A Retrospective Single-Center Study of 23 Patients to Compare Gait Before and After Total Hip Arthroplasty Using the S-ROM Modular Hip System.

BACKGROUND This retrospective study used the Harris hip score (HHS) and range of motion (ROM) to compare gait before and after total hip arthroplasty using the modular S-ROM® hip prosthesis in 23 patients treated at a single center. MATERIAL AND METHODS For this study, 23 patients with severe hip deformity, who were treated with a total hip replacement using the S-ROM® modular hip system by DePuy® in the period from 2003 until 2008, underwent a 3-dimensional gait analysis before and after surgery. Values were compared with a validated data set of healthy subjects. Gait analysis took place using 8 infrared cameras (50 Hz), 2 force platforms of Advanced Medical Technology, Inc. (AMTI)®, and the Vicon® 512 system. The HHS and ROM of the hip joint were determined preoperatively and postoperatively. The follow-up period was 16.7±15.2 months. RESULTS A healthy gait pattern is characterized by an even stride length of both legs in relation to body height and relative symmetry of certain gait phases. These characteristics are influenced by the range of motion of the hip joint and by pelvic tilt. Symmetry could be achieved postoperatively in the stance phase and in the single-leg stance phase. However, the gait phases could not normalize. HHS and ROM improved postoperatively. CONCLUSIONS The findings from this retrospective study showed that ROM and HHS significantly improved following THA with the S-ROM® hip prosthesis, but that gait did not completely return to normal.

Background

The aim of endoprosthetic joint replacement is to free patients from pain and to restore mobility so that they can cope with everyday stresses and strains and, in young patients, regain the ability to work [1]. Often, amazing athletic performance can be achieved again [1]. Total hip arthroplasty (THA) is successful in reducing pain and hip-related symptoms [2]. The hip disability and Osteoarthritis Outcome Score (HOOS) function in activities of daily living (ADL) and function in sports and recreation subscales also displays large effect sizes, indicating that surgery has a huge impact not only on pain and symptoms, but also on patient-reported function [2].

There are several studies that have shown an improvement in gait after THA [2-8]. In these studies, deformities were excluded and non-modular prosthesis systems were used [4,6,7,9]. The S-ROM prosthesis was developed for primary arthroplasty of deformities of the hip joint and is widely used [10-12]. Gait analyses of primary arthroplasty patients with deformities have not yet been published.

The aim of this study was to find out to what extent the gait pattern of patients with complex hip deformities approximates, after implantation of a modular hip endoprosthesis of the S-ROM type, the gait pattern of a healthy comparison collective, and in what aspects of the gait pattern the values of the comparison collective can be achieved.

The standard follow-up of patients after THA takes the form of X-ray checks and clinical examination, in which both passive and active mobility is tested [13]. Each clinical examination of the hip joint also includes a visual assessment of gait pattern [14,15]. The gait pattern reflects the function of the joint under everyday load. This examination technique is very subjective and therefore not comparable but can be objectified through 3-D analysis of gait [16]. This procedure is thus being adopted in an increasing number of clinical studies [5,6,17].

The DePuy S-ROM system is a cementless modular implant system introduced in 1984 [18]. Innovations include a fully polished distal stem, different neck lengths and offset variations on the stem, the possibility of independent rotational alignment of the stem and sleeve, and a variety of different proximal sleeves to ensure optimal adaptation to the proximal femur [10]. Bone ingrowth into a cementless prosthesis can be achieved by both porous and hydroxyapatite coatings [19]. Bolognesi et al compared the performance of a hydroxyapatite-coated proximal sleeve and a porous bead-coated sleeve in 52 patients performed with an S-ROM stem in a revision arthroplasty and found no difference with respect to bone ingrowth [19]. In this study, the S-ROM stem with a porous coated sleeve was used.

Distal stability without fixation is provided by the ribs and polished distal portion of the prosthesis. The slotting in the coronal plane helps to reduce bending stiffness. It also prevents distal fixation, which can lead to stress shielding. The proximal sleeves have a stepped surface to load the bone in compression and protect it from unphysiological ring stress. The steps counteract subsidence of the prosthesis without having to rely on a collar [10].

The independence of the sleeve and stem allows correction of excessive anteversion, as occurs with dysplasia and retroversion deformities, in revision hip arthroplasty. The fluted distal segment provides stability when a corrective osteotomy is required. The stem is a versatile tool in THA with unusual deformities and in revision surgery [20].

The S-ROM system is a popular choice for revision arthroplasty and proximal femoral deformity due to its modularity [21]. With independent rotational alignment of stem and sleeve and a variety of different proximal sleeves, optimal adaptation to the proximal femur can be ensured and can thus respond to a highly modified anatomy [22]. In this study, the S-ROM system was used as the primary implant system for complex hip joint deformities.

With the help of today’s data processing methods, elaborate and precise procedures of gait analysis methodology are possible [23]. Spatiotemporal parameters such as step length, step speed, step time fluctuations, and average step times, as well as joint position in space, angular velocity and acceleration, can be reliably determined parallel to electromyographic (EMG) activities in relation to the stance and swing phase [23]. Kinematic data collection can be carried out with various functional analysis systems, film-video (eg, Vicon, Oxford Metrics) optoelectrical systems or goniometry [23].

In the 1960s, a system of observational gait analysis was developed in Los Angeles at Rancho Los Amigos National Rehabilitation Center under the direction of Jacqueline Perry [16]. Motion analysis was initially performed using video cameras and video recorders and was gradually supplemented by kinematic and kinetic measurement systems, such as infrared ViconTM® cameras, floor-embedded force plates, and EMG. Her book “Ganganalyse”, published in 1992, has since become the standard work on instrumental gait analysis [16].

