Preoperative opioid use is an independent risk factor for complication, revision, and increased health care utilization following primary total shoulder arthroplasty

IntroductionThe incidence of total shoulder arthroplasty (TSA) in increasing. Evidence primary hip and knee suggest that preoperative opioid use is a risk factor for postoperative complication. This relationship TSA unknown. The purpose this study was to investigate relationship.MethodsThe Truven Marketscan claims database used identify patients who underwent primary, unilateral TSA. Preoperative status then divide into cohorts based on the average daily oral morphine equivalents (OMEs) received 6-month period. included following cohorts: naïve <1, 1-5, 5-10, >10 OMEs. In total, 29,454 with 90-day follow-up were included. Of these, 21,580 8959 had 1- 3-year follow-up, respectively. Patient information complication data collected. Univariate multivariate logistic regression performed assess association outcomes. A subgroup analysis examine revision surgery at 1 3 years postoperatively.ResultsForty-four percent identified opioids, but opioid-naïve patient became more common over Multivariate demonstrated receiving OMEs (compared naïve) higher odds overdose (odds ratio [OR] 4.17, 95% confidence interval [CI] 1.57-11.08, P = .004), wound (OR 2.04, CI 1.44-2.89, < .001), superficial surgical site infection 2.33, 1.63-3.34, prosthetic joint 3.41, 2.50-4.67, pneumonia 1.95, 1.39-2.75, thromboembolic event 1.42, 1.18-1.72, .001). same group health care utilization, including extended length stay, nonhome discharge, readmission, emergency department visits (P ? Total perioperative adjusted costs than $7000 >10-OME when compared patients.DiscussionOpioid prior associated increased complications, surgery, costs. dose dependent, efforts should be made cessation surgery. relationship. postoperatively. Forty-four patients. Opioid increasing exponentially.10Day J.S. Lau E. Ong K.L. Williams G.R. Ramsey M.L. Kurtz S.M. Prevalence projections elbow United States 2015.J Shoulder Elbow Surg. 2010; 19: 1115-1120https://doi.org/10.1016/j.jse.2010.02.009Abstract Full Text PDF PubMed Scopus (434) Google Scholar,11Dillon M.T. Chan P.H. Inacio M.C.S. Singh A. Yian E.H. Navarro R.A. Yearly trends elective arthroplasty, 2005-2013.Arthritis Care Res (Hoboken). 2017; 69: 1574-1581https://doi.org/10.1002/acr.23167Crossref (73) Scholar,15Harjula J.N.E. Paloneva J. Haapakoski Kukkonen Aarimaa V. Finnish Arthroplasty Registry Group Increasing Finland—a nationwide registry study.BMC Musculoskelet Disord. 2018; 245https://doi.org/10.1186/s12891-018-2150-3Crossref (30) Scholar,18Issa K. Pierce C.M. T.P. Boylan M.R. Zikria B.A. Naziri Q. et al.Total demographics, incidence, complications—a inpatient sample study.Surg Technol Int. 2016; 29: 240-246PubMed Scholar,22Jain N.B. Yamaguchi contribution reverse utilization arthroplasty.J 2014; 23: 1905-1912https://doi.org/10.1016/j.jse.2014.06.055Abstract (65) Scholar,40Wagner E.R. Farley K.X. Wilson J.M. Higgins I. Daly C. Gottschalk M.B. arthroplasty: rise future 2020; 2601-2609https://doi.org/10.1016/j.jse.2020.03.049Abstract (27) Scholar Consequently, prevalence now ever before anticipated continue rise.12Farley KX, JM, Kumar A, MB, C, Sanchez-Sotelo J, Wagner ER. Submission, paper submission not yet accepted publication.Google Given increase number procedures being performed, identification factors, particularly modifiable ones, complications paramount. One potential use. epidemic well established expected continue.1American Society Addiction MedicineOpioid addiction: 2016 facts figures. American Medicine, Chevy Chase, MD2016Google Scholar,32Seth P. Scholl L. Rudd Bacon S. Overdose deaths involving cocaine, psychostimulants—United States, 2015-2016.MMWR Morb Mortal Wkly Rep. 67: 349-358https://doi.org/10.15585/mmwr.mm6712a1Crossref (472) Scholar,39Vowles K.E. McEntee Julnes P.S. Frohe T. Ney J.P. van der Goes D.N. Rates misuse, abuse, addiction chronic pain: systematic review synthesis.Pain. 2015; 156: 569-576https://doi.org/10.1097/01.j.pain.0000460357.01998.f1Crossref (573) Scholar,45Yarnell Li MacGrory B. Trevisan Kirwin Substance disorders later life: synthesis literature an emerging public concern.Am J Geriatr Psychiatry. 