip a not fracture predict patients hemiarthro- SAEs while and neither does the use of femoral Level of Evidence: Conclusion: fracture. Age ( use of canalmedical pressurization SAE. ( Authors for a complete description of( levels of evidence. Rudolf W. Poolman, MD, PhD, on behalf of the HEALTH Investigators b a P rst fi -Aventis, outside the submitted work. S. Sprague reports editorial or governing board fi J Orthop Trauma UPPLEMENT S Department of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden University David R. W. MacDonald, MRCS (Glas), d a Division of Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway; and e Mohit Bhandari, MD, PhD, FRCSC, c = 0.03) were associated with a ict of interest. P fl Patients Treated With Arthroplasty Daniel Axelrod, MD, MSc (Cand), b s Health, employment from Global Research Solutions Inc., and employment from McMaster University, outside the submitted work. The ’ and Iain M. Stevenson, FRCS Tr & Orth We performed a multivariable Cox regression analysis. One thousand four hundred forty-one patients were Department of Trauma and Orthopaedic Surgery, Aberdeen Royal In

a www.jorthotrauma.com rst 90 days after hip fracture with 35.4% occurring in the Patients with hip fractures are often frail with multiple David Neilly, FRCS Tr & Orth, | ﬁ a Bzovsky, MSc, Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Diane Heels-Ansdell, MSc,

Medical Center, Leiden, the Netherlands; Department of Surgery, McMaster University, Hamilton, ON, Canada; McMaster University, Hamilton, ON, Canada; for the Canadian Orthopaedic Association,International personal Society fees for from Fracture DePuy,member Repair, board for personal or fees the committee from Orthopaedic membermember ITS, Trauma for editorial for Association, the or Hip editorial the governing Society, or board Osteosynthesis board governing for or and board the committee for Trauma member the for Care the Orthopaedic Foundation, Trauma Association personal International, fees board from or commit Pentopharm, personal fees from Sano (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. Research, interpretation The of funding the sources data; had or no the role preparation, in review, design or or conduct approval of of the the study; manuscript. the collection, managemen Surgery, University of Western Ontario, London, ON, Canada. 2ZN (e-mail: [email protected]). support from Acumed,Pendopharma, LLC, and research research supportfor support BMS and from Women personal Aphria, feesremaining research from authors report Sano support no from con Ferring Pharmaceuticals, research support and personal fees from Saunders/Mosby-Elsevier, personal feesand from personal Smith fees & fromresearch Nephew, Zimmer, support outside personal from the Lima, feespersonal submitted and fees from work. research from support Springer, R. Smith from W. & personal Link Nephew, Poolman fees Orthopaedics, personal reports fees outside from board from the Stryker, or Synthes, submitted and committee personal work. personal member F. fees fees for Frihagen from from the reports Zimmer, outside Swemac, personal Dutch the fees Orthopaedic submitted from Association, work. Amgen M. Co Bhandari reports researc ﬁ (SAEs). We investigatedascertain the predictors HEALTH of SAEs. trial patient population to comorbidities and at risk of medical serious adverse events S42 Accepted for publication AugustFrom 11, the 2020. Results: Methods: Aim:

So Predictors of Medical Serious Adverse Events in Hip Fracture Occurrence of SAEs was31 included as potential the prognostic dependentvariables. variable factors with being included as independent compared with BMI between 25= and 0.03] 29.9 and [hazard receiving ratio (HR) alar 1.32, total hemiarthroplasty hip arthroplasty (HR compared 1.36, with a bipo- (14.1%, n = 77).in The the majority of SAEs30 (50.8%, days n (n = = 131). 188) Body occurred mass index (BMI) between 18.5 and 24.9 included in thissuffered analysis. Three from hundred an(38.4%, seventy SAE. n (25.6%) = The patients 105), most respiratory common (20.8%, events n were = cardiac 77), and neurological higher risk of a medical SAE within 24 months of femoral neck The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National InstitutesE. of H. Health Schemitsch reports personal fees from Acumed, LLC, personal fees from Amgen Co, research support from Biocomposites, board or committee member Reprints: Iain M. Stevenson, FRCS Tr & Orth, DepartmentCopyright of © Trauma 2020 and Wolters OrthopaedicDOI: Kluwer Surgery, 10.1097/BOT.0000000000001935 Health, Aberdeen Royal Inc. In All rights reserved.

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Adverse Events in Hip Fracture Patients definitive treatment for hip fracture is surgical, medical months of initial surgery in the HEALTH trial. Our outcome comorbidities, as well as advancing age, have historically was a single composite measure which included any type of been reported to increase the risk of death after hip fracture.1–3 medical SAE within 24 months of initial surgery in the The ability to more accurately predict medical SAEs in this HEALTH trial. Thirty-one prognostic factors were included patient group would be a useful tool for the surgeon to as the independent variables and were selected based on guide clinical decision making regarding the most appropriate biological rationale, previous literature, and expert opinion. surgical treatment for each individual. These factors included the variables used in the minimization The primary objective of this secondary analysis was to allocation process for the HEALTH trial: age, prefracture determine the factors predictive of medical SAEs in patients living setting (institutionalized or not institutionalized), from the HEALTH trial aged 50 years or older who were prefracture functional status (using assistive device for treated with either hemiarthroplasty or total hip arthroplasty ambulation or able to ambulate without assistive device), after a femoral neck fracture.4,5 The secondary objectives and American Society of Anesthesiologists (ASA) class (class were to determine the timing of the medical SAEs and to I/II or III/IV/V). For these covariates, we used values that determine the proportion of medical SAEs that occurred after were entered into the minimization system at the time of a fracture-related complication or revision surgery. enrollment. Regression results were presented as hazard ratios (HRs) with 95% confidence intervals (CIs), and P values. All tests were 2-tailed with alpha = 0.05. Statistical analysis was METHODS conducted using R (version 3.6.1, R Foundation for Statistical Medical SAEs were diagnosed by physicians at the Computing, Vienna, Austria). clinical sites and documented during the course of the HEALTH trial.4,5 An SAE was defined as any adverse event that was fatal, life-threatening, required or prolonged hospital RESULTS stay, resulted in persistent or significant disability or incapacity, In the HEALTH trial, 1441 patients were randomized to a congenital anomaly or birth defect, or an important medical treatment. At 24 months, 1243 patients were living, and event, symptom, sign, illness, or experience that developed or complete follow-up was achieved for 1058 (85.1%). In the 24 worsened in severity during the study. SAEs included neuro- months after hemiarthroplasty and total hip arthroplasty logical, respiratory, cardiac, renal, vascular, multiorgan failure, treatment for femoral neck fracture, there was an overall sepsis, anemia and other blood issues, dehydration, gastroin- 25.6% incidence of medical SAEs (n = 370). The incidence of testinal (GI) bleed, ulcer, other GI problems, and prolonged the various SAEs are summarized in Table 1. The timing of hospitalization. Mortality itself was not considered an SAE. medical SAEs is summarized in Table 2. The majority For the purposes of this secondary analysis, we did not con- (50.8%, n = 188) occurred in the first 90 days after fracture. sider any fracture-related complications as medical SAEs. The most common events were cardiac (38.4%, n = 105), In this secondary analysis, we presented the overall respiratory (20.8%, n = 77), and neurological (14.1%, n = 52). incidence of medical SAEs that occurred in the HEALTH For patients with fracture-related complications, respi- trial, the timing of the SAEs (0–30 days after fracture, 31–60 ratory and neurological were equally the most common SAE days after fracture, 61–90 days after fracture, and .90 days observed (53.8% n = 14). There were equal cases of cardiac, after fracture), the incidence of medical SAEs that occurred neurological, renal, vascular, sepsis, and other GI problem after a fracture-related complication, and the incidence of SAEs (81.8% n = 18) in patients undergoing revision surgery. medical SAEs that occurred after a revision surgery. The incidence of SAEs after fracture-related complications Fracture-related complications were defined as compli- and revision surgery is summarized in Tables 3 and 4, cations related to the hip fracture and initial surgery as per the respectively. original HEALTH trial, such as periprosthetic fracture, hip Body mass index (BMI) between 18.5 and 24.9 as instability or dislocation, implant failure (loosening/ compared to a BMI 25–29.9 (HR 1.32, 95% CI 1.03–1.70; P subsidence and breakage), wound-healing problems (includ- = 0.03), receiving a total hip arthroplasty as compared to ing superficial/deep infection, wound necrosis), another soft- receiving a bipolar hemiarthroplasty (HR 1.36, 95% CI tissue procedure, clinically important heterotopic ossification, 1.03–1.81; P = 0.03), not using preoperative traction (HR abductor failure, implant wear and corrosion, osteolysis, 1.89, 95% CI 1.18–3.03; P , 0.01), ASA classification class neurovascular injury, decreased function, and pain. III/IV/V as compared to class I/II (HR 1.70, 95% CI 1.32– Revision surgery was defined as per the original 2.19; P , 0.001), receiving treatment for lung disease (HR HEALTH trial as any unplanned secondary procedure to 1.52, 95% CI 1.14–2.03; P = 0.004), receiving treatment for treat a fracture-related complication, such as closed and open anemia or another blood disease (HR 1.95, 95% CI 1.30– reductions of a hip dislocation, open reduction of a fracture, 2.95; P = 0.001), not receiving treatment for ulcers or stom- full or partial implant exchange, implant removal, implant ach disease (HR 1.54, 95% CI 1.02–2.32; P = 0.04), receiving adjustment, soft-tissue procedure, excision of heterotopic treatment for kidney disease (HR 1.58, 95% CI 1.0–2.42; P = ossification, insertion of an antibiotic spacer, and other events 0.04), receiving treatment for heart disease (HR 1.30, 95% CI as determined by an independent central adjudication 1.03–1.65; P = 0.03), baseline use of steroids (HR 1.70, 95% committee. CI 1.09–2.63; P = 0.02), preoperative thromboprophylaxis We also performed a multivariable Cox regression to (HR 1.58, 95% CI 1.21–2.05; P , 0.01), anterolateral/ determine factors predictive of medical SAEs within 24 lateral surgical approach as compared to posterior/

TABLE 1. Overall Incidence of Medical SAEs Within the 24 Months After Arthroplasty Incidence in the THA Group, N = Incidence in the HA Group, N = Type of Serious Medical Event Total Incidence, N = 1,441, n (%) 718, n (%) 723, n (%) Cardiac 105 (7.3) 53 (7.4) 52 (7.2) Respiratory 77 (5.3) 42 (5.9) 35 (4.8) Neurological 52 (3.6) 27 (3.8) 25 (3.5) Renal 43 (3.0) 23 (3.2) 20 (2.8) Vascular 38 (2.6) 22 (3.1) 16 (2.2) Sepsis 17 (1.2) 9 (1.3) 8 (1.1) GI bleed 11 (0.8) 5 (0.7) 6 (0.8) Other GI problem 10 (0.7) 6 (0.8) 4 (0.6) Multiorgan failure 5 (0.3) 1 (0.1) 4 (0.6) Dehydration 4 (0.3) 2 (0.3) 2 (0.3) Anemia 3 (0.2) 1 (0.1) 2 (0.3) Other blood issue 2 (0.1) 1 (0.1) 1 (0.1) Ulcer 2 (0.1) 2 (0.3) 0 (0.0) Prolonged hospitalization 1 (0.1) 0 (0.0) 1 (0.1)

GI, gastrointestinal; HA, hemiarthroplasty; THA, total hip arthroplasty. posterolateral (HR 1.65, 95% CI 1.27–2.16; P , 0.01), (20.8% n = 77), and neurological (14.1% n = 52). For patients and revision surgery (HR 9.89, 95% CI 5.96–16.4; P , with fracture-related complications, respiratory and neurolog- 0.01) were associated with a higher risk of a medical SAE ical SAEs were most commonly observed (each 26.3% n = 5). within 24 months of femoral neck fracture. Age (P =0.09), However, vascular SAE was the most common (20% n = 6) in use of femoral cement (P = 0.59), and use of canal pres- patients undergoing revision surgery. Receiving a total hip surization (P = 0.37) were not associated with a medical arthroplasty was associated with an increased rate of SAEs SAE. Table 5 summarizes the predictors of SAEs in this compared with receiving a bipolar hemiarthroplasty. The patient group. patient groups for these interventions were randomized as per the HEALTH trial methodology, so we believe that this ﬁnding is accurate and could be used to guide practice. This DISCUSSION could be attributed to the acetabular reaming and implantation Our subanalysis of the HEALTH trial data has shown or due to the longer operative time and larger physiological an overall incidence of medical SAE after hip fracture to be insult to the patient; however, we did not see any effect of 25.6%. The most common type of medical SAE in this patient length of surgery or intraoperative blood loss on postopera- group was cardiac (38.4% n = 105), followed by respiratory tive SAEs when analyzed independently.

TABLE 2. Timing of Incidence of Medical SAEs After Arthroplasty Type of Serious Medical 0–30 Days After 31–60 Days After 61–90 Days After .90 Days After Event Fracture, N = 131, n (%) Fracture, N = 35, n (%) Fracture, N = 22, n (%) Fracture, N = 182, n (%) Cardiac 44 (33.6) 7 (20.0) 4 (18.2) 50 (27.5) Respiratory 24 (18.3) 9 (25.7) 6 (27.3) 38 (20.9) Neurological 14 (10.7) 5 (14.3) 3 (13.6) 30 (16.5) Vascular 17 (13.0) 6 (17.1) 3 (13.6) 12 (6.6) Renal 13 (9.9) 4 (11.4) 3 (13.6) 23 (12.6) GI bleed 7 (5.3) 0 (0.0) 0 (0.0) 4 (2.2) Multiorgan failure 3 (2.3) 0 (0.0) 0 (0.0) 2 (1.1) Sepsis 3 (2.3) 3 (8.6) 2 (9.1) 9 (5.0) Anemia 2 (1.5) 0 (0.0) 0 (0.0) 1 (0.6) Prolonged hospitalization 1 (0.8) 0 (0.0) 0 (0.0) 0 (0.0) Other blood issue 1 (0.8) 0 (0.0) 1 (4.5) 0 (0.0) Ulcer 1 (0.8) 0 (0.0) 0 (0.0) 1 (0.6) Other GI problem 1 (0.8) 1 (2.9) 0 (0.0) 8 (4.4) Dehydration 0 (0.0) 0 (0.0) 0 (0.0) 4 (2.2)

GI, gastrointestinal.

TABLE 3. Incidence of Medical SAEs after Fracture-Related Complications No Fracture-Related Fracture-Related Complication*,N Type of Serious Medical Event Complication, N = 339, n (%) = 26, n (%) Cardiac 100 (29.5) 3 (11.5) Respiratory 69 (20.4) 7 (26.9) Neurological 44 (13.0) 7 (26.9) Renal 42 (12.4) 1 (3.8) Vascular 35 (10.3) 3 (11.5) Sepsis 16 (3.4) 1 (3.8) GI bleed 10 (3.0) 1 (3.8) Other GI problem 10 (3.0) 0 (0.0) Multiorgan failure 5 (1.5) 0 (0.0) Dehydration 4 (1.2) 0 (0.0) Ulcer 2 (0.6) 0 (0.0) Anemia 1 (0.3) 1 (3.8) Prolonged hospitalization 1 (0.3) 0 (0.0) Other blood issue 0 (0.0) 2 (7.7)

*Total N is less than 370 because 5 participants experienced a medical SAE before their fracture-related complication. GI, gastrointestinal.

There has been a recent trend in using total hip even hip fracture surgeons who favor the posterior approach replacement over hemiarthroplasty for treatment of neck of will tend to use the anterolateral approach for hip fracture femur fractures due to reported improved functional out- surgery in the frailer patient due to the reported lower comes,6 despite the reported higher dislocation rate.7–9 Most dislocation rate. The National Institute for Clinical studies of this nature record complications related to the oper- Excellence (NICE) UK guidelines for hip fracture manage- ation such as hip dislocation or revision surgery,10 but few ment advise surgeons to consider the anterolateral approach in large scale studies have recorded medical SAEs related to this favor of the posterior approach for hemiarthroplasty12; how- patient group. Hansson et al11 looked at 664 hip fracture ever, a recent review of the literature found the evidence too patients and recorded pneumonia and additional falls as the heterogeneous to make a recommendation.13 The draft NICE most common complication but did not group the patients by evidence review for approaches in total hip arthroplasty from operation. October 2019 does not make any recommendations to favor The use of a laterally based approach was associated one approach over the other, but the committee agreed that with an increased risk of SAEs compared with posterior the surgeon undertaking the approach must have experience approaches. This may have been due to selection bias because and competence in that particular approach to get consistently

TABLE 4. Incidence of Medical SAEs After Revision Surgery No Revision Surgery, N = 314, n Type of Serious Medical Event (%) Revision Surgery, N = 22, n (%) Cardiac 93 (29.6) 3 (13.6) Respiratory 68 (21.7) 2 (9.1) Neurological 46 (14.6) 3 (13.6) Renal 37 (11.8) 3 (13.6) Vascular 28 (8.9) 3 (13.6) Sepsis 13 (4.1) 3 (13.6) GI bleed 8 (2.5) 1 (4.3) Other GI problem 7 (2.2) 3 (13.6) Multiorgan failure 5 (1.6) 0 (0.0) Dehydration 4 (1.3) 0 (0.0) Anemia 2 (0.6) 0 (0.0) Ulcer 2 (0.6) 0 (0.0) Other blood issue 1 (0.3) 1 (4.3) Prolonged hospitalization 0 (0.0) 0 (0.0)

Total N is less than 370 because 34 participants experienced a medical SAE before their revision surgery. GI, gastrointestinal.

TABLE 5. Continued TABLE 5. Factors Associated With a Medical SAE (n = 1,244, ( ) Factors Associated With a Medical SAE 368 Events) (n = 1,244, 368 Events) P Variable HR (95% CI) P Variable HR (95% CI) BMI (kg/m2) Patient receiving treatment for depression ,18.5 vs. 25–29.9 1.58 (0.99–2.52) 0.055 Yes vs. no 0.81 (0.54–1.21) 0.31 18.5–24.9 vs. 25–29.9 1.32 (1.03–1.70) 0.03 Baseline use of opioid medications $30 vs. 25–29.9 0.97 (0.67–1.42) 0.88 Yes vs. no 0.93 (0.61–1.40) 0.74 ASA classification Time from injury to surgery (h) 0.998 (0.996–1.002) 0.07 Class III/IV/V vs. Class I/II 1.70 (1.32–2.19) ,0.001 Additional injuries Patient receiving treatment for lung – disease Yes vs. no 0.94 (0.51 1.74) 0.86 – Yes vs. no 1.52 (1.14–2.03) 0.004 Length of surgery (min) 1.001 (0.998 1.003) 0.50 Patient receiving treatment for Type of anesthesia anemia or another blood disease Regional/other vs. general 1.23 (0.96–1.57) 0.10 Yes vs. no 1.95 (1.30–2.95) 0.001 Intraoperative blood loss (mL) 1.001 (0.999–1.002) 0.27 Patient receiving treatment for ulcers Use of femoral cement or stomach disease Yes vs. no 1.17 (0.64–2.14) 0.59 No vs. yes 1.54 (1.02–2.32) 0.04 Use of canal pressurization as 1.30 (0.72–2.34) Patient receiving treatment for kidney cementing technique disease Yes vs. no 0.37 Yes vs. no 1.58 (1.03–2.42) 0.04 Femoral canal prepared with 0.67 (0.45–1.003) Patient receiving treatment for heart pulsatile lavage disease Yes vs. no 0.051 Yes vs. no 1.30 (1.03–1.65) 0.02 Use of acetabular cement 0.77 (0.55–1.09) Baseline use of steroids Yes vs. no 0.14 Yes vs. no 1.70 (1.09–2.63) 0.02 Fracture-related complication before Preoperative traction medical SAE No vs. yes 1.89 (1.18–3.03) 0.02 Yes vs. no 1.46 (0.45–4.70) 0.52 Preoperative thromboprophylaxis ASA, American Society of Anesthesiologists; BMI, body mass index; CI, Yes vs. no 1.58 (1.21–2.05) ,0.01 confidence interval; HR, hazard ratio; SAE, serious adverse event. Type of surgical approach Direct anterior vs. 1.42 (0.71–2.82) 0.31 posterior/posterolateral 14 Anterolateral/lateral vs. 1.65 (1.27–2.16) ,0.01 good results. However, there is a randomized controlled posterior/posterolateral trial of 500 participants underway that is comparing the direct Implant received lateral versus posterolateral approach for hemiarthroplasty to 15 Total hip arthroplasty vs. bipolar 1.36 (1.03–1.81) 0.03 answer this question. hemiarthroplasty The use of bone cement in the treatment of hip fractures Total hip arthroplasty vs. 1.34 (0.97–1.85) 0.08 has created concerns16 due to bone cement implantation syn- monopolar hemiarthroplasty drome.17 This study adds to the existing weight of evidence18 Revision surgery before medical SAE that the use of bone cement is safe in hip fracture surgery. Yes vs. no 9.89 (5.96–16.4) ,0.01 Despite the higher risk patient population, pressurization of Age 1.01 (0.99–1.03) 0.09 the canal and the use of femoral or acetabular cement was not Prefracture living setting associated with an increased rate of SAEs. Institutionalized vs. not 1.42 (0.88–2.30) 0.15 Interestingly, using preoperative skin traction seemed to institutionalized lower the incidence of medical SAEs. This could be due to Prefracture functional status traction delivering better pain control for these patients and – Use of walking aid vs. independent 1.03 (0.80 1.31) 0.82 reducing the risk of fat embolism due to less movement at the ambulator fracture site. However, few institutions in the HEALTH trial Patient receiving treatment for cancer used preoperative traction, so this finding could be attribut- Yes vs. no 1.19 (0.76–1.88) 0.45 able to sampling bias. Patient receiving treatment for rheumatoid arthritis Treatment for existing medical conditions, such as lung Yes vs. no 0.50 (0.18–1.36) 0.17 disease, heart disease, diabetes, anemia, and the use of Patient receiving treatment for steroids, was all associated with an increased risk of SAEs. diabetes This correlates with clinical practice and existing studies.19,20 Yes vs. no 1.30 (0.98–1.72) 0.07 A higher ASA grade was associated with an increased risk of SAEs. Smith et al conducted a meta-analysis of studies

