Abstract
Background
Robotic-assisted total hip arthroplasty (R-THA) affords precision yet uncertain clinical benefits. This study compares dislocation rates and related revisions between R-THA and manual total hip arthroplasty (M-THA). Secondarily we evaluated cup position, patient-reported outcome measures (PROMs), and postoperative complications.
Methods
A three-surgeon cohort study was conducted on 2247 consecutive patients (1724 M-THA and 523 R-THA) who received a primary THA between January 2014 and June 2020 at a single hospital. Demographics, PROMs, emergency department visits, readmissions, and 90-day complications were collected via the Michigan Arthroplasty Registry Collaborative Quality Initiative. Chart review yielded instability occurrence with an average follow-up of 4 years. Multivariate regression analysis was performed, and a sample of 368 radiographs, including all dislocations, were assessed.
Results
There were significantly lower rates of dislocation in R-THA (0.6%) vs M-THA (2.5%; Multivariate odds ratio 3.74, P < .046). All cases of unstable R-THA were successfully treated conservatively, whereas 46% of unstable M-THA were revised for recurrent instability. Cup anteversion (25.6° ± 5.4° R-THA vs 20.6° ± 7.6° M-THA) was greater, and cup inclination (42.5° ± 5.3° R-THA vs 47.0° ± 6.7° M-THA) was lower in the R-THA group (P < .05). No significant differences were noted for demographics, PROMs, or other complications (P > .05).
Conclusion
R-THA resulted in less than one-fourth the dislocation rate compared to M-THA and no revision for instability. It was associated with no difference in PROMs or other early complications. The influence of R-THA on stability goes beyond simply cup positioning and deserves further study.
Keywords
Total hip arthroplasty (THA) has been cited as the surgery of the century for its effectiveness at reducing pain and improving function [
[1]
]. Despite the success of THA for more than half a century, mechanical complications such as instability and aseptic loosening continue to occur secondary to component positioning, soft tissue balance, or component failure [2
, 3
, 4
]. These complications have an established effect on not only patient outcomes but also economic productivity and overall health care costs.Robotic-assisted total hip arthroplasty (R-THA) seeks to address these complications by offering increased technical precision and information for which the surgical team to act upon with a goal to limit variability. The precision, accuracy, and overall cost of robotic-assisted surgery continue to be a well-studied topic in current arthroplasty literature [
5
, 6
, 7
, 8
, 9
, 10
, 11
, 12
, 13
]. The initial added cost of R-THA has been a subject of debate, as recent literature has demonstrated excellent reproducible component positioning with conflicting data regarding benefit for outcomes and overall costs [14
, 15
, 16
, 17
]. The current controversy with robotic-assisted surgery stems from limited literature that demonstrates improved early outcomes in small series but little longer term data.The primary purpose of this study is to compare dislocation rates and their corresponding revisions between R-THA and manual total hip arthroplasty (M-THA). Secondarily, the study investigated acetabular cup position, available postoperative patient-reported outcome measures, and 90-day postoperative complications.
Materials and Methods
We performed a retrospective cohort analysis of 2247 consecutive patients that received a primary THA by 3 adult reconstruction fellowship-trained specialist surgeons at a single suburban teaching hospital between January 2014 and June 2020 [
[18]
]. All surgeons were at least 2 years into their own adult reconstruction practice at the start of this timeframe, which should minimize learning curve concerns. At the midpoint of the study period, their average time in independent practice was 7 years. Inclusion criteria included patients undergoing primary THA within the study period. Exclusion criteria were revision THA, THA secondary to trauma or hardware failure, and hip hemiarthroplasty.Two surgeons performed R-THAs while all three surgeons contributed to the M-THA cohort. One surgeon performed the vast majority of the R-THA as it became his standard procedure for all primary arthroplasty from 2017 onward. R-THAs were performed with a single system (MAKO, Stryker, Kalamazoo, MI) via a minimally invasive posterior approach to the hip in the lateral decubitus position. The express robotic workflow was utilized that does not rely upon navigated femoral implant assessment and simply helps place the cup and assess the length and offset with planned 3-dimensional template sizing. Similar posterior-based approaches were used for all M-THAs, and the acetabular component was placed before the femoral component. The combined version goal during broaching and trialing was 45-50 degrees in the majority of cases, along with stability during range of motion assessment. All surgeons typically used press-fit, uncemented implants with single-tapered stems unless otherwise indicated. Resident physicians were present in nearly all of the procedures. The majority of the robotic THAs were done with a senior resident while the manual THAs were evenly split between a postgraduate year 2 and postgraduate year 5. Teaching style and autonomy for the procedure was variable based on the seniority and skill of the resident, as well as senior staff involved. All surgeries were performed under spinal anesthesia unless otherwise contraindicated, and employed the same postoperative pain control regimen. Postoperatively all patients followed the same standardized rehabilitation protocol at that time. All clinical data, both inpatient and outpatient, was stored in the same electronic medical record for the entire study period.
Data collection was conducted following institutional review board approval. The study developed a patient list by utilizing registry data to ensure accuracy. Patient demographics, postoperative emergency department visits, readmissions, and 90-day complications were queried through the Michigan Arthroplasty Registry Collaborative Quality Initiative (MARCQI) prospective database. Registry data was entered by full-time abstractors assigned to such collections to ensure consistency. Patient-reported outcome measures were available in the registry beginning late 2017 onward, with a capture rate of over 95% during that period. As MARCQI data was limited to 90-day follow-up, further manual electronic medical record review was conducted to document the incidence of hip dislocations, the number of dislocations, dislocation secondary to mechanical failure, and revision secondary to instability for all patients. This query was strengthened beyond the review of clinic notes by a newly available electronic medical record search function (EPIC Systems, Verona, WI). A global search utilizing the terms ‘dislocation’ and ‘dislocated’ was used to capture every encounter (ie, emergency department, radiology, and operative notes) that mentioned a dislocation event to minimize omitted documentation. The minimum follow-up was 6 months, with an average of over 4 years across the cohort. Incidence of lumbar fusion prior to THA was also recorded by reviewing all operative notes within our health system.
