This Article Abstract Full Text (PDF) Free Letters to the Editor: Submit a response Alert me when this article is cited Alert me when Letters to the Editor are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Rights and Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by PATERNO, S. A. Articles by KELLEY, S. S. Search for Related Content PubMed PubMed Citation Articles by PATERNO, S. A. Articles by KELLEY, S. S. Social Bookmarking                   What's this?

The Influence of Patient-Related Factors and the Position of the Acetabular Component on the Rate of Dislocation after Total Hip Replacement*

STEVEN A. PATERNO, M.D., PAUL F. LACHIEWICZ, M.D. and SCOTT S. KELLEY, M.D., CHAPEL HILL, NORTH CAROLINA

Investigation performed at the University of North Carolina School of Medicine Chapel Hill

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The senior one of us performed 438 primary and 181 revision total hip arthroplasties with use of so-called modern prostheses between July 1983 and January 1994. Fifty-three patients, who had had forty-seven primary procedures and twelve revisions, either died or were lost to follow-up less than two years after the operation and were excluded from the study. The results for the remaining 446 patients (391 primary procedures and 169 revisions) were analyzed to determine the influence of patient-related and operative factors and the position of the acetabular component on the rate of dislocation. Dislocation occurred after thirty-two (6 per cent) of the 560 total hip arthroplasties: seventeen (4 per cent) of the 391 primary procedures and fifteen (9 per cent) of the 169 revisions (p = 0.046). There was no relationship between the variables of age, gender, obesity, or preoperative diagnosis and dislocation after either primary or revision arthroplasty. Seven (23 per cent) of the thirty arthroplasties in the patients who had a history of excessive intake of alcoholic beverages (more than 2.1 liters [seventy-two ounces] of beer or more than 0.2 liter [six ounces] of other alcoholic beverages a day) were followed by a dislocation compared with twenty-five (5 per cent) of the 530 arthroplasties in the patients who did not have such a history. This difference was significant for the patients who had had a revision arthroplasty (p = 0.00005), but with the numbers available we could not detect a difference for those who had had a primary arthroplasty (p = 0.264). Radiographic analysis was performed for thirty-two hips that had dislocated and thirty-two that had not (seventeen primary procedures and fifteen revisions in each group), matched exactly according to the type of prosthesis and the operative approach (but not age). We detected no association between either the version or the abduction angle of the acetabular component (within the range of 39 to 56 degrees for the primary prostheses and 38 to 57 degrees for the revision prostheses) and the risk of dislocation. Thirty of the thirty-two hips in each group had an abduction angle of the acetabular component that was in the so-called safe range of 30 to 50 degrees.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Dislocation is a common and distressing complication after total hip arthroplasty, second in frequency only to aseptic loosening^25,29. The prevalence of dislocation has been reported to range from less than 1 to greater than 9 per cent in series ranging from 427 to 10,500 hips^9-11,25-27. Variables such as a previous operation, the operative approach, trochanteric avulsion, and gender have been identified as factors that predispose the hip to dislocation³². The prevalence of recurrent dislocation necessitating operative treatment has been reported to range from 13 to 42 per cent in series ranging from eleven to thirty-nine hips^8,11. In a review of the literature, Morrey reported that thirty-one (22 per cent) of 142 hips that had recurrent dislocations needed operative treatment²⁵. Advances in the prevention and treatment of dislocation are therefore desirable.

The continued development of total hip arthroplasty during the last decade has led to notable changes in the procedure, including the use of implants designed to be inserted without cement and modular prostheses as well as new techniques for exposure and implantation^13,17,18. Investigators have addressed the prevalence and etiology of dislocation after total hip arthroplasties performed in the late 1970s and the early 1980s^{3-12,20,22-27,31-33}; however, few have analyzed dislocation after total hip arthroplasties performed with use of so-called modern techniques and prostheses²⁹. Therefore, a clinical study and a matched-set radiographic analysis were performed to determine the factors that contribute to dislocation after total hip arthroplasty.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
A total of 619 total hip arthroplasties (in 499 patients) were performed by the senior one of us (P.F.L.) at one hospital from July 1983 to January 1994. All patients were examined or were contacted by telephone in early 1995 and again in January 1996 to determine the prevalence of dislocation. Fifty-three patients (fifty-nine arthroplasties) were excluded from the study: thirty-three (thirty-seven arthroplasties), because they had died before the two-year follow-up evaluation, and twenty (twenty-two arthroplasties), because they would not return for evaluation or they had been lost to follow-up. There were two dislocations in this group of fifty-nine hips that had been followed for less than two years. Of the 560 hips (446 patients) included in the study, 391 (317 patients) had a primary total hip arthroplasty and 169 (149 patients) had a revision arthroplasty. Eighteen patients had one primary and one revision arthroplasty, one patient had one primary and two revision arthroplasties, and one patient had three revision arthroplasties.

A variety of combinations of so-called modern prostheses were employed, reflecting the evolution of total hip arthroplasty over the last ten years (Table I). Of the 391 primary hip arthroplasties, seventy-four were performed with insertion of Triad acetabular and femoral components with cement (Johnson and Johnson, New Brunswick, New Jersey); 106, with Harris-Galante porous-coated acetabular and femoral components (Zimmer, Warsaw, Indiana); 109, with a Harris-Galante porous-coated acetabular component and insertion of a Precoat femoral component with cement (both Zimmer); forty, with a Harris-Galante porous-coated acetabular component and insertion of a Centralign femoral component with cement (both Zimmer); thirty, with a Harris-Galante porous-coated acetabular component (Zimmer) and insertion of a Triad femoral component with cement (Johnson and Johnson); twenty-nine, with a Harris-Galante porous-coated acetabular component and insertion of a Multilock femoral component without cement (both Zimmer); and one procedure each, with a Harris-Galante porous-coated acetabular component and insertion of a CDH Precoat femoral component with cement, a Harris-Galante porous-coated acetabular component and insertion of an Iowa Precoat femoral component with cement, and insertion of a Triad acetabular component with cement and a Harris-Galante porous-coated femoral component (all Zimmer). The diameter and angle of the Morse taper were identical for all of the modular femoral components from one manufacturer (Zimmer). The Triad is a non-modular femoral component with a titanium-alloy stem and a cobalt-chromium head placed with a shrink-fit by the manufacturer (Johnson and Johnson).

