Over the past decade, there have been major advances in. Multicenter Study of Complications Following Surgical Dislocation of the Hip

Transcription

1 1132 COPYRIGHT Ó 2011 BY THE OURNAL OF BONE AND OINT SURGERY, INCORPORATED Multicenter Stuy of Complications Following Surgical Dislocation of the Hip Ernest L. Sink, MD, Paul E. Beaulé, MD, FRCSC, Daniel Sucato, MD, Young-o Kim, MD, PhD, Michael B. Millis, MD, Michael Dayton, MD, Robert T. Trousale, MD, Rafael. Sierra, MD, Ira Zaltz, MD, Perry Schoenecker, MD, Amy Monreal, BA, an ohn Clohisy, MD Backgroun: Surgical hip islocation enables complete exposure of the hip joint for treatment of various hip isorers. There is limite information regaring the complications associate with this proceure. Our purpose is to report the incience of complications associate with surgical islocation of the hip in a large, multicenter patient cohort. Methos: A retrospective, multicenter analysis of patients who ha unergone surgical hip islocation was performe. Patients who ha unergone a simultaneous osteotomy were exclue. Complications were recore, with specific assessment for osteonecrosis, trochanteric nonunion, femoral neck fracture, nerve injury, heterotopic ossification, an thromboembolic isease. We grae complications with a valiate classification scheme that inclues five graes base on the treatment require to manage the complication an any long-term morbiity. With this classification, a Grae-I complication is one that requires no change in the routine postoperative course, Grae II requires a change in outpatient management, Grae III requires invasive surgical or raiologic management, Grae IV is associate with long-term morbiity or is life-threatening, an Grae V results in eath. Results: The stuy inclue 334 hips in 302 patients seen at eight ifferent North American centers. There were eighteen complications (5.4%) that were classifie as Grae I (not clinically relevant an require no eviation from routine postoperative care). There were six complications (1.8%) classifie as Grae II (treate on an outpatient basis or with close observation an resolve). There were nine complications (2.7%) classifie as Grae III (treatable an resolve with surgery or inpatient management). There was one complication (0.3%) classifie as Grae IV (resulting in a long-term eficit). A total of thirty hips ha one or more complications, for an overall incience of 9%. Excluing heterotopic ossification, the complication rate was sixteen (4.8%) of 334. Conclusions: Surgical hip islocation is a safe proceure with a low complication rate. Many of the complications were clinically unimportant heterotopic ossification. There were no cases of femoral hea osteonecrosis or femoral neck fracture, an, with the exception of one sciatic neurapraxia that partially resolve, no other complication resulte in longterm morbiity. Level of Evience: Therapeutic Level IV. See Instructions to Authors for a complete escription of levels of evience. Over the past ecae, there have been major avances in the unerstaning of femoroacetabular impingement as a mechanism for hip pain an later osteoarthritis. Surgical islocation of the hip is one of the many approaches to the treatment of hip impingement isorers 1-5, along with arthroscopy 6-8 an arthroscopy with a limite open anterior approach 9. Surgical islocation of the hip, as originally escribe by Ganz et al. 1, has increase the knowlege of the pathomechanics of hip impingement synromes an provie a unique surgical strategy for treatment. The surgical approach enables 360 visualization an access to the acetabulum an the proximal part of the femur 5, allowing the surgeon to aress not only impingement, but a variety of hip isorers. With this approach, surgeons are able to perform labral repair, resection of the acetabular rim, femoral hea-neck osteochonroplasty, relative femoral neck lengthening, trochanteric avancement, fracture reuction, reuction of a slippe capital femoral epiphysis, proximal femoral osteotomies, an tumor resection. Although surgical hip islocation is becoming more commonplace worlwie, concerns regaring the potential risks an complications associate with it remain. While reports in the past ecae have suggeste that surgical hip islocation is an effective proceure, the ata collecte, particularly regaring complications, have not been uniform or stanarize in these reports 10. Also, except for Disclosure: The authors i not receive any outsie funing or grants in support of their research for or preparation of this work. Neither they nor a member of their immeiate families receive payments or other benefits or a commitment or agreement to provie such benefits from a commercial entity. Bone oint Surg Am. 2011;93: oi: /bs