A representative, randomized sample of 368 anteroposterior pelvis post-THA radiographs, which included all dislocations, was assessed for acetabular implant positioning. Radiographs were obtained through the hospital picture archiving and communication system and analyzed using the ellipse method within our electronic templating system (Orthoview Digital Planning Software, Materialise, Belgium). All radiographs were assessed for cup anteversion and inclination by two orthopedic surgery residents to ensure reproducibility and precision of measurements. Cup inclination was measured using the profile of the ischial tuberosities to set the horizontal axis. Cup anteversion was measured using the OrthoView anteversion smart tool on the anteroposterior radiograph for angle computation. Any outlier measurements were corroborated and corrected by a senior resident.
Statistical Analysis
Continuous data are described using mean and median and are compared between the R-THA and M-THA groups using independent two-group t-tests or Wilcoxon rank-sum tests based on distribution. An average of the two measurements for cup anteversion and inclination was used for comparison between R-THA and M-THA. All categorical data are presented using counts and relative percentages and compared between the two groups using chi-square or Fisher’s exact tests based on cell counts. Multivariate logistic regression models controlling for gender, race, body mass index, age, preoperative American Society of Anesthesiologists score, surgeon, and history of lumbar surgery were used to determine any independent predictors of overnight admission, readmission, emergency department visit, 90-day complication, or subsequent dislocation event with results presented as odds ratio with 95% confidence interval and respective P values. All analyses were performed using SAS 9.4 (SAS Institute Inc, Cary, NC).
Results
A cohort with 2247 patients underwent primary THA within the study period. There were 1724 (76.7%) M-THAs and 523 (23.3%) R-THAs included in the analysis. Univariate analyses (Table 1) demonstrated both groups had comparable demographics with the exception of the lesser mean age in the R-THA group (64.47 vs 65.63 years, P = .030) and lesser incidence of femoral component cement fixation in the M-THA group (0.19% vs 1.35%, P = .028) (Table 1). Intraoperative and postoperative outcomes were mostly comparable, though R-THA patients were found to have significantly less operating room (OR) time (69.98 vs 90.13 minutes, P < .001), lesser average postoperative length of hospital stay (days) (P < .001), and lower periprosthetic dislocation rate without mechanical failure (0.57% vs 2.49%, P = .007) (Table 2). All robotic dislocations were successful with conservative treatment without recurrence (0 of 3), whereas 46% (20 of 43) of traditional dislocators were ultimately revised for recurrent instability. There were no significant differences between available R-THA and M-THA in preoperative or postoperative patient-reported outcomes measurement information system global health, mental health, and physical health (PROMIS-GH, PROMIS-MH, PROMIS-PH), as well as hip disability and osteoarthritis outcome score (HOOS, JR) (Table 3).
Table 1Univariate Comparison of Demographics Between Robotic and Nonrobotic Total Hip Arthroplasty.
| Covariate | Statistics | Level | Robotic | P Value | |
|---|---|---|---|---|---|
| No (N = 1724) | Yes (N = 523) | ||||
| Sex | N (%) | Female | 1006 (58.35) | 293 (56.02) | .345 |
| N (%) | Male | 718 (41.65) | 230 (43.98) | ||
| Age (y) | N | 1724 | 523 | .030 | |
| Mean | 65.63 | 64.47 | |||
| Median | 66 | 65 | |||
| Race | N (%) | Asian | 8 (0.46) | 3 (0.57) | .262 |
| N (%) | Black | 397 (23.03) | 100 (19.12) | ||
| N (%) | Caucasian | 1179 (68.39) | 388 (74.19) | ||
| N (%) | Native American | 2 (0.12) | 0 (0) | ||
| N (%) | Native Hawaiian-Pacific Islander | 1 (0.06) | 0 (0) | ||
| N (%) | Other | 21 (1.22) | 6 (1.15) | ||
| N (%) | Unknown | 116 (6.73) | 26 (4.97) | ||
| Body mass index (kg/m2) | N | 1724 | 523 | .731 | |
| Mean | 30.72 | 30.62 | |||
| Median | 30.1 | 29.9 | |||
| Diabetes mellitus | N (%) | No | 1400 (81.21) | 418 (79.92) | .486 |
| N (%) | Yes - type 1 | 3 (0.17) | 0 (0) | ||
| N (%) | Yes - type 2 | 321 (18.62) | 105 (20.08) | ||
| Bleeding disorder | N (%) | No | 1719 (99.71) | 521 (99.62) | .740 |
| N (%) | Yes | 5 (0.29) | 2 (0.38) | ||
| History of deep venous thrombosis or pulmonary embolism | N (%) | No | 1612 (93.5) | 485 (92.73) | .690 |
| N (%) | Unknown | 1 (0.06) | 0 (0) | ||
| N (%) | Yes | 111 (6.44) | 38 (7.27) | ||
| Use of assistive device preoperatively | N (%) | No | 974 (56.5) | 303 (57.93) | .094 |
| N (%) | Unknown | 15 (0.87) | 0 (0) | ||
| N (%) | Yes | 735 (42.63) | 220 (42.07) | ||
| American Society of Anesthesiologists score | N (%) | I | 50 (2.9) | 24 (4.59) | .281 |
| N (%) | II | 759 (44.03) | 221 (42.26) | ||
| N (%) | III | 888 (51.51) | 269 (51.43) | ||
| N (%) | IV | 27 (1.57) | 9 (1.72) | ||
| History of lumbar surgery | N (%) | No | 1618 (93.85) | 479 (91.59) | .069 |
| N (%) | Yes | 106 (6.15) | 44 (8.41) | ||
| Method of femoral component fixation | N (%) | Cemented | 11 (1.35) | 1 (0.19) | .028 |
| N (%) | Uncemented | 801 (98.65) | 521 (99.81) | ||
Bolded text indicates statistical significance (P < .05).