28: 226-236https://doi.org/10.1016/j.jagp.2019.06.005Abstract (17) pain medication treatment pain—both acute chronic—has become common, approximately 25% rate misuse 10% addiction.39Vowles Therefore, it surprising undergoing pain-driven indications (such as arthroplasty) narcotics increasing.43Wilson Aizpuru M. Bradbury T.L. Guild G.N. Impact Use Disorder Complications Costs Primary Hip Knee Arthroplasty.Adv Orthop. 2019; : 1-7https://doi.org/10.1155/2019/9319480Crossref (21) However, potentially problematic multiple studies have complications,2Bell Shohat N. Goswami Tan Kalbian Parvizi opioids periprosthetic after Arthroplasty. 33: 3246-3251.e1https://doi.org/10.1016/j.arth.2018.05.027Abstract (41) Scholar,4Blevins Peratikos Weeks H.L. Pisansky A.J.B. Yong R.J. Stringer E.A. Effect adverse outcomes, medical spending, persistent States: large retrospective cohort administrative data.Pain Med. 21: 521-531https://doi.org/10.1093/pm/pnz083Crossref (33) Scholar,20Jain Brock J.L. Malik A.T. Phillips F.M. Khan S.N. Prediction duration use: major replacement lumbar fusion.J Bone Joint Surg Am. 101: 384-391https://doi.org/10.2106/JBJS.18.00502Crossref (48) Scholar,27Nguyen L.C. Sing D.C. Bozic K.J. reduction 31: 282-287https://doi.org/10.1016/j.arth.2016.01.068Abstract (127) Scholar,30Rozell J.C. Courtney P.M. Dattilo J.R. Wu C.H. Lee G.C. opiate independently predicts narcotic consumption 32: 2658-2662https://doi.org/10.1016/j.arth.2017.04.002Abstract (83) consumption,30Rozell readmission,4Blevins Scholar,42Weick Bawa H. Dirschl D.R. Luu H.H. readmission rates 100: 1171-1176https://doi.org/10.2106/JBJS.17.01414Crossref (77) revision,3Ben-Ari Chansky Rozet early male treated Veterans Affairs system.J 99: 1-9https://doi.org/10.2106/JBJS.16.00167Crossref (101) Scholar,36Starr Ben-Ari calculator using prediction arthroplasty.Clin Pain. 34: 328-331https://doi.org/10.1097/AJP.0000000000000544Crossref (14) costs,4Blevins dissatisfaction.13Goplen Verbeek W. Kang S.H. Jones C.A. Voaklander Churchill T.A. al.Preoperative worse outcomes meta-analysis.BMC 20: 234https://doi.org/10.1186/s12891-019-2619-8Crossref (57) regarding TSA, however, primarily limited small, single-institution investigations.8Cheah J.W. McLaughlin D. Feeley B.T. Ma C.B. Zhang A.L. effects outcomes.J 26: 1908-1914https://doi.org/10.1016/j.jse.2017.05.016Abstract Scholar,24Morris B.J. Laughlin M.S. Elkousy H.A. Gartsman G.M. Edwards T.B. 24: 11-16https://doi.org/10.1016/j.jse.2014.05.002Abstract (95) Scholar,25Morris Sciascia A.D. Jacobs anatomic 25: 619-623https://doi.org/10.1016/j.jse.2015.09.017Abstract (87) These consumption,8Cheah inferior clinical satisfaction patients.7Carducci M.P. Zimmer Z.R. Jawa Predictors unsatisfactory 8: 2113-2120https://doi.org/10.1016/j.jse.2019.04.009Abstract (16) Scholar,37Thompson K.M. Hallock J.D. Smith Brolin T.J. Azar Throckmorton T.W. radiographic analysis.J Am Acad Orthop 27: 177-182https://doi.org/10.5435/JAAOS-D-16-00808Crossref (10) one small series, there no difference or stay those opioids.8Cheah national constituted just 4% population, making interpretation these difficult.4Blevins need further investigation impact study, we sought between As secondary outcome, determine prescribing trends. We hypothesized would rates. additionally prescriptions less Health MarketScan Commercial Claims Encounters Medicare Supplemental Coordination Benefit databases (Truven Health, Ann Arbor, MI, USA). includes deidentified data. Since 1995 has amassed 240 million Information all facets included, hospital stays, outpatient visits, pharmaceutical information.38Truven AnalyticsThe services researchers: thought leadership white paper. Analytics, Atlanta, GA2019Google also filled codes medications National Drug Codes (NDCs). present NDCs prescribed There precedence databases.4Blevins by Current Procedural Terminology (CPT) (CPT 23472). Those hemiarthroplasty Additionally, excluded fracture International Classification Diseases (ICD) diagnosis codes. queried 