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Adverse Events in Hip Fracture Patients looking at preoperative indicators for mortality after hip frac- to use this technique when appropriate. Age alone is not a ture surgery. Four studies with a total of 1559 participants predictor of medical SAEs; therefore, each hip fracture patient were included assessing risk of death based on ASA grade should receive individualized care. We hope that the findings and found that patients with an ASA grade of 3 or 4 were at a of this subanalysis can aid the clinician when deciding the 44% increased risk of death at 12 months compared to those most appropriate surgical treatment for this vulnerable group. with a grade of 1 or 2.21 The Hailer et al22 review using the Swedish joint register investigated 24,699 patients who had undergone total hip arthroplasty for neck of femur fracture ACKNOWLEDGMENTS fi and found that 90-day mortality was signi cantly higher in The authors thank the HEALTH Investigators (http:// patients with ASA 3 compared with ASA 1 (HR = 9.5). links.lww.com/JOT/B244). Another interesting finding of the subanalysis was that age alone does not predict medical SAEs and neither does REFERENCES 23 prefracture functional or living status. Schnell et al reported 1. Sexson SB, Lehner JT. Factors affecting hip fracture mortality. J Orthop patient demographics and outcomes of 758 patients treated in Trauma. 1987;1:298–305. their program over a 4-year period. They showed that 1-year 2. White BL, Fisher WD, Laurin CA. Rate of mortality for elderly patients ’ mortality was highest for patients who were older than 90 after fracture of the hip in the 1980 s. J Bone Joint Surg Am. 1987;69: – – 1335–1340. years compared with those 70 79 or 80 89 years of age; 3. Zuckerman JD. Hip fracture. N Engl J Med. 1996;334:1519–1525. however, they did not specifically report on the incidence 4. Bhandari M, Devereaux PJ, Einhorn TA, et al. Hip fracture evaluation of SAEs. Although it is logical that older hip fracture patients with alternatives of total hip arthroplasty versus hemiarthroplasty are less likely to survive at 1 year compared with younger (HEALTH): protocol for a multicentre randomised trial. BMJ Open. 2015;5:e006263. patients, age does not seem to predict medical SAEs. This 5. Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or may be a result of the patients who are living to an older age hemiarthroplasty for hip fracture. N Engl J Med. 2019;381:2199– before sustaining a hip fracture having less medical comor- 2208. bidities, and in clinical practice, it is often the case that the 6. Hopley C, Stengel D, Ekkernkamp A, et al. Primary total hip arthroplasty older hip fracture patient is “fitter.” versus hemiarthroplasty for displaced intracapsular hip fractures in older patients: systematic review. BMJ. 2010;340:c2332. A higher BMI may be protective for medical SAEs in 7. Van den Bekerom MPJ, Hilverdink EF, Sierevelt IN, et al. A comparison hip fracture patients. The CDC (Centers for Disease Control of hemiarthroplasty with total hip replacement for displaced intracapsular and Prevention) categorize BMI into 4 universal categories fracture of the femoral neck: a randomised controlled multicentre trial in —,18.5 underweight, 18.5–24.9 normal weight, 25–29.9 patients aged 70 years and over. J Bone Joint Surg Br. 2010;92:1422– . 24 1428. overweight, and 30 obese. Sherrif et al reviewed the 8. Burgers PT, Van Geene AR, Van den Bekerom MP, et al. Total hip National Surgical Quality Improvement Program database arthroplasty versus hemiarthroplasty for displaced femoral neck fractures for hip fracture patients, and of the 15,108 patients, they in the healthy elderly: a meta-analysis and systematic review of random- found that patients with very low BMI had a higher mortality ized trials. Int Orthop. 2012;36:1549–1560. rate than patients with a very high BMI, but that those at the 9. Lewis DP, Wæver D, Thorninger R, et al. Hemiarthroplasty vs total hip arthroplasty for the management of displaced neck of femur fractures: a extreme ends of the BMI spectrum suffered the most compli- systematic review and meta-analysis. J Arthroplasty. 2019;34:1837– cations. Transfusion rates were observed to decrease as BMI 1843. increased.25 10. Blomfeldt R, Törnkvist H, Eriksson K, et al. A randomised controlled The strength of our study is that the data used here are trial comparing bipolar hemiarthroplasty with total hip replacement for displaced intracapsular fractures of the femoral neck in elderly patients. J sourced from a large-scale trial which had broad inclusion Bone Joint Surg Br. 2007;89:160–165. criteria. Recruitment also occurred from a large number of 11. Hansson S, Rolfson O, Åkesson K, et al. Complications and patient- hospitals in diverse health care systems, therefore, giving a reported outcome after hip fracture. A consecutive annual cohort study more accurate view of SAEs in hip fracture patients of 664 patients. Injury. 2015;46:2206–2211. encountered in day-to-day practice, as well as increasing the 12. Practice. CgCNGfC. Hip Fracture Management. Available at: https:// fi www.nice.org.uk/guidance/cg124/chapter/Recommendations#surgical- generalizability of the external validity of our ndings. procedures. Accessed April 15, 2020. A limitation of our subanalysis is the loss of patients over 13. Fullam J, Theodosi PG, Charity J, et al. A scoping review comparing two the 24 months of follow-up in the original study (14.9%). common surgical approaches to the hip for hemiarthroplasty. BMC Surg. However, the loss to follow-up was consistent with or better 2019;19:32. 25 14. National Institute for Health and Care Excellence. NICE guideline inter- than other trials including hip fracture patients. The vention evidence review for hip replacement approach: Joint replacement HEALTH trial had unavoidable variables across the patient (primary): hip, knee and shoulder [Internet]. London, United Kingdom: group, such as patient positioning, surgical exposure, use of National Institute for Health and Clinical Excellence; 2019. Available at: traction, type of anesthetic, and physiotherapy and rehabilita- https://www.nice.org.uk/guidance/GID-NG10084/documents/evidence- tion protocols, which may have affected results. Finally, as the review-13. Accessed April 15, 2020. 15. ZonMw. Surgical Approach of Hemiarthroplasty after Femoral Neck data used here were from subanalysis of another trial, there is Fracture: Posterolateral or Direct Lateral [Internet]. Amsterdam, The not as detailed information collected on each SAE, than if it Netherlands: ZonMw; 2020. Available at: https://www.zonmw.nl/nl/ had been a trial designed with SAE as the primary outcome. onderzoek-resultaten/doelmatigheidsonderzoek/programmas/project-detail/ In conclusion, total hip arthroplasty is associated with evaluatieonderzoek-zegg/surgical-approach-of-hemiarthroplasty-after- femoral-neck-fracture-posterolateral-or-direct-lateral-1/. Accessed May more SAEs than hemiarthroplasty and care should be taken in 27, 2020. selecting patients for this treatment. Cement does not increase 16. Rutter PD, Panesar SS, Darzi A, et al. What is the risk of death or severe the incidence of SAEs, and clinicians should remain confident harm due to bone cement implantation syndrome among patients under-

going hip hemiarthroplasty for fractured neck of femur? A patient safety 21. Smith T, Pelpola K, Ball M, et al. Pre-operative indicators for mortality surveillance study. BMJ Open. 2014;4:e004853. following hip fracture surgery: a systematic review and meta-analysis. 17. Donaldson AJ, Thomson HE, Harper NJ, et al. Bone cement implantation Age Ageing. 2014;43:464–471. syndrome. Br J Anaesth. 2009;102:12–22. 22. Hailer NP, Garland A, Rogmark C, et al. Early mortality and morbidity 18. British Orthopaedic Association. Use of bone cement in hip replacement after total hip arthroplasty in patients with femoral neck fracture: a surgery following a hip fracture: A position statement from the BOA, nationwide study of 24,699 cases and 118,518 matched controls. Acta BHS and OTS [Internet]. London, United Kingdom: British Orthopaedic Orthop. 2016;87:560–566. Association; 2014. Available at: https://www.boa.ac.uk/resources/use-of- 23. Schnell S, Friedman SM, Mendelson DA, et al. The 1-year mortality of bone-cement-in-hip-replacement-surgery-following-a-hip-fracture-a-position- patients treated in a hip fracture program for elders. Geriatric Orthop statement-from-the-boa-bhs-and-ots.html. Accessed April 14, 2020. Surg Rehabil. 2010;1:6–14. 19. Todd CJ, Freeman CJ, Camilleri-Ferrante C, et al. Differences in mor- 24. Centres for Disease Control and Prevention. About Adult BMI [Internet]. tality after fracture of hip: the East Anglian audit. BMJ. 1995;310:904– Atlanta, United States: Centres for Disease Control and Prevention; 2020. 908. Available at: https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/ 20. Donegan DJ, Gay AN, Baldwin K, et al. Use of medical comorbidities to index.html#InterpretedAdults. Accessed May 2, 2020. predict complications after hip fracture surgery in the elderly. JBJS. 25. Akinleye SD, Garofolo G, Culbertson MD, et al. The role of BMI in hip 2010;92:807–813. fracture surgery. Geriatric Orthop Surg Rehabil. 2018;9:2151458517747414.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. , t, he lor sonal S49 n, VT (NIH) c s from c | Department f , d c a Bzovsky, MSc, fi predictors of revision surgery So Department of Surgery, University of Division of Orthopaedic Surgery, Oslo b e b S54) – s Health, employment from Global Research ’ Sheila Sprague, PhD, f , Prognostic Level II. See Instructions for c 2020;34:S49 ndings to an active physiologically younger femoral Daniel Axelrod, MD, MSc (Cand), Both total and partial hip replacements are successful fi revision surgery, femoral neck fracture e cantly increased the need for revision surgery in our Of the 1441 patients included in this analysis, 8.1% fi 0.05). RTICLE . A P J Orthop Trauma elderly and predominately female patientgeneralize population. our One should not neck fracture population. Key Words: Level of Evidence: Results: Conclusion: (117/1441) experienced reoperationthe within studied 24 factors were months. found( None to be of We were unable to identifythat any patient signi or surgeon-controlled factors procedures in low-energy displaced femoral neck fracture patients. Authors for a complete description of( levels of evidence. on behalf of the HEALTH Investigators Emil H. Schemitsch, MD, FRCSC, a a UPPLEMENT -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees from Smith & Nephew, fi S 50 years with ict of interest. fl $ Frede Frihagen, MD, PhD, Mohit Bhandari, MD, PhD, FRCSC, d f cant short-term differences Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada. Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; fi f c Secondary Analysis of the HEALTH Trial Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com and Patrick C. Schottel, MD HEALTH was a randomized controlled trial compar- We analyzed 9 potential factors chosen a priori that Diane Heels-Ansdell, MSc, Department of Orthopaedics and Rehabilitation, University of Vermont Medical Center, South Burlington, VT;

-Aventis, outside the submitted work. S. Sprague reports editorial or governing board for BMS Women a ﬁ Rudolf W. Poolman, MD, PhD,

Displaced Femoral Neck Fractures With Arthroplasty: A Michael Blankstein, MD, MSc, FRCSC, Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Aphria, research support from FerringSano Pharmaceuticals, research support andSolutions personal Inc, fees and from employment Pendopharma,submitted from and work. McMaster research The support University, remaining and outside authors personal the fee report submitted no work. con P. C. Schottel reports paid consultancy from Synthes, outside t personal fees from Springer,Poolman personal reports fees from boardOrthopaedics, Stryker, or outside personal fees the committee from submittedSynthes, member Swemac, and work. and for personal F. personal the fees Frihagen fees from reports Dutch from Zimmer, personal Zimmer, Orthopaedic outside outside fees Association, the the from submitted research submitted Amgen work. work. Co, support R. M. personal W. from Bhandari fees Lima, reports from research and Smith support & research from Nephew, support Acumed, personal from LLC, fees research Link from support from DePuy, board or committeeITS, member editorial for or the governing Hipgoverning board Society, for board board the for or Journal thepersonal committee of Orthopaedic member Orthopaedic fees Trauma Trauma, for from Association board the International, Pentopharm, or International board committee personal Society or member for fees for committee Fracture from the member Repair, Orthopaedic Sano for personal Trauma the fees Association, Osteosynthesis from editoria and Trauma Care Foundation (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. Research, interpretation The of funding the sources data; had or nofees the role preparation, from in design review, Amgen or or Co, conduct approval research of of support the the from study; article. Biocomposites, the collection, board managemen or committee member for the Canadian Orthopaedic Association, personal fees from 05405 (e-mail: [email protected]). University Hospital, Oslo, Norway; and Western Ontario, London, ON, Canada; of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden University Medical Center, Leiden, the Netherlands; being the primary outcome. No signi ing totaldisplaced hip femoral arthroplastyunplanned neck with revision surgery fracture within hemiarthroplasty 24 patients months in of the aged initial low-energy procedure J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Methods:

Objectives: What Factors Increase Revision Surgery Risk When Treating analysis was a multivariable Cox24 regression using months reoperation of within index surgery as the dependent variable. could be associated with revisionbody mass surgery. index, The major factors comorbidities, independent includedof ambulation, age, surgical type approach,femoral length head of size, and operation, degree use of of femoral stem femoral offset. cement, Our statistical Reprints: Michael Blankstein, MD, MSc, FRCSC, Department of Orthopaedics and Rehabilitation, Robert T Stafford Hall, 95 Carrigan Drive, Burlingto M. Blankstein reports stock or stock options from 7D Surgical, outside the reported work. E. H. Schemitsch reports personal fees from Acumed, LLC, per The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health patient and surgicalrevision factors surgery were within associated 24 with months increased after hip risk fracture. of between treatment armsthis were secondary observed. The HEALTH primary trial objective analysis of was to determine if any Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001936 Health, Inc. All rights reserved.

INTRODUCTION size (36 mm for THA vs. 28 mm for THA, 32 mm for THA vs. Displaced femoral neck fractures are commonly treated 28 mm for THA, and hemiarthroplasty vs. 28 mm for THA), with arthroplasty in the elderly population. Although arthro- and degree of femoral stem offset (high vs. standard/reduced). plasty procedures are associated with a high success rate, it is We performed a secondary post-hoc exploratory analysis uncertain whether outcomes are better with total hip arthro- where we performed the original Cox regression model but plasty (THA) or hemiarthroplasty (HA). Several studies with the following changes: (1) the femoral head size variable suggest that when treating active, physiologically younger, was recategorized to 36 mm versus 32 mm versus 28 mm and low-energy femoral neck fracture patients, a total hip (2) the degree of femoral stem offset variable was replaced replacement can provide a more durable solution with lower with the implant type (THA vs. HA). Results were reported long-term revision rates.1,2 A recent study, the hip fracture as hazard ratios (HRs) with 95% confidence intervals (CIs). All evaluation with alternatives of THA versus hemiarthroplasty tests were 2-tailed with alpha = 0.05. (HEALTH) trial, attempted to more definitively answer this question. The results demonstrated no significant short-term differences between the 2 arthroplasty treatment groups in RESULTS patients 50 years of age or older who sustained a low- A total of 1441 patients were studied. THA and HA 3 energy displaced femoral neck fracture. However, other were performed in 718 and 723 patients, respectively. Table 1 aspects such as patient and surgical factors were not analyzed displays patient demographic characteristics comparing the to determine risk for revision surgery, and their contributions 1324 cases that did not require a revision versus the 117 that to clinical outcomes remain an area of active interest. required a secondary procedure. The mean age was 79 (SD 8) The purpose of this study was to investigate what years, and 70% (1009/1441) were female. Table 2 displays patient and surgical factors increase risk of the arthroplasty patients’ different fracture characteristics. All fractures were revision surgery after THA or HA in patients 50 years of age displaced, and 95% of the fractures were either subcapital or or older who sustained a low-energy displaced femoral neck midcervical. All of the injuries were the result of low-energy fracture. This secondary HEALTH trial study was designed to trauma. perform a prognostic analysis of baseline patient characteris- Approximately two-thirds of the arthroplasty proce- tics and surgical factors associated with revision surgery dures were performed with a cemented femoral stem. The within 24 months of the index arthroplasty procedure. lateral approach was most commonly used in 65.5% of cases, Identifying factors that are associated with increased revision followed by the posterolateral approach (32%). The anterior surgery can assist surgeons with treatment decisions when approach was only used 2.5% of the time. With THA, 32-mm caring for low-energy femoral neck fracture patients. heads were used in a majority of cases (54%), followed by 36- mm heads (29%) and 28-mm heads (17%). High offset METHODS femoral stems were only implanted 18% of the time. After 24 months from the index arthroplasty procedure, The HEALTH study randomized patients 50 years of 8.1% (117/1441) of patients required revision surgery. age or older with a displaced femoral neck fracture into Table 3 displays the difference in surgical factors between surgical treatment with either THA or HA. Surgeons were the 1324 patients who did not require a revision surgery expected to meet a procedural experience threshold and were versus the 117 who required a secondary procedure. Tables allowed to perform their preferred surgical approach to the 4 and 5 display our analysis of predictors for revision surgery. hip. This included anterior, anterolateral, lateral, or We were unable to identify any patient or surgical factors posterolateral/posterior approaches. In addition, surgeons associated with increased risk of revision surgery at 24 were allowed to choose at their own discretion either a months. Age, BMI, major comorbidities, independent ambu- cemented or uncemented femoral stem, the THA femoral lation status, surgical approach, operative time, use of ce- head size, and standard or high offset femoral stem. mented or uncemented femoral components, femoral head We performed 2 multivariable Cox regression analyses size, and femoral stem offset were not found to be predictors using the HEALTH primary outcome of reoperation within of revision surgery (P . 0.05). 24 months of initial surgery as the dependent variable. The number of independent variables and corresponding levels included in the 2 analyses were based on Peduzzi et al4’s recommendations, where having fewer than 10 events for DISCUSSION each predictor variable can result in overfitted, unstable mod- The top 3 reasons for revision surgery in the HEALTH els. In our first model, 9 potential prognostic factors with 11 trial were dislocations, soft-tissue procedures, and implant parameters were identified a priori that could be associated exchanges (reason unspecified). The rate of a secondary with revision surgery. Factors were selected based on biolog- procedure was relatively low at 8% in both the THA and HA ical rationale and expert opinion. The studied factors were age groups. The results of our secondary analysis did not identify (continuous), body mass index (BMI) (continuous), patients any patient-controlled or surgeon-controlled factors that limited by major comorbidities (yes vs. no), independent significantly increased the need for revision surgery. All ambulation (yes vs. no), surgical approach (direct anterior patient factors, surgical techniques, approaches, and implant vs. posterior/posterolateral vs. anterolateral/lateral), length types used did not affect these overall reproducible and of operation, use of femoral cement (yes vs. no), femoral head successful arthroplasty outcomes.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Treating Displaced Femoral Neck Fractures with Arthroplasty

TABLE 1. Patient Demographics TABLE 2. Fracture Characteristics No Revision Revision No Revision Revision Surgery, Surgery, Variable Surgery Surgery Variable N = 1324 N = 117 Fractured hip, n (%) N = 1324 N = 117 Age, mean (SD) 78.9 (8.5) 78.2 (8.2) Left 708 (53.6) 64 (54.7) Sex, n (%) N = 1323 Right 612 (46.4) 53 (45.3) Male 400 (30.2) 31 (26.5) Level of the fracture line, n (%) N = 1320 Female 923 (69.8) 86 (73.5) Subcapital 806 (61.1) 84 (71.8) Ethnicity, n (%) Midcervical 451 (34.2) 30 (25.6) Indigenous 3 (0.2) 0 (0.0) Basal 63 (4.8) 3 (2.6) South Asian 8 (0.6) 1 (0.9) Garden classification, n (%) N = 1320 East Asian 13 (1.0) 1 (0.9 Garden III (displaced) 574 (43.5) 57 (48.7) Hispanic/Latino 11 (0.8) 2 (1.7) Garden IV (displaced) 746 (56.5) 60 (51.3) White 1255 (95.1) 112 (95.7) Pauwels’ classification, n (%) N = 1318 Black 27 (2.0) 0 (0.0) Type I 106 (8.0) 11 (9.4) Middle Eastern 3 (0.2) 1 (0.9) Type II 708 (53.7) 63 (53.8) BMI (kg/m2), n (%) N = 1285 Type III 504 (38.2) 43 (36.8) Underweight (,18.5) 69 (5.4) 4 (3.4) Mechanism of injury, n (%) N = 1320 Normal weight (18.5–24.9) 641 (49.9) 52 (44.4) Fall from standing 1283 (97.2) 113 (96.6) Overweight (25–29.9) 413 (32.1) 47 (40.2) Spontaneous fracture 27 (2.0) 3 (2.6) Obese (30–39.9) 146 (11.4) 14 (12.0) Fall from small height 10 (0.8) 1 (0.9) Morbidly Obese ($40) 16 (1.2) 0 (0.0) Prefracture living setting, n (%) ASA, American Society of Anesthesiologists; BMI, body mass index. Institutionalized 53 (4.0) 4 (3.4) Not institutionalized 1271 (96.0) 113 (96.6) ,30. It is important to appreciate that this hip fracture pop- Prefracture functional status, n (%) ulation is different from many other studies that have evalu- Use of Aid 337 (25.5) 32 (27.4) ated predictors for primary THA revision, which are mostly Independent Ambulator 987 (74.5) 85 (72.6) performed for osteoarthritis and not a femoral neck fracture. ASA classification, n (%) In primary THA series in the literature, the patients are gen- Class I/II 595 (44.9) 57 (48.7) erally younger, healthier, and have a higher BMI. Peters et al Class III/IV/V 729 (55.1) 60 (51.3) studied an arthroplasty registry cohort of 218,214 primary Major comorbidities, n (%) N = 1320 THAs performed in patients with osteoarthritis in the Osteopenia 53 (4.0) 5 (4.3) Netherlands between 2007 and 2018.5 In contrast to our Osteoporosis 208 (15.8) 16 (13.7) results, the authors found higher revision rates at 1 year after Lung disease 223 (16.9) 26 (22.2) primary THA in patients with BMI .40, ASA scores of III- Diabetes 257 (19.5) 23 (19.7) V, patients older than 75 years of age, and men. At 3 years, Ulcers or stomach disease 110 (8.3) 6 (5.1) high BMI, previous hip surgery, Charnley score C, men, and Kidney disease 126 (9.5) 12 (10.3) a high ASA score were independently associated with an Anemia or other blood diseases 97 (7.3) 6 (5.1) increased risk for revision. Most common causes for revisions Depression 137 (10.4) 17 (14.5) were infections, dislocations, and periprosthetic fractures. Cancer 134 (10.2) 11 (9.4) Importantly, only one-third of their patients were older than Osteoarthritis, degenerative 186 (14.1) 16 (13.7) 75 years or had a BMI ,30, and 13.4% of patients had ASA Arthritis scores III-IV. This highlights the discrepancy in study partic- Back pain 122 (9.2) 13 (11.1) ipants from our investigation and those studying THA for Rheumatoid arthritis 32 (2.4) 2 (1.7) osteoarthritis. Heart disease 460 (34.8) 36 (30.8) A recent systematic review and meta-analysis concluded High blood pressure 801 (60.7) 76 (65.0) that THA is superior to HA for reoperation rates and quality of ASA, American Society of Anesthesiologists; BMI, body mass index. life and should be the recommended intervention for displaced femoral neck fractures in younger patients (,80) and in those whose life expectancy is greater than 4 years.6 Nevertheless, dislocation rates were slightly higher in patients with THA. In Our results revealed that age, BMI, major comorbid- the HEALTH study, THA outcomes were only marginally ities, and independent ambulation were not predictive of better for quality of life. Dislocation rates, however, were high- revision surgery. As expected, the HEALTH patient popula- er in the THA as compared with the HA group (4.7% as tion has significant comorbidities as noted in Table 1. The opposed to 2.4%). Previous published studies have found that mean age was 79 years and 55% percent of the patients were dislocations after primary THA have been linked to extremes American Society of Anesthesiologists (ASA) classification of age, BMI .30, lumbosacral pathology, surgeon experience, III-V. Furthermore, 99% of the HEALTH patients had a BMI and femoral head size. There is conflicting evidence regarding