Table 2Univariate Comparison of Early Postoperative Outcomes Between Robotic and Nonrobotic Total Hip Arthroplasty.
| Covariate | Statistics | Level | Robotic | P Value | |
|---|---|---|---|---|---|
| No (N = 1724) | Yes (N = 523) | ||||
| Operative time (minutes) | 1724 | 523 | <.001 | ||
| Mean | 90.13 | 69.98 | |||
| Median | 88 | 68 | |||
| Postoperative hospital length of stay | N (%) | 63 (3.65) | 155 (29.64) | <.001 | |
| N (%) | 1 | 999 (57.95) | 315 (60.23) | ||
| N (%) | 2 | 491 (28.48) | 36 (6.88) | ||
| N (%) | 3 | 120 (6.96) | 11 (2.1) | ||
| N (%) | 4 | 29 (1.68) | 4 (0.76) | ||
| N (%) | 5 | 5 (0.29) | 2 (0.38) | ||
| N (%) | 6 | 5 (0.29) | 0 (0) | ||
| N (%) | 7 | 4 (0.23) | 0 (0) | ||
| N (%) | 8 | 4 (0.23) | 0 (0) | ||
| N (%) | 9 | 1 (0.06) | 0 (0) | ||
| N (%) | 10 | 3 (0.17) | 0 (0) | ||
| Dislocation | N (%) | 0 | 1681 (97.51) | 520 (99.43) | .007 |
| N (%) | 1 | 43 (2.49) | 3 (0.57) | ||
| Readmission | N (%) | No | 1618 (93.85) | 497 (95.03) | .316 |
| N (%) | Yes | 106 (6.15) | 26 (4.97) | ||
| Emergency department visit | N (%) | No | 1541 (89.39) | 473 (90.44) | .488 |
| N (%) | Yes | 183 (10.61) | 50 (9.56) | ||
| Fracture | N (%) | No | 1692 (98.14) | 517 (98.85) | .271 |
| N (%) | Yes | 32 (1.86) | 6 (1.15) | ||
| Deep venous thrombosis | N (%) | No | 1714 (99.42) | 521 (99.62) | .587 |
| N (%) | Yes | 10 (0.58) | 2 (0.38) | ||
| Hematoma | N (%) | No | 1711 (99.25) | 521 (99.62) | .361 |
| N (%) | Yes | 13 (0.75) | 2 (0.38) | ||
| Pulmonary embolism | N (%) | No | 1717 (99.59) | 522 (99.81) | .470 |
| N (%) | Yes | 7 (0.41) | 1 (0.19) | ||
| Periprosthetic joint infection | N (%) | No | 1711 (99.25) | 521 (99.62) | .361 |
| N (%) | Yes | 13 (0.75) | 2 (0.38) | ||
Bolded text indicates statistical significance (P < .05).
Table 3Univariate Comparison of Patient-Reported Outcomes Between Robotic and Nonrobotic Total Hip Arthroplasty.
| Covariate | Statistics | Robotic | P Value | |
|---|---|---|---|---|
| No (N = 1724) | Yes (N = 523) | |||
| Preoperative PROMIS Global Score | N | 521 | 407 | .257 |
| Mean | 46.66 | 47.78 | ||
| Median | 47 | 50 | ||
| Preoperative PROMIS Mental Health Score | N | 512 | 401 | .75 |
| Mean | 50.22 | 50.04 | ||
| Median | 51 | 51 | ||
| Preoperative PROMIS Physical Function Score | N | 512 | 400 | .906 |
| Mean | 40.62 | 40.67 | ||
| Median | 40 | 40 | ||
| Preoperative KOOS, JR. Score | N | 521 | 406 | .568 |
| Mean | 46.66 | 47.78 | ||
| Median | 47 | 50 | ||
| 2-16 Weeks PROMIS Global Score | N | 474 | 383 | .233 |
| Mean | 72.68 | 71.4 | ||
| Median | 73 | 70 | ||
| 2-16 Weeks PROMIS Mental Health Score | N | 443 | 356 | .874 |
| Mean | 52.44 | 52.34 | ||
| Median | 53 | 53 | ||
| 2-16 Weeks PROMIS Physical Function Score | N | 441 | 357 | .154 |
| Mean | 46.98 | 46.17 | ||
| Median | 48 | 45 | ||
| 2-16 Weeks KOOS, JR. Score | N | 475 | 385 | .568 |
| Mean | 72.68 | 71.4 | ||
| Median | 73 | 70 | ||
| 4-6 Months PROMIS Global Score | N | 62 | 58 | .814 |
| Mean | 80 | 80.81 | ||
| Median | 81 | 85 | ||
| 4-6 Months PROMIS Mental Health Score | N | 64 | 58 | .701 |
| Mean | 52.3 | 52.95 | ||
| Median | 53 | 53 | ||
| 4-6 Months PROMIS Physical Function Score | N | 64 | 58 | .921 |
| Mean | 49.58 | 49.74 | ||
| Median | 51 | 51 | ||
| 4-6 Months KOOS, JR. Score | N | 62 | 58 | .438 |
| Mean | 80 | 80.81 | ||
| Median | 81 | 85 | ||
| 1 Year PROMIS Global Score | N | 57 | 44 | .599 |
| Mean | 83.26 | 81.36 | ||
| Median | 85 | 85 | ||
| 1 Year PROMIS Mental Health Score | N | 56 | 45 | .174 |
| Mean | 54.04 | 51.89 | ||
| Median | 53 | 53 | ||
| 1 Year PROMIS Physical Function Score | N | 56 | 45 | .722 |
| Mean | 49.75 | 50.31 | ||
| Median | 51 | 51 | ||
| 1 Year KOOS, JR. Score | N | 57 | 45 | .667 |
| Mean | 83.26 | 81.36 | ||
| Median | 85 | 85 | ||
PROMIS, Patient-Reported Outcomes Measurement Information System.
