Utility of postoperative radiography in routine primary total shoulder arthroplasty

ARTICLE IN PRESS J Shoulder Elbow Surg (2016) ■■, ■■–■■

www.elsevier.com/locate/ymse

ORIGINAL ARTICLE

Utility of postoperative radiography in routine primary total shoulder arthroplasty Ian J. Dempsey, MD, MBA, Michelle E. Kew, MD, Jourdan M. Cancienne, MD, Brian C. Werner, MD, Stephen F. Brockmeier, MD* Department of Orthopaedic Surgery, University of Virginia Health System, Charlottesville, VA, USA Background: The medical relevance and cost-benefit of routine radiographs after primary anatomic total shoulder arthroplasty (TSA) up to a year postoperatively are unknown. This study was performed to assess the medical relevance and cost-benefit of radiography after TSA during the first postoperative year. Methods: During the period 2010 to 2015, 160 consecutive patients undergoing anatomic TSA by a single fellowship-trained surgeon had radiographs obtained at 2 weeks, 6 weeks, 4 months, and 1 year postoperatively. Radiographs and clinic notes were assessed to determine if a change in postoperative care happened because of radiographic findings, including postoperative fracture, hardware complication, or any concerning radiographic feature. Cost data and amount billed were obtained. Results: Patients underwent radiography at 1.8 ± 0.2 weeks, 6.5 ± 1.2 weeks, 14.9 ± 2.9 weeks, and 46.8 ± 19.5 weeks postoperatively. Findings on the radiologist’s reading were normal/unremarkable for 100.0%, 96.8%, 95.9%, and 95.2%, respectively, at each visit. Results were documented in the note for 92.5%, 97.4%, 98.0%, and 92.4%, respectively, at each visit. Review of the radiographs yielded no change in management based on these parameters. The amount billed for radiographs was $284,281 ($1776.76 per patient). Conclusions: A lack of clinically meaningful impact from routine postoperative radiography does not justify the per-patient expense, as routine imaging did not cause a change in postoperative management. The available data suggest that routine radiographs after primary anatomic TSA may be unnecessary or perhaps the described frequency in which radiographs are obtained is in excess. Level of evidence: Level IV; Economic Study without Sensitivity Analysis © 2017 Keywords: Primary total shoulder arthroplasty; radiography; X-ray; cost-effectiveness; utility; health care economics

This study meets exempt criteria of our Institutional Review Board for Health Sciences Research (UVA IRB-HSR) for the following reasons, as described on the UVA IRB-HSR website: 5: Research involving materials (data, documents, records, or specimens) that have been collected solely for non-research purposes (such as medical treatment and/or diagnosis). *Reprint requests: Stephen F. Brockmeier, MD, Department of Orthopaedic Surgery, University of Virginia, PO Box 800159, Charlottesville, VA 22908, USA. E-mail address: [email protected] (S.F. Brockmeier). 1058-2746/$ - see front matter © 2017 http://dx.doi.org/10.1016/j.jse.2016.11.035

During the last decade, there has been a significant increase in the incidence of shoulder arthroplasty, with an associated improvement in techniques and outcomes.3 There is a lack of evidence supporting routine postoperative radiography obtained after total shoulder arthroplasty (TSA).12 Declining complication rates call into question the medical relevance of repeated radiography during the first postoperative year.4 Previous studies have demonstrated robust survivorship of modern TSA for the first 5 years postoperatively.1,12

ARTICLE IN PRESS 2 Other orthopedic subspecialties have critically assessed the use of repeated postoperative imaging. Total hip and total knee arthroplasty literature has found limited value in using immediate postoperative imaging.1,6,10,11,14,18 In addition, studies in the spine literature have demonstrated minimal value in repeated radiographs following spine procedures in asymptomatic patients.7,9,15,16,19 Cost-effective resource allocation is a crucial issue that is pivotal in providing high-quality orthopedic care. Excessive routine radiography is a costly variable that might be removed without negatively affecting clinical outcomes. The purpose of this study was to evaluate the medical relevance and cost-benefit of routine radiography after primary anatomic TSA within 1 year postoperatively. The hypothesis of the current study states that there is limited medical relevance of routine radiographs with substantial avoidable costs.