2 1133 T HE OURNAL OF B ONE &OINT S URGERY BS. ORG symptomatic impingement, there are no clear stanarize inications for this proceure. The contrainications specifically, avance osteoarthritis are clearer 11. Aitionally, these previous reports have been on relatively small cohorts of patients treate by a single surgeon. Documentation of complications following this proceure is important for patient counseling, improve surgical care, an comparison with alternative, lessinvasive surgical techniques. Currently, there is no universal efinition, graing system, or classification for the magnitue of orthopaeic complications. Often, complications have been classifie as major, moerate, or minor 12, a system that is not objective an oes not inicate the nee for unplanne treatment or of any future morbiity. Therefore, there is a major nee for improve methoologies for reporting an categorizing complications. As joint preservation surgery of the hip assumes a larger role in orthopaeics, it is critically important to establish uniform methos for analyzing clinical outcomes an reporting complications. The purpose of this stuy was to etermine the incience an character of complications associate with surgical islocation of the hip in a large, multicenter patient cohort. We hypothesize that this information will soliify the belief that surgical hip islocation is a safe surgical approach for treating multiple hip isorers. We utilize a complication classification scheme that has been valiate for general surgical proceures 13,14 an shoul assist in stanarizing the reporting of complications of hip preservation surgery. Materials an Methos Aretrospective analysis was performe on 355 hips (323 patients) treate by eleven surgeons at eight ifferent North American centers that participate in the ANCHOR (Acaemic Network for Conservational Hip Outcomes Research) group, which is stuying the outcomes of hip preservation surgery. Patients inclue in the stuy ha unergone a surgical islocation in the perio from October 2001 to December 2007 an ha been followe for a minimum of one year. Patients were exclue if they ha ha a simultaneous intertrochanteric, femoral neck, or acetabular osteotomy or ha receive treatment for an unstable slippe capital femoral epiphysis. Since we specifically reviewe the recors for complications of the surgical islocation approach, we inclue patients who ha ha the proceure for many ifferent iagnoses, incluing femoroacetabular impingement, Legg-Calvé-Perthes isease in a skeletally mature hip, trauma, an eformity following a slippe capital femoral epiphysis. The inications establishe for surgery at the eight stuy centers were ebilitating hip joint pain, usually in the anterior aspect of the groin; limitations in hip motion, particularly flexion an internal rotation; an raiographic evience of an osseous or softtissue eformity that correlates with the symptoms an results of the physical examination. Raiographically evient eformities inclue a eficient offset at the anterior an anterolateral aspects of the femoral hea an neck junction, a nonspherical femoral hea, or a labral/chonral tear seen on magnetic resonance imaging (MRI). Institutional review boar approval was obtaine from each of the participating centers. Specific emographic ata that may affect complications, incluing patient age, sex, iagnosis, an the year of the surgery, were obtaine. Pertinent surgical information inclue the size an number of screws use for trochanteric fixation an whether a labral repair or trimming of the acetabular rim with labral fixation ha been performe. The latter was inclue in the analysis because labral fixation may require more muscle retraction an affect the rate of heterotopic ossification. Complications were categorize with use of an aaptation of the valiate Dino-Clavien classification of surgical complications (see Appenix) 13,14. A chi-square or t test was use for univariate analysis to examine the association of complications with age, sex, iagnosis, year of the surgery, iameter an