Multivariate logistic regression models (Table 4) demonstrated multiple significant correlations with postoperative outcomes. R-THA was found to be significantly correlated to lesser rates of primary periprosthetic dislocation (P = .046) and overnight admission (P < .001) in univariate analysis. Multivariate analysis subsequently controlling for surgeons further demonstrated improved stability in the R-THA group. In addition, a patient history of lumbar spine surgery was found to be significantly correlated with the increased incidence of periprosthetic dislocation, as well as overnight admission, readmission, postoperative emergency department visit, and all-cause 90-day complications, illustrating the anticipated concerns regardless of the arthroplasty technique.
Table 4Multivariate Analysis of Outcomes versus Demographics and Use of Robotic Assistance.
| Outcome | Covariate | Level | Odds Ratio (95% CI) | P Value |
|---|---|---|---|---|
| Dislocations | Robotic surgery | No | 3.74 (1.03-13.60) | .046 |
| Yes | ||||
| Gender | Female | 0.77 (0.42-1.39) | .381 | |
| Male | ||||
| Age | 1.01 (0.98-1.05) | .347 | ||
| Race | Asian | 0.00 (0.00-.) | .992 | |
| Black | 1.80 (0.39-8.38) | .453 | ||
| Caucasian | 1.45 (0.34-6.18) | .613 | ||
| Native American | 0.00 (0.00-.) | .997 | ||
| Native Hawaiian - Pacific Islander | 0.00 (0.00-.) | .998 | ||
| Other | 0.00 (0.00-.) | .986 | ||
| Body mass index | 0.98 (0.93-1.03) | .441 | ||
| ASA | I | 0.00 (0.00-.) | .977 | |
| II | 0.61 (0.08-4.83) | .636 | ||
| III | 0.83 (0.11-6.44) | .860 | ||
| IV | - | - | ||
| History of lumbar surgery | No | 0.28 (0.13-0.59) | <.001 | |
| Yes | ||||
| Surgeon | Surgeon 1 | 0.61 (0.29-1.31) | .205 | |
| Surgeon 2 | 0.65 (0.30-1.40) | .271 | ||
| Surgeon 3 | - | - | ||
| Readmissions | Robotic surgery | No | 0.90 (0.50-1.61) | .723 |
| Yes | ||||
| Gender | Female | 1.46 (0.99-2.15) | .056 | |
| Male | ||||
| Age | 1.02 (1.00-1.04) | .067 | ||
| Race | Asian | 6.89 (1.16-40.94) | .034 | |
| Black | 0.84 (0.32-2.19) | .715 | ||
| Caucasian | 1.43 (0.61-3.37) | .415 | ||
| Native American | - | - | ||
| Native Hawaiian - Pacific Islander | - | - | ||
| Other | - | - | ||
| Body mass index | 1.02 (0.99-1.06) | .182 | ||
| ASA | I | 0.14 (0.02-0.75) | .022 | |
| II | 0.20 (0.08-0.52) | <.001 | ||
| III | 0.28 (0.11-0.70) | .007 | ||
| IV | - | - | ||
| History of lumbar surgery | No | 0.21 (0.13-0.33) | <.001 | |
| Yes | ||||
| Surgeon | Surgeon 1 | 0.84 (0.53-1.33) | .458 | |
| Surgeon 2 | 0.51 (0.30-0.88) | .015 | ||
| Surgeon 3 | - | - | ||
| Overnight admission | Robotic surgery | No | 12.83 (8.09-20.35) | <.001 |
| Yes | ||||
| Gender | Female | 1.86 (1.35-2.57) | <.001 | |
| Male | ||||
| Age | 1.03 (1.02-1.05) | <.001 | ||
| Race | Asian | 621,544.4 (0.00-.) | .985 | |
| Black | 1.66 (0.75-3.64) | .208 | ||
| Caucasian | 0.78 (0.39-1.55) | .473 | ||
| Native American | 0.03 (0.00-0.62) | .023 | ||
| Native Hawaiian - Pacific Islander | 195,375.1 (0.00-.) | .997 | ||
| Other | 1.04 (0.19-5.72) | .960 | ||
| Body mass index | 1.02 (0.99-1.05) | .116 | ||
| ASA | I | 0.08 (0.01-0.73) | .025 | |
| II | 0.17 (0.02-1.38) | .098 | ||
| III | 0.26 (0.03-2.06) | .202 | ||
| IV | - | |||
| History of lumbar surgery | No | 3.25 (1.96-5.39) | <.001 | |
| Yes | ||||
| Surgeon | Surgeon 1 | 0.69 (0.40-1.22) | .203 | |
| Surgeon 2 | 1.02 (0.56-1.86) | .943 | ||
| Surgeon 3 | - | - | ||
| Emergency department visit | Robotic surgery | No | 1.32 (0.88-1.97) | .182 |
| Yes | ||||
| Gender | Female | 1.35 (1.01-1.81) | .043 | |
| Male | ||||
| Age | 1.02 (1.01-1.03) | .006 | ||
| Race | Asian | 0.00 (0.00-.) | .984 | |
| Black | 1.92 (0.91-4.05) | .085 | ||
| Caucasian | 1.55 (0.76-3.14) | .226 | ||
| Native American | 0.00 (0.00-.) | .993 | ||
| Native Hawaiian - Pacific Islander | 0.00 (0.00-.) | .995 | ||
| Other | 0.48 (0.06-4.02) | .498 | ||
| Body mass index | 0.98 (0.96-1.01) | .162 | ||
| ASA | I | 0.19 (0.04-1.03) | .055 | |
| II | 0.45 (0.18-1.15) | .096 | ||
| III | 0.88 (0.35-2.20) | .784 | ||
| IV | - | - | ||
| History of lumbar surgery | No | 0.35 (0.23-0.53) | <.001 | |
| Yes | ||||
| Surgeon | Surgeon 1 | 1.06 (0.73-1.54) | .755 | |
| Surgeon 2 | 1.26 (0.87-1.84) | .224 | ||
| Surgeon 3 | - | - | ||
| All-cause 90-d complications | Robotic surgery | No | 1.09 (0.77-1.54) | .631 |
| Yes | ||||
| Gender | Female | 1.46 (1.14-1.87) | .003 | |
| Male | ||||
| Age | 1.02 (1.01-1.03) | .004 | ||
