Comparison of Outcomes in Dual-Mobility Versus Fixed-Bearing Implants in Primary Total Hip Arthroplasty: A Systematic Review and Meta-Analysis

Total hip arthroplasty (THA) is often regarded as one of the most successful surgical techniques developed in the twenty-first century. However, it is associated with complications such as prosthetic instability, dislocations, or infections. Dual-mobility (DM) implants have been developed with the goal of reducing the incidence of dislocations by increasing the femoral head-neck ratio, maximising hip stability, and improving the range of motion (ROM) before impingement and dislocation. This systematic review aims to comprehensively compare the safety and efficacy of DM versus fixed-bearing (FB) implants in primary THA patients. A comprehensive search strategy of PubMed, Embase, Scopus, and Web of Science Core Collection databases was executed to identify pertinent literature comparing DM and FB implants in THAs. Eligible studies underwent independent screening, and data were systematically extracted. The analysis employed pooled risk ratios (RR) for dichotomous outcomes and mean differences (MD) for continuous variables, each accompanied by their respective 95% confidence intervals (CI). Our systematic review and meta-analysis included nine studies encompassing 22,277 patients. The DM group had a significantly reduced incidence of dislocations compared to the FB group (RR 0.25, 95%CI [0.13, 0.47]; p-value <0.0001) and a significantly shorter length of stay (MD −9.92, 95%CI [−15.53, −4.32]; p-value = 0.0005). The FB group, however, had a significantly shorter operative time compared to the DM group (MD 10.41, 95%CI [7.64, 13.17]; p-value < 0.00001). We did not identify any significant statistical differences between the DM and FB groups regarding patient-reported outcome measures, the incidence of all-cause readmissions, the incidence of peri-prosthetic fractures, the incidence of infections, or the incidence of groyne pain.


Introduction And Background
Although total hip arthroplasty (THA) is often regarded as one of the most successful surgical techniques developed in the twenty-first century, it is also associated with complications such as prosthetic instability, dislocations, or infections [1].The risk of dislocation is 1% within one month of surgery and 2% within one year of surgery, increasing by 1% every five years up to 7% at 25 years post-surgery [2].Recurrent dislocations are problematic because they reduce a patient's quality of life and may require revision surgery [3].Instability and dislocations as a mode of failure after a THA are multifactorial problems that have been attributed to surgical factors such as surgical approach, component positioning, and component design, and patient factors such as abnormal spinopelvic mobility or spinopelvic impairment, neuromuscular disorders, cognitive dysfunction, age >75 years, abductor insufficiency, dysplastic hip, and previous surgical history [4][5][6].Surgical techniques and modalities such as large prosthetic heads, trochanteric advancement, constrained liners, modular components, and dual-mobility (DM) components have thus been developed with the goal of reducing the incidence of dislocation [5].DM implants were first developed in 1974 by Gilles Bousquet and Andre Rambert in France and feature a small femoral head articulating within a movable polyethylene liner that itself articulates within a larger acetabular shell [7].The DM implant's two articulations combine the principles of Charnley's low-friction arthroplasty [8] and McKee-Farrar's concept of hip stability [9] to increase the femoral head-neck ratio, which maximises hip stability and improves range of motion (ROM) before impingement and dislocation [10].Anatomic dual mobility (ADM) systems and modular dual mobility (MDM) polyethylene liners were developed to increase impingement-free ROM and the effective implant head size, thereby decreasing the risk of dislocations that often arise via primary impingement between the neck of the prosthesis and articular component and secondary impingement between the bony femur and pelvis [6,11].
The benefits of DM implants compared to fixed-bearing (FB) implants were demonstrated by Stroh et al., who demonstrated dislocation rates of 0.1% and 3.5% for primary and revision DM THA, respectively, versus 2-7% and 16% for primary and revision FB THA, respectively [12].However, the use of DM implants remains limited and is selectively used for patients with a high dislocation risk or high risk of post-operative instability, such as elderly patients, neuromuscular deficient patients, dysplastic patients, post-traumatic patients, patients with increased ligamentous laxity, patients with smaller anatomy, and patients with spinal abnormalities that reduce pelvic flexion, such as advanced degenerative spinal pathology and spinal fusion [10,13].This hesitancy to implement a more widespread use of DM implants has been attributed to previous reports of complications such as intra-prosthetic dissociation (IPD) and accelerated wear of polyethylene acetabular liners [6,14,15].
Further investigations are thus needed to provide more comprehensive insights into the effectiveness of DM implants compared to FB implants.As such, our study aims to comprehensively compare the safety and efficacy of DM versus FB implants in primary THA patients.