Of the 391 primary total hip arthroplasties, 384 (98 per cent) were performed through a posterior approach and seven (2 per cent), through a transtrochanteric approach. The 384 procedures performed through a posterior approach routinely included repair of the posterior aspect of the capsule and the external rotator tendons to the greater trochanter with non-absorbable sutures. The anterior part of the capsule was not excised. Seven primary arthroplasties included an osteotomy of the greater trochanter because exposure was difficult or the greater trochanter needed advancement. Of the 169 revision arthroplasties, eighty-three (49 per cent) were performed through a posterior approach and eighty-six (51 per cent), through a transtrochanteric approach. The choice of the approach depended on multiple factors, including the difficulty of the exposure and removal of the component and the presence of a fracture of the femur. The site of the trochanteric osteotomy was repaired with four Vitallium (cobalt-chromium-alloy) wires (two horizontal and two vertical) when the femoral component had been inserted with cement and with three horizontal Vitallium wires when it had been inserted without cement.

A twenty-eight-millimeter femoral head was used in 385 (98 per cent) of the 391 primary arthroplasties. In five hips that had a large acetabular component a thirty-two-millimeter femoral head was used, and in one small hip a twenty-two-millimeter femoral head was used. Of the 169 revision arthroplasties, 146 (86 per cent) were performed with a twenty-eight-millimeter femoral head; seventeen (10 per cent), a thirty-two-millimeter femoral head; three (2 per cent), a twenty-six-millimeter femoral head; and three (2 per cent), a twenty-two-millimeter femoral head. The size of the femoral head that was used in the revision arthroplasties was selected to match the component that was not revised.

Postoperatively, the patients were managed with use of balanced slings, and walking was begun on the third or fourth day. They were taught precautions against dislocation, and these were reinforced by both a physical therapist and an occupational therapist before the patients were discharged from the hospital, which was usually on the seventh to twelfth postoperative day. The patients were furnished with an elevated toilet seat and with instruments for putting on socks and shoes. They were advised to limit flexion to 90 degrees for the first six weeks after the operation.

Dislocation after total hip arthroplasty was studied in two manners. First, the hips that had had a dislocation after a primary arthroplasty were compared with those that had not had a dislocation after a primary procedure, to determine whether any demographic or operative factors had contributed to the dislocation. Similarly, the hips that had had a dislocation after a revision arthroplasty were compared with those that had not had a dislocation after a revision. Second, because it has been suggested that the position of the acetabular component plays a crucial role in the development of dislocation^19,24, the radiographic abduction and version angles of the acetabular component were compared in a matched set of hips that had and had not had a dislocation after a primary or revision procedure.

Multiple factors were analyzed to determine if they had had any influence on the rate of dislocation. Demographic factors included age, gender, body-mass index, history of excessive intake of alcoholic beverages, and preoperative diagnosis. Operative factors included operative approach, type of prosthesis, size of the femoral head, size of the femoral neck (short, medium, long, or extra-long), offset (in millimeters) of the femoral component according to the manufacturer's catalogue, use of an elevated-rim liner, and trochanteric avulsion. The body-mass index was calculated by dividing the patient's weight (in kilograms) by the square height (in meters). The index then was compared with the age of the patient. Obesity was defined as a body-mass index of greater than thirty kilograms per square meter². The patient's medical history, prospectively recorded both at the preoperative evaluation by the attending surgeon and on admission to the hospital, was carefully reviewed for a notation of excessive intake of alcoholic beverages. This parameter was considered positive if the patient reported daily consumption equivalent to 2.1 liters (seventy-two ounces) of beer³⁰ or 0.2 liter (six ounces) of other alcoholic beverages. This information was supplemented by interviews with family members, referring physicians, and primary-care physicians. Twenty-three patients who had a primary arthroplasty had a history of excessive consumption of alcoholic beverages, and thirteen of them had osteonecrosis. Seven patients (seven arthroplasties) who had a revision arthroplasty had a history of excessive consumption of alcohol, but the primary diagnosis could not be determined for these patients.

The statistical analysis was done in two stages: unadjusted analyses were performed to determine significant factors, and the analyses then were repeated with adjustment for the significant factors and examination of all other outcomes. The dislocation status was examined for the two groups with use of two sample statistical methods. The dislocation statuses for categorical measures were compared with use of Fisher exact, chi-square, and extended Mantel-Haenszel tests. Fisher exact tests were used when the categorical measures were dichotomous (for example, gender); chi-square tests, when the categorical measures were nominal with more than two categories (for example, type of prosthesis); and extended Mantel-Haenszel tests, for adjustment across categorical measures. Homogeneity tests were used to assess the consistency of relationships across adjustment-variable levels. Two-sample t tests, Wilcoxon rank-sum tests, and multiple-regression analyses were used to compare the dislocation status for continuous measures. Bonferroni multiple-comparison adjustments within families of measures were used, with critical values based on a* = a/k, where k is the number of measures in a family and a is the usual type-I error rate of 0.05. A power analysis was performed for the analysis of patient-related factors that were not determined to be significant and is described as 1-ß after the p value.