| Race | Asian | 1.75 (0.34-9.06) | .503 | |
| Black | 1.53 (0.85-2.76) | .159 | ||
| Caucasian | 1.37 (0.79-2.39) | .260 | ||
| Native American | 0.00 (0.00-.) | .981 | ||
| Native Hawaiian - Pacific Islander | 0.00 (0.00-.) | .986 | ||
| Other | 0.26 (0.03-2.11) | .209 | ||
| Body mass index | 0.99 (0.97-1.01) | .306 | ||
| ASA | I | 0.13 (0.04-0.47) | .002 | |
| II | 0.28 (0.13-0.60) | .001 | ||
| III | 0.49 (0.24-1.03) | .059 | ||
| IV | - | - | ||
| History of lumbar surgery | No | 0.28 (0.19-0.40) | <.001 | |
| Yes | ||||
| Surgeon | Surgeon 1 | 1.04 (0.76-1.41) | .819 | |
| Surgeon 2 | 0.91 (0.66-1.26) | .585 | ||
| Surgeon 3 | - | - |
Bolded text indicates statistical significance (P < .05).
ASA, American Society of Anesthesiologists score.
Of the 368 sample radiographs included for cup position analysis, 141 had R-THA and 227 had M-THA. Univariate comparison of the measured acetabular component anteversion and inclination showed that following R-THA, cup anteversion was significantly greater (25.6° ± 5.4° vs 20.6° ± 7.6°) and cup inclination was significantly lesser (42.5° ± 5.3° vs 47.0° ± 6.7°) than M-THA (P < .05). Taking into consideration the classic Lewinnek [
[2]
] safe zone, 44.4% of the R-THA met the criteria. Of the R-THAs reviewed, 91.7% met inclination criteria, and 45.8% met version criteria. M-THA had 57.3% that met Lewinnek safe zone with 77.1% that met inclination, and 72.2% that met anteversion criteria. Of note, the robotic acetabular cup placement was usually templated for 22-25° anteversion by practice pattern. Regarding dislocations, we noted no difference in cup position compared to the random sampling of stable hips. In utilizing the same sample size, the surgeon’s head size choices did not differ between cohorts. The head sizes ranged from 32 to 44 millimeters in diameter. There were 193 (80.4%) large head sizes (≥36 mm) in the M-THA cohort and 123 (84.2%) in the R-THA cohort, not a statistically significant difference.Discussion
Recent literature has garnered considerable interest in defining where robotic-assisted surgery may provide sufficient advantages to justify the higher initial cost [
[19]
]. The purpose of this study was to add to the limited but expanding body of literature comparing R-THA and M-THA, primarily investigating dislocation and subsequent revision and secondarily evaluating immediate postoperative outcomes, patient-reported outcomes, as well as cup positioning.Following both a univariate and multivariate regression model, this retrospective review of prospectively collected registry data found that R-THA demonstrated a lower rate of primary periprosthetic dislocation compared to M-THA. Interestingly, 46% of the M-THA dislocations went on to recurrent instability leading to revision surgery, and none of the R-THA dislocations required more than conservative management. Such costly episodes of subsequent care are important to consider in the overall value equation. For secondary findings, this investigation determined that R-THA had lower operative times and hospital lengths of stay with an advantage toward outpatient discharge. Finally, the radiographic randomized sample investigation found that R-THA exhibited greater anteversion and less inclination with a continued demonstration of accuracy but notably more cups outside of the historic Lewinnek safe zone. All other metrics, including patient-reported outcomes, were similar between the two groups.
With a moderately low complication rate after THA, mechanical complications like dislocation have created a premise for robotic-assisted surgery to aid in the controlled replication of anatomic implant positioning and limb length. Recent large database studies utilizing the Australian registry and Medicare Part A claims have found that computer navigation was associated with lower dislocation rates and revisions related to the acetabular component [
[20]
,[21]
]. Much like earlier generation navigation systems, THA performed with robotic assistance has consistently demonstrated superior cup placement and mechanical alignment compared to conventional techniques [- Agarwal S.
- Eckhard L.
- Walter W.L.
- Peng A.
- Hatton A.
- Donnelly B.
- et al.
The use of computer navigation in total hip arthroplasty is associated with a reduced rate of revision for dislocation: a study of 6,912 navigated THA procedures from the Australian orthopaedic association national joint replacement registry.