number of screws use for trochanteric fixation, an whether a labral repair (with or without trimming of the acetabular rim) ha been performe. Furthermore, logistic regression analysis was use to examine the interplay of the factors performe in the univariate analysis. Precision of incience was estimate with the 95% confience interval calculate with the exact metho. Source of Funing No external funing was receive for this stuy. Results The recors on 355 hips in 323 patients seen at eight ifferent North American centers were retrospectively reviewe. Of the 355 hips, 334 (94%) in 302 patients were followe for a minimum of twelve months an twenty-one (6%), for less than twelve months. We attempte to contact these twenty-one patients by telephone an but coul not accomplish the require minimum twelve-month followup. This subgroup of patients was followe for an average of 5.8 months (range, two to eleven months), an no complications were reporte uring this short time. The remainer of the analyses were performe on the 334 hips with a minimum of twelve months of follow-up. The mean age was twenty-six years (range, eight to sixty-one years). Of the hips stuie, 179 were in males an 155 were in females. The surgical hip islocation approach was use to treat femoroacetabular impingement (288 hips), eformity following a slippe capital femoral epiphysis (thirty-three), eformity relate to Legg-Calvé-Perthes isease (twenty-one), hip ysplasia (seven), synovial chonromatosis (three), loose boy (one), osteochonroma (one), pigmente villonoular synovitis (one), femoral hea chonral amage (one), an trauma (one). The meian uration of follow-up of the hips with a minimum uration of follow-up of twelve months was thirty-six months (range, twelve to eighty-eight months). A labral repair or a labral repair along with trimming of the acetabular rim was performe in 178 (53.3%) of the 334 hips. Trochanteric fixation was performe with two 4.5-mm screws in 189 hips (56.6%), three 4.5-mm screws in forty-five (13.5%), three 3.5-mm screws in seventyfive (22.5%), an two 3.5-mm screws in twenty-four (7.2%). Thirty hips ha one complication or more, for an overall incience of 9% (95% confience interval [CI]: 5.5% to 11.5%) (see Appenix). Four patients ha more than one complication. Eighteen (60%) of the thirty hips with a complication ha Brooker 15 Grae-I or II heterotopic ossification with no clinical relevance. Excluing Grae-I an II heterotopic ossification, the complication rate was sixteen of 334, or 4.8% (95% CI: 2.4% to 6.9%), an all but one were treatable Grae-II or III complications (see Appenix). There was one Grae-IV complication (95% CI: 0% to 1%): sciatic sensory an motor sciatic nerve paralysis immeiately following the surgery. This improve to slight numbness/tingling on the orsum of the foot an some pain extening over the peroneal istribution of the sciatic nerve. At forty-seven months after the surgery, the patient limpe occasionally. There were nine Grae-III complications (2.7%; 95%

3 1134 T HE OURNAL OF B ONE &OINT S URGERY BS. ORG TABLE I Incience of Complications in Relation to Demographic an Clinical Factors All Complications Grae-I Complications Grae-II, III, or IV Complications Factor No. of Hips No. (%) of Hips P Value No. (%) of Hips P Value No. (%) of Hips P Value Chi-square test Sex Male (12.3) (7.8) (4.5) 0.98 Female (5.8) 4 (2.6) 7 (4.5) Labral repair Yes (10.2) (6.8) (4.5) 0.99 No (8.3) 6 (3.8) 7 (4.5) Yr of op (11.0) (5.5) (6.6) (10.7) 10 (6.7) 6 (4.0) (5.4) 3 (3.2) 3 (3.2) Logistic regression analysis Screw iameter 3.5 mm * * * mm * 2.7* 3.7* Age *The values represent the incience, which was the least-squares estimate calculate with logistic regression analysis ajuste for the number of screws use. The moel showe that the incience increase with the number of screws, but the increase was not significant (p > 0.05). Age was treate as a continuous variable in the univariate logistic regression analysis. The values are given as the os ratio for each five-unit (year) increase. CI: 1.6% to 5.5%) that were treatable an that resolve. There were six Grae-II complications (1.8%; 95% CI: 0.3% to 2.9%) an eighteen Grae-I complications (5.4%; 95% CI: 3.03% to 7.9%). There were no cases of osteonecrosis, femoral neck fracture, or any complication leaing to long-term morbiity, with the exception of the one sciatic nerve injury, which