The seventeen hips that had a dislocation after a primary procedure and the fifteen that had a dislocation after a revision were computer-matched with seventeen hips that did not have a dislocation after a primary procedure and fifteen that did not have a dislocation after a revision. These two groups were then analyzed to determine the influence of two radiographic measurements—the angle of version of the acetabular component and the angle of abduction of the acetabular component—on the prevalence of dislocation. These hips were matched carefully with regard to type of prosthesis, operative approach, diagnosis, and history of excessive intake of alcoholic beverages. Differences in the dislocation status for dichotomous measures (for example, gender) were assessed with use of exact McNemar tests²¹, which involve discordances within pairs; differences in the dislocation status for ordered categorical factors (for example, size of the femoral head) were assessed with use of Wilcoxon signed-rank tests; and differences in the status for continuous measures (for example, radiographically derived data) were analyzed with use of paired-data t tests, signed-rank tests, and multiple-regression analyses. Bonferroni adjustments were used as explained previously.

The version and abduction angles of the acetabular component were measured on the postoperative radiographs of the thirty-two dislocated hips and the thirty-two matched-pair controls. The cup was designated as anteverted, neutral, or retroverted (no cup was retroverted) on the basis of the angle formed by a line drawn tangential to the face of the cup and a line drawn perpendicular to the horizontal plane, on a postoperative cross-table lateral radiograph^7,32. This method is not accurate enough to determine the exact angle of anteversion. Special studies^15,16 that allow for such a determination were performed only on a small number of patients who had had a revision arthroplasty. The abduction angle of the acetabular component was measured, on a standardized anteroposterior radiograph of the pelvis centered over the pubis, as the number of degrees between a line formed by connecting the ischial tuberosities and a line drawn tangential to the face of the cup^7,32.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Of the 560 arthroplasties (in 446 patients), thirty-two (in thirty-two patients) were followed by at least one dislocation, an over-all prevalence of 6 per cent. Dislocation occurred after seventeen (4 per cent) of the 391 primary hip arthroplasties and after fifteen (9 per cent) of the 169 revisions; this difference was significant (p = 0.046). Thus, the remaining patient-related and operative factors and the matched-set radiographic study of the acetabular components were analyzed separately for the primary and revision arthroplasties.

The direction of dislocation was posterior after thirteen primary arthroplasties and anterosuperior after four. The direction of dislocation was posterior after thirteen revisions and anterosuperior after two. The initial dislocation occurred a mean of eighteen weeks (range, 1.7 to 152 weeks) after the seventeen primary arthroplasties and eleven weeks (range, zero to 109 weeks) after the fifteen revisions. Of the seventeen dislocations after a primary procedure, four occurred while the patient was still in the hospital; eight, within three months; four, between three and twelve months; and one, after twelve months. Of the fifteen dislocations after a revision, three occurred while the patient was still in the hospital; eleven, within three months; and one, after twelve months.

The mean duration of follow-up after the 374 primary arthroplasties that were not followed by a dislocation was six years (range, two to twelve years) compared with five years (range, two to nine years) after the seventeen primary procedures that were followed by a dislocation. The mean duration of follow-up after the 154 revisions that were not followed by a dislocation was five years (range, two to ten years) compared with four years (range, two to ten years) after the fifteen revisions that were followed by a dislocation.

No modular femoral or acetabular component dissociated with the dislocation or the reduction¹⁴.

Patient-Related Factors
The mean age at the time of the seventeen primary procedures that were followed by a dislocation was 61.2 years (range, thirty-six to eighty-five years) compared with 58.8 years (range, eighteen to eighty-nine years) at the time of the 374 primary procedures that were not followed by a dislocation. After adjustment with the Mantel-Haenszel test for patients who had a history of excessive intake of alcoholic beverages, we could not detect a significant difference between these ages, with the numbers available (p = 0.4425; 1-ß < 0.80). After stratification of the patients into four age-groups (less than fifty years old [eighty-four patients who had ninety-eight arthroplasties], fifty to sixty years old [sixty-eight patients who had eighty-four arthroplasties], sixty to seventy years old [103 patients who had 131 arthroplasties], and more than seventy years old [sixty-two patients who had seventy-eight arthroplasties]), we could detect no significant difference, with the numbers available, in the rate of dislocation among the age-groups (p = 0.66; 1-ß < 0.80). The mean age at the time of the fifteen revision procedures that were followed by a dislocation was 65.8 years (range, forty-nine to eighty-three years) compared with 58.8 years (range, twenty-two to eighty-seven years) at the time of the 154 revision procedures that were not followed by a dislocation. After adjustment with the Mantel-Haenszel test for patients who had a history of excessive intake of alcohol, this difference in age was significant (p = 0.039). However, after stratification of the 149 patients (169 revisions) into the four age-groups (less than fifty years old [forty-one patients who had forty-four revisions], fifty to sixty years old [thirty-six patients who had forty revisions], sixty to seventy years old [forty-one patients who had forty-eight revisions], and more than seventy years old [thirty-one patients who had thirty-seven revisions]), we could detect no significant difference in the rate of dislocation among the age-groups, with the numbers available (p = 0.086; 1-ß < 0.80).

Of the 391 primary arthroplasties, 125 (32 per cent) were performed in men and 266 (68 per cent), in women. There were six dislocations (5 per cent) in men and eleven (4 per cent) in women who had had a primary arthroplasty; with the numbers available, this difference was not found to be significant (p = 0.793; 1-ß < 0.80). Of the 169 revisions, eighty-three (49 per cent) were done in men and eighty-six (51 per cent), in women. There were nine dislocations (11 per cent) in men and six (7 per cent) in women who had had a revision; with the numbers available, this difference was not found to be significant (p = 0.42; 1-ß < 0.80).

Measurements of the patient's height and weight were available after 380 primary procedures. Of these 380 procedures, 120 were done in obese patients (a body-mass index of greater than thirty kilograms per square meter) and 260 were done in patients who were not obese. Four (3 per cent) of the procedures done in obese patients were followed by dislocation compared with thirteen (5 per cent) of those done in patients who were not obese. With the numbers available, this difference could not be shown to be significant (p = 0.598; 1-ß < 0.80). Measurements of the patient's height and weight were available after 161 revisions, fifty-one of which were done in obese patients and 110 of which were done in patients who were not obese. Three (6 per cent) of the revisions in obese patients and twelve (11 per cent) of those in patients who were not obese were followed by a dislocation. With the numbers available, this difference also could not be shown to be significant (p = 0.392; 1-ß < 0.80).

Twenty-three primary arthroplasties were done in patients who were known to be abusing alcohol or to have a history of excessive intake of alcoholic beverages. These patients tended to be younger (mean age, fifty-two years; range, thirty-three to seventy-three years) and were more often male (70 per cent) compared with the entire group of patients who had a primary hip arthroplasty. Of these twenty-three arthroplasties, two (9 per cent) were followed by a dislocation compared with fifteen (4 per cent) of the 368 in patients who had no such history of alcohol abuse. With the numbers available, this difference could not be shown to be significant (p = 0.264; 1-ß < 0.80). Both dislocations were in men, who were seventy-one and seventy-three years old. The mean age was fifty years (range, thirty-three to sixty-eight years) for the patients who had a history of excessive intake of alcoholic beverages and did not have a dislocation.

Seven revision arthroplasties were performed in patients who were known to be abusing alcohol or had abused it in the past. These patients tended to be older (mean age, sixty-one years; range, fifty to eighty-three years), and five were men. Five of the seven revisions performed in patients who had a history of excessive intake of alcohol were followed by a dislocation compared with ten (6 per cent) of the 162 revisions done in patients who had no such history; with the numbers available, this difference could not be shown to be significant (p = 0.00005). Four of these five revisions were done in men, who had a mean age of sixty-four years (range, fifty to eighty-three years). The mean age was fifty-two years (range, fifty-one to fifty-four years) for the patients who had a history of alcohol abuse and did not have a dislocation.

Of the 391 primary arthroplasties, 216 were done because of osteoarthrosis; eighty-nine, rheumatoid arthritis; fifty-eight, osteonecrosis; thirteen, post-traumatic osteoarthrosis; ten, fracture; four, ankylosing spondylitis; and one, cancer. Of the 169 revisions, 151 were done because of aseptic loosening; nine, periprosthetic fracture; and nine, infection. The risk of dislocation was associated more with a previous arthroplasty than with any diagnosis. The rates of dislocation associated with post-traumatic osteoarthrosis (8 per cent) and avascular necrosis (7 per cent) were higher than the rates associated with the other diagnoses but, with the numbers available, we could detect no significant differences (p > 0.05; 1-ß < 0.80).

The power analysis for the patient-related factors that were not found to be significant (for example, age, gender, obesity, and diagnosis) revealed that a minimum sample size of 334 in each group would be needed to detect a difference.

Operative Factors
With the numbers available, we could detect no significant difference between the rate of dislocation after the 384 primary arthroplasties performed through a posterior approach (4 per cent) and the rate after the seven primary procedures done through a transtrochanteric approach (0 per cent) (p = 1.00; 1-ß < 0.80). The difference between the rate of dislocation after the eighty-three revisions that were done through a posterior approach (11 per cent) and the eighty-six done through a transtrochanteric approach (7 per cent) also could not be shown to be significant (p = 0.427; 1-ß < 0.80). The over-all rates of dislocation after the 467 procedures performed through a posterior approach and the ninety-three done through a transtrochanteric approach were both 6 per cent (p > 0.8; 1-ß < 0.80).

Five hundred and twenty-eight femoral components were implanted during a primary hip arthroplasty, a revision of both components, or a revision of the femoral component only; we could detect no significant differences among the rates of dislocation associated with the different types of components, with the numbers available (Table I). Thirty (7 per cent) of the 459 arthroplasties performed with a Harris-Galante porous-coated acetabular component were followed by a dislocation compared with two (2 per cent) of the ninety-three arthroplasties in which the acetabular component was inserted with cement. After adjustment for revision arthroplasty with the Mantel-Haenszel test, the difference could not be shown to be significant (p = 0.147; 1-ß < 0.80). A dislocation occurred after three (14 per cent) of the twenty-one arthroplasties in which a Harris-Galante acetebular component with a 10-degree-elevated-rim liner had been used compared with twenty-nine (7 per cent) of the 438 arthroplasties in which a Harris-Galante acetabular component without an elevated-rim liner had been employed. After adjustment for revision arthroplasty with use of the Mantel-Haenszel test, the difference could not be shown to be significant (p = 0.333; 1-ß < 0.80).

A twenty-eight-millimeter-diameter femoral head was used in 531 (95 per cent) of the 560 arthroplasties, and all but one of the thirty-two dislocations involved such a head. Therefore, little information can be obtained from the analysis of the head size as a risk factor for dislocation. The long femoral neck was associated with the highest rate of dislocation (ten [11 per cent] of ninety-five), but this difference was not significant, with the numbers available, after adjustment for revision arthroplasty with the Mantel-Haenszel test (p > 0.3; 1-ß < 0.80). The mean offset (according to the manufacturer's specifications) of the femoral components that dislocated was 42.4 millimeters (range, thirty-three to seventy-one millimeters), and the mean offset of those that did not dislocate was 40.7 millimeters (range, twenty-four to seventy-one millimeters). After adjustment for revision arthroplasty and for a history of an excessive intake of alcoholic beverages with the Mantel-Haenszel test, we detected no significant difference (p = 0.641; 1-ß < 0.80).

Of the ninety-three arthroplasties (seven primary and eighty-six revision) in which the transtrochanteric approach was used, seventy-three (78 per cent) were followed by healing at the site of the osteotomy of the greater trochanter without complication. Of these seventy-three arthroplasties, only one (1 per cent) was followed by a dislocation. After the other twenty arthroplasties (22 per cent), the osteotomy site did not heal and was classified as either a non-union in situ or an avulsion with proximal migration. Four (20 per cent) of these twenty arthroplasties were followed by a dislocation. After adjustment for revision arthroplasty with the Mantel-Haenszel test, this difference remained highly significant (p = 0.002).

The power analysis for the operative factors that were not significant (for example, type of component, use of an elevated-rim liner, and offset of the femoral component) demonstrated that a sample size of 474 in each group would be required to detect a difference.

Hips with a Dislocation Compared with a Matched Set of Hips without a Dislocation
Two sets of hips, one with and one without a dislocation, were matched exactly with respect to the type of arthroplasty (revision or primary), type and offset of the femoral component, length of the femoral neck, and operative approach. An attempt was made to also match the patients' ages exactly, but this was not possible: the mean difference in age between the patients who had a dislocation and those who did not was nine years (range, zero to thirty-six years) at the time of the primary arthroplasty and ten years (range, zero to twenty-seven years) at the time of the revision. A history of excessive intake of alcoholic beverages was documented for two of the seventeen patients who had a dislocation and for one of the seventeen patients who did not have a dislocation after a primary procedure. Such a history was documented for five of the fifteen patients who had a dislocation and for none of the fifteen patients who did not have a dislocation after a revision. In the entire series, only two patients (two hips) who had had a revision and did not have a history of excessive intake of alcoholic beverages had a dislocation, but these patients could not be matched with regard to the type of prosthesis or other factors.

No acetabular component appeared to be retroverted on the lateral radiographs. Of the seventeen hips that had a dislocation after a primary arthroplasty, three appeared to have neutral version of the acetabular component and fourteen, anteversion. Of the seventeen matched hips that did not dislocate after a primary arthroplasty, three appeared to have neutral version of the acetabular component and fourteen, anteversion. The mean angle of abduction of the acetabular component in the seventeen hips that had a dislocation after a primary arthroplasty was 44 degrees (range, 39 to 56 degrees) compared with 42 degrees (range, 32 to 54 degrees) in the seventeen matched hips that did not have a dislocation after a primary procedure; with the numbers available, this difference also was not found to be significant (p = 0.2967). When the angle of abduction of the acetabular component was categorized into three groups (less than 35 degrees, 35 to 45 degrees, and more than 45 degrees), we could detect no significant difference, with use of the McNemar test (p = 0.285), between the hips that had a dislocation and the matched set of hips that did not have a dislocation.

Of the fifteen hips that had a dislocation after a revision, two appeared to have neutral version and thirteen appeared to have anteversion of the acetabular component. Of the fifteen hips that had a revision without a dislocation, three appeared to have neutral version and twelve appeared to have anteversion of the acetabular component. With the numbers available, this difference was not found to be significant (p = 0.651). The mean angle of abduction of the acetabular component was 46 degrees (range, 38 to 57 degrees) in the fifteen hips that had a dislocation after revision compared with 44 degrees (range, 35 to 52 degrees) in the fifteen hips that had a revision without a dislocation; with the numbers available, this difference was not found to be significant (p = 0.4432). After categorization of the angles of abduction into three groups, we again could detect no significant difference, with use of the McNemar test (p = 1.00), between the hips that had a dislocation after a revision and the matched set that did not have a dislocation.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Dislocation is a major complication after total hip arthroplasty, and the risk factors include patient-related, operative, and miscellaneous variables. Lack of compliance with postoperative instructions can lead to improper positioning of the limb. Inadvertent operative malpositioning of the acetabular or femoral component, injury to the abductor mechanism or its nerve supply, and inadequate myofascial tension also can lead to instability of the hip. Trauma to the hip, such as that sustained in a motor-vehicle accident or a fall, can cause a sudden dislocation of an optimally positioned total hip prosthesis. Rarely, extreme weight loss or chronic low-grade infection can lead to instability. With advances in prosthetic design, fixation, and operative technique of total hip arthroplasty, identification of risk factors for dislocation is helpful in the prevention and treatment of the complication.

Contrary to the observations made in previous investigations^8,22,25,32, we found increasing age to be a significant risk factor (p = 0.039) for dislocation of the hip after a revision. The disparity in age between the patients who did and those who did not have a dislocation was especially great in the group with a history of excessive intake of alcoholic beverages (mean age, sixty-four years for the patients who had a dislocation compared with fifty-two years for those who did not). The exact cause of this association is uncertain, but it may reflect diminished proprioception, mental acuity, or muscle mass with increasing age.

We were not able to detect, with the numbers available, any significant difference in the prevalence of dislocation between the men and the women in this study. This finding contrasts with those of other studies^6,20,22,32, in which a ratio of dislocation as high as three to one was reported for female compared with male patients. However, the power analysis demonstrated that at least 334 patients would have been required to detect a significant difference. While the high number of dislocations in men who had a history of excessive intake of alcoholic beverages is partly responsible, accounting for this factor only equalizes the ratio of the rate of dislocation between male and female patients. These results mitigate against the influence of muscle mass on the rate of dislocation³².

To our knowledge, the influence of excessive intake of alcoholic beverages, in the present or the past, on the prevalence of dislocation after total hip arthroplasty has not been reported previously. It is possible that we did not identify all patients who had such a history, despite careful questioning and the information obtained from the referring physicians. This finding emphasizes the importance of obtaining a complete history of substance abuse during the preoperative evaluation. Perhaps perioperative detoxification or abstinence would decrease the prevalence of dislocation in this population of patients. In addition, consideration could be given to prophylactic bracing of the hip to limit motion postoperatively.

The important contribution of a previous total hip arthroplasty or hemiarthroplasty to the prevalence of dislocation has been described previously^{4,8,12,20,25,32}. Although the exact mechanism is unknown, the increased dissection with loss of soft-tissue restraints and a weakened abductor mechanism are at least partially responsible. A previous operation was the predominant factor in the rates of dislocation in this study.

In contrast to the findings of other reports^9,12,28,32, use of a posterior operative approach as compared with use of a transtrochanteric approach was not associated with an increase in the rate of dislocation in the current study. Robinson et al.²⁸ reported a decrease in the rate of dislocation with use of a posterior approach when the procedure was changed to include repair of the posterior aspect of the capsule and the external rotator tendons to the femur with non-absorbable sutures, as was usually done in the procedures performed through the posterior approach in the current study. The prevalence of trochanteric non-union after the revisions performed through the transtrochanteric approach in the current study (eighteen [21 per cent] of eighty-six hips) is greater than that reported in other studies^1,25,31. As dislocation is strongly associated with avulsion at the site of a trochanteric osteotomy, this may have influenced the over-all ratio of dislocation associated with use of the posterior approach compared with that associated with use of the transtrochanteric approach. Lower overall rates of dislocation have been reported with use of a direct lateral^13,17 or an anterolateral³² approach, but there is a risk of other serious complications with these approaches¹³.

Several aspects of the design of the femoral component were analyzed as potential risk factors for dislocation. Turner²⁹ reported that prostheses with a larger-diameter femoral neck were more prone to dislocation. The diameters of the femoral necks in the current study were identical. We believed, before this study, that an inadequate or smaller offset of the femoral component would predispose a hip to dislocation. The mean offset of the femoral components that dislocated was slightly larger than that of those that did not dislocate (42.4 compared with 40.7 millimeters), but this difference was not found to be significant, with the numbers available. The long neck was associated with the highest rate of dislocation (11 per cent [8 per cent after the primary arthroplasties and 13 per cent after the revisions]), but this difference also was not found to be significant, with the numbers available. However, the issue of the relationship between the length of the femoral neck and the rate of dislocation warrants additional study.

When the matched sets of hips that did and did not dislocate was created in order to compare the angles of version and abduction of the acetabular component, the greatest emphasis was placed on attempting to match exactly the factors of revision or primary arthroplasty, type of femoral component, length of the femoral neck, offset, and operative approach. Preoperative diagnosis, gender, obesity, age, and duration of follow-up also were considered but were less important factors. Such a technique for comparison of the radiographic results of total hip arthroplasty has not, to our knowledge, been described previously. In the current study, there were no retroverted acetabular components in either the hips that dislocated or the matched hips that did not dislocate. We also detected no significant difference between the mean angle of abduction of the hips that had a dislocation and that of the matched hips that did not dislocate, either after the primary arthroplasties or after the revisions. Over-all, there was good placement of the acetabular component, with the angles of abduction in thirty of the thirty-two hips in each group within the previously described so-called safe range of 30 to 50 degrees^23,24. Thus, the precise angle of abduction of the acetabular component seemed to have little effect on the prevalence of instability after either the primary procedures or the revisions, at least within the ranges found in this study. The importance of the angle of abduction of the acetabular component as a risk factor for dislocation may have been overstated in earlier studies.

The over-all mean duration of follow-up in the matched-set radiographic analysis was six years after the primary arthroplasties and five years after the revisions. One of the seventeen hips that had had a primary procedure and one of the fifteen that had had a revision dislocated more than one year postoperatively. Twenty-six (81 per cent) of the thirty-two dislocations occurred within the first three months after the operation, a finding that is consistent with those of other studies in which the highest rates of instability occurred in the early postoperative period^{9,20,23,31,32}. Some of our patients may have been managed for dislocation at other hospitals by other physicians; however, the senior one of us (P. F. L.) was contacted routinely, both by patients and by treating physicians, when this occurred. Therefore, we believe that our data on the prevalence of dislocation are correct and valid.

	Footnotes

 