For the gait analysis system by Vicon®, (Oxford Metrics Ltd. Oxford, Great Britain) used in this study, a very high reliability level has been demonstrated before [24].

The American Academy of Orthopedic Surgeons and the Société Internationale de Chirurgie Orthopèdique et de Traumatologie recommend that an assessment of clinical complications, a physical examination of the hip, radiographic studies, and an assessment of well-being (pain, gait, some activities of daily living, and overall satisfaction) as reported by the patient to be included in any outcome studies. A disease-specific measure should be included in all studies of outcome of the hip arthroplasty [25]. Harris introduced a rating scale with a maximum of 100 points, including the domains of pain, function, deformity, and motion [26]. The Harris hip score was compared with the Larson and Shepard system, and it was found to be “reproducible and reasonably objective” [26]. Therefore, the Harris hip score is one of the most widely used scoring systems [27].

Material and Methods

The Medical Ethics Committee of the University of Duisburg-Essen, Germany, approved the study on the following basis:

Gait analysis and clinical examinations are performed. Data from patient records and data from X-rays are collected and evaluated, with the aim of finding out whether reconstruction of the anatomical conditions at the hip joint leads to more physiological movement patterns, whether, in the case of a measurable improvement in passive mobility, this is also exploited under the everyday load of walking or whether the old movement pattern continues and whether consequences for rehabilitation can be derived from the results.

The patients were informed about the study and the procedure of the gait analysis during the presurgical preparations and agreed to participate by giving their written consent.

Patients were selected from the outpatient clinic population in case of matching the inclusion criteria listed below and being willing to participate.

The data were anonymized and forwarded to a statistician at the University Hospital in Düsseldorf for statistical analysis. Access to the data was given to the first author of the study and to the employees of the gait laboratory.

We defined inclusions criteria for this investigation as follows:

The indication for THA on one site together with a complex deformity, the use of an S-ROM shaft system, surgery to be performed by the corresponding author. Figure 1 shows the preoperative X-ray of a patient treated in this study. Figure 2 shows the postoperative X-ray of this patient. Exclusion criteria were: Patients over the age of 70, previous major orthopedic surgery in the lower limbs with exception of the hip joint to be operated on, other lower-extremity joint pain or severe back pain, rheumatoid arthritis, neurologic disease and/or other conditions affecting walking ability.

Figure 1

Preoperative X-ray.

Figure 2

Postoperative X-ray.

In the period from 2003 until 2008 23 Patients matched those criteria.

Before and after surgery, with a definition of after surgery as on average 16.7 months (±15.2 standard deviation), a clinical examination including ROM, X-ray, HHS, and gait analysis was performed.

Gait analysis took place using 8 infrared cameras (50 Hz) by Vicon® (Oxford Metrics Ltd. Oxford, Great Britain) and 2 force platforms by Advanced Medical Technology, Inc., Watertown, USA (AMTI)® situated at the mid-point of the 10-m-long level walkway and the Vicon 512 software system [28]. 22 Pearl Hard Reflective Marker by Vicon Oxford Metrix [29] in a diameter of 25 mm were fixed with double-sided tape at the points described by Kadaba et al [30]. Spatiotemporal gait parameters and kinetic and kinematic parameters were obtained.

Spatiotemporal gait parameters include relative stride length, as well as the duration of the stance phase, single-leg stance phase, and loading response. The duration of the gait phases is given as a percentage of the gait cycle.

The kinematic parameters include the angular position of the hip joint as well as the pelvic tilt and the position of the pelvis in relation to the position of the upper body at each phase of the gait cycle. The values are given in angular degrees. The measurement was made in sagittal, frontal, and transversal planes. The maximum measured value and the ROM were used for evaluation.

The kinetic parameters include the generated and absorbed power, which is determined by means of force plates and attributed to the hip joint, the knee joint and the ankle joint on the basis of the acting force vector.

The affected side was compared preoperatively with postoperatively, the affected side preoperatively with the opposite side preoperatively, the affected side postoperatively with the opposite side postoperatively, and the affected side and the opposite side with the healthy control group both preoperatively and postoperatively.

The HHS is a disease-specific test used to provide an evaluation system for various hip disabilities and methods of treatment [26]. This rating system is staff-administered, not self-administered. The Harris hip score gives a maximum of 100 points and the domains include pain, function, deformity, and motion. In the HHS, pain and function are the 2 basic considerations and receive the greatest weighting, with 44 and 20 points, respectively. ROM and deformity are seldom of primary importance and thus receive 5 and 4 points, respectively. Function is subdivided into activities of daily living (14 points) and gait (33 points) [27].

The Harris hip score was assessed and the preoperative and postoperative passive ROM of the hip joint were measured on the day of gait analysis.

The Wilcoxon test for dependent samples was used for the comparison between preoperative and postoperative data. It indicates whether 2 dependent samples differ significantly in their central difference [31]. For the comparison of the patients with the normal collective, the Mann-Whitney U test for independent samples was used. This test tests the null hypothesis that 2 independent samples come from ‘identically shaped’ distributed populations with identical median values [31].

SPSS Predictive Analysis Software statistics 18 for Windows by International Business Machines Corporation (SPSS PASW by IBM) was used for the statistical analyses.

A significance level α≤0.05 is assumed. The maximum tolerable risk for a wrong decision in favor of the alternative hypothesis is thus 5%. The significance level α is compared with the calculated p-value. If p≤0.05, the null hypothesis of equality can be discarded [31].