Search Strategy
This study was performed according to the Preferred Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [16].A literature search was carried out on the PubMed, Embase, Scopus, and Web of Science Core Collection databases from their respective inceptions through April 24, 2024.The terms used were (total hip replacement OR total hip arthroplasty) AND (dual-mobility OR dual mobility) AND (fixedbearing OR fixed bearing).Boolean operators (AND, OR) were used to combine terms and narrow search results.Reference lists of eligible articles and previous meta-analyses were examined manually to identify relevant citations.

Eligibility Criteria
During title and abstract screening, animal studies, case reports, technical reports, systematic reviews, letters to the editor, and conference abstracts were excluded.Non-comparative studies were also excluded.Studies that did not address DM versus FB implants in THA were excluded.During the full-text screening, only full-length manuscripts focused on comparing DM versus FB implants in THA were included.This process was carried out by two reviewers, and discrepancies were resolved by discussion between the two reviewers.

Study Selection
Studies were uploaded as abstracts to the online systematic review management platform Covidence (Melbourne, Australia), from which duplicate studies were automatically identified and deleted.Two reviewers evaluated the retrieved studies independently.Discrepancies were resolved by discussion between the two independent reviewers.

Data Collection
Two reviewers independently extracted study characteristics, which were recorded using Microsoft Excel sheets (Microsoft® Corp., Redmond, WA, USA).The information extracted included the author, year, country of origin, study design, number of patients, ages, gender distribution, body mass index (BMI), American Society of Anaesthesiology (ASA) score, follow-up duration, outcomes, surgical approaches, and descriptions of implants used.
The primary outcome of this review was to compare the following between primary THA with DM implants and primary THA with FB implants: (i) operative time; (ii) length of stay; (iii) patient-reported outcome measures (PROMs); (iv) incidence of all-cause readmission; (v) incidence of dislocations; (vi) incidence of infection; (vii) incidence of periprosthetic fracture; and (viii) incidence of groyne pain.

Data Analysis
Data were analysed using RevMan Web (Cochrane, London, UK).Mean differences (MD) with standard deviations were pooled for continuous outcomes.Risk ratios (RR) with 95% confidence intervals (CIs) were employed using the Mantel-Haenszel method for dichotomous outcomes.A fixed-effect model was used for homogeneous studies.A random-effects model was used for heterogeneous studies.Statistical heterogeneity was evaluated through the I2 and Chi2 tests.A p-value < 0.10 indicated heterogeneity, and I2 ≥ 50% indicated high heterogeneity.

Assessment of the Quality of Studies
Two reviewers used the Newcastle-Ottawa Scale (NOS) to assess the methodological quality and risk of bias in eligible studies [17].This system assesses the quality of cohort studies based on selection, comparability, and outcome/exposure criteria.Scores range from zero to nine, and higher NOS scores indicate a lower risk of bias.The following variables were identified to likely confound associations between the type of implant and outcomes: (i) BMI, (ii) ASA score, and (iii) reason for primary hip replacement.Observational studies that attempted to control for one or more of these theoretical confounds were at a decreased risk of comparability bias based on NOS criteria.

Results
The literature search provided 135 potentially relevant articles from PubMed (n = 37), Embase (n = 35), Scopus (n = 32), and the Web of Science Core Collection (n = 31) (Figure 1).After excluding 84 duplicates, 51 records were available for title and abstract screening.Forty publications were removed after reviewing the titles and abstracts.Of the residual corpus of literature, full-text reviews were performed.This identified nine relevant studies for inclusion in this systematic review and meta-analysis.

Surgical Approach
Regarding surgical approaches, two studies utilised posterior approaches [18,19], four utilised posterolateral approaches [10,[20][21][22], one used both posterior and anterior approaches [24], and two did not specify the surgical approach used [13,23].Rowan et al. further elaborated that a capsular and posterior short-external rotator repair was subsequently undertaken [10].Moore et al. specified that a posterior approach was used in 173 DM cases and 2,378 FB cases, while an anterior approach was used in 5 DM cases and 1,022 FB cases [24].

Operative Time
The mean operative time in minutes was assessed in two studies, including a total of 15,265 patients [13,23].
Our analysis showed a substantial difference in operative time between the DM and FB groups (MD 10.41, 95%CI [7.64, 13.17]; p-value<0.00001).Pooled studies were homogenous, so a fixed effect model was used.I2 was 0%, and Chi2-p was 0.33. Figure 2 depicts the forest plot for operative time.

Length of Stay
Two studies assessed the length of stay in hours, including 4,015 patients [23,24].Our analysis showed a substantial difference in length of stay between the DM and FB groups (MD −9.92, 95%CI [−15.53,−4.32]; pvalue = 0.0005).Pooled studies were homogenous, so a fixed-effect model was used.I2 was 0%, and Chi2-p was 0.87. Figure 3 depicts the forest plot for the length of the stay.