TABLE 3. Surgical Characteristics TABLE 4. Prognostic Factors for Reoperation (n = 1,137, 98 Events) No Revision Revision Surgery, Surgery, Hazard Ratio Variable N = 1324 N = 117 Variable (95% CI) P Implant received, n (%) N = 1310 N = 116 Age (10-y increase) 0.87 (0.68–1.11) 0.26 Total hip arthroplasty 621 (47.4) 57 (49.1) BMI (5-point increase) 0.98 (0.80–1.21) 0.87 Monopolar hemiarthroplasty 315 (24.0) 27 (23.3) Patient limited by comorbidities Bipolar hemiarthroplasty 374 (28.5) 32 (27.6) Yes vs. no 1.15 (0.75–1.75) 0.54 Time from injury to surgery, 52.3 (75.3) 68.6 (116.0) Independent ambulation mean (SD) (h) Yes vs. no (any aid used) 1.21 (0.73–1.89) 0.42 Length of procedure, 86.7 (40.3) 88.0 (37.5) Type of surgical approach mean (SD) (minutes) Direct anterior vs. 1.05 (0.32–3.51) Preoperative traction, n (%) N = 1319 9 (7.7) posterior/posterolateral 100 (7.6) Anterolateral/lateral vs. 0.77 (0.51–1.17) Overall: 0.24 Preoperative thromboprophylaxis, n (%) N = 1317 83 (70.9) posterior/posterolateral 930 (70.6) Length of operation (minutes) 0.999 (0.995–1.01) 0.99 Postoperative thromboprophylaxis, n (%) N = 1311 116 (99.1) Use of femoral cement 1303 (99.4) Yes vs. no 1.11 (0.70–1.74) 0.67 Type of Anaesthesia, n (%) N = 1318 Femoral head size Overall: 0.83 General 496 (37.6) 39 (33.3) 32 mm for total hip arthroplasty 0.93 (0.44–2.01) Regional 872 (66.2) 81 (69.2) vs. 28 mm for total hip Sedation 13 (1.0) 1 (0.9) arthroplasty – Neurolept 14 (1.1) 2 (1.7) 36 mm for total hip arthroplasty 0.78 (0.31 1.93) vs. 28 mm for Total hip Intraoperative blood loss, 292.0 (183.7) 305.3 (211.9) arthroplasty mean (SD) (mL) Hemiarthroplasty vs. 28 mm 0.92 (0.45–1.87) Use of femoral cement, n (%) N = 1312 for total hip Arthroplasty Yes 839 (63.9) 79 (67.5) Offset of chosen stem No 473 (36.1) 38 (32.5) High vs. standard/reduced 1.57 (0.87–2.94) 0.13 Type of surgical approach, n (%) N = 1311 N = 116 Direct anterior 33 (2.5) 3 (2.6) BMI, body mass index; CI, confidence interval. Anterolateral/lateral 869 (66.3) 67 (57.8) Posterior/posterolateral 409 (31.2) 46 (57.8) Femoral head size, n (%) N = 1290 N = 114 28 mm for total hip arthroplasty 104 (8.1) 11 (9.6) The direct lateral approach reduced the risk of reoperation after 32 mm for total hip arthroplasty 324 (25.1) 30 (26.3) hip fractures treated with HA in patients over 75 years of age. 36 mm for total hip arthroplasty 175 (13.6) 14 (12.3) However, a later study of 20,908 patients from the Norwegian Hemiarthroplasty 687 (53.3) 59 (51.8) Hip Fracture Registry specifically looking at patient-reported Offset of chosen stem, n (%) N = 1087 N = 100 outcome measures after HA for femoral neck fractures, found Standard/Reduced 890 (81.9) 88 (88.0) that the posterior approach caused less pain, better ambulation, High 197 (18.1) 12 (12.0) and better satisfaction and overall quality of life than the lateral approach.9 The risk of reoperation was similar with both SD, standard deviation. approaches. Finally, the muscle sparing direct anterior approach has recently become popular for primary arthroplasty.10 It has been to shown to be linked to slightly faster functional recovery and improved short-term outcomes.11 Disadvantages with the the effect of neuromuscular disease and surgical approach on anterior approach include a trend toward more major complica- THA instability.7 Interestingly, our results demonstrated that tions, particularly femoral component failure12–14, although the factors often associated with instability such as surgical posterior approach continues to be associated with higher dislo- approach, head size, and femoral stem offset were not associ- cation rates in both THA and HA.12,15 A systematic review and ated with an increased rate of secondary revisions. meta-analysis of the direct anterior approach for HA for femoral Surgical approach to the hip continues to be a controver- neck fracture suggests superior early functional mobility and sial topic in hip arthroplasty. In the HEALTH study, the lateral significantly lower dislocation rates compared with posterior approach was most commonly used in 65.5% of cases, followed approaches for HA.16 The influence of surgical approach by the posterolateral approach 32% of the time. Our results did and/or use of dual mobility components on dislocations not find surgical approach as a predictor of revision surgery. The when THA is performed for fractures are currently being Norwegian and Swedish Hip Fracture Registries reviewed investigated worldwide. 33,205 hip fractures in patients 60 years of age and older who The femoral head size has been shown to reduce were treated with modular HAs.8 The posterior approach was dislocations in THA. This is predominately related to shown to increase the risk of reoperation because of dislocations. increased head to neck ratio and jump distance. The

observational data, they found that the risk factors for revision TABLE 5. Secondary Post-Hoc Exploratory Analysis for in both the THA and HA groups were male sex, age younger Prognostic Factors for Reoperation (n = 1,000, 79 Events*) than 80 years, posterolateral approach, and uncemented com- Hazard Ratio ponents. The group concluded that when performing arthro- Variable (95% CI) P plasty for hip fractures, both a posterolateral approach and Age (10-y increase) 0.93 (0.71–1.22) 0.61 uncemented femoral stems have higher risks for revision sur- BMI (5-point increase) 1.07 (0.85–1.34) 0.58 gery compared with an anterolateral approach and cemented Patient limited by comorbidities stems. A recent systematic review and meta-analysis of modern Yes vs. no 1.23 (0.76–2.00) 0.41 stems concluded that cemented stems for HA result in fewer Independent ambulation implant related complications with similar mortality rates.20 Yes vs. no (any aid used) 1.31 (0.78–2.18) 0.31 Likewise, Grosso et al retrospectively reviewed the results of Type of surgical approach a consecutive cohort of 686 patients who underwent HA for the Direct anterior 1.18 (0.27–5.11) treatment of femoral neck fractures at their institution with a vs. posterior/posterolateral minimum of 2-year follow-up. The overall revision rate was – Anterolateral/Lateral 0.92 (0.57 1.49) Overall: 0.71 low at 5.6%. Relatively higher conversion rates to THA were vs. posterior/posterolateral seen in the younger population and fewer periprosthetic frac- Length of operation (minutes) 0.996 (0.988–1.01) 0.34 tures with the use of cemented stems.21 Use of femoral cement Femoral stem offset when used appropriately can Yes vs. no 1.12 (0.68–1.82) 0.66 influence abductor moment arm, strength, and overall hip Femoral head size function. The leg length and offset dependent soft-tissue 32 mm vs. 28 mm 0.91 (0.53–1.55) tension can in turn affect hip stability. A patient’s ideal ana- 36 mm vs. 28 mm 0.72 (0.35–1.46) Overall: 0.38 tomical offset can only be determined by preoperative tem- Implant type plating.22 In our study, high offset stems were rarely used and Total hip arthroplasty 0.96 (1.05–1.67) 0.88 vs. hemiarthroplasty were not found to be predictive of revision surgery. A recent study investigated the influence of femoral component offset *Femoral head sizes in most HA participants were .36 mm. These participants on revision rates for primary THA in the New Zealand Joint were not assigned any value under the femoral head size category, and when running the 23 model, they were automatically removed. For that the reason, the overall n value and Registry. Both high and low offset femoral component number of events decreased as compared with the first model (Table 4). stems, as compared with standard offset, were found to affect fi BMI, body mass index; CI, con dence interval. the overall all-cause revision rate of cemented stems. Our study has several limitations. First, it is important to remember that the surgeons in the HEALTH study might dislocation rates in the HEALTH study were 4.7% with THA differ from the typical trauma or general orthopaedic and 2.4% with HA. In our analysis, the femoral head size did surgeon. Nearly all participating surgeons (97.8%) met the not seem to be associated with the need for revision surgery. thresholds for surgical expertise. To enroll patients in the This could possibly be due to the fact that 83% of the THA HEALTH study, surgeons had to be comfortable performing cases used 32 mm and 36 mm heads. An analysis of 166,231 HA or THA routinely for hip fractures. Surgeons who primary THAs performed in the Dutch Arthroplasty Registry, regularly perform THA could theoretically be more versatile examined the effect of the femoral head size and other with press-fit and cemented implants, better at assessing surgical factors on revision rates.17 They reported that for all soft-tissue tension and likely have lower dislocation rates surgical approaches, 32-mm heads reduced the risk of revi- because of better component position. This expertise could sion for dislocation compared with small head sizes. They theoretically decrease the generalizability of our findings. In also found that 36 mm heads reduced the risk of revision addition, although nearly 1500 patients were enrolled in the for dislocation only with the posterolateral approach. With HEALTH study, arthroplasty registries can sometimes be the anterior approach, 36-mm heads increased the risk of more informative given their larger sample size. revision for other reasons. Observational data from numerous arthroplasty registries The AAOS clinical practice guidelines for the manage- cited throughout this article have often reached different ment of hip fractures in the elderly recommend the use of conclusions than those reported in our study. We are unable cemented femoral stems when performing an arthroplasty to definitively explain why our results differ. Registry data procedure for displaced femoral neck fractures.18 Nearly two- can identify trends and outliers, but it does not allow thirds of the arthroplasty procedures in our study were per- determination of cause and effect relationships. The low formed with a cemented femoral stem. This could account revision rate in HEALTH potentially limited our ability to for the relatively low revision rate because of fewer peripros- detect any predictive factors for revision surgery. Further thetic fractures, and it might provide an explanation of why studies are needed to definitively determine the true cementless fixation was not predictive of revision surgery. Data influence of various patient and surgical factors on the need from the Dutch Arthroplasty Registry evaluated 30,830 for revision surgery. Finally, the study is limited by the patients with hip fractures who were treated with THA and relatively short-term 2-year clinical follow-up. It is possible HA.19 Revision rates at 1 year for HA were 1.6% and 2.4% that particular patient and surgical factors result in a higher for THAs. Dislocation was the most common reason for revi- revision surgery rate that will only be observed with longer sion in both groups (HA 29% and THA 41%). Based on clinical follow-up. Further follow-up is therefore necessary.

In conclusion, both THA and HA are successful 10. Patel NN, Shah JA, Erens GA. Current trends in clinical practice for the direct procedures with low reoperation rates when treating low- anterior approach total hip arthroplasty. JArthroplasty.2019;34:1987–1993.e3. energy displaced femoral neck fracture patients. We were 11. Wang Z, Hou JZ, Wu CH, et al. A systematic review and meta-analysis of direct anterior approach versus posterior approach in total hip arthro- unable to identify any patient-controlled or surgeon- plasty. J Orthop Surg Res. 2018;13:229. controlled factors that significantly increased the need for 12. Meneghini RM, Elston AS, Chen AF, et al. Direct anterior approach: risk revision surgery in our elderly predominately female patient factor for early femoral failure of cementless total hip arthroplasty: a population. One should not generalize our findings to an multicenter study. J Bone Joint Surg Am. 2017;99:99–105. active younger femoral neck fracture population. 13. Angerame MR, Fehring TK, Masonis JL, et al. Early failure of primary total hip arthroplasty: is surgical approach a risk factor? J Arthroplasty. 2018;33:1780–1785. 14. Pincus D, Jenkinson R, Paterson M, et al. Association between surgical ACKNOWLEDGMENTS approach and major surgical complications in patients undergoing total The authors thank the HEALTH Investigators (http:// hip arthroplasty. JAMA. 2020;323:1070–1076. links.lww.com/JOT/B231). 15. van der Sijp MPL, van Delft D, Krijnen P, et al. Surgical approaches and hemiarthroplasty outcomes for femoral neck fractures: a meta-analysis. J REFERENCES Arthroplasty. 2018;33:1617–1627.e9. 16. Kunkel ST, Sabatino MJ, Kang R, et al. A systematic review and 1. Hauer G, Heri A, Klim S, et al. Survival rate and application number of total meta-analysis of the direct anterior approach for hemiarthroplasty hip arthroplasty in patients with femoral neck fracture: an analysis of clinical – for femoral neck fracture. Eur J Orthop Surg Traumatol. 2018;28: studies and national arthroplasty registers. J Arthroplasty. 2020;35:1014 1022. – 2. Hansson S, Nemes S, Kärrholm J, et al. Reduced risk of reoperation after 217 232. treatment of femoral neck fractures with total hip arthroplasty. Acta 17. Zijlstra WP, De Hartog B, Van Steenbergen LN, et al. Effect of femoral Orthop. 2017;88:500–504. head size and surgical approach on risk of revision for dislocation after 3. HEALTH Investigators, Bhandari M, Einhorn TA, Poolman RW, et al. total hip arthroplasty. Acta Orthop. 2017; 88:395–401. Total hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 18. Brox WT, Roberts KC, Taksali S, et al. The American academy of 2019;381:2199–2208. orthopaedic surgeons evidence-based guideline on management of hip 4. Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number fractures in the elderly. J Bone Joint Surg Am. 2015;97:1196–1199. of events per variable in logistic regression analysis. J Clin Epidemiol. 19. Moerman S, Mathijssen NMC, Tuinebreijer WE, et al. Hemiarthroplasty 1996;49:1373–1379. and total hip arthroplasty in 30,830 patients with hip fractures: data from 5. Peters RM, Peters RM, van Steenbergen LN, et al. Patient characteristics the Dutch Arthroplasty Register on revision and risk factors for revision. influence revision rate of total hip arthroplasty: American society of Acta Orthop. 2018;89:509–514. anesthesiologists score and body mass index were the strongest predic- 20. Veldman HD, Heyligers IC, Grimm B, et al. Cemented versus cementless tors for short-term revision after primary total hip arthroplasty. J hemiarthroplasty for a displaced fracture of the femoral neck: a system- – Arthroplasty. 2020;35:188 192 e2. atic review and meta-analysis of current generation hip stems. Bone Joint 6. Lewis DP, Wæver D, Thorninger R, et al. Hemiarthroplasty vs total hip J. 2017;99-B:421–431. arthroplasty for the management of displaced neck of femur fractures: a 21. Grosso MJ, Danoff JR, Murtaugh TS et al. Hemiarthroplasty for dis- systematic review and meta-analysis. JArthroplasty.2019;34:1837–1843 e2. 7. Rowan FE, Benjamin B, Pietrak JR, et al. Prevention of dislocation after placed femoral neck fractures in the elderly has a low conversion rate. – total hip arthroplasty. J Arthroplasty. 2018;33:1316–1324. J Arthroplasty. 2017;32:150 154. 8. Rogmark C, Fenstad AM, Leonardsson O, et al. Posterior approach and 22. Lecerf G, Fessy MH, Philippot R, et al. Femoral offset: anatomical fi uncemented stems increases the risk of reoperation after hemiarthroplas- concept, de nition, assessment, implications for preoperative templating ties in elderly hip fracture patients. Acta Orthop. 2014;85:18–25. and hip arthroplasty. Orthop Traumatol Surg Res. 2009;95:210–219. 9. Kristensen TB, Vinje T, Havelin LI, et al. Posterior approach compared 23. Wyatt MC, Kieser DC, Kemp MA, et al. Does the femoral offset affect to direct lateral approach resulted in better patient-reported outcome after replacements? The results from a National Joint Registry. Hip Int. 2019; hemiarthroplasty for femoral neck fracture. Acta Orthop. 2017;88:29–34. 29:289–298.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. h t, c , ees a ports S55 (NIH) dence | urnal of fi 9.91%) ns of this 22.12%, – f – 8.84%). A – Department of g c 34.02%, respec- – The Michael G. DeGroote National b Frede Frihagen, MD, a Bzovsky, MSc, fi e So 2V9, 1280 Main St. West, Hamilton, ON, Canada – c , 24.35%; ARI 12.5%, 95% CI 2.85% a -Aventis, outside the submitted work. The remaining – fi Among hip fracture patients undergoing arthro- Mohit Bhandari, MD, PhD, FRCSC, hip arthroplasty, hemiarthroplasty, femoral neck frac- 79-year-old; ARI 6.3%; 95% CI 2.67% a # Department of Orthopedic and Trauma Surgery, OLVG, Amsterdam 17.64%). 25.89%; ARI 21.1%; 95% CI 8.23% e – – RTICLE A ger ageand ( higher prefacture3.80% functional status (ARI 10.7%; 95% CI respectively] and prefracture5.41% opioid usetively). (ARI Female 15.6%, sex was 95%sistent associated CI with pain an at increasedgreater risk risk 1 of of persistent year per- pain at (ARI 2 years 6.2%, was associated 95% with youn- CI 3.53% associated withfracture reporting [absolute moderate-to-severe riskinterval hip (CI) increase 6.44% pain (ARI) before 15.3%, 95% con Conclusions: plasty, approximately one inpain 10 up will to experience moderate-to-severe 2functioning prefracture, years living after with surgery.prescription hip Younger opioids pain were age, prefracture, predictive female and of use sex, persistent of Key higher pain. Words: ture, predictors, pain on behalf of the HEALTH Investigators Sheila Sprague, PhD, h , b , Division of Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway; g f , a The Chronic Pain Centre of Excellence for Canadian Veterans, Hamilton, ON, Canada. 2) at 12 b h , s Health, employment from Global Research Solutions Inc, and employment from McMaster ’ Rudolf W. Poolman, MD, PhD, a $ UPPLEMENT Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, c d S Diane Heels-Ansdell, MSc, c Moderate-to-severe pain (at least 2 One thousand four hundred forty-one hip ict of interest. s Web site (www.jorthotrauma.com). fl ’ Volume 34, Number 11 Supplement, November 2020 www.jorthotrauma.com Total hip arthroplasty or hemiarthroplasty. Atefeh Noori, MSc, PhD (Cand), and Jason W. Busse, DC, PhD To identify factors associated with the development of -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees from Smith & Nephew, personal fees from Springer, personal fees from fi Department of Surgery, University of Western Ontario, London, ON, Canada; d Of 840 and 726 patients with complete baseline data and Eighty hospitals in 10 countries. Secondary analysis of a randomized controlled trial. Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton,Ontario, Canada; Emil H. Schemitsch, MD, FRCSC, a Predictors of Long-Term Pain After Hip Arthroplasty in Patients With Femoral Neck Fractures: A Cohort Study Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. McMaster Surgical Associates, andanalysis, Stryker or Orthopaedics. interpretation The of funding the sources data; had or no the role preparation, in design review, or or conduct approval of of the the study; manuscript. the collection, managemen (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw) (17088.2503), Sophies Minde Foundation for Orthopaedic Research, Canada; Pain Center, McMaster University, Hamilton, ON, Canada; L8S 4K1 (e-mail: [email protected]). support and personal fees from Pendopharma, and research support and personal fees from Sano Stryker, personal fees from Swemac,Dutch and Orthopaedic personal Association, fees from research Zimmer, supportpersonal outside from fees the Lima, submitted from and work. Amgen research R.work. W. support Co, M. Poolman from personal reports Bhandari Link board fees Orthopaedics, reports or outside from committee research the member Smith support submitted for & the from work. Nephew, F. Acumed, Frihagen personal LLC, re fees research from support Synthes, from and Aphria, personal research fees support from from Zimmer, Ferring outside Pharmaceuticals, the researc submitted Society, board or committee memberOrthopaedic for the Trauma, International board SocietyAssociation for or Fracture International, Repair, committee board personal or member feesfrom committee from for Sano member ITS, editorial the for or the Orthopaedic governing Osteosynthesis board Trauma and for Trauma the Association, Jo Care editorial Foundation, or personal fees governing from board Pentopharm, for personal f the Orthopaedic Trauma University, outside the submittedBiocomposites, work. board E. or H. committee Schemitsch member reports for personal the Canadian fees Orthopaedic from Association, Acumed, personal LLC, fees personal from fees DePuy, from board Amgen or Co, committee member research for support the from Hip and Leiden University Medical Center, Leiden, the Netherlands; Anesthesia, McMaster University, Hamilton, ON, Canada; and authors report no con article on the journal Daniel Axelrod, MD, MSc (Cand), J Orthop Trauma Accepted for publication AugustFrom 11, the 2020. Results: Main Outcome Measures: Patients/Participants: Interventions: Design: Setting: Objectives: prolonged pain after hip fracture surgery. (11.0%) reportedincreasedriskofpainatboth1and2yearsaftersurgerywas moderate-to-severe pain, respectively. An outcomes at 1-year and 2-year follow-up, 96 (11.4%) and 80 S. Sprague reports editorial or governing board for BMS Women The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health fracture patients in the HEALTH trial. and 24 months after hip arthroplasty. activities onOsteoarthritis the questionnaire Western pain subscale Ontario with and scores McMaster Universities Reprints: Jason W. Busse, DC, PhD, Michael G. DeGrooteCopyright National Pain © Centre, 2020 McMaster Wolters University,DOI: Kluwer HSC 10.1097/BOT.0000000000001929 Health, Inc. All rights reserved. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versio

Level of Evidence: Prognostic Level II. See Instructions for pain (a score of 2 or more for at least 2 questions on activities Authors for a complete description of levels of evidence. using the WOMAC pain subscale) or no moderate-to-severe – pain (a score of 0 or 1 for at least 3 items on activities, which (J Orthop Trauma 2020;34:S55 S63) we used as the reference group in our analysis) at 12 and 24 15–17 INTRODUCTION months after hip surgery. Hip fracture is a severe and frequent event affecting a large number of older adults around the world.1 Predictors – The first group of predictors consisted of demographic Approximately 52,000 hip fracture related hospital admis- # . sions were reported among the population aged 80 years and variables, including age ( 79 vs. 79 years. 79 was the older between 2011 and 2015 in Ontario, Canada, alone.2 Hip median age of all included participants) and sex (women vs. fractures are associated with impaired mobility, loss of quality men). We also considered prefracture variables including of life, and one-year mortality rate estimations between 14% those reporting moderate-to-severe hip pain before the occur- and 58% after injury.3–7 rence of the fracture (yes vs. no), use of opioid medications Displaced femoral neck fractures are common injuries, before fracture (yes vs. no), prefracture functional status (am- which are often managed with arthroplasty, especially for bulating without assistance vs. use of an assistive device), patients $80 years.8 With an aging population globally, the prefracture living status (institutionalized vs. noninstitutional- demands for arthroplasty are predicted to increase ized), previous surgery affecting hip (yes vs. no), having substantially.9 Although symptom relief is an important major comorbidities including back pain, osteoarthritis, cancer, rheumatoid arthritis, and depression (yes vs. no), body goal of hip arthroplasty, many patients experience persistent 2 2 10 mass index (BMI, 25–34.9 kg/m vs. #24.9 kg/m ), and the postsurgical pain. For instance, one study of patients with a fi displaced subcapital hip fracture found that 13% reported American Society of Anesthesiologists (ASA) classi cation moderate-to-severe persistent pain at one year after the (III/IV/V vs. I/II). In addition, procedure-related factors, surgery.11 Some evidence suggests that total hip including type of arthroplasty surgery (THA vs. bipolar or arthroplasty (THA) may provide better long-term pain relief unipolar hemiarthroplasty) and type of surgical approach than hemiarthroplasty12; however, the HEALTH randomized (anterolateral/direct anterior vs. posterior/posterolateral), were controlled trial results did not find a benefit of THA over included. We also considered 4 postoperative variables, all of 13 which were measured up to 30 days after surgery, including hemiarthroplasty in reducing persistent pain. Improved ’ understanding of factors associated with long-term pain patient s weight-bearing status (non/partial vs. full), use of after hip arthroplasty could facilitate targeting of high-risk any physiotherapy or rehabilitation (yes vs. no), serious patients in an effort to improve prognosis. adverse events (yes vs. no), and fracture-related complica- In this study, we aimed to investigate which factors tions (complication with or without revision surgery vs. no (modifiable and nonmodifiable) are associated with moderate- complication). We chose 30 days after surgery as our cut-off to-severe pain at 12 and 24 months after hip arthroplasty in mark to avoid concerns regarding the temporality of the pre- participants aged 50 years and older with a displaced femoral dictors and the study outcomes. neck fracture. Statistical Analysis We reported the mean and SD of continuous variables, MATERIAL AND METHODS and absolute and relative frequencies for categorical variables. We used logistic regression analyses to determine Participants predictors of moderate-to-severe pain at 12 and 24 months This study used data from 1441 participants enrolled in after hip fracture repair. We excluded patients from our the HEALTH trial, a randomized controlled trial that explored analyses if their pain score, measured using the WOMAC, the effect of THA or hemiarthroplasty for patients $50 years before fracture and at follow-up times (12 and 24 months), with a displaced femoral neck fracture.13 were unavailable. We first ran univariable models to estimate the unadjusted odds ratios (ORs), then we constructed Outcome Measure multivariable models for both visits. We selected 12 inde- In the HEALTH trial, the Western Ontario and pendent factors for multivariable models previously reported McMaster Universities Osteoarthritis (WOMAC) question- as important predictors18 or variables judged to be associated naire was administered to assess hip-related pain status of with persistent pain by experts, including age, sex, BMI, patients before their fracture occurred (completed at time of prefracture moderate-to-severe hip pain, opioid use, func- study enrollment) and at 12 months and 24 months after tional status, depression, weight-bearing status, use of any surgery. The WOMAC has been shown to be valid, reliable, physiotherapy or rehabilitation, and serious adverse events. responsive, and feasible in hip fracture patients.14 The ques- We also adjusted for the type of arthroplasty surgery and tionnaire contains 5 different activity questions for pain inten- surgical approach that were used in the HEALTH trial. We sity assessment with options of “no,”“mild,”“moderate,” excluded independent variables with fewer than 50 observa- “severe,” and “extreme” pain. We summed these scores for tions per category, unless we were able to collapse them with all 5 questions and calculated the overall pain score (range 0– other related variables to exceed this threshold (Table 1). We 20) for each patient. In keeping with previous studies, we then applied the Hosmer–Lemeshow (H-L) goodness of fit defined our outcome as the presence of moderate-to-severe test for these adjusted models19 and reported the statistics.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Long-Term Pain After Hip Arthroplasty