partially resolve. Of the nine Grae-III complications, six were trochanteric nonunions (a 1.8% rate of trochanteric nonunion). All unite after repeat open reuction an internal fixation. Three of these patients ha ha 4.5-mm screws for trochanteric fixation an three, 3.5-mm screws. There were two cases of eep venous thrombosis in the calf. One was in a patient with a familial coagulation abnormality an was treate with inpatient anticoagulation followe by Coumain (warfarin) for six weeks. The other was in a forty-year-ol an was manage with inpatient anticoagulation followe by Coumain for six weeks. Both patients ha ha mechanical an pharmacologic prophylaxis postoperatively. There was no consistent metho of prophylaxis against eep venous thrombosis among the eight centers inclue in this series. The one eep infection in this series was manage effectively with surgical incision an ebriement. No Grae-III complications resulte in long-term morbiity. The six complications that resolve with observation or outpatient management (Grae II) inclue two superficial woun infections that resolve with oral antibiotics, one transient sciatic neurapraxia that resolve in five weeks, two trochanteric elaye unions (elaye for more than three months) that heale without invasive treatment, an one trochanteric fracture, sustaine in a fall, that unite without surgery. The eighteen Grae-I complications were all cases of Brooker Grae-I or II heterotopic ossification that require no treatment an were not associate with clinical symptoms. Table I shows the rates of complications groupe by emographic an clinical variables. There was no significant ecrease in complications over time from 2001 to Males ha a significantly lower incience of Grae-I heterotopic ossification (7.8% versus 2.6% in females, p = 0.03), but there was no ifference in the rates of Grae-II, III, or IV complications between the sexes (p = 0.98). Multiple logistic regression analysis with sex, labral fixation, screw size, an number of screws as preictors showe the same results as univariate analysis (Table I). Discussion The purpose of this stuy was to etermine the incience of complications associate with surgical islocation of the hip in a large, multicenter patient cohort. Our ata suggest that this technique is a safe approach to the hip joint to treat femoroacetabular impingement an other complex hip isorers. It has a low complication rate that is uniform among multiple North American centers. Many of the complications consiste of clinically unimportant heterotopic ossification. There were

4 1135 T HE OURNAL OF B ONE &OINT S URGERY BS. ORG no cases of osteonecrosis of the femoral hea or femoral neck fractures. There was one complete sciatic nerve paralysis with incomplete recovery. None of the other complications resulte in long-term morbiity. The overall complication rate was 9% in this retrospective stuy of 334 hips. What is reporte as a complication is of critical importance in etermining this rate. This was a rigorous analysis of complications that inclue even asymptomatic ones that require no treatment. The complication rate was 4.8% if Grae-I an II heterotopic ossification is exclue. Recent investigators note that the reporting of surgical complications was not homogeneous, well efine, or stanarize in orthopaeic prospective ranomize stuies 16. Reporting of complications has been inconsistent in nonranomize stuies of surgical treatment for femoroacetabular impingement an hip preservation surgery 10. One of our aims was to stanarize the graing an efinition of complications of hip preservation surgery. If the complication graing is base on the treatment require or the therapeutic consequences of the complication, incluing long-term isability, it becomes less subjective 14. The general surgical literature oes escribe establishe methoologies for reporting complications. For our stuy, we aapte the Dino-Clavien classification, which has been valiate for other surgical subspecialties. The classification system has five graes (see Appenix) base on the treatment require to manage the complication an on the long-term morbiity. The aaptations that we introuce for this stuy mae the system more applicable to the relatively young, healthy population being treate for femoroacetabular impingement an receiving osteotomies to treat ysplasia. For example, Grae IV in the general surgery