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

Department of Orthopaedics, University of North Carolina School of Medicine, 242 Burnett-Womack Building, Campus Box 7055, Chapel Hill, North Carolina 27599-7055. Please address requests for reprints to Dr. Lachiewicz.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Amstutz, H. C., and Maki, S.: Complications of trochanteric osteotomy in total hip replacement. J. Bone and Joint Surg., 60-A: 214-216, March 1978.[Abstract/Free Full Text]
2. Bray, G. A., and Gray, D. S.: Obesity. Part I—Pathogenesis. Western J. Med., 149: 429-441, 1988.[Medline]
3. Brien, W. W.; Salvati, E. A.; Wright, T. M.; and Burstein, A. H.: Dislocation following THA: comparison of two acetabular component designs. Orthopedics, 16: 869-872, 1993.[Medline]
4. Carlsson, Å. S., and Gentz, C.-F.: Postoperative dislocation in the Charnley and Brunswik total hip arthroplasty. Clin. Orthop., 125: 177-182, 1977.
5. Chandler, R. W.; Dorr, L. D.; and Perry, J.: The functional cost of dislocation following total hip arthroplasty. Clin. Orthop., 168: 168-172, 1982.
6. Coventry, M. B.: Late dislocations in patients with Charnley total hip arthroplasty. J. Bone and Joint Surg., 67-A: 832-841, July 1985.[Abstract/Free Full Text]
7. Daly, P. J., and Morrey, B. F.: Operative correction of an unstable total hip arthroplasty. J. Bone and Joint Surg., 74-A: 1334-1343, Oct. 1992.[Abstract/Free Full Text]
8. Dorr, L. D.; Wolf, A. W.; Chandler, R.; and Conaty, J. P.: Classification and treatment of dislocations of total hip arthroplasty. Clin. Orthop., 173: 151-158, 1983.
9. Eftekhar, N. S.: Dislocation and instability complicating low friction arthroplasty of the hip joint. Clin. Orthop., 121: 120-125, 1976.
10. Ekelund, A.; Rydell, N.; and Nilsson, O. S.: Total hip arthroplasty in patients 80 years of age and older. Clin. Orthop., 281: 101-106, 1992.
11. Etienne, A.; Cupic, Z.; and Charnley, J.: Postoperative dislocation after Charnley low-friction arthroplasty. Clin. Orthop., 132: 19-23, 1978.
12. Fackler, C. D., and Poss, R.: Dislocation in total hip arthroplasties. Clin. Orthop., 151: 169-178, 1980.
13. Foster, D. E., and Hunter, J. R.: The direct lateral approach to the hip for arthroplasty. Advantages and complications. Orthopedics, 10: 274-280, 1987.[Medline]
14. Friedman, R. J.: Displacement of an uncemented femoral component after dislocation of a total hip replacement. A case report. J. Bone and Joint Surg., 71-A: 1406-1407, Oct. 1989.[Free Full Text]
15. Ghelman, B.: Radiographic localization of the acetabular component of a hip prosthesis. Radiology, 130: 540-542, 1979.[Medline]
16. Ghelman, B.: Three methods for determining anteversion and retroversion of a total hip prosthesis. AJR: Am. J. Roentgenol., 133: 1127-1134, 1979.[Abstract]
17. Hardinge, K.: The direct lateral approach to the hip. J. Bone and Joint Surg., 64-B(1): 17-19, 1982.
18. Harris, W. H., and Sledge, C. B.: Total hip and total knee replacement (first of two parts). New England J. Med., 323: 725-731, 1990.[Medline]
19. Herrlin, K.; Selvik, G.; Pettersson, H.; Kesek, P.; Onnerfalt, R.; and Ohlin, A.: Position, orientation and component interaction in dislocation of the total hip prosthesis. Acta Radiol., 29: 441-444, 1988.[Medline]
20. Khan, M. A. A.; Brakenbury, P. H.; and Reynolds, I. S. R.: Dislocation following total hip replacement. J. Bone and Joint Surg., 63-B(2): 214-218, 1981.
21. Koch, G. G., and Edwards, S.: Clinical efficacy trials with categorical data. In Biopharmaceutical Statistics for Drug Development, pp. 403-457. Edited by K. E. Peace. New York, Dekker, 1988.
22. Kristiansen, B.; Jorgensen, L.; and Holmich, P.: Dislocation following total hip arthroplasty. Arch. Orthop. and Traumat. Surg., 103: 375-377, 1985.
23. Lewinnek, G. E.; Lewis, J. L.; Tarr, R.; Compere, C. L.; and Zimmerman, J. R.: Dislocations after total hip-replacement arthroplasties. J. Bone and Joint Surg., 60-A: 217-220, March 1978.[Abstract/Free Full Text]
24. McCollum, D. E., and Gray, W. J.: Dislocation after total hip arthroplasty. Causes and prevention. Clin. Orthop., 261: 159-170, 1990.
25. Morrey, B. F.: Instability after total hip arthroplasty. Orthop. Clin. North America, 23: 237-248, 1992.[Medline]
26. Rao, J. P., and Bronstein, R.: Dislocations following arthroplasties of a hip. Incidence, prevention, and treatment. Orthop. Rev., 20: 261-264, 1991.[Medline]
27. Ritter, M. A.: Dislocation and subluxation of the total hip replacement. Clin. Orthop., 121: 92-94, 1976.
28. Robinson, R. P.; Robinson, H. J., Jr.; and Salvati, E. A.: Comparison of the transtrochanteric and posterior approaches for total hip replacement. Clin. Orthop., 147: 143-147, 1980.
29. Turner, R. S.: Postoperative total hip prosthetic femoral head dislocations. Incidence, etiologic factors, and management. Clin. Orthop., 301: 196-204, 1994.
30. Urbaniak, J. R.; Coogan, P. G.; Gunneson, E. B.; and Nunley, J. A.: Treatment of osteonecrosis of the femoral head with free vascularized fibular grafting. A long-term follow-up study of one hundred and three hips. J. Bone and Joint Surg., 77-A: 681-694, May 1995.[Abstract/Free Full Text]
31. Williams, J. F.; Gottesman, M. J.; and Mallory, T. H.: Dislocation after total hip arthroplasty. Treatment with an above-knee spica cast. Clin. Orthop., 171: 53-58, 1982.
32. Woo, R. Y. G., and Morrey, B. F.: Dislocations after total hip arthroplasty. J. Bone and Joint Surg., 64-A: 1295-1306, Dec. 1982.[Abstract/Free Full Text]
33. Wroblewski, B. M.; Hodgkinson, J. P.; and Maltby, B.: Structural changes in the socket of a Charnley low friction arthroplasty associated with recurrent subluxation. Internat. Orthop., 13: 231-233, 1989.[Medline]