TABLE 1. Demographics and Clinical Characteristics at 12- and 24-Month Follow-up Visits in Patients With Femoral Neck Fractures 12 months 24 months Total, No Moderate-to- Moderate-to-Severe Total, No Moderate-to- Moderate-to-Severe Characteristics n = 840 Severe Pain, n = 744 Pain, n = 96 n = 726 Severe Pain, n = 646 Pain, n = 80 Demographic factors Age, n (%) Below median 399 (47.5) 346 (86.7) 53 (13.3) 362 (49.9) 315 (87.0) 47 (13.0) (#79 years) Above median 441 (52.5) 398 (90.2) 43 (9.8) 364 (50.1) 331 (91.0) 33 (9.0) (.79 years) Sex, n (%) Female 618 (73.6) 538 (87.0) 80 (13.0) 546 (75.0) 484 (88.6) 62 (11.3) Male 222 (26.4) 206 (92.8) 16 (7.2) 180 (25.0) 162 (90.0) 18 (10.0) Preoperative factors BMI, n (%) #24.9 431 (52.0) 384 (89.1) 47 (10.9) 368 (51.0) 332 (90.0) 36 (10.0) 25–34.9 399 (48.0) 352 (88.2) 47 (11.8) 351 (49.0) 308 (88.0) 43 (12.0) Prefracture moderate- to-severe hip pain, n (%)* No 696 (89.7) 631 (90.7) 65 (9.3) 605 (90.7) 552 (91.2) 53 (8.8) Yes 80 (10.3) 60 (75.0) 20 (25.0) 62 (9.3) 45 (72.6) 17 (27.4) Prefracture opioid use, n (%) No 774 (92.1) 695 (89.8) 79 (10.2) 670 (92.2) 607 (90.6) 63 (9.4) Yes 66 (7.9) 49 (74.2) 17 (25.8) 56 (7.7) 39 (69.6) 17 (30.3) Prefracture functional status, n (%) Uses an assistive 673 (80.1) 603 (89.6) 70 (10.4) 589 (81.1) 535 (91) 54 (9) device Ambulate without 167 (19.9) 141 (84.4) 26 (15.6) 137 (18.8) 111 (81) 26 (19) assistance Prefracture living status, n (%)† Not institutionalized 822 (97.9) 731 (89.0) 91 (11.0) 717 (98.7) 639 (89.1) 78 (10.8) Institutionalized 18 (2.1) 13 (72.2) 5 (27.8) 9 (1.2) 7 (77.7) 2 (22.2) Previous surgery to affected hip n (%)† No 836 (99.6) 740 (88.5) 96 (11.5) 723 (99.7) 644 (89.0) 79 (10.9) Yes 3 (0.4) 3 (100) 0 (0) 2 (0.28) 1 (50.0) 1 (50.0) ASA class, n (%) I/II 416 (49.5) 370 (89.0) 46 (11.0) 389 (53.5) 354 (91) 35 (9) III/IV/V 424 (50.4) 374 (88.2) 50 (11.8) 337 (46.4) 292 (87) 45 (13) Preoperative comorbidities Rheumatoid arthritis, n (%) No 817 (97.3) 725 (88.8) 92 (11.2) 704 (96.9) 627 (89) 77 (11) Yes 23 (2.7) 19 (82.6) 4 (17.4) 22 (3.0) 19 (86) 3 (14) Osteoarthritis, n (%) No 714 (85) 643 (90.0) 71 (10.0) 623 (85.8) 556 (90) 67 (10) Yes 126 (15) 101 (80.1) 25 (19.9) 103 (14.1) 90 (87) 13 (13) Cancer, n (%) No 757 (90.1) 674 (89.0) 83 (11.0) 654 (90.0) 583 (89) 71 (11) Yes 83 (9.9) 70 (84.3) 13 (15.7) 72 (9.9) 63 (87.5) 9 (12.5)

(continued on next page)

TABLE 1. (Continued) Demographics and Clinical Characteristics at 12- and 24-Month Follow-up Visits in Patients With Femoral Neck Fractures 12 months 24 months Total, No Moderate-to- Moderate-to-Severe Total, No Moderate-to- Moderate-to-Severe Characteristics n = 840 Severe Pain, n = 744 Pain, n = 96 n = 726 Severe Pain, n = 646 Pain, n = 80 Depression, n (%) No 738 (87.9) 658 (89.1) 80 (10.9) 639 (88.0) 570 (89) 69 (11) Yes 102 (12.1) 86 (84.0) 16 (16.0) 87 (11.9) 76 (87) 11 (13) Back pain, n (%) No 759 (90.3) 679 (89.4) 80 (10.6) 655 (90.2) 594 (91) 61 (9) Yes 81 (9.7) 65 (80.2) 16 (19.8) 71 (9.7) 52 (73) 19 (27) Procedure-related factors Surgical approach, n (%) Posterior/ 315 (37.5) 284 (90.1) 31 (9.9) 280 (38.6) 249 (89) 31 (11) posterolateral Anterolateral/direct 524 (62.46) 459 (87.6) 65 (12.4) 446 (61.4) 397 (89) 49 (11) anterior Surgery type, n (%) THA 414 (49.4) 372 (89.9) 42 (10.1) 363 (50.0) 331 (91.2) 32 (8.8) Bipolar 212 (25.3) 186 (87.8) 26 (12.2) 189 (26.0) 164 (86.8) 25 (13.2) hemiarthroplasty Unipolar 212 (25.3) 184 (86.8) 28 (13.2) 173 (23.8) 150 (86.8) 23 (13.2) hemiarthroplasty Postoperative factors Early weight-bearing status, n (%)‡ Non or partial 355 (42.2) 317 (89.3) 38 (10.7) 295 (40.7) 258 (87.4) 37 (12.6) Full 485 (57.8) 427 (88.0) 58 (12.0) 431 (59.3) 388 (90.0) 43 (10.0) Early physiotherapy or rehabilitation, n (%)‡ No 61 (7.3) 51 (83.6) 10 (16.4) 53 (7.3) 44 (83.0) 9 (17.0) Yes 774 (92.7) 690 (89.1) 84 (10.9) 669 (92.7) 599 (89.5) 70 (10.5) Serious adverse events, n (%)§ No 672 (80.0) 596 (88.7) 76 (11.3) 594 (81.8) 530 (89.2) 64 (10.8) Yes 168 (20.0) 148 (88.1) 20 (11.9) 132 (18.2) 116 (87.9) 16 (12.1) Fracture-related complications, n (%)†§ No complication 774 (92.2) 686 (88.6) 88 (11.4) 673 (92.7) 600 (89.2) 73 (10.8) Complication 29 (3.4) 27 (93.1) 2 (6.9) 22 (3.0) 19 (86.4) 3 (13.6) without revision surgery Complication with 37 (4.4) 31 (83.8) 6 (16.2) 31 (4.3) 27 (87.1) 4 (12.9) revision surgery

*The WOMAC questionnaire was administered and asked patients to indicate the severity of hip pain before their fracture occurred. †Variables with very low frequency in one subcategory were not included in the regression models. ‡These variables were collected immediately after surgery. §These variables were collected within 30 days after surgery. ASA, American Society of Anesthesiologists; BMI, body mass index; THA, total hip arthroplasty.

We assessed the impact of influential observations by statistic, with values between 0.7 and 0.8 indicating accept- calculating the deviance residual and explored for multi- able classification performance of a model. We reported collinearity among the independent variables using the adjusted ORs (aORs) along with their 95% confidence inter- variance inflation factor, with values greater than 5 indicating vals (95% CIs). In addition, we reported the absolute risk a possible issue regarding collinearity.20 Discriminability of increase (ARI) for each significant predictor in the selected the models was checked using the area under the curve adjusted models by estimating the baseline risk of outcome

(for both 12- and 24-month follow-up) among patients who Patient Characteristics did not have any significant risk factors. We also tested for an At the 12-month follow-up visit, 96 of the 840 patients interaction between opioid use and functional status, hypoth- (11.4%) and at 24 months, 80 of the 726 patients (11.0%) esizing that patients who used opioids and used assistive reported moderate-to-severe pain. Patients had a median age devices to ambulate before fracture may be more likely to of 79 years, most were women (73.0% at 12 months and experience moderate-to-severe persistent pain. Variables with 75.0% at 24 months), not institutionalized (approximately P values less than 0.05 in the adjusted models were consid- 98.0% at both visits) before the fracture occurred and did not ered statistically significant. All analyses were performed in have a previous surgery affecting the hip (.99.0%). Stata version 15. Demographic, preoperative/prefracture factors, comorbidities, procedure related, and postoperative characteristics among patients with and without moderate-to-severe pain are pre- RESULTS sented in Table 1. At 12- and 24-month follow-up visits, 1127 and 1122 patients, respectively, were eligible for our study. At the 1- Multivariable Logistic Model at 12-Month year follow-up visit, 64 (5.7%) and by 2 years postopera- Follow-Up Visit tively, 113 (10%) participants had died. Of the 1063 and 1009 Among 165 patients without identified risk factors for patients who were alive, 840 (20% missing data) and 726 pain (significant predictors in the adjusted model that reduced (28% missing data) had follow-up data available for the the risk of moderate-to-severe pain at 12-month visit), 8 analysis of pain at 12 and 24 months follow-up, respectively. reported moderate-to-severe pain at 1 year for a baseline risk Details regarding participant flow and the reasons for of 4.8%. In the adjusted model (Table 2), female patients (OR exclusion are provided in Fig. 1. 1.85; 95% CI 1.0–3.45; ARI 6.2%, 95% CI 3.53%–8.84%;

FIGURE 1. Cohort of patients’ selection ﬂow.

P = 0.049), patients with prefracture hip pain (OR 2.68; 95% reported moderate-to-severe pain at 2 years for a baseline risk CI 1.46–4.90; ARI 15.4%, 95% CI 6.44%–24.35%; P = of 5.05%. Patients with prefracture hip pain (OR 2.52; 95% 0.001), and opioid use (OR 2.66; 95% CI 1.35–5.24; ARI CI 1.25–5.06; ARI 12.5%, 95% CI 2.85%–22.12%; P = 15.7%, 95% CI 5.41%–25.90%; P = 0.005) had a signifi- 0.01), prefracture opioid use (OR 4.32; 95% CI 2.08–8.95; cantly higher risk of experiencing moderate-to-severe pain ARI 21.1%; 95% CI 8.23%–34.02%; P , 0.001), aged #79- at 1-year after surgery (Table 2). year-old (OR 1.92; 95% CI 1.09–3.39; ARI 6.3%; 95% CI 2.67%–9.91%; P = 0.025), and who ambulated without assis- Multivariable Logistic Model at 24-Month tive devices before surgery (OR 2.38; 95% CI 1.28–4.44; ARI Follow-Up Visit 10.7%; 95% CI 3.80%–17.64%; P = 0.006) were more likely Among 218 patients without identified risk factors for to report moderate-to-severe pain after 2 years (Table 3). pain (significant predictors in the adjusted model that reduced The interaction between prefracture opioid use and the risk of moderate-to-severe pain at 24-months visit), 11 functional status was not statistically significant (P = 0.32).

TABLE 2. Factors Associated With Moderate-to-Severe Pain (vs. Mild or No Pain) at 12 Months in Patients With Femoral Neck Fractures Factor Unadjusted OR (95% CI) P Adjusted OR (95% CI) P ARI %(95% CI) Age Above median (.79 years) Reference Reference Below median (#79 years) 1.41 (0.92–2.17) 0.109 1.42 (0.86–2.32) 0.167 — Sex Male Reference Reference Female 1.91 (1.09–3.35) 0.023 1.85 (1.0–3.45) 0.049 6.2 (3.53–8.84) BMI #24.9 Reference — Reference 25–34.9 1.09 (0.71–1.67) 0.691 0.96 (0.6–1.55) 0.874 — Prefracture moderate-to-severe hip pain No Reference — Reference Yes 3.23 (1.83–5.70) ,0.001 2.68 (1.46–4.90) 0.001 15.4 (6.44–24.35) Prefracture opioid use No Reference — Reference Yes 3.05 (1.67–5.55) ,0.001 2.66 (1.35–5.24) 0.005 15.7 (5.41–25.90) Prefracture functional status Uses an assistive device Reference — Reference Ambulate without assistance 1.58 (0.97–2.58) 0.062 1.52 (0.86–2.67) 0.147 — Depression No Reference Reference Yes 1.53 (0.85–2.73) 0.152 1.19 (0.6–2.33) 0.62 — Surgical approach Posterior/posterolateral Reference — Reference Anterolateral/direct anterior 1.29 (0.82–2.04) 0.260 1.1 (0.64–1.89) 0.722 — Surgery type THA Reference — Reference Bipolar hemiarthroplasty 1.24 (0.74–2.08) 0.421 1.05 (0.59–1.87) 0.858 — Unipolar hemiarthroplasty 1.35 (0.81–2.24) 0.251 1.35 (0.74–2.43) 0.325 — Weight-bearing status Full Reference — Reference Non or partial 0.88 (0.57–1.36) 0.573 0.94 (0.57–1.57) 0.826 — Early physiotherapy or rehabilitation No Reference — Reference Yes 0.62 (0.30–1.26) 0.191 0.9 (0.39–2.07) 0.809 — Serious adverse events No Reference — Reference Yes 1.05 (0.62–1.79) 0.828 0.76 (0.4–1.43) 0.388 —

Adjusted model: the Hosmer–Lemeshow (HL) test P value: 0.833; C-statistics: 0.67. AR, absolute risk increase; BMI, body mass index; OR, odds ratio; THA, total hip arthroplasty.

TABLE 3. Factors Associated With Moderate-to-Severe Pain (vs. Mild or No Pain) at 24 Months in Patients With Femoral Neck Fractures Factor Unadjusted OR (95% CI) P Adjusted OR (95% CI) P ARI % (95% CI) Age Above median (.79 years) Reference Reference Below median (#79 years) 1.49 (0.93–2.39) 0.093 1.92 (1.09–3.39) 0.025 6.2 (2.67–9.90) Sex Male Reference Reference Female 1.15 (0.66–2.00) 0.615 1.20 (0.63–2.28) 0.561 — BMI #24.9 Reference Reference 25–34.9 1.29 (0.81–2.06) 0.291 1.09 (0.64–1.86) 0.743 — Prefracture moderate-to-severe hip pain No Reference Reference Yes 3.93 (2.11–7.35) ,0.001 2.52 (1.25–5.06) 0.01 12.4 (2.85–22.11) Prefracture opioid use No Reference Reference Yes 4.20 (2.25–7.85) ,0.001 4.32 (2.08–8.95) ,0.001 21.1 (8.22–34.02) Prefracture functional status Uses an assistive device Reference Reference Ambulate without assistance 2.32 (1.39–3.87) 0.001 2.38 (1.28–4.44) 0.006 10.7 (3.80–17.63) Depression No Reference Reference Yes 1.20 (0.61–2.36) 0.606 0.91 (0.41–1.99) 0.807 — Surgical approach Posterior/posterolateral Reference Reference Anterolateral/direct anterior 0.99 (0.61–1.59) 0.972 0.88 (0.49–1.58) 0.658 — Surgery type THA Reference Reference Bipolar hemiarthroplasty 1.58 (0.90–2.75) 0.108 1.42 (0.76–2.66) 0.27 — Unipolar hemiarthroplasty 1.59 (0.89–2.80) 0.112 1.97 (1–3.88) 0.052 — Weight-bearing status Full Reference Reference Non or partial 1.29 (0.81–2.06) 0.279 1.50 (0.85–2.64) 0.161 — Early physiotherapy or rehabilitation No Reference Reference Yes 0.57 (0.26–1.21) 0.148 0.79 (0.32–1.96) 0.609 — Serious adverse events No Reference Reference Yes 1.14 (0.63–2.04) 0.655 1.06 (0.54–2.08) 0.87 —

Adjusted model: the Hosmer–Lemeshow (HL) test P value: 0.229; C-statistics: 0.71. AR, absolute risk increase; BMI, body mass index; OR, odds ratio; THA, total hip arthroplasty.

We did not find any variables with a variance inflation factor In our adjusted analyses, patients reporting pain and use greater than 1.50, and no variables had a deviance residual of opioids before their fracture occurred were more than twice more than 3, which may suggest no evidence of serious multi- as likely to report persistent moderate-to-severe pain after hip collinearity or influential observations in our data. surgery, which is consistent with findings of other studies exploring prognosis after THA and total knee arthroplasty.21,22 Evidence suggests that chronic use of opioids 23 DISCUSSION may produce opioid-induced hyperalgesia, which can result We found that approximately one in 10 hip fracture in pain sensitivity and in developing persistent pain after 24,25 patients experienced moderate-to-severe pain 2 years after surgery. In addition, opioids are commonly prescribed arthroplasty. Younger age, female sex, higher functioning for chronic pain; pre-existing pain in another part of the body before fracture, living with hip pain, and use of prescription may influence pain severity of replaced joints.26,27 As such, opioids were predictive of persistent pain after surgery. there is a possibility that patients reported persistent

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Noori et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 moderate-to-severe pain after surgery because of an underly- years after hip arthroplasty. Patients with prefracture hip pain, ing chronic pain condition.28 opioid use, independent ambulatory status, female sex, and Our results did not indicate that surgery type (eg, THA younger ages (#79 vs. greater 79 years) are more likely to vs. bipolar or unipolar hemiarthroplasty) is a significant have long-term moderate-to-severe pain after arthroplasty sur- predictor of pain at 1 and 2 years after surgery. We also gery for management of femoral neck fractures. This infor- found no association between receiving early physiotherapy/ mation will further inform health care providers and patients rehabilitation or weight-bearing status and pain at 12- and 24- and allow for better understanding of the expected benefits of month follow-up. Consistent with our findings, a retrospec- hip arthroplasty in this patient population. tive study did not find a significant difference between a full versus no weight-bearing status and pain among hip fracture 29 patients who were treated with surgery, although other stud- Acknowledgments ies have shown that no weight-bearing status is related to poor 30,31 fi The authors thank the HEALTH Investigators (http:// functioning after surgery. Our ndings also revealed that links.lww.com/JOT/B233). patients who were able to ambulate without assistive devices experienced more moderate-to-severe pain 2 years after hip arthroplasty. However, these results are in contrast to some REFERENCES 1. Carnevale V, Fontana A, Scillitani A, et al. Incidence and all-cause studies that reported loss in long-term functional ability was mortality for hip fracture in comparison to stroke, and myocardial infarc- inversely associated with higher prefracture functional sta- tion: a fifteen years population-based longitudinal study. Endocrine. tus.31,32 These results suggest that the complex relationship 2017;58:320–331. between pain and function after hip fracture surgery requires 2. Clemens KK, Ouedraogo A, Speechley M, et al. Hip fractures in older adults in Ontario, Canada—monthly variation, insights, and implications. further evaluation. Can Geriatr J. 2019;22:148. Female sex was also associated with persistent hip pain 3. Peeters CM, Visser E, Van de Ree CL, et al. Quality of life after hip 1 year after surgery. Other researchers have reported similar fracture in the elderly: a systematic literature review. Injury. 2016;47: findings and have stated that female sex is related to increased 1369–1382. chronic pain,33,34 whereas other research has found no asso- 4. Jongjit J, Komsopapong L, Songjakkaew P, et al. Health-related quality 35 # . of life after hip fracture in the elderly community-dwelling. Southeast ciation. Younger ages ( 79 vs. 79 years) predicted more Asian J Trop Med Public Health. 2003;34:670–674. moderate-to-severe pain 2 years after hip arthroplasty, which 5. Bentler SE, Liu L, Obrizan M, et al. The aftermath of hip fracture: is consistent with previous studies that assessed predictors of discharge placement, functional status change, and mortality. Am J chronic pain after surgery.36–38 Explanations for this result Epidemiol. 2009;170:1290–1299. 6. Roche J, Wenn RT, Sahota O, et al. Effect of comorbidities and post- may relate to higher expectations of recovery among younger operative complications on mortality after hip fracture in elderly people: patients or biological differences, with younger patients prob- prospective observational cohort study. BMJ. 2005;331:1374. ably having greater neuron system responsiveness as com- 7. Brauer CA, Coca-Perraillon M, Cutler DM, et al. Incidence and mortality pared to older patients.39 Nevertheless, these findings differ of hip fractures in the United States. JAMA. 2009;302:1573–1579. from some studies40,41 that failed to show age as a significant 8. Bhandari M, Devereaux P, Tornetta P III, et al. Operative management of displaced femoral neck fractures in elderly patients: an international predictor of persistent pain after surgery. survey. J Bone Joint Surg Am. 2005;87:2122–2130. Our study has some limitations, including a high rate of 9. Kurtz SM, Lau E, Ong K, et al. Future young patient demand for primary incomplete data for the outcome, especially at the 2-year and revision joint replacement: national projections from 2010 to 2030. follow-up; however, our rate of missing data is consistent Clin Orthop Relat Res. 2009;467:2606–2612. 42,43 10. Yoo JI, Ha YC, Lim JY, et al. Early rehabilitation in elderly after arthro- with previous studies in this area. Baseline characteristics plasty versus internal fixation for unstable intertrochanteric fractures of of patients with loss to follow-up data for the outcome were femur: systematic review and meta-analysis. J Korean Med Sci. 2017;32: similar to patients with complete data for all significant pre- 858–867. dictors except for prefracture hip pain and functional status 11. Ravikumar KJ, Marsh G. Internal fixation versus hemiarthroplasty versus — related to the 24-month visit (see Table 1, Supplemental total hip arthroplasty for displaced subcapital fractures of femur 13 year results of a prospective randomised study. Injury. 2000;31:793–797. Digital Content 1, http://links.lww.com/JOT/B205). Some 12. Liao L, min Zhao J, Su W, et al. A meta-analysis of total hip arthroplasty relevant comorbidities (eg, anxiety and substance use disor- and hemiarthroplasty outcomes for displaced femoral neck fractures. der) or the indication and dose of opioids that patients were Arch Orthop Trauma Surg. 2012;132:1021–1029. using before were not collected in the trial, but for future 13. Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019;381:2199–2208. studies, these variables should be considered. The strengths 14. Burgers PT, Poolman RW, Van Bakel TM, et al. Reliability, validity, and of our study include using a large, international, representa- responsiveness of the Western Ontario and McMaster Universities tive sample that increases the generalizability of our findings, Osteoarthritis Index for elderly patients with a femoral neck fracture. J investigating several clinically relevant independent variables Bone Joint Surg Am. 2015;97:751–757. in the prediction model, and using a validated instrument for 15. McConnell S, Kolopack P, Davis AM. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC): a review of its pain assessment. utility and measurement properties. Arthritis Rheum. 2001;45:453–461. 16. Kapstad H, Hanestad BR, Langeland N, et al. Cutpoints for mild, moderate and severe pain in patients with osteoarthritis of the hip or knee CONCLUSION AND ready for joint replacement surgery. BMC Musculoskelet Disord. 2008;9: 55. FUTURE RECOMMENDATIONS 17. Goggins J, Baker K, Felson D. What WOMAC pain score should make a The results of this study provide insights into those at a patient eligible for a trial in knee osteoarthritis? J Rheumatol. 2005;32: higher risk of having moderate-to-severe pain at one and 2 540–542.

18. Xu BY, Yan S, Low LL, et al. Predictors of poor functional outcomes 32. McGilton KS, Chu CH. Factors influencing outcomes of older adults and mortality in patients with hip fracture: a systematic review. BMC after undergoing rehabilitation for hip fracture. J Am Geriatr Soc. Musculoskelet Disord. 2019;20:568. 2016;64:1601–1609. 19. Hosmer DW, Hosmer T, Le Cessie S, et al. A comparison of goodness‐of‐ 33. Strauss A, Rommelspacher Y, Nouri B, et al. Long-term outcome of total fi – t tests for the logistic regression model. Statistics Med. 1997;16:965 980. hip arthroplasty in patients with haemophilia. Haemophilia. 2017;23: fl 20. Akinwande MO, Dikko HG, Samson A. Variance in ation factor: as a 129–134. condition for the inclusion of suppressor variable (s) in regression anal- 34. Sylliaas H, Thingstad P, Wyller TB, et al. Prognostic factors for ysis. Open J Statistics. 2015;5:754. self-rated function and perceived health in patient living at home 21. Morris BJ, Sciascia AD, Jacobs CA, et al. Preoperative opioid use asso- three months after a hip fracture. Disabil Rehabil. 2012;34:1225– ciated with worse outcomes after anatomic shoulder arthroplasty. J Shoulder Elbow Surg. 2016;25:619–623. 1231. 22. Rakel BA, Blodgett NP, Zimmerman MB, et al. Predictors of postoper- 35. MacWilliam CH, Yood MU, Verner JJ, et al. Patient-related risk factors ative movement and resting pain following total knee replacement. Pain. that predict poor outcome after total hip replacement. Health Serv Res. 2012;153:2192–2203. 1996;31:623. 23. Kum E, Buckley N, de Leon-Casasola O, et al. Attitudes towards and 36. Nilsdotter AK, Lohmander LS. Age and waiting time as predictors of management of opioid-induced hyperalgesia: a survey of chronic pain outcome after total hip replacement for osteoarthritis. Rheumatology. practitioners. Clin J Pain. 2020;36:359–364. 2002;41:1261–1267. 24. Mercadante S, Ferrera P, Villari P, et al. Hyperalgesia: an emerging 37. Röder C, Parvizi J, Eggli S, et al. Demographic factors affecting long- iatrogenic syndrome. J Pain Symptom Manage. 2003;26:769–775. term outcome of total hip arthroplasty. Clin Orthop Relat Res. 2003;62– 25. Angst MS, Clark JD. Opioid-induced hyperalgesia: a qualitative system- 73. atic review. Anesthesiology. 2006;104:570–587. 38. Kessler S, Mattes T, Cakir B, et al. The impact of preoperative function 26. Nikolajsen L, Brandsborg B, Lucht U, et al. Chronic pain following total and pain on early patient-centred outcome after total hip arthroplasty. Z hip arthroplasty: a nationwide questionnaire study. Acta Anaesthesiol Orthop Unfall. 2007;145:563–567. – Scand. 2006;50:495 500. 39. Gibson SJ, Helme RD. Age-related differences in pain perception and 27. Rat AC, Guillemin F, Osnowycz G, et al. Total hip or knee replacement report. Clin Geriatr Med. 2001;17:433–456. v-vi. for osteoarthritis: mid-and long-term quality of life. Arthritis Care Res. 40. Pagé MG, Katz J, Escobar EMR, et al. Distinguishing problematic from – 2010;62:54 62. nonproblematic postsurgical pain: a pain trajectory analysis after total 28. Croft P, Dunn KM, Von Korff M. Chronic pain syndromes: you can’t knee arthroplasty. Pain. 2015;156:460–468. have one without another. Pain. 2007;131:237–238. 41. Sieberg CB, Klajn J, Wong C, et al. Predictors and trajectories of chronic 29. Baer M, Neuhaus V, Pape HC, et al. Influence of mobilization and weight bearing on in-hospital outcome in geriatric patients with hip postoperative pain following hip preservation surgery. J Hip Preserv – fractures. SICOT J. 2019;5:4. Surg. 2017;4:45 53. 30. Ariza-Vega P, Jiménez-Moleón JJ, Kristensen MT. Non-weight-bearing 42. Riddle DL, Wade JB, Jiranek WA, et al. Preoperative pain catastrophiz- status compromises the functional level up to 1 yr after hip fracture ing predicts pain outcome after knee arthroplasty. Clin Orthop Relat Res. surgery. Am J Phys Med Rehabil. 2014;93:641–648. 2010;468:798–806. 31. Mariconda M, Costa GG, Cerbasi S, et al. Factors predicting mobility 43. Roth ML, Tripp DA, Harrison MH, et al. Demographic and psychosocial and the change in activities of daily living after hip fracture: a 1-year predictors of acute perioperative pain for total knee arthroplasty. Pain prospective cohort study. J Orthop Trauma. 2016;30:71–77. Res Manag. 2007;12:185–194.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. t, = and n& P (NIH) ehta@ s from ns of this = 0.03). = 0.03), and a P P on behalf of d , a cantly higher odds c fi Lippincott Williams & — = 0.02) than those treated by Department of Health Research The main outcome was an d P cantly higher risk of PJI in patients ict of interest. fl fi fellowship-trained surgeons ( – Matthew K. Stein, MD, a and Samir Mehta, MD Sheila Sprague, PhD, a c Arthroplasty for hip fracture can be performed by all There was a signi Department of Surgery, University of Western Ontario, London, ON, s Health, employment from Global Research Solutions Inc, and b ’ Volume 34, Number 11 Supplement, November 2020 -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees fi RTICLE A treated by surgeons without0.01), fellowship training surgeons in with arthroplasty unknownsurgeons ( with fellowship no training fellowship training ( an ( arthroplasty-trained surgeon. There wereof signi being discharged tounderwent a surgery facility by rather apared than with surgeon home arthroplasty with in no patients fellowship who training com- Conclusions: Main Outcome Measurements: Results: orthopaedic surgeons with equivalent reoperation rates. Infection unplanned secondary procedure atincluded 24 months. death, Secondary serious outcomes (PJI), adverse dislocation, events, dischargedevices prosthetic postoperatively. disposition, joint and infection use of ambulatory a Bzovsky, MSc, nancial or material support from Wolters Kluwer Health fi fi So b J Orthop Trauma UPPLEMENT fellowship S ’ Derek J. Donegan, MD, MBA, the HEALTH Investigators d , Gregory T. Minutillo, MD, MPH, c a mizedtoeitherHAor Analysis of the HEALTH Database One thousand four hundred forty-one s Web site (www.jorthotrauma.com). ’ Patients were rando Emil H. Schemitsch, MD, FRCSC, This study compares outcomes for patients with Who Did the Arthroplasty? Hip Fracture Surgery Ryan D. DeAngelis, MD, Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; and c 50 years with low-energy hip fractures requiring surgical Eighty clinical sites across 10 countries. Retrospective review of HEALTH trial data. Department of Orthopaedic Surgery, University of Pennsylvania, Philadelphia, PA; -Aventis, outside the submitted work. S. Mehta reports research support from Acumed, LLC, board or committee member for AO Foundation, paid $

Canada; paid consultant and research support for Synthes, publishing royalties, Aphria, research support from FerringSano Pharmaceuticals, research support andpresenter personal or fees speaker from for Pendopharma,Johnson Bioventus, and Company, editorial board research or or support committee governing and member personal board for fee the for Orthopaedic Current Trauma Opinion Association, paid in consultant Orthopaedics, and paid paid presenter presenter or or speaker speaker for for Smith & DePuy Nephew, and Johnso Trauma Care Foundation, personalfrom fees Smith from & Pentopharm, Nephew,submitted personal personal fees fees work. from from S. Springer, Sano employment personal Sprague from fees reports McMaster from University, Stryker, editorial personal outside or fees the from governing submitted Swemac, board work. and personal M. for fees Bhandari BMS from reports Zimmer, Women research outside the support from Acumed, LLC, research support from Acumed, LLC, personal fees frompersonal Amgen fees Co, research from support DePuy, frompersonal board Biocomposites, fees or board or committee from committee memberAssociation, member ITS, for for editorial editorial the the or Canadian or Hip Orthopaedic governing Society, governing Association, board board board or for for committee the member the Orthopaedic for Journal Trauma the of Association International Orthopaedic International, Society for board Trauma, Fracture or board Repair, committee or member committee for member the for Osteosynthesis the Orthopaedic Trauma (1UM1AR063386-01), ZorgOnderzoek Nederland-medische wetensehappen (ZonMw)McMaster (17088.2503), Surgical Sophies Associates, Minde and Foundationanalysis, for Stryker or Orthopaedic Orthopaedics. Research, interpretation The of funding the sources data; had or no the role preparation, in review, design or or conduct approval of of the the study; manuscript. the collection, managemen Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada. pennmedicine.upenn.edu). Wilkins, and research support from Zimmer, outsidearticle the on submitted the work. journal The remaining authors report no con compared. THA groupstraining in was the ascertained initial retrospectively, and data outcomes set. were Surgeons S64 Accepted for publication AugustFrom 11, the 2020. Intervention: Patients/Participants: Design: Setting: Objectives: Mohit Bhandari, MD, PhD, FRCSC, patients intervention.

displaced femoral(HA) neck or total fracturesfellowship training. hip undergoing arthroplasty hemiarthroplasty (THA) by surgeons of different Reoperation Rates are Not Affected by Type of Training—An Reprints: Samir Mehta, MD, Department of Orthopaedic Surgery, University of Pennsylvania, 3737 Market St, Philadelphia, PA 19104 (e-mail: Samir.M D. J. Donegan reports paid consultant for DePuy and Johnson & Johnson Company, outside the submitted work. E. H. Schemitsch reports personal fees from The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health Copyright © 2020 WoltersDOI: Kluwer 10.1097/BOT.0000000000001931 Health, Inc. All rights reserved. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versio

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Hip Fracture Surgery Reoperation Rates prevention strategies and use of “care pathways” by arthroplasty- that there is no difference in outcomes at 2 years for HA or fellowship-trained surgeons may account for the lower risk of PJI THA after femoral neck fractures based on fellowship and higher rate of discharge to home. The authors advocate for the training. use of evidence-based infection prevention initiatives and standardized care pathways in this patient population. METHODS Key Words: displaced femoral neck fracture, total hip arthroplasty, Our study was a retrospective review of the data from hemiarthroplasty, fellowship training the HEALTH trial. Level of Evidence: Prognostic Level II. See Instructions for Both the primary surgeon and supporting institution Authors for a complete description of levels of evidence. were listed for every case in the database. Several resources were used to identify each surgeon’s subspecialty training. – (J Orthop Trauma 2020;34:S64 S69) The surgeon’s profile on their hospital web site was the primary source of information. Several surgeons were contacted INTRODUCTION directly through telephone and/or e-mail, when their informa- Femoral neck fractures are common injuries in the tion could not be found otherwise. Professional networking geriatric population, with an annual incidence projected to web sites, such as LinkedIn and Doximity, were used as well. surpass 6.26 million worldwide by 2050.1 This increasing The surgeons’ level of training was classified in to 5 incidence underscores the necessity for treatment strategies major groups as follows: trauma-fellowship trained, that optimize patient outcomes from a surgical and medical arthroplasty-fellowship trained, fellowship training outside perspective. Despite the overall success in the current treat- of trauma or arthroplasty, unknown fellowship status, or no ment of femoral neck fractures, devastating complications fellowship training. Fellowship was defined as subspecialized persist. These complications include, but are not limited to, training completed after completion of an orthopaedic surgery thromboembolic events, infection, implant failures prompting residency. If a surgeon had completed individual fellowships revision surgery, and one-year mortality of approximately in both trauma and arthroplasty or a combination fellowship 30%.2–5 focusing on trauma and arthroplasty, that surgeon was Nondisplaced femoral neck fractures are amenable to counted toward the arthroplasty group because the interven- closed reduction and percutaneous pinning, but displaced tion in the HEALTH trial was arthroplasty. Surgeons for fractures are typically treated with either hemiarthroplasty whom there was no available information regarding their (HA) or total hip arthroplasty (THA). The decision to perform training were included in the unknown fellowship group. A a HA or THA is driven by several factors, such as previous surgeon was only listed as having no fellowship if their train- hip pain, activity level, risk for dislocation, and surgeon ing record clearly stated their residency training and did not comfort or training. Although most displaced femoral neck identify a fellowship. fractures are still treated with HA, current studies have Participant demographics, baseline characteristics, demonstrated an increasing trend and a potentially improved major comorbidities, and outcomes of the population were outcome in THAs.6–9 summarized by fellowship training and treatment group. The field of orthopaedics has become progressively Descriptive statistics were used to summarize these data. more subspecialized over time with .90% of the US ortho- Means and SDs were used for continuous data, and categor- paedic residency graduates completing a fellowship as of ical data were presented as frequencies and percentages. 2013.10,11 Recent literature has studied femoral neck fractures Cox regression analyses were then performed to undergoing HA and compared outcomes based on surgeon investigate the association between fellowship training and fellowship training: arthroplasty, trauma, or general orthopae- 5 HEALTH data outcomes in all participants, using fellow- dics.12 The data show decreased operative time with arthro- ship training as the independent variable in each model. We plasty surgeons, higher complication rates with general also included randomized treatment as a covariate and orthopaedists, and higher mortality rates with trauma sur- surgeon as a random effect in each model. We used the geons. However, this was a single-center series that included following outcomes as the dependent variable in each only 298 hip fractures. respective model: (1) the HEALTH trial primary outcome The Hip Fracture Evaluation with Alternatives of THA of unplanned secondary hip procedures within 24 months of versus Hemiarthroplasty (HEALTH) trial found no difference initial surgery (yes vs. no), (2) hip dislocation (yes vs. no), (3) in outcomes in HA versus THA for femoral neck fractures at death (yes vs. no), (4) serious adverse events (yes vs. no), and 2-year follow-up.13 This prospective randomized multicenter (5) prosthetic joint infection (PJI) (yes vs. no). We also study included patients undergoing HA and THA performed performed the 5 abovementioned analyses including only by surgeons with various training backgrounds. There are no participants in the THA group and only participants in the HA studies examining the HEALTH data outcomes of patients group, separately. Randomized treatment was removed as a treated by fellowship-trained orthopaedic surgeons versus covariate when we performed analyses on the separate those treated by non–fellowship-trained orthopaedic sur- treatment groups. Results were reported as hazard ratios geons. This study serves to investigate a potential difference (HRs) with 95% confidence intervals (CIs). T tests were 2- in outcomes for HA or THA after femoral neck fractures tailed with alpha = 0.05. treated by fellowship-trained orthopaedic surgeons and non– We also performed 2 logistic regression analyses to fellowship-trained orthopaedic surgeons. Our hypothesis is investigate the association between fellowship training and 2

HEALTH data outcomes in all participants, using fellowship higher risk of PJI in the no fellowship group as compared to training as the independent variable in each model. We also the arthroplasty-trained group (HR 4.49, 95%, CI 1.20–16.81; included randomized treatment as a covariate and surgeon as P = 0.03). In the HA patients, there was no difference in PJI a random effect in each model. We used the following among the various fellowship groups (P . 0.05). outcomes as the dependent variable in each respective model: There were significantly higher odds of being dis- 1) discharge disposition (discharged to a facility vs. discharged to a facility rather than home postoperatively for the charged home) and 2) use of ambulatory devices postoper- patients who underwent surgery with a surgeon with no atively (use of an ambulatory device vs. nonuse of an fellowship training (OR 1.43, 95% CI 1.02–1.99; P = 0.03) as ambulatory device). Prefracture living status and prefracture compared to undergoing surgery with an arthroplasty– functional status were also included as covariates in these 2 fellowship-trained surgeon. This difference was not demon- models, respectively. We also performed the 2 abovemen- strated when looking at HA and THA groups separately (P . tioned logistic regression analyses including only patients in 0.05 for all). There was no difference in the use of assistive the THA group and only patients in the HA group, separately. devices postoperatively, when comparing the fellowship Randomized treatment was removed as a covariate when we training groups (P . 0.05 for all). These results are sum- performed analyses on the separate treatment groups. Results marized in Tables 1–3. were reported as odds ratios (ORs) with 95% CIs. T tests were 2-tailed with alpha = 0.05. All data analyses were conducted using R (version 4.0.0, R Foundation for Statistical Computing, Vienna, DISCUSSION Austria). We found no difference relative to surgical training when comparing the risk of unplanned secondary procedure, dislocation, death, and serious adverse events for patients RESULTS who underwent HA or THA for displaced femoral neck The HEALTH trial enrolled 1441 patients with 723 fractures. However, there was a significantly higher risk of patients randomized to HA and 718 patients randomized to PJI in patients treated by surgeons who completed fellow- THA. Of these 1441 procedures, 281 cases were performed ships outside of arthroplasty, surgeons with unknown fellow- by orthopaedic surgeons who completed a fellowship in ship training, and surgeons with no fellowship training than orthopaedic trauma, 394 cases were performed by orthopae- those treated by an arthroplasty-trained surgeon. Specifically, dic surgeons who completed a fellowship in arthroplasty, and in the THA group, being treated by a surgeon with no 139 cases were performed by orthopaedic surgeons who fellowship training was associated with a higher risk of PJI as completed fellowship in specialties other than trauma or compared to those treated by an arthroplasty-trained surgeon. arthroplasty. There were 308 cases completed by orthopaedic Previous literature has suggested a difference in surgeons who did not complete any further training after their numerous outcomes related to surgical training; our data did residency and 319 cases completed by orthopaedic surgeons not support this. Mabry et al reviewed 298 displaced femoral whose fellowship training status was unknown. Both HA and neck fractures treated with HA and studied outcomes THA cases were performed by each training group. There was comparing surgeons trained in arthroplasty, trauma, or no no difference in the body mass index, prefracture living fellowship (“generalists”). Their series demonstrated that setting, prefracture functional status, American Society of arthroplasty-trained surgeons had the shortest operative time, Anesthesiologists (ASA) score, or major comorbidities when generalists had the highest overall complication rate, and the patients were grouped by surgical training (P . 0.05 for trauma-trained surgeons had the highest mortality rate at all). There were differences among the groups about age, sex, one year.12 In addition, there was no difference in PJI in their and ethnicity (P , 0.05 for all). These demographic data are study. Our study also showed no difference in PJI for patients detailed in Supplemental Digital Content 1 (Appendix 1, undergoing HA for hip fracture. However, there was a signif- http://links.lww.com/JOT/B210). icantly higher risk of PJI in the THA patients (Table 3). There was no difference in the incidence of unplanned Considering there was no overall difference in reoperation secondary procedure, dislocation, death, or serious adverse rates among groups, PJIs in the THA patients were not caus- events among the different groups (P . 0.05 for all). When ing a significant rate of return to the operating room, and thus, analyzing the HA and THA patients separately, there was still the clinical implications of this finding in our study are uncer- no difference in these end points (P . 0.05 for all) between tain. Given this finding, it would be rather drastic to recom- the fellowship training groups. mend that THA for hip fracture be avoided by surgeons with There was a difference in PJI among the fellowship no fellowship training. Instead, our group recommends all training groups (P = 0.02). The patients treated by surgeons surgeons performing these procedures follow the most current who completed fellowships outside of trauma or arthroplasty infection prevention guidelines. (HR 4.06, 95% CI 1.46–11.33; P = 0.01), surgeons with There were several limitations in the Mabry et al study unknown fellowship training (HR 3.69, 95% CI 1.11– as it was a single-center series of only 298 patients with 35 12.27; P = 0.03), and surgeons with no fellowship training surgeons performing only hip HA. The HEALTH trial (HR 3.42, 95% CI 1.20–9.74; P = 0.02) had a higher risk of included nearly 5 times the number of patients in 80 centers PJI than those treated by an arthroplasty-trained surgeon. across 10 countries undergoing both HA and THA.13 With a When looking specifically at the THA patients, there was a much larger and more diverse study group of both patients

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 Hip Fracture Surgery Reoperation Rates

TABLE 1. Association Between Fellowship Training and TABLE 2. Association Between Fellowship Training and HEALTH Outcomes in all Arthroplasty Patients HEALTH Outcomes in HA Patients Outcome HR (95% CI) P Outcome HR (95% CI) P Unplanned secondary procedure Overall: 0.82 Unplanned secondary procedure Overall: 0.14 Trauma vs. arthroplasty 1.11 (0.64–1.95) Trauma vs. arthroplasty 0.78 (0.36–1.65) Other vs. arthroplasty 1.35 (0.80–2.26) Other vs. arthroplasty 1.39 (0.63–3.08) Unknown vs. arthroplasty 1.34 (0.70–2.57) Unknown vs. arthroplasty 0.91 (0.45–1.84) None vs. arthroplasty 0.81 (0.45–1.46) None vs. arthroplasty 0.34 (0.12–0.95) Dislocation Overall: 0.66 Dislocation Overall: 0.54 Trauma vs. arthroplasty 0.88 (0.38–2.04) Trauma vs. arthroplasty 0.50 (0.12–2.01) Other vs. arthroplasty 1.02 (0.47–2.22) Other vs. arthroplasty 0.66 (0.13–3.44) Unknown vs. arthroplasty 0.76 (0.25–2.32) Unknown vs. arthroplasty 0.60 (0.16–2.24) None vs. arthroplasty 0.82 (0.37–1.82) None vs. arthroplasty 0.58 (0.13–2.52) Death Overall: 0.85 Death Overall: 0.15 Trauma vs. arthroplasty 1.14 (0.74–1.75) Trauma vs. arthroplasty 1.39 (0.69–2.77) Other vs. arthroplasty 1.24 (0.83–1.86) Other vs. arthroplasty 2.12 (0.99–4.54) Unknown vs. arthroplasty 1.15 (0.68–1.97) Unknown vs. arthroplasty 1.85 (0.97–3.54) None vs. arthroplasty 1.14 (0.75–1.73) None vs. arthroplasty 1.49 (0.71–3.15) Serious adverse event Overall: 0.18 Serious adverse event Overall: 0.39 Trauma vs. arthroplasty 1.02 (0.80–1.31) Trauma vs. arthroplasty 0.87 (0.67–1.39) Other vs. arthroplasty 1.02 (0.80–1.30) Other vs. arthroplasty 1.31 (0.86–2.00) Unknown vs. arthroplasty 1.18 (0.87–1.60) Unknown vs. arthroplasty 0.90 (0.63–1.29) None vs. arthroplasty 0.98 (0.77–1.26) None vs. arthroplasty 0.86 (0.57–1.28) Prosthetic joint infection Overall: 0.02 Prosthetic joint infection Overall: 0.20 Trauma vs. arthroplasty 3.77 (0.91–8.37) 0.07 Trauma vs. arthroplasty 1.69 (0.31–9.26) Other vs. arthroplasty 4.06 (1.46–11.33) 0.01 Other vs. arthroplasty 2.57 (0.42–15.52) Unknown vs. arthroplasty 3.69 (1.11–12.27) 0.03 Unknown vs. arthroplasty 3.48 (0.75–16.14) None vs. arthroplasty 3.42 (1.20–9.74) 0.02 None vs. arthroplasty 2.03 (0.36–11.37)