Results

Spatiotemporal Gait Parameters

The one-legged stance phase of the affected side preoperatively was significantly shorter compared with the healthy control group (35.95%±3.14% vs 40.27%±1.37% [P<0.001]). The step length was also shorter (0.33±0.04 m vs 0.39±0.02 m). The loading response was significantly longer (12.2%±3.18% vs 9.69%±1.16% [P<0.001]). The total stance phase of the affected side shows normal values preoperatively (Table 1).

Table 1

Spatiotemporal gait parameters.

	PreOP (n=23)	PostOP (n=23)	Control (n=39)	Opposite side preOP (n=23)	Opposite side postOP (n=23)	p1	p2	p3	p4	p5	p6	p7	p8
Stance phase (% GC)	60.67 (3.87)	61.34 (2.03)	59.70 (1.03)	63.97 (3.04)	62.29 (3.78)	0.409	0.000	0.153	0.000	0.181	0.000	0.000	0.018
One-legged stance phase (% GC)	35.95 (3.14)	37.70 (3.32)	40.27 (1.23)	38.91 (4.02)	38.86 (2.09)	0.000	0.000	0.004	0.001	0.114	0.332	0.005	0.715
Loading response phase (% GC)	12.2 (3.18)	11.50 (1.78)	9.69 (1.16)	12.80 (3.66)	12.10 (3.32)	0.000	0.000	0.362	0.784	0.648	0.000	0.000	0.224
Relative step length (meter)	0.33 (0.04)	0.34 (0.05)	0.39 (0.02)	0.32 (0.05)	0.35 (0.04)	0.000	0.000	0.089	0.260	0.412	0.000	0.000	0.002

PreOP – postoperative; PostOP – preoperative; GC – gait cycle. p1: p-value preoperative in comparison with healthy control group (Mann-Whitney-U Test); p2: p-value for the test postoperative in comparison with control group (Mann-Whitney-U Test); p3: p-value for test preoperative in comparison with postoperative (Wilcoxon Test); p4: p-value for test preoperative in comparison with opposite side (Mann-Whitney-U Test); p5: p-value test postoperative in comparison with opposite side (Mann-Whitney-U Test); p6: p-value for test opposite side preoperative in comparison with control group (Mann-Whitney-U Test); p7: p-value for test opposite side postoperative in comparison with control group (Mann-Whitney-U Test); p8: p-value for test opposite side preoperative in comparison with opposite side postoperative (Wilcoxon Test).

On the affected side, the one-legged stance phase was observed to be longer postoperatively (37-7%±3.32 vs 35.95%±3.14% [P=0.004]). Loading response, total stance and relative step length did not significantly changed when comparing preoperatively to postoperatively (Table 1). The values of the healthy control group could not be achieved.

The single-leg stance phase and the relative step length were shorter, the total stance phase and the loading response were longer on the affected side (cf. Table 1).

Compared with the opposite side the affected side showed that stance phase and the one-legged stance were preoperatively shorter (61.34%±2.03% and 37.70%±3.32% [P =0.001]) than on the opposite side (63.97%±3.04% and 38.91%±4.02% respectively [P <0.001]). Postoperatively there was no significant difference between the 2 sides (61.34%±2.03% and 37.70%±3.32% on the operated side compared to 62.29%±3.78% and 38.86%±2.09% on the opposite side) (Table 1).

The opposite side displayed longer stance (P<0.001) and loading response (P<0.001) and shorter step length (P<0.001) and one-legged stance phase (P=0.005) before and after THA than the healthy control group.

Kinematics

In the kinematic examination of the pelvis, pelvic tilt describes the relative movement of the pelvis in 3 dimensions. For anatomical reasons, it is not possible to perform an isolated pelvic tilt of one side. The opposite side automatically undergoes the same tilting of the reference side, so that only the values of the affected side are examined during evaluation. Differences in the measured values of the opposite side are evaluated as measurement inaccuracy.

In the sagittal plane, preoperatively and postoperatively, a significantly enlarged anteversion was found (21.92°±8.95° preoperatively with P <0.001 and 18.78°±7.25° postoperatively with P<0.001) compared with the healthy control group (12.62°±7.25°). After THA, the anteversion was reduced with an insignificant result (Table 2). The maximum pelvic tilt on the sagittal plane preoperatively and postoperatively was significantly higher (7.3°±4.48° preoperatively (P<0.001) and 4.93°±2.35° postoperatively (P<0.001)) than in the healthy control group (1.85°±0.87°) (Table 2). Examining the pelvic tilt in the sagittal plane in comparison preoperatively to postoperatively, a significantly lower pelvic tilt postoperatively was found (P=0.008).

Table 2

Kinematics of the pelvis and the hip joint.