22
, 23
, 24
, 25
]. There is limited data comparing a primary outcome of dislocation when investigating R-THA vs M-THA. Illgen et al. demonstrated a lower dislocation rate in R-THA, but this was not statistically significant (0% to 3%; P > .05) [[23]
]. This current cohort is the first to demonstrate that R-THA had a 3.47 times lower dislocation rate compared to M-THA after controlling for gender, race, body mass index, age, preoperative American Society of Anesthesiologists score, surgeon, and history of lumbar surgery. For the surgeon performing the majority of the R-THA, there were limited robotic dislocations with many more in his manual cohort during this study period, which accounts for the multivariate finding still holding up for a stability benefit. Revision surgery for instability was not encountered with the robotic cohort while it approached almost half of the patients with manual dislocations. This was despite the fact that cup positions were similar on average, again showing contemporary dislocations are often within safe zones illustrating the multifactorial nature of the problem [[26]
,[27]
].Another theme comparing R-THA and M-THA that has generated a lot of debate has been the notion that, similar to navigation, robotic-assisted surgery increases surgical time with questionable clinical benefits [
[28]
]. Domb et al. found that R-THA had a higher mean OR time compared to M-THA (110 vs 102 minutes; P = .08) [- Sweet M.
- Borrelli G.
- Manawar S.
- Miladore N.
Comparison of outcomes after robotic-assisted or conventional total hip arthroplasty at a minimum 2-year follow-up.
JBJS Rev. 2021; 9https://doi.org/10.2106/JBJS.RVW.20.00144
[20]
]. This concept of longer OR time has been a common trend with the introduction of robotic arthroplasty surgery [29
, 30
, 31
, 32
]. Our data with respect to surgical time illustrates that with experience, the extra time required for registration may be balanced with reaming and trialing efficiencies afforded by this haptic technology. The added intraoperative technology, once effectively incorporated into an optimized workflow, does not necessarily have to lengthen OR time and may actually yield more immediate episodes of care savings. While operative times are certainly multifactorial, our 20-minute shorter average robotic procedure duration suggests that time burden does not need to be a deterrent even in a teaching hospital when considering the adoption of robotic technology for THA.Improper cup positioning has been correlated to a higher rate of periprosthetic dislocation, for which multiple methods of cup implantation have aimed to improve upon in recent years [
[2]
,[33]
,[34]
]. M-THA can demonstrate a 38%-47% rate of acetabular implant malpositioning [[35]
,[36]
]. Therefore, for many surgeons, the current intraoperative tools may be inadequate to reproducibly implant the acetabular component from case to case. This may be truer for those that do not regularly perform hip arthroplasty, as fellowship-trained, high-volume arthroplasty specialists tend to have less variation given continuous refinement of technique. Nonetheless, outliers still exist in every practice [[37]
]. It is important to note that we did not evaluate offset or leg length. While we did have precise navigated offset and length data recorded in the majority of the robotic cohort, there was no comparable method to evaluate the manuals. Lack of ideal femoral rotation during positioning for standard radiographs can often underestimate the offset reproduction and vary even in the same patient depending on the date of image capture. CT would be ideal as the only way to truly acquire accurate numbers but was beyond the scope of our series.A benefit many have noticed with R-THA is comparable postoperative radiographs in every patient. The reproducibility of technological assistance for cup placement across multiple platform options enables controlled positioning based on surgeon preference. Multiple authors have previously shown that superior cup positioning and offset, which were found with R-THA, compared to manual techniques [
22
, 23
, 24
, 25
,[38]
]. While this is likely a contributing factor for the increased stability noted in our series, it is clearly not the only factor. Technology is only as good as the input data it is instructed to replicate. Rudimentary estimates of ‘safe-zones’ do not account for combined anteversion, biomechanical restoration, and the rapidly evolving concept of the hip-spine relationship or functional positioning. By utilizing our small random sample size, this cohort demonstrated that R-THA had significantly less inclination and more anteversion by choice. The safe zone criteria demonstrated similar findings as previous data that demonstrated improved precision in R-THA than M-THA [22
, 23
, 24
, 25
]. More data available for surgeons to act upon may factor into decisions on head size and offset and which side of the implant construct to add or remove anteversion. The fact that our manual surgeries were more accurate at restoring a classic Lewinnek safe zone for anteversion than R-THA (57.3% vs 44.4%) illustrates the confines of antiquated boundaries. The decision to target 25° of anteversion in the majority of the R-THA naturally pushed beyond the upper limit of that definition, which may in itself be a protective factor to consider [[26]
]. The precision to hit a target, whether that be a predetermined range or functional hip-spine adjustments, is a clear advantage of the current generation of technology assistance. The definition of what to target may be more elusive.Although technology in hip arthroplasty allows for improved precision, this does not always translate into clinical relevance. R-THA may allow for safer minimally invasive surgery since direct visualization is not as essential, which may theoretically present as improved PROMs, though this has not always been borne out in the literature. Much like the debate over the optimal surgical approach to the hip, the treatment of what is equally recognized as soft tissue surgery is dependent on surgeon preferences with the tools available at that time. If the definition of success is patient-reported outcome measures, our current study once again found no difference in scores, which included hip-specific and global health PROMs. This was echoed in another large series by Singh et al who looked at both a robotic and navigation cohort compared to manual [