Materials and methods A retrospective chart review was performed of all patients who underwent TSA (Current Procedural Terminology code 23472) from August 2010 to February 2015 by a single fellowship-trained orthopedic surgeon. Inclusion criteria for this study included (1) having undergone primary TSA between August 2010 and February 2015 and (2) a minimum of 1 year of follow-up. In addition, a lesser tuberosity osteotomy was used for all selected patients for subscapularis management and glenoid exposure. Patients undergoing reverse TSA or revision shoulder arthroplasty were excluded. Any patients with a noted precipitating factor that would require a radiograph, such as a fall or a recent trauma, were likewise excluded. A total of 170 patients were identified; 10 patients were excluded because of a precipitating factor, leaving 160 available for final data analysis. Of the 10 excluded patients, 8 sustained a ground-level fall and 2 sustained a twisting injury to the operative arm that prompted a clinic visit and nonroutine radiographs. Of these 10 patients, 2 had abnormal findings on radiography; 1 of the 8 patients sustained a greater tuberosity fracture with associated rotator cuff insufficiency requiring conversion to a reverse TSA; 1 of the 2 patients sustained a nondisplaced glenoid fracture that was treated conservatively. The following variables were obtained: demographic data, clinic notes, complications, and radiologist reports. Patient characteristics and implant data are presented in Table I. A retrospective chart and imaging review was performed evaluating all postoperative radiographs, radiologist interpretations, and clinic notes. In this clinical practice, a single postoperative Grashey anteroposterior (AP) view of the shoulder was obtained in the postanesthesia care unit. Subsequent to discharge, patients were evaluated and routine radiographs were obtained at approximately 2 weeks, 6 weeks, 4 months, and 1 year postoperatively. Radiographs were obtained at the radiology department adjacent to the clinic before being seen in the form of a Grashey AP view at neutral humeral rotation, an external rotation Grashey AP view, and a true axillary view of the shoulder. These were obtained postoperatively to accurately assess component placement, alignment, and version. Evaluation of each radiograph included the radiologist’s reading and clinic notes, which included the surgeon’s reading of the radiographs at the time of the visit, to determine if it was a “normal” or “abnormal” study. All films were reviewed by a radiologist after they were obtained

I.J. Dempsey Table I

Cohort characteristics

Patient demographics Age (years) Male gender Body mass index (kg/m2) Laterality Primary osteoarthritis Comorbidities (%) Diabetes mellitus Active smoker Chronic obstructive pulmonary disease Hypertension Hyperlipidemia Congestive heart failure Coronary artery disease Chronic kidney disease Depression Chronic pain

66.7 ± 10.9 100 (62.5) 30.1 ± 6.8 96 (60.0) 156 (97.5) 25.9 15.4 3.7 63.0 35.8 1.2 6.8 4.9 22.8 5.6

Operative details Component type DePuy Biomet Stryker Cemented (%) Glenoid Humerus

72 (45.0) 83 (51.9) 5 (3.1) 100 4.4

Categorical variables are presented as number (%) except as noted. Continuous variables are presented as mean ± standard deviation.

within a 24-hour period. However, immediate review of radiographs could be obtained by the surgeon using the on-site radiologist if there were any immediate questions about the radiographs. Next, each follow-up clinic note was reviewed to determine if the radiology results were documented. The clinic note was then examined to decide if the radiographic finding caused a change in clinical management. Management was changed if any the following radiologic parameters were met: humeral or scapular fracture; hardware complication on the humeral side, including early loosening or frank component dissociation; hardware complication on the glenoid side, including early loosening or frank component dissociation; and any other concerning radiographic feature to include shoulder dislocation or displacement of lesser tuberosity osteotomy. Cost was determined by obtaining Medicare reimbursement data from the Centers for Medicare and Medicaid Services from the years 2010 to 2015. Reimbursement was determined from the Current Procedural Terminology (CPT) code 73030 (radiograph 2 views of the shoulder). Cost data included total cost, professional fee, and technical fee. Actual amount billed was obtained from the billing records of our institution’s department of radiology. The medical center billed for the radiographs and the radiologist billed for the reading according to typical practice at our medical center.

Results There were 160 patients who had routine shoulder radiography at an average of 1.8 ± 0.2 weeks, 6.5 ± 1.2 weeks,

ARTICLE IN PRESS Utility of postoperative radiography in primary TSA 14.9 ± 2.9 weeks, and 46.8 ± 19.5 weeks postoperatively. Normal postoperative findings were reported by a musculoskeletal radiologist for 100.0%, 96.8%, 95.9%, and 95.2% at each of these follow-up visits. The radiologist’s assessments for these reports were documented 92.5%, 97.5%, 98.0%, and 92.4% of the time. The DePuy Global Total Shoulder System (DePuy Synthes, Warsaw, IN, USA) was used in 45.0% (72/160) of patients, the Biomet Comprehensive Total Shoulder System (Zimmer Biomet, Warsaw, IN, USA) was used in 51.9% (83/160) of patients, and the Stryker ReUnion Total Shoulder System (Stryker, Kalamazoo, MI, USA) was used in 3.1% (5/160) of patients; 100% of glenoid components were cemented, and 4.4% of humeral components were cemented. No management changes were made on the basis of these routine radiographs. The most frequent radiologist reading as abnormal was a lesser tuberosity “fracture.” This notation was found in 50% of patients. Of those patients, 90% noted that the fracture was actually the healing lesser tuberosity osteotomy without interval displacement. The most frequent reason that a radiograph reading was not documented in the note related to the author of the note. A total of $284,281.00 was billed to the patients in this cohort. This computes to $1776.76 billed per patient.

Discussion Cost-conscious behavior has become a necessary entity for complete health care. Appropriate allocation of resources has become paramount for the long-term survival of any orthopedic practice in the health care system, especially as bundled care continues to be implemented throughout the country. In addition, radiation safety continues to be a relevant topic in specialties that regularly use radiography. It is in the patient’s best interest to limit radiography as this reduces overall radiation exposure and reduces the overall cost to the patient. The data presented in this study found that routine radiography after primary TSA is not medically beneficial from a cost standpoint. Almost all radiographs in this study were normal, and repeated imaging failed to reveal any meaningful changes in postoperative care. Current literature in shoulder arthroplasty notes that immediate postoperative radiography adds little clinically meaningful information. Namdari et al found that it is technically difficult to obtain high-quality films in the postoperative recovery area after TSA; 283 patients underwent radiography immediately postoperatively, whereas 241 patients underwent radiography at their first postoperative visit. They found the postanesthesia care unit radiographs to be inadequate and added little to overall medical utility. This added a cost of $64,524 to this patient group. They concluded that routine postanesthesia care unit radiographs are unnecessary.12 The current study advances these findings to the period within the first year after TSA. Not only are immediate

3 postoperative radiographs unnecessary, in the absence of concerning clinical findings up to a year postoperatively, routine radiographs do not appear to be of any medical value. Given the continued innovation and evolution of TSA systems, survivorship continues to increase, which also plays a factor in routine radiography.3,4 Churchill et al found minimal increase in lucency in the partially cemented glenoid component at 5 years of follow-up in 20 patients who underwent primary TSA. They also demonstrated no significant change in Simple Shoulder Test or Constant scores during this time in comparing the patients who demonstrated increased lucency vs. the ones who did not.1 This was further elucidated in a much larger series of 151 patients by Fox et al, who found that glenoid radiolucencies were seldom seen early and glenoid radiolucencies develop with notable changes 5 years or more after surgery. This further confirms that routine radiography may be unnecessary even up to 5 years in the absence of concerning clinical findings, although our data are presented only up to 1 year postoperatively.5 Multiple studies in spine surgery have found little utility in routine postoperative radiography. One study demonstrated that postoperative radiography was not required for up to 1 year without concerning history or physical examination findings.15 Yamashita et al demonstrated no change in management from routine postoperative radiography in 63 patients undergoing lumbar instrumented fusion in the setting of a benign clinical examination during follow-up.19 In cervical spine surgery, Shau et al found that regardless of technique, routine postoperative radiographs provided low medical utility in guiding treatment in asymptomatic patients.16 The adult reconstruction literature has demonstrated similar results regarding postoperative radiography. Niskanen found in 100 total hip arthroplasties and 100 total knee arthroplasties that radiography postoperatively at 2 to 3 months and 2 years may be unnecessary. He found that limiting the number of follow-up radiographs does not compromise the quality of care given to each patient.14 Moskal and Diduch found that routine postoperative radiographs can be omitted in 98% of total knee arthroplasties and can be delayed without compromising patient care until even the initial postoperative office visit.10 This has been further substantiated by other research that also showed minimal medical benefit for routine total knee and hip arthroplasty imaging before discharge.6,7,11,13,18 A potential benefit of using routine radiographs after primary TSA is to evaluate component positioning. At the initial postoperative visit, 100% of radiographs were read as normal and 92.5% were documented in the note. The most frequent reason for an abnormal report was associated with a lesser tuberosity fracture. This lesser tuberosity fracture was in fact the lesser tuberosity osteotomy that the primary surgeon used to mobilize the subscapularis muscle during exposure and placement of the humeral and glenoid components. This “abnormality” was noted 50% of the time. Of those

ARTICLE IN PRESS 4 patients, 90% of the notations stated that it was in fact the lesser tuberosity fracture that demonstrated no interval displacement and continued interval healing if it had a comparison view. These radiographs could also be used to demonstrate appropriate union of the lesser tuberosity osteotomy. The most frequent reason for lack of documentation of the radiograph related to the author of the note. A nonorthopedic clinician was more likely to omit radiographic findings than an orthopedic clinician who was evaluating the patient in the clinic. One limitation of this study is the homogeneous nature of the sample examined. This cohort of patients who underwent only primary TSA came from a single fellowshiptrained surgeon. It is uncertain if these findings would be applicable to a more heterogeneous multisurgeon group. It is also difficult to translate these findings to hemiarthroplasty or reverse TSA. Reverse TSA changes the overall biomechanics of the shoulder, and previous designs have been fraught with early failure, likely making it necessary for increased radiographic surveillance. One example of this is glenosphere component failure secondary to scapular notching in early designs.2,8,17 Another limitation is that we did not perform a costeffective analysis for this paper. It would be challenging to estimate the overall cost of any other diagnoses that may be found from this routine radiography. In addition, no comparison group was available to see if management differed. This was simply due to the fact that standard of care at our institution at this time is to obtain routine radiography at each of the follow-up visits. An additional limitation to this study is the lack of available evidence-based protocols for postoperative radiography in TSA. The timing for postoperative radiography was created by the surgeon on the basis of his individual experience and previous fellowship training in caring for these specific patients. The sample size in this analysis is also modest, and the parameters for which a routine radiograph would have changed management are uncommon, thus requiring a larger sample size to confidently say that routine radiographs would not benefit patients undergoing primary TSA. On the basis of the results of this study, we have evolved our clinical practice and substantially decreased the frequency of routine radiographs obtained after primary TSA at our institution. However, we have not entirely eliminated them from our postoperative protocol as we think that appropriate radiographic surveillance is still an integral part of follow-up, even if there is a remote chance of an abnormality. In our current protocol, we now recommend obtaining postoperative radiographs at 2 weeks to confirm initial component positioning and to evaluate any early complications (ie, nondisplaced fracture), at 4 months to assess definitive healing of the lesser tuberosity osteotomy, and at 1 year for long-term surveillance visits. This is with the obvious caveat that if any clinical concern arises that would require radiographs, the clinician should be prepared to obtain these to aid in the appropriate

I.J. Dempsey care of the patient. A common example of this is a precipitating factor like a recent fall or trauma.

Conclusion Routine postoperative radiography after primary TSA does not appear to add clinical utility up to a year postoperatively. The lack of clinically meaningful change from routine radiography does not justify the per-patient expense, as almost all postoperative radiographs were read as normal and did not alter postoperative management. These data suggest that in the setting of an uncomplicated primary TSA, routine radiography may be unnecessary or perhaps the frequency in which these radiographs are obtained is excessive and not cost-efficient.

Disclaimer The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.