classification has subcategories, such as single-organ an multi-organ ysfunction, that are less likely in our patient population. However, we kept the basic concept; therefore, Grae IV is given for a complication involving any long-term orthopaeic isability or that is life-threatening, such as a permanent nerve injury, pulmonary embolus, osteonecrosis, or amission to the intensive care unit. The Dino-Clavien classification was teste in a cohort of 6336 patients who ha unergone elective surgery 14. The complexity of the surgery an the length of the hospital were strongly correlate with the graes of the classification system 14. A five-year review of the Dino-Clavien classification, analyzing its use in the literature an examining interobserver variability in graing scenarios among patients, octors, an nurses, was recently reporte 13. There was an 89% agreement in the ratings an an increase in use of ifferent forms of the classification in the literature across ifferent subspecialties since its publication in One Grae-IV complication (either life-threatening or with the potential for permanent ysfunction) occurre in our series. This complete sciatic paralysis partially resolve, but the patient was left with numbness an pain. There were nine Grae-III complications (requiring invasive treatment): six trochanteric nonunions that heale after repeat fixation, one eep infection that require surgical ebriement, an two eep venous thromboses in the calf that resolve with meical therapy. All of these Grae-III complications resolve with treatment. The complications observe in this review are similar to those reporte in smaller series from iniviual centers, in whichtherewasabouta1.0%to1.5%rateoftrochanteric screw failure, heterotopic ossification, an rare cases of partial neurapraxia (see Appenix). Ganz et al. originally escribe this approach, in 213 patients, an reporte no cases of osteonecrosis, two cases of partial neurapraxia, three cases of faile trochanteric fixation (a rate of 1.5%, which was similar to the rate in this review), an seventy-nine cases of heterotopic ossification 1. Peters an Erickson reporte on forty-two surgical islocations, none of which were followe by infection, osteonecrosis, hematoma, or harware failure 3.Beaulé et al. reporte on thirty-seven hips, with one failure of trochanteric fixation, one case of Grae-IV heterotopic ossification, an painful bursitis ue to trochanteric fixation in nine hips 17. Both 3.5 an 4.5-mmiameter screws were use for trochanteric fixation in Beaulé s series, with surgeon preference the eciing factor. Failure was seen in association with both screw sizes, so no conclusion on the effect of screw iameter on the risk of trochanteric nonunion coul be provie. The only significant association ientifie in our review was a higher rate of heterotopic ossification (a Grae-I complication) in males. We i not see a significant increase in the incience of complications with age or evience of a change in complications over the years uring which the surgery was performe. This stuy ha limitations. The ata are retrospective, so it is possible that some complications were misse or were not ocumente in the meical recor. The intent of the stuy was to evaluate a consecutive series treate between October 2001 an December 2007, but some patients were lost to follow-up. Although there were no cases of osteonecrosis, osteonecrosis coul present after longer-term follow-up. Any reporte complication other than osteonecrosis shoul have been present within six months. Also, we stuie complications that ha occurre at centers with experience with the technique, an it may be ifferent when surgical hip islocation is performe at centers with less experience. Also, although stuies examining the reliability of the Dino-Clavien complication classification have been publishe in the general surgery literature 13,14, there are no stuies of the reliability of this classification specifically for use in orthopaeic surgery. In conclusion, these ata inicate that surgical islocation of the hip is a safe surgical approach with a minimal risk of long-term morbiity. Trochanteric nonunion after 1.8% of the surgical proceures was the most serious complication particular to surgical islocation that woul not have occurre with less invasive approaches. Nevertheless, surgeons nee to recognize that the islocation approach also provies extensive exposure of the acetabulum an the proximal part of the femur that cannot be obtaine with less invasive techniques. This stuy provies unique an important information for the expaning fiel of joint preservation surgery. We have confirme the safety of surgical hip islocation an utilize a complication