[39]
] but contradicted the findings from Domb et al [[15]
]. This study utilized the Forgotten Joint Score that has been used in other studies to more effectively tease out subtle nuances in PROMs. While instability is an outcome of importance for patients, our patient-reported outcome scores did not reflect a difference at any time point. This metric may be better suited to evaluate more responsive differences since subtle postoperative differences from variation in intraoperative technology may demonstrate response bias that causes scale attenuation effects.Limitations for the study must be acknowledged. While registry data are prospectively collected, a retrospective review of such data with subgroups presents room for an inherent bias. The fact that most R-THAs were performed by one surgeon may limit the generalizability of the data generated in the R-THA cohort in this study, although multivariate analysis accounted for the surgeon. Similarly, although the goal of the study was to compare overall dislocation rates over the 6-year study period and the dislocation rate of the pooled surgeon cohort matches other data in the literature for incidence of dislocation after posterior approach M-THA [
[40]
], another limitation is the absence of individualized dislocation rate by surgeon and year for the analysis. We did note a higher dislocation rate early in the study period. An attempt at subgroup analysis using just the latter half of the time period (2017-2020) with an improved steady-state dislocation rate still yielded a 2.8 times higher dislocation rate with M-THA, but then multivariate control was not possible given one surgeon’s conversion to all R-THA. Due to the academic nature of the health system studied, residents within the case do add to the variability of the procedures performed. Since teaching methods vary based on the experience of the resident, as well as the teaching allowance of the senior staff, levels of resident involvement can vary by case. For example, the majority of the case could be done by a senior resident, whereas only select portions can be done by junior residents, potentially affecting outcomes and length of surgery. All surgeries within the cohort were done with the attending surgeon scrubbed and either closely supervising or actively performing the case. Operative time efficiency is likely correlated to the surgeon and would not be expected to be a result of the technology despite some technical efficiencies made possible by its usage. The reliance on the express robotic workflow without navigation of the femoral version also makes comparisons to other series heterogeneous if not well defined.Combined anteversion during a THA is utilized as an objective marker for component placement and hip stability. Since postoperative radiographs do not provide a reliable mode of measurement for the femoral component version, this objective data could not be collected during the radiographic review. The so-called ‘Ranawat sign’, including combined anteversion, was used for intraoperative estimation in the majority of the cases, but even the robotic cohort did not use the technology for discrete femoral version, so this major variable remains unaccounted for in the analysis [
[41]
]. Particularly for cup positioning and head size, our evaluation represents only a randomly selected sample of controls along with those that had dislocation episodes, and our series did not have reproducible offset and leg length measurements that are equally known to influence stability. Although the retrospective review controlled for surgeons in the analysis, many M-THA surgeries utilized different implant systems than the robotic cohort, which introduces other factors that can contribute to primary dislocation. However, all implants were typical constructs widely equivalent in most registries, and no surgeon relied primarily upon anything more than standard heads and liners. Finally, although the cohorts demonstrated a statistically favorable ratio of 3:1, the secondary outcomes such as length of stay should be carefully interpreted since they are not appropriately powered to make clinically relevant conclusions, especially in light of other factors, including the push toward more ambulatory surgery in the middle of the study period. The length of follow-up available for analysis in the R-THA cohort is not as robust as the follow-up available for the M-THA cohort, and it is noted that although only three dislocations were noted in the R-THA cohort, the incidence of late dislocations is not picked up as readily in this particular dataset given this limitation in follow-up.Conclusions
In this cohort study of prospectively collected registry data, R-THA demonstrated improved postoperative outcomes with regard to instability with one-fourth the risk. It adds to the literature that R-THA continues to produce reliable cup positioning that could contribute to this cohort's decreased dislocation rate. However, the multifactorial nature of instability points to the added value of additional information available for intraoperative surgical decisions that may ultimately be the primary benefit of such technology. Robotic assistance was found to not only protect against dislocation but also revision for subsequent recurrence, and also benefit for earlier and outpatient discharge home. The study demonstrates that even at an academic surgical program, R-THA does not have to increase the operative time. Robotic-assisted hip surgery continues to increase its adoption, and continued high-level studies are necessary to define its significant advantages.
Acknowledgments
We would like to acknowledge the hard work of Yueren Zhou and Stephanie Stebens, MLIS, AHIP, for their contributions to the data analytics present in this investigation.
Appendix A. Supplementary Data
- Conflict of Interest Statement for Lindsay-Rivera
- Conflict of Interest Statement for Jiang
- Conflict of Interest Statement for Davis
- Conflict of Interest Statement for Shaw
- Conflict of Interest Statement for Wesemann
- Conflict of Interest Statement for Rahman
References
- The operation of the century: total hip replacement.Lancet. 2007; 370: 1508-1519https://doi.org/10.1016/S0140-6736(07)60457-7
- Dislocations after total hip-replacement arthroplasties.J Bone Joint Surg Am. 1978; 60: 217-220