All-Cause Readmissions
All-cause readmissions were assessed in three studies [13,18,23].At 30 days, all-cause readmissions were only assessed in one study encompassing 882 patients [18].There was no significant statistical difference in the incidence of 30-day readmission between DM and FB groups (RR 0.38, 95%CI [0.13, 1.13]; p-value = 0.08).At 60 days, all-cause readmissions were only assessed in one study encompassing 882 patients [18].There was no significant statistical difference in the incidence of 60-day readmission between the DM and FB groups (RR 0.66, 95%CI [0.25, 1.73]; p-value = 0.39).At 90 months, all-cause readmissions were assessed in three studies, encompassing a total of 16,147 patients [13,18,23].There was no significant statistical difference in the incidence of 90-day readmission between the DM and FB groups (RR 1.10, 95%CI [0.85, 1.41]; p-value = 0.47).Pooled studies reporting 90-day all-cause readmissions were homogenous, so a fixed effect model was used.I2 was 8%, and Chi2-p was 0.34. Figure 4 depicts the forest plot for all-cause readmissions.

FIGURE 4: Forest plot of all-cause readmissions
DM: dual mobility; FB: fixed bearing
HHS was reported in two studies encompassing 1,143 total patients [18,22].Epinette recorded HHS at a minimum of two years of follow-up [22], while Dubin et al. recorded HHS at a mean of 2.09 years and 1.83 years in the DM and FB groups, respectively [18].There was no significant statistical difference in HHS between the DM and FB groups (MD 1.34, 95%CI [−2.84, 5.53]; p-value = 0.53).The pooled studies reporting HHS were heterogenous, indicating the complexity of the research.A random effect model was used, with I2 at 91% and Chi2-p at 0.001.Figure 5 depicts the forest plot for HHS, further illustrating the heterogeneity in the studies.

Infections
The incidence of infections was recorded in five studies, encompassing 16,359 total patients [10,13,18,20,22].Our analysis showed no significant statistical difference in the incidence of infections between DM and FB groups (RR 0.87, 95%CI [0.48, 1.57]; p-value = 0.64).Pooled studies were homogenous, so a fixed effect model was used.I2 was 33%, and Chi2-p was 0.20.Figure 9 depicts the forest plot for the incidence of infections.

Discussion
This systematic review and meta-analysis aimed to compare the outcomes between DM implants and FB implants used in THA.Nine studies, encompassing a total of 22,277 patients, were included in the final analysis.The DM group had a significantly reduced incidence of dislocations compared to the FB group (pvalue <0.0001) and a significantly shorter LOS (p-value = 0.0005).The FB group, however, had a significantly shorter operative time compared to the DM group (<0.00001).We did not identify any significant statistical differences between the DM and FB groups regarding HHS and mHHS, the incidence of all-cause readmissions, the incidence of peri-prosthetic fractures, the incidence of infections, or the incidence of groyne pain.
The two distinct articulations in DM implants were introduced with the aim of decreasing the risk of dislocation after a THA by using a larger effective head diameter, allowing for a greater range of motion before complete dislocation can occur, increasing stability, and reducing impingement between the neck and cup rim [25].The results of our study demonstrated this reduced risk of dislocation in DM implants compared to FB implants.These findings also align with previous literature [25][26][27] [26].Lastly, although patients with hip dysplasia have a high risk of dislocation due to anatomical deformities, DM implants have been found to lower the risk of post-operative dislocations [21].
Regarding mean operative time, our analysis showed a substantial difference in operative time between the DM and FB groups.Simcox et al. attributed this to the time required to assemble the DM implant and surgeons utilising a DM implant after failing to achieve desired stability with a trial FB implant [13].
Similarly, a study of DM implants in revision THAs by Harwin et al. also did not find any statistically significant improvement in pre-operative and post-operative HHS scores [27].Conversely, a systematic review of DM implants in patients younger than 55 years old by Zampogna et al. found that the mean value of the HHS score increased from 50.9 pre-operatively to 91.6 post-operatively [26].This may suggest that DM implants may be more well tolerated in young patients undergoing a primary THA compared to patients undergoing a revision THA.
Regarding other post-operative complications, we found no significant statistical difference between the DM and FB groups regarding the incidence of all-cause readmissions, the incidence of peri-prosthetic fractures, the incidence of infections, or the incidence of groyne pain.In addition to these complications that were analysed in this study, other complications that may arise from a THA include venous thromboembolism [13,20], pneumonia [20], urinary tract infections [20], enteritis with ileus [20], rheumatoid flares [20], and post-operative anaemia [13].Other complications of DM implants that must be noted are intra-prosthetic dissociation (IPD), which can lead to wear, metallisation, and collision of the femoral head [21].
In addition to better outcomes when using DM implants versus FB implants, DM implants also appear to be more cost-effective than FB implants [28].A medicoeconomic study in France carried out by Epinette et al. found that, when comparing the costs between the two types of implants in a cohort of 100,000 patients while assuming a relative risk of dislocation of 0.4 with DM implants versus FB implants, DM implants avoid 3,283 dislocations, 882 revisions, and 932 rehabilitation unit admissions while potentially saving 28.3 million euros per year for 100,000 THAs performed annually and 283 euros per patient [28].An analysis of 235,857 revision THAs by Bozic et al. [1] reported that revision THAs had an average cost of $23,130 ± 36,643 per hospitalisation and a higher length of stay for revision THAs (mean ± SD, 5.8 days ± 14.0 days) than revision TKAs (mean ± SD, 4.8 days ± 10.5 days).Therefore, the cost-effectiveness of DM implants should be considered when choosing between DM and FB implants for a THA.
This study was not without limitations.The literature search utilised four databases (PubMed, Embase, Scopus, and Web of Science Core Collection), which may not have included all relevant studies that may have been available on other databases.The search was limited to articles in English and grey literature was not examined, so relevant articles that fall under these categories would have been excluded.Confounding factors in baseline characteristics such as BMI and ASA scores across studies could have influenced observed outcomes.Sample sizes between studies also ranged from 126 [20] to 14,818 [13].Simcox et al. also featured unequal arms (DM vs. FB) with a substantial ratio disparity (1:28), which could introduce bias and impact the meta-analysis [13].Follow-up times also ranged from three months [13,24] to six years [22], which may have influenced the comparative analysis given that there is no fixed timing at which post-operative complications can arise.

Conclusions
Our review showed that patients who had undergone a primary THA with DM implants had a significantly reduced incidence of dislocations and a significantly shorter length of hospital stay compared to those who had received FB implants.However, THAs that use FB implants have a significantly shorter operative time.There were no significant differences in PROMs such as the HHS and mHHS, the incidence of all-cause readmissions, the incidence of peri-prosthetic fractures, the incidence of infections, or the incidence of groyne pain.DM implants also appear to be more cost-effective than FB implants.These factors that we have discussed in our review should, therefore, be considered and evaluated when choosing between DM and FB implants for a primary THA.

FIGURE 1 :
FIGURE 1: PRISMA flow chart PRISMA: Preferred Items for Systematic Reviews and Meta-Analysis

FIGURE 2 :
FIGURE 2: Forest plot of operative time DM: dual mobility; FB: fixed bearing

FIGURE 3 :
FIGURE 3: Forest plot of length of stay DM: dual mobility; FB: fixed bearing

FIGURE 8 :
FIGURE 8: Forest plot of incidence of peri-prosthetic fractures DM: dual mobility; FB: fixed bearing

FIGURE 9 :
FIGURE 9: Forest plot of incidence of infections DM: dual mobility; FB: fixed bearing

FIGURE 10 :
FIGURE 10: Forest plot of groin pain DM: dual mobility; FB: fixed bearing

Table 2 . Author (year) Study arm Acetabular component Acetabular liner Femoral head Femoral head diameter (mm) Acetabular cup diameter (mm) Femoral implant
[20,21] Moon et al. used Microplasty, VerSys, and Wagner cone implants (Zimmer Biomet, Inc., Warsaw, IN, USA), with the Wagner cone being used in patients with severely narrow bone marrows[20,21].A summary of the implants used can be found in

TABLE 3 : Newcastle-Ottawa Scale Scores
NOS: Newcastle-Ottawa Scale [25]study byHarwin et al.comparing outcomes of DM implants versus FB implants in revision THA found one dislocation among 85 patients who underwent revision THA using DM implants and six dislocations among 170 patients who underwent revision THA using FB implants (1.18% versus 3.53%)[27].A previous systematic review and meta-analysis by Romagnoli et al. encompassing 2,408 patients and comparing the efficacy of DM implants in preventing dislocation after THA compared to FB implants concluded that DM implants resulted in lower incidences of dislocations (RR 0.16, 95%CI [0.09, 0.28]; p-value <0.00001)[25].Another systematic review by Zampogna et al., encompassing 1,530 patients and analysing the outcomes and complications of DM implants in patients younger than 55 years old, reported 46 (2.7%) cases of dislocations where the PE liner dissociated from the femoral head