5 1136 T HE OURNAL OF B ONE &OINT S URGERY BS. ORG graing scheme that may be aopte for future reports on hip joint preservation proceures as well as on other orthopaeic proceures. Appenix Tables showing the aapte Dino-Clavien complication classification, a summary of complications at single institutions, an an overview of the complications in the present series are available with the online version of this article at jbjs.org. n NOTE: The authors acknowlege Pan Zhaoxing an Lauren St. ohn for their contributions. Ernest L. Sink, MD Amy Monreal, BA The Chilren s Hospital, East 16th Avenue, Aurora, CO aress for E.L. Sink: sinke@hss.eu Paul E. Beaulé, MD, FRCSC Ottawa Hospital, 501 Smyth Roa, CCW 1646, Ottawa, ON K1H 8L6, Canaa Daniel Sucato, MD Texas Scottish Rite Hospital, 2222 Welborn Street, Dallas, TX Young-o Kim, MD, PhD Michael B. Millis, MD Chilren s Hospital Boston, 300 Longwoo Avenue, Hunnewell 225, Boston, MA Michael Dayton, MD University of Colorao Hospital, East 17th Place, Aurora, CO Robert T. Trousale, MD Rafael. Sierra, MD Mayo Clinic, 200 First Street S.W., Rochester, MN Ira Zaltz, MD Oaklan Orthopaeic Surgeons, Woowar Avenue, #100, Royal Oak, MI Perry Schoenecker, MD ohn Clohisy, MD Washington University in St. Louis, 4921 Parkview Place, Suite A, St. Louis, MO References 1. Ganz R, Gill T, Gautier E, Ganz K, Krügel N, Berlemann U. Surgical islocation of the ault hip a technique with full access to the femoral hea an acetabulum without the risk of avascular necrosis. Bone oint Surg Br. 2001;83: Beck M, Leunig M, Parvizi, Boutier V, Wyss D, Ganz R. Anterior femoroacetabular impingement: part II. Miterm results of surgical treatment. Clin Orthop Relat Res. 2004;418: Peters CL, Erickson A. Treatment of femoro-acetabular impingement with surgical islocation an ébriement in young aults. Bone oint Surg Am. 2006;88: Beaulé PE, Harvey N, Zaragoza E, Le Duff M, Dorey F. The femoral hea/neck offset an hip resurfacing. Bone oint Surg Br. 2007;89: Espinosa N, Beck M, Rothenfluh DA, Ganz R, Leunig M. Treatment of femoroacetabular impingement: preliminary results of labral refixation. Surgical technique. Bone oint Surg Am. 2007;89 Suppl 2(Pt 1): Byr W, ones KS. Prospective analysis of hip arthroscopy with 10-year followup. Clin Orthop Relat Res. 2010;468: Philippon M, Briggs KK, Yen YM, Kuppersmith DA. Outcomes following hip arthroscopy for femoroacetabular impingement with associate chonrolabral ysfunction: minimum two-year follow-up. Bone oint Surg Br. 2009;91: Larson CM, Giveans MR. Arthroscopic management of femoroacetabular impingement: early outcomes measures. Arthroscopy 2008;24: Clohisy C, Zebala LP, Nepple, Pashos G. Combine hip arthroscopy an limite open osteochonroplasty for anterior femoroacetabular impingement. Bone oint Surg Am. 2010;92: Clohisy C, St ohn LC, Schutz AL. Surgical treatment of femoroacetabular impingement: a systematic review of the literature. Clin Orthop Relat Res. 2010; 468: Tannast M, Siebenrock KA. [Open therapy of femoroacetabular impingement]. Oper Orthop Traumatol. 2010;22:3-16. German. 12. Davey P, Santore RF. Complications of periacetabular osteotomy. Clin Orthop Relat Res. 1999;363: Clavien PA, Strasberg SM. Severity graing of surgical complications. Ann Surg. 2009;250: Dino D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients an results of a survey. Ann Surg. 2004;240: Brooker AF, Bowerman W, Robinson RA, Riley LH r. Ectopic ossification following total hip replacement. Incience an a metho of classification. Bone oint Surg Am. 1973;55: Golhahn S, Sawaguchi T, Auigé L, Muni R, Hanson B, Bhanari M, Golhahn. Complication reporting in orthopaeic trials. A systematic review of ranomize controlle trials. Bone oint Surg Am. 2009;91: Beaulé PE, Le Duff M, Zaragoza E. Quality of life following femoral hea-neck osteochonroplasty for femoroacetabular impingement. Bone oint Surg Am. 2007;89: Espinosa N, Rothenfluh DA, Beck M, Ganz R, Leunig M. Treatment of femoroacetabular impingement: preliminary results of labral refixation. Bone oint Surg Am. 2006;88: