Spinal Surgery and Patient Safety: A Systems Approach Abstract In every spinal procedure, identifying the specific patient, proper surgical site, and pathologic lesion for surgical intervention are crucial patient safety concerns. The 1999 Institute of Medicine report “To Err is Human” identified adverse events associated with surgery. Both the American Academy of Orthopaedic Surgeons and the North American Spine Society have had voluntary, systems- based programs in place for several years to address wrong patient, wrong procedure, and wrong site surgery. Beginning July 1, 2004, the Joint Commission on the Accreditation of Healthcare Organizations mandated that hospitals comply with the JCAHO Universal Protocol. In addition to surgical site marking, the protocol incorporates additional factors, such as calling a time out before skin incision to do a final systems check of patient identification, surgical site, records, imaging studies, equipment, and review of preoperative medication. C rucial to patient safety in any spinal surgery is ensuring that the correct patient, the proper surgi- cal site, and the extent of the patho- logic lesion are all properly and pre- cisely identified. Although the initial reaction to this assertion is likely to be that these issues are in- herent in the performance of any surgical procedure, on reflection, these processes constitute a prime example of a complex medical sys- tem 1,2 requiring a systems solution in order to reach a consistent and correct outcome. The American Academy of Orthopaedic Surgeons (AAOS) and the North American Spine Society (NASS) site-marking programs have evolved into more comprehensive systems-based inter- ventions to help surgeons avoid not only wrong site but also wrong pa- tient and wrong procedure surgery. Why a Systems Approach? The 1999 Institute of Medicine (IOM) report “To Err is Human” 1 contained the alarming assertion that “[a]t least 44,000 people, and perhaps as many as 98,000” Americans die in hospi- tals every year as the result of pre- ventable medical errors. This state- ment spurred an immediate reaction from the media, regulatory agencies, medical societies, and patients. At- tention was focused on so-called ad- verse events occurring during the course of treatment. The evaluation of contributing factors led to the de- scription of medicine as a prime model of a complex interactive sys- tem. As such, a systems methodol- ogy may be applied in order to frame both an analysis of existing problems and strategies for potential solutions. David A. Wong, MD, MSc, FRCSC Dr. Wong is Director, Advanced Center for Spinal Microsurgery, Presbyterian/ St. Luke’s Medical Center, Denver, CO. Dr. Wong is also Past President, North American Spine Society; Chairman, AAOS Patient Safety Committee; and Chairman, NASS Patient Safety Task Force. Dr. Wong or the department with which he is affiliated has received research or institutional support from Stryker, Arcus Centerpulse. Dr. Wong or the department with which he is affiliated has received nonincome support (such as equipment or services), commercially derived honoraria, or other non-research–related funding (such as paid travel) from Stryker. Dr. Wong or the department with which he is affiliated serves as a consultant to or is an employee of Stryker. Reprint requests: Dr. Wong, Denver Spine Center, Suite 4000, 1601 E 19th Avenue, Denver, CO 80218. J Am Acad Orthop Surg 2006;14:226- 232 Copyright 2006 by the American Academy of Orthopaedic Surgeons. 226 Journal of the American Academy of Orthopaedic Surgeons Systems Issues Involved in Spinal Surgery Several adverse events with general application to spinal surgery, ortho- paedic surgery, and neurosurgery were identified in the IOM report. Incidents were initially categorized as nonsurgical or surgical. Nonsurgical Adverse Events Medication error was the most frequent nonsurgical adverse event identified in the two studies forming the basis of the IOM report; 3,4 such errors represent approximately 19% of the total number of adverse events. 3,4 Interestingly, these inci- dents were considered “negligent” (therefore, likely preventable) in 17.7% of the New York cases 3 and in 35.1% of the cases in Colorado and Utah. 4 By a large margin, antibiotics were the largest medication class of concern. The most consistent oppor- tunity for problems regarding antibi- otics for spinal surgery is in the se- lection and timely administration of preoperative prophylactic antibiot- ics. A specific history of patient al- lergy to antibiotics should be ob- tained from the medical record and solicited from the patient before per- forming any surgical procedure. 3,4 The AAOS recently published an advisory statement on prophylactic antibiotics, outlining guidelines for medication selection and timing of preoperative antibiotic prophylaxis. 5 Although joint arthroplasty is the pri- mary focus of the advisory statement, the literature review encompasses in- formation relevant to surgery in gen- eral, particularly procedures involv- ing metal implants. A summary of the literature evaluation for various regimens is included in the support document. Several of the recommen- dations are firmly grounded in evidence-based medicine. Intravenous cephazolin is the general choice of an- tibiotic for the uncomplicated case. A lower rate of wound infection has been found when the antibiotic is ad- ministered before the skin incision is made. The suggested duration of pro- phylaxis is 24 hours. No additional benefit has been found when antibi- otics are administered for longer than 24 hours. Extended antibiotic use raises concerns of contributing to the development of antibiotic-resistant strains of bacteria. 5 Surgical Adverse Events The largest proportion of total ad- verse events identified in the IOM report were surgical in origin (47.7% of the total in New York and 44.9% in Colorado and Utah). 1 Reviews of medical records indicated that ap- proximately 17% of the adverse events in both series were consid- ered negligent. 3,4 Technical errors during surgery made up the largest category of sur- gical adverse events. The IOM report did not break down technical errors more specifically, but other studies have specified the most frequent technical complications in spine surgery, including dural tears, 6-11 bat- tered root syndrome, 12,13 incomplete decompression, 14-17 development of instability, 18 and wrong level sur- gery. 19 Antonacci and Eismont 11 sug- gested that the key to minimizing the risk of many types of adverse events—particularly those involving the neurologic system—is attention to the essential aspects of identifying anatomy and tissue handling. Avoiding Wrong Site Spinal Surgery: A Systems Solution The multi-tiered and complex na- ture of health care delivery necessi- tates a systems solution in order for the surgeon to appropriately identi- fy and expose the appropriate patient surgical site. The AAOS 20 and NASS 21 have developed voluntary programs—Sign Your Site (SYS) and Sign, Mark and X-Ray (SMaX), re- spectively—that outline a reason- able routine for distinguishing the appropriate patient and anatomic surgical site. Although both SYS and SMaX began as surgical site marking programs only, they now include more comprehensive systems solu- tions to comply with the recently mandated Joint Commission on the Accreditation of Healthcare Organi- zations (JCAHO) Universal Protocol for avoiding wrong patient, wrong site, and wrong procedure surgery. 22 Wrong Site Surgery: Historical Perspective In North America, the issue of wrong site surgery was first high- lighted in a closed claims analysis by the Canadian Medical Protective As- sociation, Canada’s national mal- practice insurance carrier. The data were additionally reviewed by a committee of the Canadian Ortho- paedic Association chaired by Paul Wright, MD. 23 In the United States, Terry Canale, MD, chaired the AAOS Task Force on Wrong Site Surgery, which performed a closed claims review of wrong site surgeries in the United States. 20 Additional data have come from analysis of the Sentinel Event program of the JCAHO. 24 The AAOS closed claims data contained several wrong site spinal surgeries, with 11 closed claims in- volving wrong level spinal sur- gery. 20 All claims were found to be indefensible. The 11 cases all con- cerned a one-level diskectomy for herniated lumbar disks. In 10 of the 11 cases, surgery was performed one level higher than intended. These evaluations led the Canadi- an Orthopaedic Association to de- velop the Operate Through Your Ini- tials program and the AAOS to develop the Sign Your Site initiative. NASS developed a similar program (SMaX) specifically related to the spine. These initial proposals repre- sented the first steps in a systems so- lution to the problem of wrong site surgery. In Canada, the Canadian David A. Wong, MD, MSc, FRCSC Volume 14, Number 4, April 2006 227 Medical Protective Association has noted a 65% reduction in the num- ber of threats of lawsuits and actual suits filed related to wrong site sur- gery since the inception of the vol- untary Operate Through Your Ini- tials program. 23 Sign, Mark and X-Ray Program The SMaX program 21 contains three key elements performed in three steps: (1) reviewing the medical records, including imaging studies, to confirm patient identification, the site of surgery, and the procedure to be performed; (2) specifying the side of surgery as part of the site marking process; and (3) confirming the spinal level of pathology intraoperatively (Table 1). The steps involved in checking these key elements are: (1) completing the preoperative checklist, (2) signing the surgeon’s initials on the patient’s extremity or trunk to identify the side to be ap- proached as well as to serve as an ini- tial approximation of level, and (3) in- traoperatively marking a fixed, bony landmark and taking a radiograph for final identification of the spinal level. The SMaX checklist (Figure 1), which is analogous to the preflight checklist used in aviation, provides an organized method of reviewing the relevant information necessary for a successful and safe surgery. For spinal surgery, the review includes correlating the patient’s spoken and armband identification with various medical records to confirm the pa- tient’s identity. Clinical notes and imaging studies are checked to cor- roborate the surgical procedure to be performed. These facts should be checked against the surgical consent and verbally verified with the pa- tient. The surgeon can then “sign” the checklist to fulfill the first step in the SMaX program. Currently, this first step is made particularly complex by the obliga- tion to coordinate information from multiple sources (eg, office/clinic notes, imaging reports, electromyo- gram and nerve conduction velocity studies) and geographic locations (eg, office/clinic, imaging center, store- front urgent care facility, hospital). Recent movement toward electron- ic medical records and digital imag- ing has made this task even more daunting. A reliable system must be established at each treatment center to ensure that relevant records are available at the time of surgery. The second step in the SMaX pro- gram, identifying the side of surgery, is literally a “sign” intervention. The trunk or neck on the side of the surgical procedure is marked with the surgeon’s initials to establish the definitive side of surgical interven- tion. This usually correlates with the patient’s side of dominant symp- toms. With a bilateral procedure, ini- tialing the general location on either side will suffice. The last component of SMaX is performed in the operating room. Fi- nal documentation of levels is con- firmed by radiographic examination. A radiopaque marker (eg, needle, towel clip, Kocher clamp) is used to identify a bony landmark (eg, pedicle, spinous process). A needle positioned just below the facet will generally lo- calize the top of the pedicle. When the spinous process is employed, cor- relation with preoperative plain ra- diographs will help localize the vari- able superior/inferior relation of the spinous process. This is particularly true in the cervical and thoracic spine. The appropriate interlaminar space may then be anatomically re- lated to the bony landmark. Identifying the correct level in the transitional lumbar and sacral verte- brae is particularly challenging. The anteroposterior (AP) view may not easily correlate with either the lateral radiograph or the magnetic resonance image (MRI). In the presence of anomalies, the surgeon should define a numbering system that is consis- tent with both advanced imaging and plain radiography in order to accu- rately define where surgery should be performed. Ideally, this adjudication should take place before entering the operating room. Intraoperative con- sultation with a radiologist or repeat radiographs may be needed to ensure accuracy. In the presence of inade- quate surgical findings, the surgeon should reassess and review the preop- erative imaging and consider taking repeat radiographs with markers lo- cated at pedicles or over disk spaces. Identification in the thoracic spine presents unique difficulties. Before any surgical procedure, the surgeon must ensure that the pathologic level is accurately identified. This is most reliably performed by obtaining an MRI spanning the thoracic pathol- ogy, extending either up to the base Table 1 The Identification Process According to the North American Spine Society Sign, Mark and X-ray (SMaX) Program Correct patient Verbal name check with patient Confirm ID bracelet Correlate medical records: written notes, laboratory reports, imaging studies, consent forms Proper side of surgery With patient participation, when feasible Surgeon or credentialed provider (eg, fellow, resident, physician assistant), who is a member of the patient’s surgical team, marks skin for first site marking with indelible ink, on the neck or torso, in the general area of surgery, and on the proper side of surgery (when a unilateral approach is planned) Appropriate spinal level Final localization of surgical level intraoperatively via radiograph, bony landmark, metal marker Spinal Surgery and Patient Safety: A Systems Approach 228 Journal of the American Academy of Orthopaedic Surgeons of the skull or down to the sacrum. Levels may then be counted directly. Intraoperatively, when the length of a radiograph is limited, two lateral films (lumbar and thoracic) may be necessary in order to correlate double markers. One marker is placed around L1 so that the exact level may be identified on the lumbar film by counting up from the sacrum. On the thoracic film, both the marker placed around L1 and the second marker at the estimated surgical site should be visible. Direct correlation of levels may be made by integrating informa- tion from the two films. Using an AP radiograph to count up from the lowest rib is also an ac- curate option if the lowest rib level has been recognized on preoperative imaging. It is important to remem- ber that, in the thoracic spine, the spinous process may angle down suf- ficiently to be opposite the body and pedicle of the level below. The ability to obtain an AP radio- graph intraoperatively may be an ad- vantage in any situation in which the surgeon anticipates difficulty in obtaining good-quality lateral imag- ing (eg, in the obese patient). One should plan ahead to use a radiolu- cent graphite operating frame, which offers the option of both AP and lat- eral views. In the cervical spine, lateral radio- graphs are the most commonly used in both anterior and posterior ap- proaches for correlating a marked level with reliably identified upper cervical segments. Where the shoul- ders would obscure a marker in the lower cervical area, the exposure may have to be extended up to allow marking of an unobstructed level. Direct counting down to the point of pathology may then be dependably performed. JCAHO Sentinel Event Program: Not Just Wrong Site Since 1995, the JCAHO Sentinel Event program has monitored the in- cidence of “wrong site” surgeries. From its inception through 2003, 278 incidents of wrong site surgery were reported. 24 On further review of the data, it was found that this cate- gory, wrong site, in fact comprised several clinically important subcat- egories (Figure 2). Root cause analy- sis of the Sentinel Events showed that 59% were wrong side proce- dures. Somewhat surprisingly, 12% were the wrong patient. In 10%, the wrong procedure was performed, and 19% represented other wrong sites (eg, wrong digit, wrong finger joint). Based on these data, the JCAHO mandated that hospitals comply with their Universal Protocol 22 for patient surgical site identification beginning July 1, 2004. JCAHO Universal Protocol Elements of the JCAHO Universal Protocol include: (1) verifying patient Figure 1 Sign, Mark & X-ray (SMaX) preoperative checklist. (© 2001 North American Spine Society. Reprinted with permission from NASS.) David A. Wong, MD, MSc, FRCSC Volume 14, Number 4, April 2006 229 identity, (2) marking the surgical site, and (3) taking a brief time out immediately before skin incision for a surgical team conference to recon- firm patient identity; surgical site; planned surgical procedure; adminis- tration of preoperative medications (eg, antibiotics); and the presence of appropriate medical records, imag- ing, and equipment (Table 2). Both the AAOS and NASS had consider- able input in the development of the JCAHO Universal Protocol, and many of the elements of the SYS and SMaX programs were incorporated into the JCAHO methodology. Both the AAOS and NASS have endorsed the Universal Protocol. Site Marking is Not Enough: Why a Systems Approach is Needed In the following example, site mark- ing alone likely would not have pre- vented an adverse event. Figure 3, A is a sagittal, T2-weighted MRI scan of a man whose chief complaint was sciatica. The radiologist identified a rudimentary disk at the lumbosacral junction, judged it to be a transition- al level, and designated this segment as S1-2. According to this numbering system, the herniated disk is at the next level cephalad, L5-S1. Using a grease pencil, the radiologist num- bered the vertebrae accordingly on the film. Unfortunately, the official report did not contain a description of the transitional level and the numbering decisions that had been made. The report conclusion was “herniated disk L5-S1.” On the AP plain radiograph (Figure 3, B), the T12 rib and a right sacrotransverse joint at the transitional level can be clearly seen. The transitional level is otherwise segmented, with a well- defined interlaminar space. Desig- nating the lowest rib-bearing verte- brae as T12, the transitional segment would be L5-S1 (not S1-2, as designated on the MRI report); there- fore, the level of the herniated disk would be the true L4-5 (not L5-S1, as on the report). The systems failure occurred when the MRI, which was done at an outside imaging center, was not available in the operating room at the time of surgery. Thus, nomenclature of levels and location of pathology could not be correlated between the MRI scans and plain radiographs while fulfilling the preoperative checklist. Having retrospectively reviewed the nomenclature issues illustrated in Figures 3, A and B, the concerns raised by Figure 3, C—the intraoper- ative localization radiograph—can be anticipated. The surgeon’s dictat- ed surgical report indicates that a laminotomy/diskectomy was per- formed at the L5-S1 level, which is the location of the herniated disk de- scribed in the written MRI report from the radiologist. However, the intraoperative radiograph shows the Taylor retractor and the probe mark- ing the disk at the transitional level. In fact, were the radiologist’s num- bering scheme to be used, this is the S1-2 segment. The herniated disk is actually at the level above. Surgical site marking alone would not have avoided this wrong level surgery. This example represents an instance of systems failure within the framework of medicine as a complex interactive system. Simi- larly, site marking would not avoid a Figure 2 Joint Commission on Accreditation and Healthcare Organizations analysis of Sentinel Events reported under the wrong site surgery category. Table 2 Elements of the Joint Commission on the Accreditation of Healthcare Organizations Universal Protocol Patient identification Verbal check with patient, if possible Check armband Confirm medical records Surgical site marking Sign initials near the incision site so the mark is visible after prep/drape Time out In the operating room, prior to incision, members of the surgical team confirm patient identification; surgical site; planned surgical procedure; administration of preoperative medications; presence of appropriate medical records, imaging, and equipment Spinal Surgery and Patient Safety: A Systems Approach 230 Journal of the American Academy of Orthopaedic Surgeons wrong site surgery if the wrong pa- tient were brought into the operat- ing room or if there were a mix-up in the procedure to be performed. Such concerns are precisely why NASS and the AAOS have added a surgical checklist to their SMaX and SYS programs. The time out in the JCAHO Universal Protocol serves a similar function. The additional steps represent opportunities to rec- tify issues such as patient identifica- tion, availability of appropriate med- ical records and imaging studies, confirming patient medication aller- gies, and ensuring that antibiotics have been given and that required equipment is available for the case. Table 3 summarizes the issues to be addressed in both preoperative hold- ing and the operating room. PRE- PARE, the acronym developed by John M. Purvis, MD, was the win- ning submission in the 2004 AAOS Patient Safety Tip contest. 25 Table 3 Systems Approach to Avoiding Wrong Patient, Wrong Site, Wrong Level Surgery Preoperative Holding (PREPARE) P Procedure/ Plan Discuss everything with the operating room team R Radiology Images in the room, equipment requested E Equipment Implants and supplies available and in working order P Patient Correct patient: check ID bracelet, surgical site marked A Anesthesia Be aware of the surgical plan, positioning, special needs (eg, hypotension) R Rx given Prophylactic antibiotics, patient-specific medications E Exceptions Any special considerations Surgery/Operating Room (TOR) TO Time out Check patient identity, records, imaging, surgical procedure, site marking R Radiograph Confirm level with an intraoperative radiograph Figure 3 A, Sagittal T2-weighted lumbar MRI scan of a man whose chief complaint was sciatica. The lowest disk level was designated S1-2. By this numbering system, the herniated disk is at L5-S1. B, Anteroposterior scout radiograph of the lumbar spine. The twelfth thoracic rib (A) is visible. There are five true lumbar vertebrae. The transitional level with (B) a right sacrotransverse joint (congenital transverse process—sacral ala articulation) is at the true L5-S1 level. C, Intraoperative lateral localization radiograph demonstrating the Taylor retractor (A) and the tip of the mosquito clamp (B) at the transitional level (true L5-S1, but S1-2 by the numbering on the MRI scan in panel A). The herniated disk is actually at the level above. David A. Wong, MD, MSc, FRCSC Volume 14, Number 4, April 2006 231 Summary Medicine has become an increasing- ly complex interactive system, with accompanying systems problems in- herent in coordinating and clarifying sometimes-conflicting clinical infor- mation. The AAOS and NASS have found that their initial patient safe- ty focus on surgical site marking re- quired expansion to include a broad- er systems approach in order to address the multiple issues inherent in wrong site surgery. A systems so- lution, such as the updated and more comprehensive AAOS Sign Your Site and NASS Sign, Mark and X-ray pro- grams, provides surgeons with the tools to reduce the incidence of wrong site surgery. Following the steps outlined in those programs also ensures compliance with the JCAHO Universal Protocol. References Citation numbers printed in bold type indicate references published within the past 5 years. 1. Why do er rors happen? in Institute of Medicine: To Err is Human: Building a Safer Health System. Washington, DC: National Academy Press, 1999, pp 49-68. Available at http://www.nap.edu/books/03090683 71/html/49.html. Accessed Novem- ber 21, 2005. 2. WongDA: It’s more than humanerror: A systems approach to patient safety. SpineLine May/June 2002;3:20-21. 3. Brennan TA, Leape LL, Laird NM, et al: Incidence of adverse events and negli- gence in hospitalized patients: Results of the Harvard Medical Practice Study I. N Engl J Med 1991;324:370-376. 4. Thomas EJ, Studdert DM, Burstin HR, et al: Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care 2000;38:261- 271. 5. AAOS Advisory Statement: Recom- mendations for the use of intravenous antibiotic prophylaxis in primary total joint arthroplasty. Rosemont, IL: American Academy of Orthopaedic Surgeons, June 2004. Available at http://www .aaos.org/wordhtml/papers/ advistmt/1027.htm. Accessed February 23, 2006. 6. Hoffman RM, Wheeler KJ, Deyo RA: Surgery for herniated lumbar discs: A literature synthesis. J Gen Intern Med 1993;8:487-496. 7. Barrios C, Ahmed M, Arrotegui J, Björnsson A, Gillström P: Microsur- gery versus standard removal of the herniated lumbar disc: A 3-year com- parison in 150 cases. Acta Orthop Scand 1990;61:399-403. 8. Turner JA,Ersek M, Herron L, DeyoR: Surgery for lumbar spinal stenosis: Attempted meta-analysis of the liter- ature. Spine 1992;17:1-8. 9. Bauer RD, Garfin SR: Complications of surgery for lumbar stenosis. Seminars in Spine Surgery 1993;5: 123-127. 10. Jones AA, Stambough JL, Balderston RA, Rothman RH, Booth RE Jr: Long- term results of lumbar spine surgery complicated by unintended inciden- tal durotomy. Spine 1989;14:443-446. 11. Antonacci MD, Eismont FJ: Neuro- logic complications after lumbar spine surgery. J Am Acad Orthop Surg 2001;9:137-145. 12. Bertrand G: The “battered” root prob- lem. Orthop Clin North Am 1975;6: 305-309. 13. Balderston RA, An HS: Complica- tions in Spinal Surgery. Philadelphia, PA: WB Saunders, 1991. 14. Herkowitz HN, Garfin SR: Decom- pressive surgery for spinal stenosis. Seminars in Spine Surgery 1989;1: 163-167. 15. Macnab I: Negative disc exploration: An analysisof the causes ofnerve-root involvement in sixty-eight patients. J Bone Joint Surg Am 1971;53:891- 903. 16. Epstein JA, Epstein BS, Rosenthal AD, Carras R, Lavine LS: Sciatica caused by nerve root entrapment in the later- al recess: The superior facet syn- drome. J Neurosurg 1972;36:584-589. 17. Burton CV, Kirkaldy-Willis WH, Yong-Hing K, Heithoff KB: Causes of failure of surgery on the lumbar spine. Clin Orthop 1981;157:191-199. 18. Johnsson KE, Redlund-Johnell I, Udén A, Willner S: Preoperative and postop- erative instability in lumbar spinal stenosis. Spine 1989;14:591-593. 19. McCulloch JA: Focus issue on lumbar disc herniation: Macro- and microdis- cectomy. Spine 1996;21:45S-56S. 20. Report of the Task Force on Wrong Site Surgery. Rosemont, IL: American Academy of Orthopaedic Surgeons, 1998. Available at http:// www.aaos.org/wordhtml/meded/task site.htm. Accessed November 21, 2005. 21. Wong D, Mayer T, Watters W, et al: Prevention of wrong-site surgery: Sign, Mark and X-ray (SMaX). La Grange, IL: North American Spine So- ciety, 2001. Available at http:// www.spine.org/smax.cfm. Accessed February 23, 2006. 22. Universal Protocol for Preventing Wrong Site, Wrong Procedure, Wrong Person Surgery. Oak Brook, IL: Joint Commission on the Accreditation of Healthcare Organi- zations, 2003. Available at http://www.jcaho.org/accredited+org anizations/patient+safety/universal+ protocol/universal_protocol.pdf. Ac- cessed November 21, 2005. 23. Lewis BD: Initial evidence: Reduced levels of wrong sided surgery. COA Bulletin ACO February/March 2002:10. 24. Sentinel Event Alert. Oakbrook, IL: Joint Commission on the Accredita- tion of Healthcare Organizations. December 5, 2001. Available at http:// www.jcaho.org/about+us/news+let ters/sentinel+event+alert/sea_24.htm. Accessed February 23, 2006. 25. Porucznik MA: Patient safety tip contest winners announced. AAOS Bulletin 2004;52:37. Available at http://www.aaos.org/wordhtml/bulle tin/apr04/acdnws2.htm. Accessed November 21, 2005. Spinal Surgery and Patient Safety: A Systems Approach 232 Journal of the American Academy of Orthopaedic Surgeons . implants. A summary of the literature evaluation for various regimens is included in the support document. Several of the recommen- dations are firmly grounded in evidence-based medicine. Intravenous cephazolin. either side will suffice. The last component of SMaX is performed in the operating room. Fi- nal documentation of levels is con- firmed by radiographic examination. A radiopaque marker (eg, needle, towel