	PreOP (n=23)	PostOP (n=23)	Control (n=39)	Opposite side preOP (n=23)	Opposite side postOP (n=23)	p1	p2	p3	p4	p5	p6	p7	p8
Sagittal plane pelvis
Maximum anteversion	21.92 (8.95)	18,78 (7.25)	12.62 (4.36)	22.19 (9.01)	18.82 (7.00)	0.000	0.000	0.068	0.082	0.733	0.000	0.000	0.052
Minimum anteversion	14.62 (6.97)	13.84 (6.36)	10.76 (4.30)	14.76 (6.92)	14.01 (6.25)	0.020	0.021	0.715	0.260	0.114	0.013	0.018	0.761
Pelvic tilt in sagittal plane	7.30 (4.48)	4.93 (2.35)	1.85 (0.87)	7.43 (4.47)	4.81 (2.12)	0.000	0.000	0.008	0.306	0.249	0.000	0.000	0.003
Frontal plane pelvis
Maximum „up”	2.21 (4.56)	3.48 (3.91)	4.32 (1.11)	3.27 (4.39)	2.75 (3.05)	0.034	0.172	0.136	0.685	0.733	0.127	0.021	0.605
Minimum „up”	−3.62 (4.43)	−2.69 (3.03)	−4.28 (1.06)	−2.81 (5.04)	−3.45 (3.75)	0.469	0.030	0.287	0.761	0.592	0.310	0.354	0.412
Pelvic tilt in frontal plane	5.82 (2.88)	6.17 (2.83)	8.60 (2.16)	6.08 (2.95)	6.2 (2.83)	0.000	0.000	0.394	0.052	0.661	0.000	0.000	0.738
Transversal plane pelvis
Maximum „forward”	4.62 (5.17)	6.34 (4.33)	5.22 (2.55)	5.43 (4.88)	5.71 (4.36)	0.303	0.246	0.107	0.758	0.733	0.739	0.750	0.484
Minimum „forward”	−4.86 (5.11)	−4.00 (3.90)	−4.75 (2.39)	−4.81 (4.80)	−5.52 (4.28)	0.696	0.784	0.484	0.758	0.661	0.675	0.297	0.429
Pelvic tilt in transversal plane	9.48 (4.88)	11.32 (3.70)	9.97 (4.90)	10.25 (4.73)	11.24 (3.65)	0.795	0.212	0.052	0.036	0.884	0.817	0.234	0.153
Sagittal plane hip joint
Maximum flexion	32.53 (9.50)	34.34 (10.21)	35.85 (5.56)	41.81 (11.41)	37.44 (9.78)	0.470	0.297	0.394	0.002	0.123	0.075	0.851	0.068
Minimum flexion	7.26 (15.09)	1.20 (11.10)	−7.76 (4.10)	−0.02 (11.66)	−2.26 (11.35)	0.000	0.000	0.021	0.001	0.033	0.007	0.022	0.648
Range of motion	25.27 (9.93)	33.14 (8.55)	43.61 (4.78)	41.83 (10.59)	39.70 (8.53)	0.000	0.000	0.000	0.000	0.022	0.369	0.062	0.101
Frontal plane hip joint
Maximum adduction	1.37 (5.20)	3.62 (5.51)	4.75 (2.18)	5.01 (4.96)	4.00 (3.89)	0.014	0.277	0.059	0.128	0.910	0.874	0.862	0.412
Minimum adduction	−4.09 (5.22)	−3.53 (4.22)	−6.14 (2.22)	−3.04 (5.72)	−3.41 (5.5)	0.212	0.001	0.693	0.761	0.910	0.006	0.004	0.543
Range of motion	5.43 (2.83)	7.15 (2.80)	10.89 (2.02)	8.05 (4.07)	7.41 (3.51)	0.000	0.000	0.011	0.004	0.685	0.000	0.000	0.260
Transversal plane hip joint
Maximum internal rotation	0.96 (11.13)	−0.22 (11.16)	6.05 (4.78)	−1.03 (8.74)	0.19 (11.45)	0.010	0.020	0.627	0.715	0.818	0.001	0.056	0.447
Minimum internal rotation	−17.84 (1.88)	−18.73 (9.95)	−7.18 (5.29)	−21.00 (11.44)	−21.24 (10.73)	0.000	0.000	0.605	0.465	0.291	0.000	0.000	0.808
Range of motion	18.80 6.61	18.52 7.79	13.23 (3.90)	19.97 (5.54)	21.431 (7.37)	0.000	0.006	0.903	0.503	0.338	0.000	0.000	0.465

PreOP – preoperative; postOP – postoperative. p1: p-value preoperative in comparison with healthy control group (Mann-Whitney-U Test); p2: p-value for the test postoperative in comparison with control group (Mann-Whitney-U Test); p3: p-value for test preoperative in comparison with postoperative (Wilcoxon Test); p4: p-value for test preoperative in comparison with opposite side (Mann-Whitney-U Test); p5: p-value test postoperative in comparison with opposite side (Mann-Whitney-U Test); p6: p-value for test opposite side preoperative in comparison with control group (Mann-Whitney-U Test); p7: p-value for test opposite side postoperative in comparison with control group (Mann-Whitney-U Test); p8: p-value for test opposite side preoperative in comparison with opposite side postoperative (Wilcoxon Test).

Kinematic examination of the hip joint showed that the maximum extension on the affected side preoperatively with 7.26° flexion (±15.09°) and postoperatively with 1.20° flexion (±11.10°) was significantly lower than in the healthy control group with −7.76° flexion (±4.10°) (P<0.0001 for the comparison of the values of the affected side preoperatively and postoperatively with the values in the control group).

The affected side had a significantly higher extension postoperatively than preoperatively (P=0.021) (Table 2).

Preoperatively and postoperatively, the opposite side also showed a significantly lower extension (−0.02° flexion ±11.66° preoperatively [P=0.007] −2.26° flexion ±11.35° postoperatively [P=0.022]) than the healthy control group.

The ROM of the affected side was preoperatively and postoperatively significantly smaller than in the control group (−25.27°±9.93° before THA [P<0.001] and 33.14°±8.55° after THA [P<0.001]).

The ROM on the affected side after THA was significantly higher than before (P<0.001) (Table 2).

In the frontal plane, the maximum abduction of the affected side (minimum adduction) was lower postoperatively (−3.53°±4.22°) in comparison with the control group (−6.14°±2.22° [P=0.001]). The opposite side displayed both before (−3.04° ±5.72°) and after THA (−3.41°±5.50°) a lower abduction in comparison with the control group (6.14°±2.22° [P=0.006 preoperatively, P=0.004 postoperatively]).