OR (95% CI) P OR (95% CI) P Discharged to facility postoperatively Overall: 0.04 Discharged to facility postoperatively Overall: 0.15 Trauma vs. arthroplasty 1.34 (0.96–1.88) 0.09 Trauma vs. arthroplasty 1.49 (0.92–2.39) Other vs. arthroplasty 1.26 (0.91–1.74) 0.17 Other vs. arthroplasty 2.21 (1.21–4.03) Unknown vs. arthroplasty 1.46 (0.95–2.24) 0.08 Unknown vs. arthroplasty 1.57 (0.98–2.52) None vs. arthroplasty 1.43 (1.02–1.99) 0.03 None vs. arthroplasty 1.47 (0.89–2.44) Use of ambulatory devices Overall: 0.11 Use of ambulatory devices Overall: 0.73 postoperatively postoperatively Trauma vs. arthroplasty 0.45 (0.04–5.05) Trauma vs. arthroplasty 1.22 (0.07–19.92) Other vs. arthroplasty 0.90 (0.01–9.01) Other vs. arthroplasty 0.72 (0.04–11.79) Unknown vs. arthroplasty 0.48 (0.03–7.95) Unknown vs. arthroplasty 0.91 (0.01–13.09) None vs. arthroplasty 0.19 (0.02–1.63) None vs. arthroplasty 0.57 (0.06–5.79)

Other, fellowship not in trauma or arthroplasty; Unknown, unknown fellowship Other, fellowship not in trauma or arthroplasty; Unknown, unknown fellowship status; None, no fellowship training. status; None, no fellowship training. Significance = P , 0.05. Significance = P , 0.05. CI, confidence interval; HR, hazard ratio; OR, odds ratio. CI, confidence interval; HR, hazard ratio; OR, odds ratio. and surgeons, our results are more generalizable to the entire high-volume group had significantly lower rates of mortality, population. dislocation, and superficial infection. Revision surgery rates Although there is an overall scarcity of literature were significantly higher for the high-volume surgeons com- focusing specifically on fellowship training and correlation pared with no-volume surgeons; however, this may simply be with surgical outcomes in the treatment of hip fractures, there due to the fact that the high-volume surgeons were more have been numerous studies examining outcomes about comfortable in performing revision surgeries. Nonetheless, surgeon volume. Ames et al compared HA for hip fracture these findings were likely not observed in our study due to outcomes in surgeons in relation to their yearly volume. The the fact that more than 95% of the surgeons who participated study team divided the surgeons ranging from no volume (0 in the HEALTH trial met the threshold for surgical exper- cases/year) to high volume (25+ cases/year).14 When compar- tise.13,15 Very few, if any, of the HEALTH trial surgeons ing no volume with high volume, this group found that the are low-volume surgeons in regard to arthroplasty.

of stay.20 In addition, the data show that these pathways lead TABLE 3. Association Between Fellowship Training and to a higher rate of discharge to home.21 Arthroplasty surgeons HEALTH Outcomes in THA Patients using their “care pathways” despite these cases not being a Outcome HR (95% CI) P part of their elective practice may explain this group’s higher Unplanned secondary procedure Overall: 0.29 rate of discharge to home postoperatively. Given the data Trauma vs. arthroplasty 1.44 (0.63–3.28) supporting the use of these pathways for elective joint arthro- Other vs. arthroplasty 0.84 (0.24–2.92) plasty patients, the authors advocate for utilization of these Unknown vs. arthroplasty 1.81 (0.86–3.80) pathways in hip fracture patients as well. None vs. arthroplasty 1.38 (0.66–2.89) There are several limitations to this study. Given that Dislocation Overall: 0.98 the HEALTH trial was international, there were numerous Trauma vs. arthroplasty 1.10 (0.40–3.06) surgeons where details regarding surgical training were Other vs. arthroplasty 0.75 (0.16–3.45) particularly difficult to ascertain. This was challenging for Unknown vs. arthroplasty 1.25 (0.49–3.22) some of the surgeons in Europe, partially because of a None vs. arthroplasty 0.95 (0.37–2.44) language barrier and a lack of hospital web sites clearly Death Overall: 0.81 displaying the surgeon’s training history. This created a large Trauma vs. arthroplasty 1.06 (0.59–1.90) “unknown fellowship” group that may have included sur- Other vs. arthroplasty 0.59 (0.23–1.49) geons with subspecialized training. In addition, the surgeons Unknown vs. arthroplasty 0.86 (0.48–1.55) labeled as “no fellowship” may be somewhat misleading and None vs. arthroplasty 1.01 (0.61–1.69) imply that these surgeons are not specialized in one area or Serious adverse event Overall: 0.52 lack expertise. In this study, completion of fellowship was Trauma vs. arthroplasty 1.08 (0.76–1.54) used as an objective marker of “expertise” in a specific sub- Other vs. arthroplasty 0.95 (0.59–1.52) specialty. Regardless of completing a fellowship, many ortho- Unknown vs. arthroplasty 1.17 (0.85–1.63) paedic surgeons focus on specific areas of practice. Some None vs. arthroplasty 1.08 (0.79–1.46) more senior surgeons did not complete a fellowship because Prosthetic joint infection Overall: 0.04 fellowship training was not as common as it is today; these Trauma vs. arthroplasty 3.94 (0.92–16.81) 0.06 surgeons are still “experts” in their field. As such, this may Other vs. arthroplasty 4.54 (0.91–22.64) 0.06 have skewed the data because some surgeons may have been Unknown vs. arthroplasty 4.00 (0.98–16.33) 0.053 “misrepresented” if their training information could be not None vs. arthroplasty 4.49 (1.20–16.81) 0.03 obtained or if they did not complete a fellowship. P Our study supports that arthroplasty for hip fracture can OR (95% CI) be performed by all orthopaedic surgeons with no difference Discharged to facility postoperatively Overall: 0.26 in reoperation rates at 2 years. The authors advocate for the Trauma vs. Arthroplasty 1.25 (0.76–2.06) – utilization of the most current infection prevention strategies Other vs. Arthroplasty 0.92 (0.50 1.69) and standardized care pathways when treating these patients. Unknown vs. Arthroplasty 0.99 (0.62–1.57) – The surgeons performing these procedures, regardless of None vs. Arthroplasty 1.43 (0.93 2.22) fellowship training, should remain up-to-date on the current Use of ambulatory devices Overall: 0.13 postoperatively operative and postoperative recommendations in caring for Trauma vs. Arthroplasty 1.90 (0.01–20.09) these patients. Other vs. Arthroplasty 2.10 (0.03–19.87) Unknown vs. Arthroplasty 1.87 (0.05–18.91) None vs. Arthroplasty 1.89 (0.07–19.54) ACKNOWLEDGMENTS

Other, fellowship not in trauma or arthroplasty; Unknown, unknown fellowship The authors thank the HEALTH Investigators (http:// status; None, no fellowship training. links.lww.com/JOT/B234). CI - confidence interval, HR - hazard ratio, OR - odds ratio Significance = p , 0.05 REFERENCES 1. Cooper C, Campion G, Melton LJ. Hip fractures in the elderly: a worldwide projection. Osteoporos Int. 1992;2:285–289. 2. Voskuijl T, Neuhaus V, Kinaci A, et al. In-hospital outcomes after hemi- Another difference noted in our study was a higher arthroplasty versus total hip arthroplasty for isolated femoral neck frac- odds of discharge to a facility rather than home postopera- tures. Arch Bone Joint Surg. 2014;2:151–156. tively for cases performed by nonarthroplasty–fellowship- 3. Cummings SR, Melton LJ. Epidemiology and outcomes of osteoporotic – trained surgeons. This finding may be explained by the fractures. Lancet. 2002;359:1761 1767. “ ” 4. Belmont PJ, Garcia EJ, Romano D, et al. Risk factors for complications utilization of standardized care pathways by arthroplasty- and in-hospital mortality following hip fractures: a study using the trained surgeons with their hip fracture patients. Many National Trauma Data Bank. Arch Orthop Trauma Surg. 2014;134: arthroplasty surgeons use these evidence-based algorithms for 597–604. their elective arthroplasty cases to optimize patient care and 5. Mundi S, Pindiprolu B, Simunovic N, et al. Similar mortality rates in hip 16–19 fracture patients over the past 31 years. Acta Orthop. 2014;85:54–59. ultimately improve outcomes and reduce cost. Previous 6. Boniello AJ, Lieber AM, Denehy K, et al. National trends in total hip studies have substantiated the implementation of these path- arthroplasty for traumatic hip fractures: an analysis of a nationwide all- ways because they significantly decrease both cost and length payer database. World J Orthop. 2020;11:18–26.

7. Avery PP, Baker RP, Walton MJ, et al. Total hip replacement and 14. Ames JB, Lurie JD, Tomek IM, et al. Does surgeon volume for total hip hemiarthroplasty in mobile, independent patients with a displaced intracapsular arthroplasty affect outcomes after hemiarthroplasty for femoral neck fracture of the femoral neck: a seven- to ten-year follow-up report of a pro- fracture? Am J Orthop. 2010;39:E84–E89. spective randomised controlled trial. J Bone Joint Surg Br. 2011;93:1045– 15. Devereaux PJ, Bhandari M, Clarke M, et al. Need for expertise based 1048. randomised controlled trials. BMJ. 2005;330:88–92. 8. Stronach BM, Bergin PF, Perez JL, et al. The rising use of total hip 16. Barbieri A, Vanhaecht K, Van herck P, et al. Effects of clinical pathways arthroplasty for femoral neck fractures in the United States. Hip Int. in the joint replacement: a meta-analysis. BMC Med. 2009;7:32–42. 2020;30:107–113. 17. Chaurasia A, Garson L, Kain ZL, et al. Outcomes of a joint replacement surgical 9. Bishop J, Yang A, Githens M, et al. Evaluation of contemporary trends in home model clinical pathway. Biomed Res Int. 2014;2014:296302–296308. femoral neck fracture management reveals discrepancies in treatment. 18. Mertes SC, Raut S, Khanduja V. Integrated care pathways in lower-limb Geriatr Orthop Surg Rehabil. 2016;7:135–141. arthroplasty: are they effective in reducing length of hospital stay? Int 10. Ruddell JH, Eltorai AEM, DePasse JM, et al. Trends in the orthopaedic Orthop. 2013;37:1157–1163. surgery subspecialty fellowship match. J Bone Joint Surg Am. 2018;100: 19. Kim S, Losina E, Solomon DH, et al. Effectiveness of clinical pathways e139(1)–e139(9). for total knee and total hip arthroplasty: literature review. J Arthroplasty. 11. Horst PK, Choo K, Bharucha N, et al. Graduates of orthopaedic residency 2003;18:69–74. training are increasingly subspecialized a review of the American Board of 20. Iorio R, Clair AJ, Inneh IA, et al. Early results of Medicare’s bundled Orthopaedic Surgery Part II database. J Bone Joint Surg Am. 2014;97:869–875. payment initiative for a 90-day total joint arthroplasty episode of care. J 12. Mabry SE, Cichos KH, McMurtrie JT, et al. Does surgeon fellowship Arthroplasty. 2016;31:343–350. training inﬂuence outcomes in hemiarthroplasty for femoral neck frac- 21. Featherall J, Brigati DP, Faour M, et al. Implementation of a total hip ture? J Arthroplasty. 2019;34:1980–1986. arthroplasty care pathway at a high-volume Health system: effect on 13. Bhandari M, Einhorn TA, Guyatt G, et al. Total hip arthroplasty or hemi- length of stay, discharge disposition, and 90-day complications. J arthroplasty for hip fracture. N Engl J Med. 2019;381:2199–2208. Arthroplasty. 2018;33:1675–1680.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. t - 7- fi ing h Hip rom tice, f the utside ciation ersonal urnal of Lippincott f dforBMS ciation, paid — d , c a Department of Orthopaedic g American, publishing royalties, -Aventis, outside the submitted Department of Health Research d fi — ict of interest. = 0.02). fl P rmary, Aberdeen, United Kingdom AB25 fi Frede Frihagen, MD, Department of Surgery, University of Calgary, b e Sheila Sprague, PhD, Emil H. Schemitsch, MD, FRCSC, ve (54.7%) were diagnosed less than 6 weeks c fi g The incidence of symptomatic VTE in hip fracture xation was the only variable associated with increased fi 2520 hip fracture patients were included in the analysis. Volume 34, Number 11 Supplement, November 2020 nancial or material support from Wolters Kluwer Health fi David Neilly, FRCS Tr & Orth, RTICLE a A patients recruited to the 2 trials was 2.5%. Although over half of the Conclusions: Results: (46%) andmolecular postoperative weight heparin. (73%)to Treatment internal thromboprophylaxis with arthroplasty was compared low Sixty-four patients (2.5%)(1.1%)]. had Thirty- a VTEpostfracture [DVT: and 29 36 (45.3%)thousand (1.4%), nine more PE: hundred than 28 ninety-three 6boprophylaxis (79%) preoperatively weeks patients and postfracture. 2502 received (99%) One phylaxis throm- received postoperatively. thrombopro- The most common method of preoperative risk of VTE (hazard ratio 2.67, Department of Orthopedic and Trauma Surgery, OLVG, Amsterdam and Leiden e rmary, Aberdeen, United Kingdom; a Bzovsky, MSc, fi fi So b on behalf of the FAITH and HEALTH Investigators J Orthop Trauma a Rudolf W. Poolman, MD, PhD, UPPLEMENT c Department of Surgery, University of Western Ontario, London, ON, Canada. Marc Swiontkowski, MD, S h d , nancial or material support for the Journal of Bone and Joint Surgery c fi Division of Orthopaedic Surgery, Oslo University Hospital, Oslo, Norway; f Services Incorporated. M. Bhandari was also funded, in part, through the Early Research Award Program, which provided funding ’ David R. W. MacDonald, MRCS (Glas), Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, ON, Canada; c Subanalysis of the FAITH and HEALTH Trials The primary objective of this study was to determine Prism S. Schneider, MD, FRCSC, Using data from the FAITH and HEALTH trials, the s Health, employment from Global Research Solutions Inc., and employment from McMaster University, outside the submitted work. R. W. Poolman ’ Department of Trauma and Orthopaedic Surgery, Aberdeen Royal In Daniel Axelrod, MD, MSc (Cand), a www.jorthotrauma.com | Venous Thromboembolism in Hip Fracture Patients: A Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. nancial or material support from Saunders/Mosby-Elsevier, and publishing royalties, 2ZN (e-mail: [email protected]). Aventis, personal fees from personal Saunders/Mosby-Elsevier, fees fees from from Smith Swemac, and & personal Nephew, fees personal fees from from Zimmer, Springer, outside personal the fees submitted from work. Stryker, The p remaining authors report no con Zimmer, outside the submitted work.Pharmaceuticals, M. research Bhandari support reports and research personal supportwork. M. fees from Swiontkowski Acumed, reports from board LLC, Pendopharma, or research andeditorial committee support member research or from for support Aphria, governing the research American and board Orthopaedic support Association, personal consultant and from to fees Ferr publishing the from Minnesota royalties, Board Sano of Medical Prac presenter/speaker for Stryker, and paidWomen presenter/speaker for Synthes, outsidereports board the or submitted committee work. member forthe S. the submitted Sprague work. Dutch F. reports Orthopaedic Frihagen Association, editorial reports research personal support or fees from from governing Lima, Amgen boar and Co, research personal support fees from from Smith Link & Orthopaedics, o Nephew, personal fees from Synthes, and personal fees f Hamilton, ON, Canada).National The Institutes HEALTH of trial HealthFoundation was for (NIH) supported Orthopaedic Research, (1UM1AR063386-01), by McMaster ZorgOnderzoekstudy; Surgical research the Nederland-medische Associates, grants and collection, wetensehappen Stryker from management, (ZonMw) Orthopaedics. the analysis, The (17088.2503), funding or Canadian Sophies sources interpretation Institutes had Minde of no of role the in Health data; design Research or or the conduct (CIHR) o preparation, (MCT-90168), review, or approval of the manuscript. Health (1R01AR055267-01A1), Stichting NutsOhra11032), (SNO-T-0602-43), and the Physicians Netherlands Organisationfor for Health the Research FAITH and study Development as (80-82310-9 well as by a Canada Research Chair in Musculoskeletal Trauma, which is unrelated to the FAITH study (McMaster University, Calgary, Canada; University Medical Center, Leiden, the Netherlands; Orthopaedic Trauma, board or committee member forInternational, the board Orthopaedic Trauma or Association, editorial or committee governing board member for for the Orthopaedic the Trauma Asso Osteosynthesis and Trauma Care Foundation, personal fees from Pentopharm, personal fees from Sano fi Williams & Wilkins, outsidefrom the Biocomposites, board submitted or committee work. member for E.Society, the board H. Canadian Orthopaedic or Association, Schemitsch personal committee fees reports member from DePuy, personal for board the fees or International committee from member Society for Acumed, for the Fracture LLC, Repair, personal personal fees fees from from ITS, Amgen editorial Co, or research governing suppor board for the Jo Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada; Surgery, University of Minnesota, Minneapolis, MN; and and Iain M. Stevenson, FRCS Tr & Orth Mohit Bhandari, MD, PhD, FRCSC, Reprints: Iain M. Stevenson, FRCS Tr & Orth, DepartmentCopyright of © Trauma 2020 and Wolters OrthopaedicDOI: Kluwer Surgery, 10.1097/BOT.0000000000001939 Health, Aberdeen Royal Inc. In All rights reserved. S70 From the Accepted for publication August 11, 2020. Methods: Background: time to surgery, and method of fracture management. incidence of VTE, includingwere DVT determined. and A PE, multivariableto Cox and regression determine the analysis timing which was ofVTE, factors used including VTE were age, treatment associated for comorbidity, with thromboprophylaxis, increased risk of given, and identifying any factors associated with VTE. the incidence ofincluding symptomatic pulmonary venous embolism(DVT), in thromboembolism the (PE) hip fracture (VTE), anddetermining population. Secondary deep objectives timing included of vein VTE thrombosis diagnosis, VTE thromboprophylaxis P. S. Schneider reports editorial or governing board for the Canadian Journal of Surgery, board or committee member for the Canadian Orthopaedic Asso The FAITH Study was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of

Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 Downloaded from http://journals.lww.com/jorthotrauma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD31kf1ZzWr4XsjVhe354jOFxMp83x8XK6YzMp5zB4l+aQ= on 10/22/2020 J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 VTE in Hip Fracture Patients cases were diagnosed within 6 weeks of fracture, VTE is still determine the incidence of symptomatic VTE, including DVT prevalent after this period. The majority of patients received and PE, in this population of 2520 hip fracture patients. In thromboprophylaxis. Treatment with arthroplasty rather than fixation both studies, incidence of symptomatic VTE was recorded as was associated with increased incidence of VTE. a secondary outcome during inpatient stay and when participants returned for follow-up (either in person or Key Words: femoral neck fracture, venous thromboembolism, pul- through telephone) at 1 and 10 weeks, and at 6, 9, 12, 18, monary embolism, deep vein thrombosis and 24 months after surgery. This allowed for both incidence Level of Evidence: Therapeutic Level II. See Instructions for of VTE in this population and the timing of VTE diagnosis to Authors for a complete description of levels of evidence. be determined and presented as frequencies and percentages. In both studies, the baseline characteristics of all – (J Orthop Trauma 2020;34:S70 S75) patients included were recorded, including age, sex, body INTRODUCTION mass index, prefracture living setting, prefracture functional status, American Society of Anesthesiologists (ASA) classi- Venous thromboembolism (VTE), which comprises fication, and comorbidity. This allowed us to describe the deep vein thrombosis (DVT) and pulmonary embolism characteristics of patients who suffered a VTE and the (PE), is a common cause of morbidity and mortality in hip characteristics of patients who did not. The preoperative fracture patients.1 Although most cases are asymptomatic, in and postoperative VTE prophylaxis given for each patient the absence of thromboprophylaxis, postoperative VTE rates was also recorded, allowing us to determine the proportion of of up to 80% have been reported.2 Symptomatic VTE is less patients receiving VTE prophylaxis and the type of VTE common, particularly when thromboprophylaxis is used. In prophylaxis given. Method of fracture management and time the presence of thromboprophylaxis, the incidence of symp- from injury to surgery were also recorded in each trial. We tomatic postoperative DVT and PE has been reported from used descriptive statistics to summarize these results (fre- – 1.18%–6% and 0.25%–4.6%, respectively.3 6 quencies and percentages for categorical variables and means Symptomatic VTE is most often diagnosed within the and ranges for continuous variables). first 30 days after hip fracture,7,8 but patients may be at A multivariable Cox regression analysis was used to increased risk for up to 1 year after surgery.9 VTE is also determine which factors were associated with increased risk commonly present before surgery in hip fracture patients, of symptomatic VTE including: age, treatment for comorbid- – although this is usually asymptomatic.10 12 Anumberof ity (yes vs. no), thromboprophylaxis (medical vs. nonmedi- factors have been associated with an increased risk of cal), time from injury to surgery, and method of fracture VTE in hip fracture patients including delay to surgery, management (arthroplasty vs. internal fixation). Factors were extracapsular fracture pattern, and pulmonary disease.8,10,13 selected based on biological rationale, previous literature, and The increased risk of VTE after hip fracture is well recog- expert opinion. Statistical analysis was conducted using R nized and VTE prevention is emphasized by many organi- (version 3.6.1, R Foundation for Statistical Computing, zations including the National Institute for Health and Vienna, Austria). Results were presented as hazard ratios Clinical Excellence (NICE),14 the Scottish Intercollegiate (HRs), 95% confidence intervals (CIs), and P-values. A P- Guidelines Network, and American College of Chest value of less than 0.05 was considered to be statistically – Physicians.14 16 These bodies recommend different combi- significant. nations of chemical and mechanical VTE prophylaxis, but all agree that VTE prophylaxis should be instituted for hip fracture patients. RESULTS Analyses of the data from the FAITH (Fixation using In the FAITH trial, 1079 patients were randomized to Alternative Implants for the Treatment of Hip fractures) and treatment. At 24 months, 923 patients were alive, and the HEALTH (Hip fracture evaluations with Alternatives of complete follow-up was achieved for 844 (91%).17 In the Total Hip Arthroplasty vs. Hemiarthroplasty) trials both HEALTH trial, 1441 patients were randomized to treat- independently and combined were performed.17,18 The pri- ment.18 At 24 months, 1243 patients were alive, and complete mary objective was to determine the incidence of symp- follow-up was achieved for 1058 (85.1%). tomatic VTE, including DVT and PE, in the population Within the FAITH trial, 17 cases of VTE were recorded of hip fracture patients included in these studies. (1.6%) and 47 cases of VTE were recorded within the Secondary objectives were to determine the timing of HEALTH trial (3.3%), resulting in a total of 64 cases of VTE diagnosis, determine the proportion of patients receiv- symptomatic VTE across the 2520 hip fracture patients from ing VTE prophylaxis and the type of prophylaxis given, both studies and an incidence of VTE of 2.5%. The baseline describe the differences in patients who suffered a VTE characteristics of patients who suffered a VTE and those who and those who did not, and to determine which factors were did not suffer a VTE are summarized in Table 1. associated with VTE. In the FAITH trial, 12 cases of DVT were recorded (1.1%) with 24 cases of DVT in the HEALTH trial (1.7%). In total, 36 cases of symptomatic DVT were recorded across the METHODS 2520 hip fracture patients from both studies giving an In this preplanned study, we analyzed the data from the incidence of symptomatic DVT of 1.4%. The FAITH trial 1079-patient FAITH and the 1441-patient HEALTH trials to recorded 5 cases of PE (0.5%) and 23 cases of PE were