CiteULike

Connotea

Del.icio.us

Facebook

Technorati

Twitter

This article has been cited by other articles:

				R.M.D Meek, D.B. Allan, G. McPhillips, L. Kerr, C.R. Howie, and From the Scottish Arthroplasty Project Late Dislocation after Total Hip Arthroplasty Clin. Med. Res., May 1, 2008; 6(1): 17 - 23. [Abstract] [Full Text] [PDF]

				J. Parvizi, E. Picinic, and P. F. Sharkey Revision Total Hip Arthroplasty for Instability: Surgical Techniques and Principles J. Bone Joint Surg. Am., May 1, 2008; 90(5): 1134 - 1142. [Full Text] [PDF]

				J. W. Babisch, F. Layher, and L.-P. Amiot The Rationale for Tilt-Adjusted Acetabular Cup Navigation J. Bone Joint Surg. Am., February 1, 2008; 90(2): 357 - 365. [Abstract] [Full Text] [PDF]

				L. Pulido, C. Restrepo, and J. Parvizi Late Instability Following Total Hip Arthroplasty Clin. Med. Res., June 1, 2007; 5(2): 139 - 142. [Abstract] [Full Text] [PDF]

				D. Guss and T. Bhattacharyya Perioperative management of the obese orthopaedic patient. J. Am. Acad. Ortho. Surg., July 1, 2006; 14(7): 425 - 432. [Abstract] [Full Text] [PDF]

				T. Kalteis, M. Handel, H. Bathis, L. Perlick, M. Tingart, and J. Grifka Imageless navigation for insertion of the acetabular component in total hip arthroplasty: IS IT AS ACCURATE AS CT-BASED NAVIGATION? J Bone Joint Surg Br, February 1, 2006; 88-B(2): 163 - 167. [Abstract] [Full Text] [PDF]

				A. W. Blom, L. Astle, J. Loveridge, and I. D. Learmonth Revision of an acetabular liner has a high risk of dislocation J Bone Joint Surg Br, December 1, 2005; 87-B(12): 1636 - 1638. [Abstract] [Full Text] [PDF]

				R. Raman, R. P. Kamath, A. Parikh, and P. D. Angus Revision of cemented hip arthroplasty using a hydroxyapatite-ceramic-coated femoral component J Bone Joint Surg Br, August 1, 2005; 87-B(8): 1061 - 1067. [Abstract] [Full Text] [PDF]

				R. Biedermann, A. Tonin, M. Krismer, F. Rachbauer, G. Eibl, and B. Stockl Reducing the risk of dislocation after total hip arthroplasty: THE EFFECT OF ORIENTATION OF THE ACETABULAR COMPONENT J Bone Joint Surg Br, June 1, 2005; 87-B(6): 762 - 769. [Abstract] [Full Text] [PDF]

				E. L. Peak, J. Parvizi, M. Ciminiello, J. J. Purtill, P. F. Sharkey, W. J. Hozack, and R. H. Rothman The Role of Patient Restrictions in Reducing the Prevalence of Early Dislocation Following Total Hip Arthroplasty. A Randomized, Prospective Study J. Bone Joint Surg. Am., February 1, 2005; 87(2): 247 - 253. [Abstract] [Full Text] [PDF]

				M. Soong, H. E. Rubash, and W. Macaulay Dislocation After Total Hip Arthroplasty J. Am. Acad. Ortho. Surg., September 1, 2004; 12(5): 314 - 321. [Abstract] [Full Text] [PDF]

				P. E. Beaule, M. J. LeDuff, F. J. Dorey, and H. C. Amstutz Fate of Cementless Acetabular Components Retained During Revision Total Hip Arthroplasty J. Bone Joint Surg. Am., December 1, 2003; 85(12): 2288 - 2293. [Abstract] [Full Text] [PDF]

				R. L. Barrack Dislocation After Total Hip Arthroplasty: Implant Design and Orientation J. Am. Acad. Ortho. Surg., March 1, 2003; 11(2): 89 - 99. [Abstract] [Full Text] [PDF]

				W. Sioen, J. P. Simon, L. Labey, and R. VanAudekercke Posterior Transosseous Capsulotendinous Repair in Total Hip Arthroplasty : A Cadaver Study J. Bone Joint Surg. Am., October 10, 2002; 84(10): 1793 - 1798. [Abstract] [Full Text] [PDF]

				P. F. Lachiewicz and S. S. Kelley The Use of Constrained Components in Total Hip Arthroplasty J. Am. Acad. Ortho. Surg., July 1, 2002; 10(4): 233 - 238. [Abstract] [Full Text] [PDF]

				P. E. Beaule, T. P. Schmalzried, P. Udomkiat, and H. C. Amstutz Jumbo Femoral Head for the Treatment of Recurrent Dislocation Following Total Hip Replacement J. Bone Joint Surg. Am., February 1, 2002; 84(2): 256 - 263. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free Letters to the Editor: Submit a response Alert me when this article is cited Alert me when Letters to the Editor are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Rights and Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by PATERNO, S. A. Articles by KELLEY, S. S. Search for Related Content PubMed PubMed Citation Articles by PATERNO, S. A. Articles by KELLEY, S. S. Social Bookmarking                   What's this?

Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online.
Copyright © 1997 by The Journal of Bone and Joint Surgery, Inc. Online ISSN: 1535-1386.
The Journal of Bone and Joint Surgery®, JBJS®, JB&JS®, and eJBJS® are registered in the U.S. Patent and Trademark Office.