References 1. Churchill RS, Zellmer C, Zimmers HJ, Ruggero R. Clinical and radiographic analysis of a partially cemented glenoid implant: five-year minimum follow-up. J Shoulder Elbow Surg 2010;19:1091-7. http:// dx.doi.org/10.1016/j.jse.2009.12.022 2. Delloye C, Joris D, Colette A, Eudier A, Dubuc JE. Mechanical complications of total shoulder inverted prosthesis. Rev Chir Orthop Reparatrice Appar Mot 2002;88:410-4. 3. Deshmukh AV, Koris M, Zurakowski D, Thornhill TS. Total shoulder arthroplasty: long-term survivorship, functional outcome, and quality of life. J Shoulder Elbow Surg 2005;14:471-9. http://dx.doi.org/10.1016/ j.jse.2005.02.009 4. Farng E, Zingmond D, Krenek L, SooHoo NF. Factors predicting complication rates after primary shoulder arthroplasty. J Shoulder Elbow Surg 2011;20:557-63. http://dx.doi.org/10.1016/j.jse.2010.11.005 5. Fox TJ, Foruria AM, Klika BJ, Sperling JW, Schleck CD, Cofield RH. Radiographic survival in total shoulder arthroplasty. J Shoulder Elbow Surg 2013;22:1221-7. http://dx.doi.org/10.1016/j.jse.2012.12.034 6. Glaser D, Lotke P. Cost-effectiveness of immediate postoperative radiographs after uncomplicated total knee arthroplasty. J Arthroplasty 2000;15:475-8. 7. Kosashvili Y, Alvi M, Mayne IP, Safir O, Gross A, Backstein D. Immediate recovery room radiographs after primary total knee arthroplasty—why do we keep doing them? Int Orthop 2010;34:1167-73. http://dx.doi.org/10.1007/s00264-009-0888-9 8. Lévigne C, Boileau P, Favard L, Garaud P, Molé D, Sirveaux F, et al. Scapular notching in reverse shoulder arthroplasty. J Shoulder Elbow Surg 2008;17:925-35. http://dx.doi.org/10.1016/j.jse.2008.02.010 9. Molinari R, Hunter J. Same-hospitalization postoperative radiographs for instrumented single-level spinal fusions: utility following intraoperative fluoroscopy. Spine J 2011;11:131S-132S. http://dx.doi.org/ 10.1016/j.spinee.2011.08.320

ARTICLE IN PRESS Utility of postoperative radiography in primary TSA 10. Moskal JT, Diduch DR. Postoperative radiographs after total knee arthroplasty: a cost-containment strategy. Am J Knee Surg 1998;11:8993. 11. Mulhall KJ, Masterson E, Burke TE. Routine recovery room radiographs after total hip arthroplasty: ineffective for screening and unsuitable as baseline for longitudinal follow-up evaluation. J Arthroplasty 2004;19:313-7. http://dx.doi.org/10.1016/j.arth.2003 .08.021 12. Namdari S, Hsu JE, Baron M, Huffman GR, Glaser D. Immediate postoperative radiographs after shoulder arthroplasty are often poor quality and do not alter care. Clin Orthop Relat Res 2013;471:1257-62. http://dx.doi.org/10.1007/s11999-012-2551-9 13. Ndu A, Jegede K, Bohl DD, Keggi K, Grauer JN. Recovery room radiographs after total hip arthroplasty: tradition vs utility? J Arthroplasty 2012;27:1051-6. http://dx.doi.org/10.1016/j.arth.2011.12.020 14. Niskanen RO. Early repetitive radiography is unnecessary after an uncomplicated cemented hip or knee arthroplasty for osteoarthritis. Acta Orthop Belg 2005;71:692-5.

5 15. Romero NC, Glaser J, Walton Z. Are routine radiographs needed in the first year after lumbar spinal fusions? Spine 2009;34:1578-80. http:// dx.doi.org/10.1097/BRS.0b013e3181a972b2 16. Shau DN, Bible JE, Samade R, Gadomski SP, Mushtaq B, Wallace A, et al. Utility of postoperative radiographs for cervical spine fusion: a comprehensive evaluation of operative technique, surgical indication, and duration since surgery. Spine 2012;37:1994-2000. http://dx.doi.org/ 10.1097/BRS.0b013e31825c0130 17. Vanhove B, Beugnies A. Grammont’s reverse shoulder prosthesis for rotator cuff arthropathy. A retrospective study of 32 cases. Acta Orthop Belg 2004;70:219-25. 18. Ververeli PA, Masonis JL, Booth RE, Hozack WJ, Rothman RH. Radiographic cost reduction strategy in total joint arthroplasty. A prospective analysis. J Arthroplasty 1996;11:277-80. 19. Yamashita T, Steinmetz MP, Lieberman IH, Modic MT, Mroz TE. The utility of repeated postoperative radiographs after lumbar instrumented fusion for degenerative lumbar spine. Spine 2011;36:1955-60. http:// dx.doi.org/10.1097/BRS.0b013e31820125b5