2009-2018 initially adult (?18 old) (anatomic prosthesis). ultimately needed continual enrollment 6 months 90 days without minimal excluded. our final cohort. Figure 1. (hydrocodone, oxycodone, oxymorphone, dihydrocodeine morphine, hydromorphone, fentanyl, methadone, meperidine, codeine) preoperatively each patient. Tramadol purposes study. conversion tables convert prescription equivalent (OME) doses. calculated dividing (in days) labeled Based data, divided (1) (no period), (2) <1 OME (received prescription, per day), (3) 1-5 (ie, ?1 <5 OMEs), (4) 5-10 ?5 <10), (5) ?10 equal greater 10 OMEs) (Table I) .Table IPatient demographics comorbiditiesCharacteristicOpioid groupP value?P value indicates any statistically significant differences groups.Opioid naïve, n (%)(N 16,529; 56.12%)<1 OME, (%)(n 1637; 5.56%)1-5 4227; 14.35%)5-10 1958; 6.65%)>10 5103; 17.33%)Demographic Age group, yr<551397 (8.45)159 (9.71)387 (9.16)194 (9.91)619 (12.13)<.00155-645774 (34.93)567 (34.64)1583 (37.45)727 (37.13)2061 (40.39)65-744919 (29.76)460 (28.10)1164 (27.54)546 (27.89)1417 (27.77)75-843895 (23.56)390 (23.82)960 (22.71)421 (21.50)874 (17.13)?85544 (3.29)61 (3.73)133 (3.15)70 (3.58)132 (2.59) SexMale8599 (52.02)833 (50.89)2041 (48.28)853 (43.56)2185 (42.82)<.001Female7930 (47.98)804 (49.11)2186 (51.72)1105 (56.44)2918 (57.18)Comorbidities Obesity1760 (10.65)194 (11.85)544 (12.87)236 (12.05)716 (14.03)<.001 Chronic kidney disease739 (4.47)100 (6.11)261 (6.17)116 (5.92)349 (6.84)<.001 Alcohol disorders105 (0.64)14 (0.86)45 (1.06)27 (1.38)81 (1.59)<.001 Tobacco use463 (2.80)71 (4.34)195 (4.61)65 (3.32)324 (6.35)<.001 Hypertension9583 (57.98)1009 (61.64)2670 (63.17)1264 (64.56)3310 (64.86)<.001 Coronary artery disease2618 (15.84)273 (16.68)755 (17.86)358 (18.28)978 (19.17)<.001 Congestive heart failure575 (3.48)73 (4.46)212 (5.02)105 (5.36)356 (6.98)<.001 Hyperlipidemia7402 (44.78)761 (46.49)1949 (46.11)881 (44.99)2148 (42.09)<.001 Rheumatoid arthritis711 (4.30)71 (4.34)210 (4.97)124 (6.33)480 (9.41)<.001 Diabetes3097 (18.74)363 (5.56)920 (21.76)462 (23.60)1220 (23.91)<.001 Depression1280 (7.74)158 (9.65)465 (11.00)256 (13.07)995 (10.71)<.001OME, equivalent.? groups. Open table new tab equivalent. After separated cohorts, collected baseline demographic comorbid variables: age, sex, comorbidities. controlled comorbidities: obesity (defined BMI ? 30), disease, hypertension, coronary congestive failure, hyperlipidemia, rheumatoid arthritis, diabetes, depression, tobacco use, alcohol disorder. characteristics chi-square analysis. data: (ED) visit, pain-related ED ?3 days), overdose, (SSI), (PJI), pneumonia, (deep vein thrombosis pulmonary embolism), myocardial infarction, stroke. examined 30-day readmission. Finally, order analysis, continuous database, Revision CPT Cost period cohorts. “cost” defined sum net plus deductibles, copays, coinsurance similar studies.19Jacobs Hawk G.S. Jochimsen K.N. Conley C.E. Vranceanu A.M. Thompson al.Depression anxiety are femoroacetabular impingement arthroscopy: database.Arthroscopy. 36: 745-750https://doi.org/10.1016/j.arthro.2019.09.048Abstract (19) recorded median interquartile range. controlling variables listed Table I, isolate financial tracked plotted annual basis time population. All statistical SAS, version 9.4 (SAS Institute, Cary, NC, <.05 considered significant. comorbidities compiled groups binomial regression, (all factors I). comparisons reference. Because right-skewed distribution costs, generalized linear model gamma logarithmic link function compare opioid-use analyze 1-year adequate follow-up. From 2009-2018, met inclusion criteria. specific breakdown can seen preceding 44% 56% opioid-naïve. largest >10-OMEs contained significantly young (<64-year-old) patients, female predominance, commonly nearly every comorbidity .001; revealed outcome .013) except infarction stroke (p-value 0.793 0.724, respectively). univariate results II. Subsequent found following: (>10 vs. opioid-naïve, SSI PJI OR Furthermore, effect It noted II).Table IIAdjusted resource dataOutcomeOpioid group<1 OME1-5 OME5-10 OME>10 OMEOR (95% CI)?P groups; usage year surgery.P valueOR