The ROM in the frontal plane was smaller both before and after THA on both sides, the affected and the opposite, in comparison with the control group (P<0.001 for the comparison of the affected side preoperatively with the control group; P<0.001 for the comparison of the affected side postoperatively with the control group; P<0.001 for comparison of the opposite side preoperatively with control group; P<0.001 for the comparison of the opposite side postoperatively with the control group) (Table 2).

In the transverse plane, greater internal rotation (P=0.010 affected side, P=0.001 opposite side) and less external rotation (P<0.001 affected side and P<0.001 opposite side) was observed preoperatively on the affected and opposite side compared to the control group.

Postoperatively, the affected side continued to have less internal rotation (P=0.02) and greater external rotation (P<0.001) compared to the control group.

On the opposite side, external rotation was greater postoperatively than in the control group (P<0.001). Internal rotation was not significantly different postoperatively on the opposite side compared to the control group.

ROM was significantly greater on the affected and opposite side than in the control group, both before and after THA (P<0.001 for the comparison of the affected side preoperatively with the control group; P=0.006 for the comparison of the affected side postoperatively with the control group; P<0.001 for the comparison of the opposite site preoperatively with the control group; P<0.001 for the comparison opposite site postoperatively with the control group) (Table 2).

The trunk in one-legged stance of the affected side in the frontal plane inclined toward the affected side both before and after THA when compared with the control group (P=0.002 preoperatively and P<0.001 postoperatively vs the control group) (Table 3).

Table 3

Kinematics of trunk, pelvis and hip joint during the one-legged stance phase.

	PreOP (N=23)	PostOP (N=23)	Control group (N=24; 25; 38; 39)	Opposite side preoperative (N=23)	Opposite side postoperative (N=23)	p1	p2	p3	p4	p5	p6	p7	p8
Frontal plane
Trunk room	−3.04 (3.99)	−2.76 (2.31)	−0.80 (0.53)	−1.10 (3.45)	−0.68 (2.88)	0.002	0.000	0.855	0.181	0.031	0.841	0.315	0.301
Trunk pelvis	−1.62 (5.43)	−2.23 (4.40)	−1.92 (0.72)	−0.82 5.63)	0.17 (4.32)	0.325	0.168	0.484	0.831	0.114	0.218	0.019	0.224
Pelvis	−1.42 (5.00)	−0.54 (3.93)	1.27 (0.57)	−0.28 (4.78)	−0.85 (3.21)	0.057	0.127	0.201	0.563	0.761	0.169	0.003	0.465
Hip (adduction)	−0.21 (5.57)	2.02 (5.44)	2.38 (2.13)	2.58 (4.60)	2.23 (3.93)	0.113	0.988	0.064	0.274	1.000	0.000	0.795	0.67

PreOP – pre-operative; PostOP – post-operative. p1: p-value preoperative in comparison with healthy control group (Mann-Whitney-U Test); p2: p-value for the test postoperative in comparison with control group (Mann-Whitney-U Test); p3: p-value for test preoperative in comparison with postoperative (Wilcoxon Test); p4: p-value for test preoperative in comparison with opposite side (Mann-Whitney-U Test); p5: p-value test postoperative in comparison with opposite side (Mann-Whitney-U Test); p6: p-value for test opposite side preoperative in comparison with control group (Mann-Whitney-U Test); p7: p-value for test opposite side postoperative in comparison with control group (Mann-Whitney-U Test); p8: p-value for test opposite side preoperative in comparison with opposite side postoperative (Wilcoxon Test).

In the single-leg stand phase of the opposite side, there was no difference in thoracic inclination preoperatively compared to the control group.

After THA, the opposite side displayed reduced inclination of the thorax in relation to the pelvis toward the same side (P=0.019). In addition, the opposite side showed increased inclination of the pelvis to the same side compared with the control group (P=0.003) (Table 3).

Kinetics

Preoperatively, work performed in the hip joint was significantly lower compared with the control group (P<0.001). Postoperatively, the work performed in the hip joint was still significantly lower (0.15±0.07 J/kg [P =0.009]) compared with the control group (0.20±0.05 J/kg). Postoperatively, work of the affected hip joint was higher compared to preoperatively (P<0.001). In comparison with the opposite side (0.17± 0.10 J/kg), the work performed before THA was lower (0.09±0.05 J/kg [P<0.001]). After THA, there was no longer a significant difference between the side operated upon (0.15±0.07 J/kg) and the opposite side (0.21±0.14 J/kg) (Table 4).

Table 4

Positive work.

Positive work	Affected side preOP (n=23)	%	Affected side postOP (n=23)	%	Control (n=39)	%	p1	p2	p3	p4	p5
Hip joint	0.09 (±0.05)	23.08	0.15 (±0.07)	31.25	0.20 (±0.05)	36.36	0.000	0.009	0.000	0.000	0.098
Knee	0.05 (±0.04)	12.82	0.06 (±0.06)	12.50	0.06 (±0.03)	10.91	0.186	0.163	0.964	0.315	0.211
Ankle	0.25 (±0.08)	64.10	0.27 (±0.10)	56.25	0.29 (±0.08)	52.73	0.078	0.633	0.329	0.377	0.234
Total	0.39	100	0.48	100	0.55	100
Positive work	Opposite side preOp (n=23)	%	Opposite side postOP (n=23)	%	Control (n=39)		%	p6	p7		p8
Hip joint	0.17 (±0.10)	33.33	0.21 (±0.14)	35.00	0.20 (±0.05)		36.36	0.116	0.686		0.086
Knee	0.07 (±0.05)	13.73	0.077(±0.06)	11.67	0.06 (±0.03)		10.91	0.761	0.988		0.988
Ankle	0.27 (±0.06)	52.94	0.32 (±0.15)	53.33	0.29 (±0.08)		52.73	0.426	0.654		0.211
Total	0.51	100	0.60	100	0.55		100

PreOP – preoperative; PostOP – postoperative. p1: p-value for the test positive work affected side preoperative in comparison with the control group (Mann-Whitney-U Test); p2: p-value for the test positive work affected side postoperative in comparison with the control group (Mann-Whitney-U Test); p3: p-value for the test positive work affected side preoperative in comparison with positive work affected side postoperative (Wilcoxon Test); p4: p-value for the test positive work affected side preoperative in comparison with positive work opposite side preoperative (Mann-Whitney-U Test); p5: p-value for the test positive work affected side postoperative in comparison with positive work opposite side postoperative (Mann-Whitney-U Test); p6: p-value for the test positive work opposite side preoperative in comparison with the control group (Mann-Whitney-U Test); p7: p-value for the test positive work opposite side postoperative in comparison with the control group (Mann-Whitney-U Test); p8: p-value for the test positive work opposite side preoperative in comparison with positive work affected side postoperative (Wilcoxon Test).

When comparing the absorbed energy, the affected side showed a significantly reduced energy absorption in hip (P=0.039) and knee (P<0.001) joints before THA, whereas in the ankle joint, more energy was absorbed (P=0.001) in comparison with the control group. After THA, the negative energy in the ankle joint remained elevated (P=0.002), whereas no difference to the control group could be detected in the hip joint (P<0.001). The affected side showed still a significantly reduced energy absorption in the knee joint postoperatively (P=0.024).

Before THA, significantly less energy was absorbed in the hip and knee joints of the affected side than on the opposite side. After THA, there was no longer a significant difference between the 2 sides (Table 5).

Table 5

Negative work.

Negative work	Affected side preOP (n=23)	%	Affected side postOP (n=23)	%	Control group	%	p1	p2	p3	p4	p5
Hip joint	−0.09 (± 0.07)	20.93	−0.15 (± 0.14)	28.85	−0.11 (± 0.04)	23.91	0.039	0.806	0.080	0.002	0.142
Knee	−0.15 (± 0.10)	34.88	−0.17 (± 0.08)	32.69	−0.21 (± 0.06)	45.65	0.000	0.024	0.170	0.014	0.273
Ankle	−0.19 (± 0.05)	44.19	−0.20 (± 0.08)	38.46	−0.14 (± 0.05)	30.43	0.001	0.002	0.247	0.893	0.823
Total	−0.43	100	−0.52	100	−0.46	100
Negative work	Opposite side preOP (n=23)	%	Opposite side postOp (n=23)	%	Control group		%	p6	p7		p8
Hip joint	−0.19 (±0.14)	32.20	−0.18 (±0.15)	31.03	−0.11 (±0.04)		23.91	0.012	0.478		0.687
Knee	−0.21 (±0.11)	35.59	−0.20 (±0.09)	34.48	−0.21 (±0.06)		45.65	0.332	0.223		0.988
Ankle	−0.19 (±0.07)	32.20	−0.20 (±0.08)	34.48	−0.14 (±0.05)		30.43	0.003	0.001		0.754
Total	−0.59	100	−0.58	100	−0.46		100.00

PreOP – preoperative; PostOP – postoperative. p1: p-value for the test negative work affected side preoperative in comparison with the control group (Mann-Whitney-U Test); p2: p-value for the test negative work affected side postoperative in comparison with the control group (Mann-Whitney-U Test); p3: p-value for the test negative work affected side preoperative in comparison with negative work affected side postoperative (Wilcoxon Test); p4: p-value for the test negative work affected side preoperative in comparison with negative work opposite side preoperative (Mann-Whitney-U Test); p5: p-value for the test negative work affected side postoperative in comparison with negative work opposite side postoperative (Mann-Whitney-U Test); p6: p-value for the test negative work opposite side preoperative in comparison with the control group (Mann-Whitney-U Test); p7: p-value for the test negative work opposite side postoperative in comparison with the control group (Mann-Whitney-U Test); p8: p-value for the test negative work opposite side preoperative in comparison with negative work affected side postoperative (Wilcoxon Test).

Passive ROM

After THA, the hip flexion (94.13°±15.86°) was significantly higher than before (82.39°±25.40° [P=0.002]). The abduction in the hip joint increased significantly after THA from 18.04° (± 7.03°) to 22.61° (±7.21°, P=0.014). The internal rotation (16.74°±12.02°) and the external rotation (16.30°±9.80°) both increased significantly after THA (27.61°±11.27° internal rotation and 23.04°±6.87° external rotation (P=0.001 for internal rotation and P =0.006 for external rotation) (Table 6).

Table 6

Passive hip range of motion, leg length difference and Harris hip score.

	N	Preoperative*	N	Postoperative*	p1
Leg length difference	23	#1.17 (0.87)	23	#0.94 (1.19)	0.198
Passive hip extension	23	#8.39 (5.80)	23	#8.61 (3.09)	0.787
Passive hip flexion	23	#82.39 (25.40)	23	#94.13 (15.86)	0.002
Passive hip abduction	23	#18.04 (7.03)	23	#22.61 (7.21)	0.014
Passive hip adduction	23	#12.83 (6.37)	23	#14.57 (5.82)	0.336
Passive hip internal rotation	23	#16.74 (12.02)	23	#27.61 (11.27)	0.001
Passive hip external rotation	23	#16.30 (9.80)	23	#23.04 (6.87)	0.006
Pain (HHS)	21	12.38 (±8.31)	21	33.81 (±10.43)	0.000
Function (HHS)	21	12.81 (±3.98)	21	19.52 (±4.09)	0.000
ROM and deformity (HHS)	21	7.29 (±1.06)	21	8.57 (±2.52)	0.004
Gait pattern (HHS)	21	14.90 (±5.03)	21	17.29 (±4.80)	0.026
HHS-total	21	47.38 (±12.25)	21	79.19 (±15.75)	0.000

The mean and simple standard deviation are given;

values are given in angle degree.

p1: p-value for the test preoperative in comparison with postoperative (Wilcoxon Test).

HHS

After THA, there were significantly higher values in all 4 areas of the Harris hip score and for the total score (Table 6). The value for pain increased from 12.38 (±8.31) preoperatively to 33.81 (±10.43) postoperatively (P<0.001). The value for function increased from preoperatively 12.81 (±3.98) to 19.52 (±4.09) postoperatively (P<0.001). The total score increased from 47.38 (±12.25) preoperatively to 79.19 (±15.75) postoperatively (P<0.001) (Table 6).

Discussion

The findings from this retrospective study showed that ROM and HHS significantly improved following THA with the S-ROM® hip prosthesis, that some parameters of gait improved postoperatively in comparison to the preoperative situation, but that gait did not completely return to normal.

Before THA, less weight is put on the affected side by significantly shortening the standing time in relation to the opposite side. This movement is seen as limping to spare the joint and has been recognized in a number of studies [32,33].

One year after THA, a significantly lengthened standing time in comparison with the control group was shown. Compared with the opposite side, no significant difference of the one-legged stance time is apparent. The load put upon both legs after THA becomes almost the same. The one-legged stance phase before THA is significantly shortened compared with both the control group and the opposite leg. One year after THA, the one-legged stance phase was still shorter than in the control group. However, in comparison with the opposite side, there was no longer a significant difference. Thus, we can see that the load put on both legs was becoming more equal after THA. The stance phase is divided into single-supported and double-supported stance phases. The single-supported stance phases include mid-stance and terminal stance. Both together form the one-legged stance phase. Double-supported stance phases include initial contact, loading response, and pre-swing [16]. As the total stance time is not significantly changed, the affected leg is clearly spared by lengthening the doubly supported phases. This can be seen distinctly in a significantly lengthened phase in the loading response in both legs, whereby changes on the opposite side spare the affected leg and lead to a more symmetrical gait. In addition, Nagariya et al were able to identify a preoperative shortening of the one-legged stance phase and a lengthening of the swing phase and of the doubly supported stance phase on the operated side compared to the control group [34]. However, to assess the symmetry of the gait, the opposite leg should also be examined. Four years after THA, no interlimbic differences between the operated and the non-operated leg for the spatiotemporal gait parameters were found by John et al, suggesting a symmetrical gait pattern after this period of time. However, they provided no information on the implant and preoperative differences [35].

Short step length automatically leads to shorter one-legged stance time and thus to relief for the leg. One year after THA, a measurable improvement in spatiotemporal gait parameters was obvious. Nevertheless, there were still distinct differences in the comparison with the control group. This was also evident in previous studies on osteoarthritis patients who were observed over a shorter period of time and with different implants [36,37].

In the terminal stance phase, a hyperextension in the hip joint physiologically takes place. Should this movement be impossible, in order to return the leg to a retroversion position, an anterior leaning of the hip with lordosis lombarde occurs. This compensatory positioning in the lumbar spine can cause long-term problems in the lumbar spine [16]. The reduced hip joint extension could be improved after THR so that both the extension and the whole ROM in the hip joint improved in the postoperative phase. The ROM was nonetheless still below that of the control group. This was also demonstrated by Queen et al and Mendiolagoitia et al [3,38]. Leijendekkers et al also showed increased anterior pelvic tilt and reduced hip extension on the operated side. They attributed this to hip flexion contracture in dysplasia patients. However, in their study, only 6 patients were examined and there were no preoperative values for comparison [17]. Stief et al showed that, even 2 years postoperatively, there was a persistently reduced hip extension in the terminal stance phase compared to the control group in 15 osteoarthritis patients without severe deformity [6]. In the period after THA, the increased pelvis anteversion was not significantly reduced in our collective. Parratte et al were able to show a significant reduction in pelvic anteversion of 3.1° at 12 months after THA [39]. Stief et al showed no difference in pelvic tilt in the sagittal plane compared to the control group 2 years postoperatively [6]. However, in both studies, patients with dysplasia and severe deformity were excluded. The increased pelvic tilt in the sagittal plane is a reaction to the limited mobility of the hip joint. The compensatory movement of the pelvis in this study is reduced but not completely eliminated. Weber et al demonstrated that THA has an influence on the change of pelvic tilt postoperatively, and suggested considering this when aligning the implant [7]. Physiologically, the pelvis and trunk reach their lowest point in the pre-swinging phase, whereas the hip joint reaches the maximum extension position in the transition from the terminal stance phase to the pre-swinging phase [16]. Reduced hyperextension in the hip joint leads to less lowering of the pelvis and shortened step length, which causes a reduction in extension in the hip joint on the opposite side.

In the frontal plane, progression of osteoarthritis leads to a significant decrease in ROM of the hip joint. This improved significantly after THA in this study. Normal ROM, however, was not attained. In the transversal plane, physiologically, the maximum internal rotation occurs in the loading response, when the leg takes the weight of the body [16]. In external rotation the capsular ligaments are relaxed and the strain on the hip joint is reduced, as shown by Weißgraeber et al [40], whereby the hip joint is spared when taking the burden of body weight. In this study, the ROM in transversal plane was significantly increased before and after THA. The increased ROM was apparent also on the opposite side and did not change in the postoperative period. An insufficiency of the abductors results in a leaning of the pelvis toward the opposite side. To compensate for this, the trunk inclines toward the standing foot in order to reduce the lowering of the pelvis [16,41]. This Trendelenburg limping was apparent in both the preoperative and the postoperative periods. The positioning of the implant seems to have a decisive influence on the hip abduction force. Mahmood et al showed that a reduction in global femoral offset (GFO) of more than 5 mm following surgical THA reduces abductor muscle strength and should be avoided. Restoration and increasing of GFO gave better results that were comparable, with no negative effects [42]. Kubota et al listed this mechanism as one of the characteristics of bilateral hip joint arthrosis [43]. In this study, after THA, the opposite side displayed less inclination of the thorax in relation to the pelvis to the same side. In addition, an increase in the inclination of the pelvis to the same side was apparent compared to the control group. Stief et al showed no difference in pelvic and thoracic obliquity in the frontal plane compared to the control group 2 years postoperatively [6,17]. Leijendekkers et al showed an increased lateral flexion in the ipsilateral trunk on the operated side in dysplasia patients. However, in their study, only 6 patients were examined and there were no preoperative values for comparison, so that no statement can be made about a possible postoperative improvement [17]. To compensate for a weakness of the hip flexors on the opposite side, the pelvis is tilted down toward the standing leg and lifted on the side of the swinging leg to lift the swinging leg from the ground. To keep one’s balance in this position, the upper body is tilted slightly to the side of the standing leg. In this study, the deviations were below one degree in comparison with the control group, so they are imperceptible to the human eye. Arnold et al showed an improvement in upper body deviation in the frontal plane after THA, but normal values were not achieved [44]. Miltner et al demonstrated that chronic pain before THA leads to more pathological movement patterns after THA [45]. During walking, energy is released and absorbed with every footstep to produce steady forward movement. Seen from a physical point of view, work is the product of force and the distance covered [46]. Seichert et al showed that from measured ground reaction forces, active and passive proportions can be differentiated by calculating the power curve and thus functional deficits can be ascertained [47]. The measured forces developed in this study in the knee ankle and opposite hip joint stayed equal to the control group. In this study, the force generated in the affected hip joint was lower than in the control group, both before and after THA. The knee, ankle, and opposite hip joint did not develop higher force. In the present study, THA led to a significant increase in force in the hip joint operated upon to a level that after THA no significant difference was seen between the force generated in the operated and the opposite hip joint, unlike Queen et al, who discovered a postoperative reduction in the ground reaction forces on the operated side, resulting in asymmetrical strain [38]. With respect to joint load bearing, Stief et al observed a shift in knee joint load distribution from medial to lateral, so that correspondingly in the hip joint greater adduction moment was observed. They concluded that patients with hip replacement may be at higher risk for abnormal joint loading and thus for the development of osteoarthritis in other lower-extremity joints [6]. John et al showed, 4 years postoperatively, still reduced strength values for hip flexion, hip extension, and hip abduction compared to the values of healthy control groups [35]. Postoperative reduced hip abduction force and hip flexion force have been demonstrated in several studies [3]. When considering the energy absorbed in this study, a preoperative reduction in energy absorption in knee and hip joints and an increase in energy absorption in the ankle joint was found, as well as in the hip and ankle joints on the opposite side. After THA, the energy absorbed in the hip and knee joint on the affected side increased in this study, so that there was now no difference from the control group. To a certain extent, the hip joint operated upon can also take up more force compared to preoperatively, equalizing the forces in both hip joints. The absorption of energy requires force and represents a strain. The ankle joint of the affected side in this study absorbed significantly more energy in the preoperative and postoperative phases in relation to the control group, which can be regarded as a compensatory reaction. Rasch et al found an improvement of muscle strength in hip and knee joints, which still continued 2 years after THA, although normal values were not attained [48].

Bahl et al showed that the ground reaction force profiles do not reflect the hip reaction force profile. Therefore, ground reaction force cannot be used as a surrogate for internal hip joint loading. However, reduced gait velocity and reduced hip flexion and extension range were significantly correlated with the prevalence of a single-peak profile [49]. After THA, passive ROM in the affected hip joint showed clear improvement. This increase in passive ROM is not reflected in gait. This illustrates that patterns of movement displayed over years are hard to correct. Behave et al used the findings gained by gait analysis and isokinetic force recordings to specifically treat functional deficits existing after THA, which led to demonstrable success [50,51]. According to Widhager et al, gait disorders can be improved by postoperative mental training, which proves that becoming accustomed to a pathological gait contributes to its continuance [52]. John et al showed reduced hip strength and active hip ROM values even 4 years after THA compared to healthy control groups [35]. This suggests that active ROM is obviously more important for walking than is passive ROM.

The Harris hip score is an appropriate and therefore an often-applied procedure to assess the clinical outcome of hip replacement. The reliability and validity of the score were already examined in studies. There was significant progress one year after THA in all parameters of the Harris hip score. The results are comparable with different studies in which patients with an S-ROM endoprosthesis were examined for a long period of time and partially after previous adjustment osteotomy for proximal femur deformity. Improvements by 39 to 49 points to a total of 83 to 90 points were achieved [53-55]. Our results with patients with complex deformities are slightly lower with an average improvement of 32 points to an average final total of 79.19 points.

Limitations of This Study Are

This was a single-center, retrospective study of a small cohort of 23 patients. The Harris hip score (HHS) is commonly used to evaluate the outcomes of primary THA in clinical trials. However, the application of the HHS may be limited in studies that evaluate and compare new techniques [56].

Conclusions

The findings from this retrospective study showed that ROM, HHS, and gait pattern of patients with severe deformity improved following THA with the modular S-ROM hip prosthesis, but gait did not completely return to normal. Further investigations are needed to show how long gait patterns change postoperatively and whether specific training influences gait parameters positively.