TABLE 1. Continued TABLE 1. Characteristics of Patients Who had a VTE ( ) Characteristics of Patients Who had a VTE Variable No VTE, N = 2456 VTE, N = 64 Variable No VTE, N = 2456 VTE, N = 64 Age, mean (SD) 75.7 (10.8) 77.3 (10.0) Medical only Sex, n (%) Male 829 (33.9) 24 (37.5) Heparin 158 (6.4) 3 (4.7) Female 1617 (66.1) 40 (62.5) Warfarin 95 (3.9) 2 (3.1) DOAC/other oral 91 (3.7) 3 (4.7) Body mass index, mean (SD) 24.8 (4.7) 25.4 (5.0) Prefracture living setting, n (%) acetylsalicylic acid 25 (1.0) 0 (0.0) Institutionalized 116 (4.7) 2 (3.1) LMWH 1804 (73.5) 51 (79.7) Not institutionalized 2340 (95.3) 62 (96.9) Mechanical only 81 (3.3) 0 (0.0) Medical and mechanical Prefracture functional status, n (%) Use of aid 581 (23.7) 19 (29.7) Heparin and mechanical 25 (1.0) 0 (0.0) Independent ambulator 1875 (76.3) 45 (70.3) Warfarin and mechanical 37 (1.5) 1 (1.6) DOAC/other oral and 13 (0.5) 0 (0.0) ASA classification, n (%) mechanical Class I/II 1302 (53.0) 34 (53.1) acetylsalicylic acid and 23 (0.9) 0 (0.0) Class III/IV/V 1154 (47.0) 30 (46.9) mechanical Receiving treatment for major LMWH and mechanical 175 (7.1) 6 (9.4) comorbidities, n (%) None 14 (0.6) 0 (0.0) Osteopenia 34 (1.4) 0 (0.0) Time from injury to surgery, mean 52.0 (75.6) 40.1 (31.8) Osteoporosis 129 (5.3) 9 (14.1) (SD) (hours) Lung disease 316 (12.9) 7 (10.9) Diabetes 328 (13.4) 10 (15.6) *In some cases, more than one type of thromboprophylaxis was used. ASA, American Society of Anesthesiologists; DOAC, direct oral anticoagulants; Ulcers or stomach disease 213 (8.7) 8 (12.5) LMWH, low-molecular-weight heparin; SD, standard deviation; VTE, venous throm- Kidney disease 112 (4.6) 5 (7.8) boembolism. Anemia or other blood disease 109 (4.4) 6 (9.4) Depression 259 (10.6) 5 (7.8) Cancer 117 (4.8) 4 (6.3) recorded in the HEALTH trial (1.6%). A total of 28 cases of Osteoarthritis, degenerative 204 (8.3) 3 (4.7) PE were recorded across the 2520 hip fracture patients from arthritis both studies giving an incidence of PE of 1.1%. Back pain 178 (7.3) 3 (4.7) The timing of VTE diagnosis is summarized in Table 2. Rheumatoid arthritis 47 (1.9) 6 (9.4) Across both studies, 3 (4.7%) cases of VTE were diagnosed Heart disease 644 (26.2) 11 (17.2) presurgery and 61 (95.3%) were diagnosed postsurgery. Of High blood pressure 1192 (48.5) 31 (48.4) VTE cases, 22 (34.4%) were diagnosed prehospital discharge Method of fracture management, n and 42 (65.6%) were diagnosed posthospital discharge. (%) Overall, 35 (54.7%) cases of VTE were diagnosed less than Internal fixation 1062 (43.2) 17 (26.6) 6 weeks postfracture and 29 (45.3%) were diagnosed more Arthroplasty 1394 (56.8) 47 (73.4) than 6 weeks postfracture. Preoperative thromboprophylaxis, n The proportion of patients given VTE prophylaxis (%)* and the type of prophylaxis used are summarized in Medical only Table 3. Across both studies, 1993 patients (79.3%) Heparin 143 (5.8) 2 (3.1) received thromboprophylaxis preoperatively and 2502 Warfarin 68 (2.8) 0 (0.0) (99.4%) received thromboprophylaxis postoperatively. DOAC/other oral 47 (1.9) 1 (1.6) The most common methods of preoperative thrombopro- acetylsalicylic acid 16 (0.7) 0 (0.0) phylaxis were low molecular weight heparin (LMWH) LMWH 1120 (45.6) 32 (50.0) (46.0%) and mechanical prophylaxis (18.0%). The most Mechanical only 433 (17.6) 9 (14.1) common method of postoperative thromboprophylaxis Medical and mechanical was LMWH (73.0%). Heparin and mechanical 23 (0.9) 0 (0.0) The results of the multivariable Cox regression analysis Warfarin and mechanical 10 (0.4) 1 (1.6) are summarized in Table 4. Advancing age, receiving treat- DOAC/other oral and 3 (0.1) 0 (0.0) mechanical ment for a comorbidity, type of postoperative thrombopro- acetylsalicylic acid and 2 (0.1) 0 (0.0) phylaxis, and increased time from injury to surgery were mechanical not associated with an increased incidence of symptomatic LMWH and mechanical 67 (2.7) 1 (2.9) VTE (P . 0.05). However, surgical treatment with arthro- None 509 (20.7) 18 (28.1) plasty was associated with an increased incidence of symp- Postoperative thromboprophylaxis, tomatic VTE within 24 months of hip fracture compared to n (%)* surgical treatment with internal fixation (HR 2.21, P = 0.02). Total hip arthroplasty and hemiarthroplasty versus internal

TABLE 2. Timing of VTE in FAITH and HEALTH Weeks From Hip Fracture FAITH, N = 17 HEALTH, N = 47 Overall, N = 64 Presurgery, n (%) 0 (0.0) 3 (6.4) 3 (4.7) Postsurgery, n (%) 17 (100.0) 44 (93.6) 61 (95.3) Prehospital discharge, n (%) 3 (17.6) 19 (40.4) 22 (34.4) Posthospital discharge, n (%) 14 (82.4) 28 (59.6) 42 (65.6) ,6 wk postfracture, n (%) 8 (47.0) 27 (57.5) 35 (54.7) $6 wk postfracture, n (%) 9 (53.0) 20 (42.5) 29 (45.3)

ﬁxation were associated with a 2.67 and 1.77 times increase is 2.5%, with the overall combined incidence of symptom- in VTE, respectively. atic DVT being 1.4%. The FAITH trial recorded 5 cases of PE (0.5%) and 23 cases of PE were recorded in the HEALTH trial (1.6%). Although over half of cases were DISCUSSION diagnosed within 6 weeks of the fracture, VTE is still This sub-analysis of data from the FAITH and prevalent after this period. The majority of patients received HEALTH trials demonstrates that the incidence of symp- thromboprophylaxis before and after surgery with LMWH tomatic VTE in hip fracture patients included in these trials given alone being the most common type of VTE

TABLE 3. VTE Prophylaxis in FAITH and HEALTH FAITH, N = 1079 HEALTH, N = 1441 Overall, N = 2520 Preoperative thromboprophylaxis, n (%)* Medical only Heparin 73 (6.8) 72 (5.0) 145 (5.7) Warfarin 37 (3.4) 31 (2.2) 68 (2.7) DOAC/other oral 29 (2.7) 21 (1.5) 50 (2.0) ASA 3 (0.3) 10 (0.7) 13 (0.5) LMWH 506 (46.9) 663 (46.0) 1169 (46.0) Mechanical only 280 (25.9) 177 (12.3) 457 (18.0) Medical and mechanical Heparin and mechanical 16 (1.5) 7 (0.5) 23 (0.9) Warfarin and mechanical 5 (0.5) 6 (0.4) 11 (0.4) DOAC/other oral and 2 (0.2) 1 (0.1) 3 (0.1) Mechanical ASA and mechanical 0 (0.0) 2 (0.2) 0 (0.0) LMWH and mechanical 41 (3.8) 27 (1.9) 2 (0.1) None 106 (9.8) 421 (29.2) 527 (20.7) Postoperative thromboprophylaxis, n (%)* Medical only Heparin 85 (7.9) 76 (5.3) 161 (6.3) Warfarin 39 (3.6) 58 (4.0) 97 (3.8) DOAC/other oral 52 (4.8) 29 (2.0) 81 (3.2) ASA 11 (1.0) 14 (1.0) 25 (1.0) LMWH 733 (67.9) 1122 (77.9) 1855 (73.0) Mechanical only 40 (3.7) 41 (2.9) 81 (3.2) Medical and mechanical Heparin and mechanical 18 (1.7) 7 (0.5) 25 (1.0) Warfarin and mechanical 10 (0.9) 28 (1.9) 38 (1.5) DOAC/other oral and 6 (0.6) 7 (0.5) 13 (0.5) mechanical ASA and mechanical 6 (0.6) 17 (1.2) 23 (0.9) LMWH and mechanical 120 (11.1) 61 (4.2) 181 (7.2) None 5 (0.5) 9 (0.6) 14 (0.6)

*In some cases, more than one type of thromboprophylaxis was used. ASA, acetylsalicylic acid; DOAC, direct oral anticoagulants; LMWH, low-molecular-weight heparin; VTE, venous thromboembolism.

days). Eighty-five percent of VTE events occurred within 5 TABLE 4. Association Between Thromboprophylaxis and VTE weeks of fracture, but patients presented up to 13 weeks Incidence Post Internal Fixation or Arthroplasty Surgery (N = after the injury. Despite the majority of cases presenting 2,231, 56 Events)* within the first 6 weeks in the current study, VTE was still Variable HR (95% CI) P prevalent after this period with 45% of cases presenting after Age (y) 0.99 (0.97–1.03) 0.94 6 weeks. Pedersen et al9 conducted a cohort study from the Receiving treatment for a 1.46 (0.76–2.78) 0.25 Danish Health registry comparing 110,563 hip fracture comorbidity yes vs. no patients to a cohort of 552,774 people from the general Postoperative thromboprophylaxis 1.27 (0.54–2.97) 0.58 population.bib9 The adjusted HR of VTE among hip frac- medical vs. not medical fi – ture patients was highest during the rst 30 days after hip Time from injury to surgery (h) 0.99 (0.986 1.002) 0.11 fracture (HR = 17.29, 95% CI: 14.74–20.28), but remained Method of fracture management 2.21 (1.15–4.26) 0.02 fi elevated during 31–365 days after hip fracture (HR = 2.13, arthroplasty vs. internal xation – Method of fracture management 95% CI: 1.95 2.32). Although this prolonged increased risk Total hip arthroplasty vs. internal 2.67 (1.33–5.38) 0.006 of VTE is consistent with this current study, our results fixation suggest that risk may be greater than previously Hemiarthroplasty vs. internal 1.77 (0.84–3.74) 0.14 demonstrated. fixation National Institute for Health and Clinical Excellence (NICE, UK), Scottish Intercollegiate Guidelines Network, *Patients with VTE occurring before the initial surgery were excluded from this analysis. CI, confidence interval; HR, hazard ratio; OR, odds ratio. and American College of Chest Physicians guidance documents recommend preoperative medical VTE prophylaxis with heparin unless contraindication exists, in addition to preoperative mechanical VTE prophylaxis for patients with hip fractures.14–16 These guidelines also recommend postop- prophylaxis given preoperatively and postoperatively. erative VTE prophylaxis with heparin, unless contraindica- Treatment with arthroplasty rather than fixation was the tion exists, in addition to mechanical VTE prophylaxis until only factor analyzed that was associated with increased the patient’s mobility is no longer significantly reduced. In incidence of symptomatic VTE in this population. our subanalysis, 61.3% of patients received preoperative med- The incidence of symptomatic VTE from the FAITH ical thromboprophylaxis, most commonly with LMWH and HEALTH trials is comparable to the incidences reported (46.1%), and 23.1% of patients received preoperative in other large studies. The ESCORTE study was a multicenter mechanical prophylaxis. 96.2% of patients received medical cohort study of 6860 hip fracture patients, 98% of whom VTE prophylaxis postoperatively, most commonly with received VTE prophylaxis with LMWH.5 The authors re- LMWH (80.2%), and 11% received postoperative mechanical ported a VTE incidence of 1.34% at 3 months of follow-up, prophylaxis. The lower rate of preoperative medical VTE with a 1.09% incidence of DVT and 0.25% incidence of PE. prophylaxis may be partially due to early surgery preventing The Pulmonary Embolism Prevention trial was a multicenter the use of anticoagulants, but our study suggests that mechan- randomized controlled trial comparing VTE thromboprophy- ical VTE prophylaxis is often not used preoperatively or post- laxis with aspirin or placebo, in addition to other thrombo- operatively despite advice from available guidelines. The use prophylaxis thought necessary in hip fracture patients.3 In the of mechanical VTE prophylaxis has been suggested to confer 6679 patients assigned aspirin, the authors reported a VTE additional benefit when used with medical prophylaxis and incidence of 1.6% at 35 days of follow-up, with a 1% inci- this is an area of practice that may be improved.19 dence of DVT and 0.7% incidence of PE. McNamara et al8 Regression analysis in this study demonstrated that age conducted a single-center cohort study of 5300 hip fracture was not associated with VTE incidence. This is consistent patients, all of whom received thromboprophylaxis with hep- with analyses performed by McNamara et al and the arin. The authors reported a 2.2% incidence of symptomatic ESCORTE study, both of which showed that age was not VTE at 1-year follow-up, with a 1.5% incidence of DVT and associated with postoperative symptomatic VTE.5,8 In addi- 0.7% incidence of PE. tion, a regression analysis by Shin et al11 demonstrated that Although the majority (55%) of symptomatic VTE age was not associated with preoperative DVT diagnosed by occurred within 6 weeks of hip fracture in the current study, multidetector CT venography. Our analysis also demonstrated asignificant percentage occurred after this period. This is in that patient comorbidity was not associated with VTE inci- keeping with the results of other large studies. Bjørnarået al dence. The analysis by Shin et al also showed that many conducted a single-center cohort study of 2420 hip fracture comorbidities including diabetes, hypertension, and cardiac patients with 6 months of follow-up.7 Thirty-six patients had disease were not associated with VTE, but suggested that a symptomatic DVT. The median time to diagnosis was 24 pulmonary disease was associated with increased risk of days (range 3–150 days). Thirty-nine patients had a symp- VTE. Our analysis demonstrated that time from injury to tomatic PE. The median time to diagnosis was 17 days surgery was not associated with increased risk of VTE. (range 1–173 days). McNamara et al8 reported a mean time Although there is limited evidence investigating the effect to VTE presentation of 24 days (range 3–91 days). Mean of time to surgery on postoperative VTE, several studies have time to DVT presentation was 25 days (range 4–91 days) suggested that delay increases the risk of preoperative and mean time to PE presentation was 20 days (range 3–81 VTE.10,20

Our regression analysis demonstrated that treatment 3. Prevention of pulmonary embolism and deep vein thrombosis with low with arthroplasty as opposed to fixation was associated with dose aspirin: pulmonary Embolism Prevention (PEP) trial. Lancet. 2002; 355:1295–1302. an over 2-fold increased incidence of VTE, and that this 4. Eriksson BI, Lassen MR. PENTasaccharide in HIp-FRActure Surgery increased risk was greater with total hip arthroplasty than Plus Investigators. Duration of prophylaxis against venous thromboem- hemiarthroplasty. No published studies have previously bolism with fondaparinux after hip fracture surgery: a multicentre, ran- compared VTE incidence for arthroplasty and fixation as domized, placebo-controlled, double blind study. Arch Intern Med. 2003; surgical treatments for a femoral neck fracture. Analyses by 163:1337–1342. 5. Rosencher N, Vielpeau C, Emmerich J, et al. Venous thromboembolism McNamara et al. and Shin et al demonstrated that intertro- and mortality after hip fracture surgery: the ESCORTE study. J Thromb chanteric and subtrochanteric fracture patterns were associ- Haemost. 2005;3:2006–2014. ated with increased incidence of postoperative VTE compared 6. Westrich GH, Rana AJ, Terry MA, et al. Thrombpembolic disease pro- to femoral neck fractures.8,11 Although this may suggest that phylaxis in patients with hip fracture: a multimodal approach. J Orthop fixation increases postoperative VTE incidence, these studies Trauma. 2005;19:234–240. 7. Bjornara BT, Gudmundsen TE, Dahl OE. Frequency and timing of clin- are not directly comparable to our study of femoral neck ical venous thromboembolism after major joint surgery. J Bone Joint fractures only. Surg Br. 2006;88:386–391. The main strength of our subanalysis is that our data are 8. McNamara I, Sharma A, Prevost T, et al. Symptomatic venous throm- sourced from 2 large-scale international studies, which both boembolism following a hip fracture: incidence and risk factors in 5300 had broad inclusion criteria with recruitment from a large patients. Acta Orthop. 2009;80:687–692. 9. Pedersen AB, Ehrenstein V, Szepligeti SK, et al. Excess risk of venous number of hospitals in diverse health care systems. These thromboembolism in hip fracture patients and the prognostic impact of results are therefore more likely to give a realistic picture of comorbidity. Osteoporos Int. 2017;28:3421–3430. symptomatic VTE in hip fracture patients. Although the use 10. Smith EB, Parvizi J, Purtill JJ. Delayed surgery for patients with femur of imaging to screen patients may identify many more cases and hip fractures-risk of deep venous thrombosis. J Trauma. 2011;70: of VTE, our study used symptomatic VTE as its outcome, E113–E116. 11. Shin WC, Woo SH, Lee SJ, et al. Preoperative prevalence of and risk which we believe makes our results more clinically relevant factors for venous thromboembolism in patients with a hip fracture: an to patients and current practice in most health care systems. indirect multidetector CT venography study. J Bone Joint Surg Am. A limitation of our subanalysis is the loss of patients 2016;98:2089–2095. over the 24 months of follow-up in both studies. However, 12. Xia ZN, Xiao K, Zhu W, et al. Risk assessment and management of the loss to follow-up in both studies was consistent with or preoperative venous thromboembolism following femoral neck fracture. 21 J Orthop Surg Res. 2018;13:291. better than other trials including hip fracture patients. Both 13. Shin WC, Lee SM, Suh KT. Recent Updates in the diagnosis and pre- studies also had unavoidable heterogeneity of variables such vention of venous thromboembolisms in patients with a hip fracture. Hip as patient positioning, surgical exposure, use of traction, type Pelvis. 2017;29:159–167. of anesthetic and physiotherapy and rehabilitation protocols, 14. National Institute for Health and Clinical Excellence. Venous which may have affected results. Thromboembolism: Reducing the Risk of Venous Thromboembolism (Deep Vein Thrombosis and Pulmonary Embolism) in Patients In conclusion, the incidence of symptomatic VTE in the Admitted to Hospital. Clinical Guideline 92 [Internet]. London, United hip fracture patients recruited to the FAITH and HEALTH Kingdom: National Institute for Health and Clinical Excellence; 2015. trials was 2.5%. Although over half of cases were diagnosed Available at: https://pubmed.ncbi.nlm.nih.gov/23346611/. Accessed within 6 weeks of surgery, VTE was still prevalent after this February 23, 2020. period with 45% of patients presenting after 6 weeks. The 15. Scottish Intercollegiate Guidelines Network. Management of Hip Fracture in Older People: A National Clinical Guideline [Internet]. majority of patients received thromboprophylaxis before and Edinburgh, United Kingdom: Scottish Intercollegiate Guidelines after surgery, but mechanical prophylaxis was not commonly Network; 2009. Available at: https://pdf4pro.com/view/part-of-nhs- used. Although several factors were examined, only treatment quality-improvement-scotland-67a27.html. Accessed February 13, 2020. with arthroplasty rather than fixation was associated with 16. Falck-Ytter Y, Francis CW, Johanson NA, et al. Prevention of VTE in increased incidence of VTE. orthopedic surgery patients: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of chest Physicians evidence- The high incidence of symptomatic VTE over 6 weeks based clinical practice guidelines. Chest. 2012;141:e278S–e325S. after surgery and the increased VTE risk with arthroplasty 17. Fixation using Alternative Implants for the Treatment of Hip fractures seen in our study have not been previously described. These (FAITH) Investigators. Fracture fixation in the operative management of results should be considered to guide future clinical decision hip fractures (FAITH): an international, multicentre, randomised con- making in hip fracture patients. trolled trial. Lancet. 2017;389:1519–1527. 18. HEALTH Investigators. Total hip arthroplasty or hemiarthroplasty for hip fracture. N Engl J Med. 2019;381:2199–2208. 19. Nam JH, Kim DH, Yoo JH, et al. Does preoperative mechanical pro- ACKNOWLEDGMENTS phylaxis have additional benefits in preventing postoperative venous — The authors thank the FAITH and HEALTH thromboembolism in elderly patients with hip fracture? retrospective case-control study. PLoS One. 2017;12:e0187337. Investigators (http://links.lww.com/JOT/B230). 20. Hefley FG Jr, Nelson CL, Puskarich-May CL. Effect of delayed admis- sion to the hospital on the preoperative prevalence of deep-vein throm- REFERENCES bosis associated with fractures about the hip. J Bone Joint Surg Am. 1. Todd C, Freeman C, Camilleri-Ferrante C, et al. Differences in mortality 1996;78:581–583. after fracture of hip: the East Anglian audit. BMJ 1995;310:904–908. 21. Bhandari M, Devereaux PJ, Swiontkowski MF, et al. Internal fixation 2. Geerts WH, Code KI, Jay RM, et al. A prospective study of venous throm- compared with arthroplasty for displaced fractures of the femoral neck. A boembolism after major trauma. NEnglJMed.1994;331:1601–1606. meta-analysis. J Bone Joint Surg Am. 2003;85-A:1673–1681.

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. l g e er rd or ee 2), g rom arch nn@ ict of fl Health uscript. script. All ma reports, a 50, treated with $ Department of Trauma Surgery, nancial or material support from e fi cial views of the National Institutes of fi Pieta Krijnen, PhD, Marc Swiontkowski, MD, b Lippincott Williams & Wilkins, outside the d s Health, employment from Global Research , ’ c — Johannes H. Hegeman, MD, PhD, Department of Orthopaedic and Trauma Surgery, Leiden b b Data were derived from the FAITH trial and Dutch Hip American, publishing royalties, Volume 34, Number 11 Supplement, November 2020 — RTICLE A Fracture Audit (DHFA). Patients with aa FNF, aged SHS,included. with Implant at failureregression was least model analyzed adjusted 3-month in forfracture living a follow-up age, setting sex, multivariable and data functional fracture logistic for mobility, displacement, and Anesthesiologists available, American pre- Class. Society were Methods: -Aventis, outside the submitted work. M. Swiontkowski reports board or committee Side? fi Bart L. Kaptein, PhD, Department of Surgery, University of Western Ontario, London, ON, Canada; and f d , J Orthop Trauma Investigators c on behalf of the Dutch Hip Fracture Audit Group and FAITH UPPLEMENT xation. a S fi Division of Orthopaedics, Department of Surgery, McMaster University, Hamilton, ON, Canada; c Mohit Bhandari, MD, PhD, FRCSC, Rudolf W. Poolman, MD, PhD, FACS, f nancial or material support from Wolters Kluwer Health a fi r patients with a left-sided -Aventis, personal fees from Saunders/Mosby-Elsevier, personal fees from Smith & Nephew, personal fees from fi linical relevance of the clockwise Sheila Sprague, PhD, c ft and right-sided FNFs after nal approval of the version to be published and are therefore accountable for the accuracy and integrity of this study. fi This study evaluated whethe Services Incorporated. M. Bhandari was also funded, in part, through the Early Research Award Program which provided funding for this study

nancial or material support for the Journal of Bone and Joint Surgery ’ ﬁ Department of Trauma Surgery, Leiden University Medical Center, Leiden, The Netherlands; Franka S. Würdemann, MD,

a www.jorthotrauma.com | a Bzovsky, MSc, Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN. Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada; R01AR055267-01A1. The content is solely the responsibility of the authors and does not necessarily represent the of (1R01AR055267-01A1), Stichting NutsOhra (SNO-T-0602-43), thePhysicians Netherlands Organisation for Healthand Research by a and Canada Development Research (80-82310-97-1103 was Chair in also Musculoskeletal supported Trauma by which is the unrelated National to Institute this study of (McMaster Arthritis University, Hamilton, and ON, Musculoskeletal Canada). and The FAITH Skin tria Diseases of the National Institutes of Health under Award Numb authors gave ﬁ g d support from Acumed,PendoPharma, LLC, and research researchmember support support for from and the personal Aphria, Americanroyalties, fees research Orthopaedic from Association, support Sano consultant from to Ferring the Pharmaceuticals, Minnesota Board research of support Medical and Practice, personal editorial or fees governing from board and publishin governing board for the Journalfor of the Orthopaedic Orthopaedic Trauma Trauma, Association boardPentopharm, International, or personal board committee fees or member committee forSpringer, from member personal the Sano for fees Orthopaedic from the Trauma Stryker, Osteosynthesis Association, personal and editorial fees Trauma or from Care governing Swemac, Foundation, boa and personal personal fees fees f from Zimmer, outside the submitted work. M. Bhandari reports rese International Radiostereometry Society, outsidemember the for submitted the work. DutchOrthopaedic Dr. Hip Surgery Hegeman Fracture & Rehabilitation, reports Audit,Co, outside research board the research or support submitted support committee from work. membercommittee from the E. member for AO H. Biocomposites, for the Schemitsch foundation, board the reports Dutch board or personal Hip Trauma or committee fees Society, Society, committ from board member and Acumed, or for editorial LLC, committee the or personal member fees governing Canadian for from board Orthopaedic the Amgen for International Association, the Society personal Geriatric for fees Fracture from Repair, DePuy, personal fees board from or ITS, editorial The Dutch Hip Fracture Audit (DHFA) GroupOrthopaedics, did outside not the receive submitted anySolutions work. funding. Inc., S. Sprague and reports employment editorial or from governing McMaster board for University, BMS outside Women the submitted work. B. L. Kaptein reports board or committee member for the Health. Research reported inoperates this as publication part was ofdesign also the or supported National conduct by of Institute The the for study; County Health the Durham Research collection, & Comprehensive management, Tees Clinical analysis, Valley or Research interpretation Comprehensive Network of Local in the Research England. data Network The or which funding the preparation, sources review, had or no approval role of in the man lumc.nl). Saunders/Mosby-Elsevier, publishing royalties, submitted work. I. B. Schippereditorial reports or editorial governing or board governing for boardinterest. injury, for and the president of British the Journal board of of Surgery, AO editorial Netherlands, or outside governing the board submitted for work. current The trau remaining authors report no con University Medical Center, Leiden, The Netherlands;

Ziekenhuisgroep Twente, Almelo-Hengelo, The Netherlands; Clockwise Torque of Sliding Hip Screws: Is There a Right Emil H. Schemitsch, MD, FRCSC, and Inger B. Schipper, MD, PhD, FACS rotational torque of the femoralthe neck rotational lag stability screw of in le a SHS, in relation to femoral neck fracture (FNF) treateda with higher a implant sliding hip failure screw rateThis (SHS) was than had performed patients to treated determine for the a c right-sided FNF. The FAITH trial was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of All authors made substantial contributions to the design of the study, and/or the analysis and interpretation of the work, did draft or revise the manu S76 Accepted for publication AugustFrom 11, the 2020. Objectives: So R. W. Poolman reports board or committee member for the Dutch Orthopaedic Association, research support from Lima, and research support from Link Reprints: Franka S. Würdemann, MD, Department of TraumaCopyright Surgery, © Leiden 2020 University Medical WoltersDOI: Center, Kluwer 10.1097/BOT.0000000000001934 Leiden, Health, The Inc. Netherlands All (e-mail: rights f.s.wuerdema reserved.

Results: One thousand seven hundred fifty patients were included, a similar instability in the bone-implant construct in left-sided of which 944 (53.9%) had a left-sided and 806 (46.1%) a right-sided FNFs compared with right-sided FNFs. Not only during FNF. Implant failure occurred in 60 cases (3.4%), of which 31 were insertion, but also during cyclic loading, eccentric forces may left-sided and 29 right-sided. No association between fracture side result in the femoral head and neck rotating around the lag and implant failure was found [odds ratio (OR) for left vs. right 0.89, screw.12,13 As the patterns of the resulting torques at the 95% confidence interval (CI) 0.52–1.52]. Female sex (OR 3.02, CI: fracture site are opposite for left and right hips, resistance 1.62–6.10), using a mobility aid (OR 2.02, CI 1.01–3.96) and a to these cyclic torques by the screw thread is also opposite displaced fracture (OR 2.51, CI: 1.44–4.42), were associated with as the screw thread is always clockwise. A dorsocaudally implant failure. applied load over the left femoral head may therefore cause Conclusions: micromovement with backward rotation of the femoral head This study could not substantiate the hypothesis that and subsequent loosening of the implant from the head.14–16 the biomechanics of the clockwise screw rotation of the SHS To illustrate this, a comparison could be made with the ped- contributes to an increased risk of implant failure in left-sided dles on a bicycle. The left and right peddle both move in a FNFs compared with right-sided fractures. forward cyclic manner while cycling and, equal to walking, Key Words: femoral neck fracture, implant failure, sliding hip go through a loaded and nonloaded phase. Opposed to the screw, side, left, right, clockwise torque, rotational stability, biome- right peddle with a “normal” clockwise thread attachment, a chanics counter-clockwise thread is used in the attachment of the left peddle to prevent loosening of the left peddle (and eventually (J Orthop Trauma 2020;34:S76–S80) falling off). The primary objective of this pre-planned analysis was INTRODUCTION to determine whether patients with a left-sided FNF have a Despite the fact that fixation techniques have been higher failure rate of SHS fixation than patients with a FNF modernized over the years, failure rates in femoral neck on their right side. fractures (FNFs), especially in the displaced ones, remain as high as 20%–40%.1,2 Some of these failures can be attributed to avascular necrosis of the femoral head, whereas in others, MATERIALS AND METHODS fracture collapse or true implant failure cause a breakdown of Data Sources the bone-implant construct.1–4 In many cases, however, it is not clear which of the underlying factors contribute most to Data were derived from the FAITH trial and The Dutch 17,18 construct failure. It is, therefore, not without reason that the Hip Fracture Audit (DHFA). The FAITH trial was a mul- fi FNF is still described as the “unsolved” fracture.5,6 ticenter, concealed randomized controlled trial comparing x- Patient and fracture-related factors that influence failure ation of FNF with cancellous screws versus a SHS. The rates have been thoroughly examined. The fracture type or current study concerns a preplanned secondary analysis of fi pattern seems to be one of the most critical determinants of patients of the SHS xation arm of the study. The DHFA is the failure of the bone-implant construct.7,8 Could it be, the Dutch nationwide multidisciplinary hip fracture audit in despite all the research on the topic, that we are overlooking which all patients with a hip fracture have been registered or underestimating the influences of the biomechanics of the since 2016. It is part of the Dutch Institute for Clinical implant? Auditing (DICA). A study using radiostereometric analysis (RSA) showed Patients that right-sided trochanteric fractures seemed to be more stable compared with left-sided trochanteric fractures after Patients aged $50 years, diagnosed with a FNF (OTA/ fixation with a sliding hip screw (SHS).9 It was hypothesized AO 31B) treated with a SHS, and with 3 months follow-up, that the clockwise rotational torque in the sagittal plane dur- data available were included in this study.19 Excluded were ing placement of the lag screw may contribute to a potentially patients with periprosthetic fractures or pathological fractures more unstable construct in left-sided trochanteric fractures and patients with no prefracture functional mobility. compared with right-sided fractures.10 It is during the screw insertion that clockwise rotational torque is imparted to the Outcome Parameters head–neck fragment, which in left-sided trochanteric fractures The primary outcome parameter was implant failure. In may cause flexion in the hip and extension of the fracture site, the FAITH trial, implant failure was defined as revision leading to a potentially unstable construct.10 These biome- surgery because of loosening or breakage of the implant or chanical theories and findings suggest that the risk of implant other reasons (mostly screw cut-out). Implant failure in the failure is higher in left-sided trochanteric fractures fixed with DHFA was defined as revision surgery because of migration a SHS compared with right-sided trochanteric fractures. A of the implant, loosening, or the implant breaking out. similar biomechanical theory has been described for lateral FNFs.11 To the best of our knowledge, no evidence or theo- Statistical Analysis ries have yet been published to support or deny this sugges- Variables recorded as “unknown” were recorded as tion for FNFs. missing. Missing values were assumed to be missing at ran- Although there are obvious biomechanical differences dom and were, therefore, left out from the analysis. The inde- between trochanteric hip fractures and FNFs, we hypothesize pendent sample t test was used for comparison of continuous

Copyright © 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Würdemann et al J Orthop Trauma Volume 34, Number 11 Supplement, November 2020 normally distributed variables, the Mann–Whitney U test for group of 1750 patients with a FNF treated with a SHS, 3.4% non-normally distributed variables, and the x2 test for cate- had a failure of the bone-implant construct. No association gorical variables. The odds ratio (OR) with 95% confidence between fracture side and implant failure was found. interval (CI) for the risk of implant failure for left-sided frac- Two studies performed by Mohan et al10 and Van tures versus right-sided fractures was calculated using multi- Embden et al9 mentioned a possible difference in rotational variable logistic regression analysis with adjustment for age, stability in left-sided and right-sided trochanteric fractures. fracture displacement (displaced or undisplaced), prefracture As a response, Pervez and Parker21 reviewed the incidence living situation (institutionalized or not institutionalized), pre- of screw cut-out of 1147 SHS in trochanteric fractures and fracture functional mobility (using a walking aid or able to did not find a difference in left-sided and right-sided fractures. ambulate without a walking aid), and American Society for Although the biomechanical theory for a potential difference in Anesthesiologists (ASA) Class (Class 1/2 or 3/4/5). P-values implant stability between left-sided and right-sided hips seems , 0.05 were regarded as statistically significant. Statistical plausible, we could not find any biomechanical studies on left- analysis was performed with R Studio Version 1.1.456 (R sided and right-sided differences in FNFs. Also, no additional Foundation for Statistical Computing, Vienna, Austria).20 evidence was found from 2 publications on large randomized hip fracture trials investigating implant failure of SHS because these did not include fracture side as a confounding factor in RESULTS In total, 1750 patients (1215 from the DHFA and 535 fi from the FAITH data set) ful lled the inclusion criteria and TABLE 1. Baseline Characteristics of Included Patients, per were analyzed. Baseline characteristics per data source are Data Source shown in Table 1. The fracture was left-sided in 944 (53.9%) Total DHFA FAITH P patients, and 806 (46.1%) patients had a right-sided FNF. The * mean age was 70.7 years (SD 12.6), 57.1% were women, who Total 1750 1215 535 were shown to have significantly more left-sided fractures Age, mean (SD) 70.7 70.2 71.9 0.02 (12.6) (12.9) (12.0) compared with men left-sided fractures (54.0% vs. 46.0%, Sex, n (%) P = 0.02). Sixty-nine percent of patients were ASA Class 1 Male 751 539 212 0.07 or 2. Before the fracture, 94.9% were not institutionalized, (42.9) (44.4) (39.6) and most patients did not use a mobility aid (80.9%). Of all Female 999 676 323 fractures, 681 were displaced (38.9%). There was a small but (57.1) (55.6) (60.4) statistically significant difference between the data sets for Prefracture living situation, most baseline characteristics. n (%) Implant failure within the first 3 months occurred in 60 Home 1561 1056 505 0.60 patients (3.4%). The incidence of implant failure was higher (94.9) (95.1) (94.4) in the FAITH trial than in the DHFA (7.9% vs. 1.5%, P , Institution 84 (5.1) 54 (4.9) 30 (5.6) 0.01; Table 1). The implant failure rates in left-sided hips (31/ Prefracture mobility, n (%) 944 = 3.3%) and in right-sided hips (29/806 = 3.6%) were Without aid 1347 931 416 0.03 similar. Fifty-five percent of implant failures occurred in (80.9) (82.4) (77.8) FNFs which were initially displaced. With aid 318 199 119 fi (19.1) (17.6) (22.2) No signi cant association between the fracture side and ASA class, n (%) implant failure was found. The OR of implant failure in left- 1,2 1180 840 340 ,0.01 sided hips compared with right-sided hips was 0.91 (95% CI: (68.6) (70.8) (63.6) 0.54–1.53, P = 0.72) in the univariable analysis and 0.89 3,4, and 5 541 346 195 (95% CI: 0.52–1.53, P = 0.66) in the multivariable analysis (31.4) (29.2) (36.4) (Table 2). Displacement, n (%) Female sex (univariable analysis: OR 2.54, 95% CI Undisplaced 1069 714 355 ,0.01 1.42–4.83, P , 0.01; multivariable analysis: OR 3.02, 95% (61.1) (58.8) (66.4) CI 1.52–6.10, P , 0.01), using a mobility aid before the Displaced 681 501 180 fracture (univariable analysis: OR 2.01, 95% CI 1.11–3.51, (38.9) (41.2) (33.6) P = 0.02; multivariable analysis: OR 2.02, 95% CI 1.01–3.96, Fracture side, n (%) P = 0.04), and having a displaced fracture (univariable anal- Right 806 551 255 0.40 (46.1) (45.3) (47.7) ysis: OR 1.97, 95% CI 1.17–3.32, P = 0.01; multivariable – , Left 944 664 280 analysis: OR 2.51, 95% CI 1.44 4.42, P 0.01) were asso- (53.9) (54.7) (52.3) ciated with implant failure (Table 2). Implant failure, n (%) No 1690 1197 493 ,0.01 DISCUSSION (96.6) (98.5) (92.1) Yes 60 (3.4) 18 (1.5) 42 (7.9) To the best of our knowledge, this is the first study that fi *Dutch Hip Fracture Audit (DHFA) versus FAITH trial. evaluates the difference in fracture xation stability using the SD, standard deviation. SHS in left-sided and right-sided FNFs. In our combined study

TABLE 2. Risk of Implant Failure of FNFs, Univariable, and Multivariable Logistic Regression Analysis Univariable Analysis Multivariable Analysis OR (95% CI) P OR (95% CI) P Side Left vs. right 0.91 (0.54–1.53) 0.72 0.89 (0.52–1.53) 0.66 Age 1.01 (0.99–1.02) 0.19 1.00 (0.98–1.02) 0.76 Sex Female vs. male 2.54 (1.42–4.83) ,0.01 3.02 (1.52–6.10) ,0.01 Prefracture living situation Institution vs. home 1.34 (0.40–3.38) 0.58 1.21 (0.34–3.35) 0.74 Prefracture mobility With vs. without aid 2.01 (1.11–3.51) 0.02 2.02 (1.01–3.96) 0.04 ASA class 3,4, or 5 vs. 1 or 2 1.18 (0.68–2.01) 0.54 1.04 (0.55–1.92) 0.89 Displacement Displaced vs. undisplaced 1.97 (1.17–3.32) 0.01 2.51 (1.44–4.42) ,0.01

CI, conﬁdence interval; OR, odds ratio.

their analyses.17,22 From these studies we may conclude that if mobility aid before the fracture had a significantly higher risk of present, rotational instability caused by the clockwise torqued implant failure. These findings are in line with previous studies. screw did not have clinical relevance in increased implant The strength of this study is the large number of failure in left-sided trochanteric fractures. Our study shows that patients included from both a clinical trial and a national this conclusion also seems to apply to FNFs. registry, so that the study group is representative of the The implant failure rate found in this study was 3.4%. general FNF patient population. A limitation of the combined Other studies found overall complication rates of SHS data sets was that some information was lost because varying between 5.2% and 16.7%. Tsang et al23 reported a variables were classified differently. For instance, the frac- surgical revision rate for mechanical causes of 2.8%. The tures in the FAITH data set were described according to the difference in implant failure rate between the FAITH trial Pauwels and Garden classification but were analyzed as data (7.9%) and the DHFA data (1.5%) may be explained displaced or undisplaced to match the fracture classification by an underestimation of the implant failure rate because of used in the DHFA. Also, data of only the first 3 months of the the retrospective design of the DHFA registry, whereas the available 24-month follow-up in the FAITH data set was used data derived from the FAITH trial were prospectively col- to match the length of follow-up to that in the DHFA data set. lected and thus of better quality. If implant failures were However, implant failure of SHS mostly occurred in the first missed in the DHFA, it is likely this would have occurred several months postsurgery,17 presumably when a patient equally in right-sided and left-sided hip fractures and would starts weight-bearing. The number of implant failures occur- therefore not have biased the results of our study. The differ- ring between 3 and 24 months in the FAITH data set was too ences in baseline characteristics of the patients from both data low to be analyzed separately. We considered the 3-month sources were small and of little clinical relevance. follow-up to be sufficient to evaluate the difference in implant Factors increasing the risk of implant failure have been failure in left-sided and right-sided SHS. thoroughly studied by several investigators. Failure because of Because of the retrospective design of this study, only nonunion or malunion is more often seen in displaced fractures variables collected in both the FAITH trial and the DHFA (eg, Garden type 3 and 4) and high shear angle fractures (eg, registry could be used in the analysis. Thus, we could not correct Pauwels type 3).8,24–26 In our study, fracture displacement con- for all risk factors known to be related to implant failure, such as firmed a statistically significant association with implant failure. bone density and body mass index.7 Details on the quality of the Associations of implant failure with other previously reduction, the quality and technique of the fixation (ie, if an found risk factors such as fracture pattern, bone quality,3,27,28 antirotation K-wire was used while inserting the screw), the accuracy of the fracture reduction and implant positioning,8,29 tip apex distance, and screw position were not available nor body mass index, smoking, concomitant injuries and comorbid- did the data contain details on the exact type of implant. ities, and therapy compliance (in weight-bearing) of the patient5,7 could not be substantiated in our study because these factors were not documented in both the data sources. Some of CONCLUSIONS these factors were collected in one of the data sources, but not The results of this study could not substantiate the the other, or vice versa. Other factors described in the literature hypothesis that the biomechanics of the clockwise screw rotation that may contribute to fixation failure are age and female sex. In of the SHS contributes to an increased risk of failure of the our study, we have found that female patients and patients using implant in left-sided FNFs compared with right-sided FNFs.

ACKNOWLEDGMENTS 14. Anderson AE, Ellis BJ, Maas SA, et al. Effects of idealized joint geom- fi The authors thank the Dutch Hip Fracture Audit etry on nite element predictions of cartilage contact stresses in the hip. J Biomech. 2010;43:1351–1357. (DHFA) Group Investigators (A. J. Arends, B. J. Blom, M. 15. Hoechel S, Alder M, Wirz D, et al. The human hip joint and its long-term van Eijk, M. J. Heetveld, M. van Heijl, M. C. Luyten, B. G. load intake - how X-ray density distribution mirrors bone strength. HIP Schutte, M. S. Slee-Valentijn, S. C. Voeten, and A. H. Calf) and Int. 2013;23:583–589. the FAITH Investigators (http://links.lww.com/JOT/B249). 16. Brown TD, Shaw DT. In vitro contact stress distributions in the natural human hip. J Biomech. 1983;16:373–384. REFERENCES 17. Fixation using Alternative Implants for the Treatment of Hip frac- 1. Queally JM, Harris E, Handoll HHG, et al. Intramedullary nails for tures (FAITH) Investigators, Nauth A, Creek AT, et al. Fracture fi extracapsular hip fractures in adults. Cochrane Database Syst Rev. xation in the operative management of hip fractures (FAITH): an 2014;9:CD004961. international, multicentre, randomised controlled trial. Lancet. 2017; – 2. Lu-Yao GL, Keller RB, Littenberg B, et al. Outcomes after displaced 389:1519 1527. fractures of the femoral neck. A meta-analysis of one hundred and six 18. Voeten SC, Arends AJ, Wouters MWJMJM, et al. The Dutch Hip published reports. J Bone Joint Surg Am. 1994;76:15–25. Fracture Audit: evaluation of the quality of multidisciplinary hip fracture 3. Heetveld MJ, Raaymakers ELFB, van Eck-Smit BL, et al. Internal fixa- care in The Netherlands. Arch Osteoporos. 2019;14:28. tion for displaced fractures of the femoral neck. Does bone density affect 19. Meinberg E, Agel J, Roberts C, et al. Fracture and dislocation classifi- clinical outcome? J Bone Joint Surg Br. 2005;87:367–373. cation compendium—2018. J Orthop Trauma. 2018;32:S1–S10. 4. Bhandari M, Devereaux PJ, Swiontkowski MF, et al. Internal fixation 20. Rstudio Team. RStudio: Integrated Development for R. RStudio, Inc. compared with arthroplasty for displaced fractures of the femoral neck: a Boston, MA: RStudio, 2019. meta-analysis. J Bone Joint Surg Am. 2003;85:1673–1681. 21. Pervez H, Parker MJ. Dynamic hip screw: does side make a difference? 5. Estrada LS, Volgas DA, Stannard JP, et al. Fixation failure in femoral Effects of clockwise torque on the right and left DHS (multiple letters) neck fractures. Clin Orthop Relat Res. 2002;399:110–118. [1]. Injury. 2002;33:93–94. 6. Li M, Cole PA. Anatomical considerations in adult femoral neck fractures: 22. Parker MJ. Sliding hip screw versus intramedullary nail for trochanteric how anatomy influences the treatment issues? Injury. 2015;46:453–458. hip fractures; a randomised trial of 1000 patients with presentation of 7. Sprague S, Schemitsch EH, Swiontkowski M, et al. Factors associated results related to fracture stability. Injury. 2017;48:2762–2767. with revision surgery after internal fixation of hip fractures. J Orthop 23. Tsang STJ, Aitken SA, Golay SK, et al. When does hip fracture surgery Trauma. 2018;32:223–230. fail? Injury. 2014;45:1059–1065. 8. Schep NWL, Heintjes RJ, Martens EP, et al. Retrospective analysis of 24. Garden RS. Low-angle fixation in fractures of the femoral neck. J Bone factors influencing the operative result after percutaneous osteosynthesis Joint Surg Br. 1961;43-B:647–663. of intracapsular femoral neck fractures. Injury. 2004;35:1003–1009. 25. Meinberg E, Agel J, Roberts C, et al. Fracture and Dislocation 9. van Embden D, Stollenwerck GANL, Koster LA, et al. The stability of Classification Compendium-2018. J Ortho Trauma. 2018; 32 Suppl 1; fixation of proximal femoral fractures. Bone Joint J. 2015;97-B:391–397. S1–S170. 10. Mohan R, Karthikeyan R, Sonanis SV. Dynamic hip screw: does side 26. Pauwels F. Der schenkelhalsbruch: ein mechanisches problem. Br J Surg. make a difference? Effects of clockwise torque on right and left DHS. 1936;23:874. Injury. 2000;31:697–699. 27. Viberg B, Ryg J, Overgaard S, et al. Low bone mineral density is not 11. Collinge CA, Mir H, Reddix R. Fracture morphology of high shear angle related to failure in femoral neck fracture patients treated with internal “Vertical” femoral neck fractures in young adult patients. J Orthop fixation. Acta Orthop. 2014;85:60–65. Trauma. 2014;28:270–275. 28. Kain MS, Marcantonio AJ, Iorio R. Revision surgery occurs frequently 12. Den Hartog BD, Bartal E, Cooke F. Treatment of the unstable intertro- after percutaneous fixation of stable femoral neck fractures in elderly chanteric fracture: effect of the placement of the screw, its angle of patients. Clin Orthop Relat Res. 2014;472:4010–4014. insertion, and osteotomy. J Bone Joint Surg Am. 1991;73:726–733. 29. Haynes RC, Pöll RG, Miles AW, et al. Failure of femoral head fixation: a 13. Kawaguchi S, Sawada K, Nabeta Y. Cutting-out of the lag screw after cadaveric analysis of lag screw cut-out with the gamma locking nail and internal fixation with the Asiatic gamma nail. Injury. 1998;29:47–53. AO dynamic hip screw. Injury. 1997;28:337–341.