- Effect of acetabular component orientation on recurrent dislocation, pelvic osteolysis, polyethylene wear, and component migration.J Arthroplasty. 1998; 13: 530-534https://doi.org/10.1016/s0883-5403(98)90052-3
- Total hip arthroplasties: what are the reasons for revision?.Int Orthop. 2008; 32: 597-604https://doi.org/10.1007/s00264-007-0364-3
- Trends and patient factors associated with technology-assisted total hip arthroplasty in the United States from 2005 to 2014.Arthroplast Today. 2020; 6: 112-117.e1https://doi.org/10.1016/j.artd.2019.12.009
- Computer-assisted orthopaedic surgery and robotic surgery in total hip arthroplasty.Clin Orthop Surg. 2013; 5: 1-9https://doi.org/10.4055/cios.2013.5.1.1
- Makoplasty and the accuracy and efficacy of robotic-assisted arthroplasty.Surg Technol Int. 2014; 24: 302-306
- Precision and accuracy of robot-assisted technology with simplified express femoral workflow in measuring leg length and offset in total hip arthroplasty.Int J Med Robot. 2020; 16: 1-6https://doi.org/10.1002/rcs.2141
- Precision and accuracy of imageless navigation versus freehand implantation of total hip arthroplasty: a systematic review and meta-analysis.Int J Med Robot. 2017; 13: e1843https://doi.org/10.1002/rcs.1843
- The current role of robotics in total hip arthroplasty.EFORT Open Rev. 2019; 4: 618-625https://doi.org/10.1302/2058-5241.4.180088
- Robot-assisted total hip arthroplasty.Expert Rev Med Devices. 2016; 13: 47-56https://doi.org/10.1586/17434440.2016.1124018
- Current concepts in robotic total hip arthroplasty.SICOT J. 2020; 6: 45https://doi.org/10.1051/sicotj/2020041
- Robotic-arm assisted direct anterior total hip arthroplasty; improving implant accuracy.Surg Technol Int. 2020; 38: 347-352https://doi.org/10.52198/21.STI.38.OS1368
- Robotic arm-assisted total hip arthroplasty is more cost-effective than manual total hip arthroplasty: a Markov model analysis.J Am Acad Orthop Surg. 2021; 29: e168-e177https://doi.org/10.5435/JAAOS-D-20-00498
- Minimum 5-year outcomes of robotic-assisted primary total hip arthroplasty with a nested comparison against manual primary total hip arthroplasty: a propensity score-matched study.J Am Acad Orthop Surg. 2020; 28: 847-856https://doi.org/10.5435/JAAOS-D-19-00328
- Robotic-assisted compared with conventional total hip arthroplasty: systematic review and meta-analysis.Postgrad Med J. 2018; 94: 335-341https://doi.org/10.1136/postgradmedj-2017-135352
- The cost of robot-assisted total hip arthroplasty: comparing safety and hospital charges to conventional total hip arthroplasty.J Am Acad Orthop Surg. 2021; 29: 609-615https://doi.org/10.5435/JAAOS-D-20-00715
- Surgeon level of expertise in adult reconstruction: a brief communication regarding the need for reporting the level of expertise.Arthroplast Today. 2021; 8: 1-4https://doi.org/10.1016/j.artd.2020.12.011
- Robots in orthopaedic surgery: past, present, and future.Clin Orthop Relat Res. 2007; 463: 31-36
- Computer-assisted navigation is associated with reductions in the rates of dislocation and acetabular component revision following primary total hip arthroplasty.J Bone Joint Surg Am. 2019; 101: 250-256https://doi.org/10.2106/JBJS.18.00108
- The use of computer navigation in total hip arthroplasty is associated with a reduced rate of revision for dislocation: a study of 6,912 navigated THA procedures from the Australian orthopaedic association national joint replacement registry.J Bone Joint Surg Am. 2021; 103: 1900-1905https://doi.org/10.2106/JBJS.20.00950
- Comparison of robotic-assisted and conventional acetabular cup placement in THA: a matched-pair controlled study.Clin Orthop Relat Res. 2014; 472: 329-336https://doi.org/10.1007/s11999-013-3253-7
- Robotic-assisted total hip arthroplasty: outcomes at minimum two-year follow-up.Surg Technol Int. 2017; 30: 365-372
- Accuracy of component positioning in 1980 total hip arthroplasties: a comparative analysis by surgical technique and mode of guidance.J Arthroplasty. 2015; 30: 2208-2218https://doi.org/10.1016/j.arth.2015.06.059
- Adoption of robotic vs fluoroscopic guidance in total hip arthroplasty: is acetabular positioning improved in the learning curve?.J Arthroplasty. 2017; 32: 125-130https://doi.org/10.1016/j.arth.2016.06.039
- The safe zone range for cup anteversion is narrower than for inclination in THA.Clin Orthop Relat Res. 2018; 476: 325-335https://doi.org/10.1007/s11999.0000000000000051
- What safe zone? The vast majority of dislocated THAs are within the Lewinnek safe zone for acetabular component position.Clin Orthop Relat Res. 2016; 474: 386-391https://doi.org/10.1007/s11999-015-4432-5
- Comparison of outcomes after robotic-assisted or conventional total hip arthroplasty at a minimum 2-year follow-up.JBJS Rev. 2021; 9https://doi.org/10.2106/JBJS.RVW.20.00144
- Primary and revision total hip replacement using the Robodoc system.Clin Orthop Relat Res. 1998; 354: 82-91https://doi.org/10.1097/00003086-199809000-00011
- Comparison of robotic-assisted and manual implantation of a primary total hip replacement. A prospective study.J Bone Joint Surg Am. 2003; 85: 1470-1478https://doi.org/10.2106/00004623-200308000-00007
- Clinical outcome following robotic assisted versus conventional total hip arthroplasty: a controlled and prospective study of seventy-one patients.Z Orthop Ihre Grenzgeb. 2005; 143 ([in German]): 391-398https://doi.org/10.1055/s-2005-836776
- Comparison of manual rasping and robotic milling for short metaphyseal-fitting stem implantation in total hip arthroplasty: a cadaveric study.Comput Aided Surg. 2013; 18: 33-40https://doi.org/10.3109/10929088.2012.744430
- The transverse acetabular ligament: an aid to orientation of the acetabular component during primary total hip replacement: a preliminary study of 1000 cases investigating postoperative stability.J Bone Joint Surg Br. 2006; 88: 883-886https://doi.org/10.1302/0301-620X.88B7.17577
- Improving the accuracy of acetabular component orientation: avoiding malposition.J Am Acad Orthop Surg. 2010; 18: 286-296https://doi.org/10.5435/00124635-201005000-00005
- Accuracy of acetabular component position in hip arthroplasty.J Bone Joint Surg Am. 2013; 95: 1760-1768https://doi.org/10.2106/JBJS.L.01704
- Factors predisposing to dislocation after primary total hip arthroplasty: a multivariate analysis.J Arthroplasty. 2002; 17: 282-288https://doi.org/10.1054/arth.2002.30286
- The John Charnley Award: risk factors for cup malpositioning: quality improvement through a joint registry at a tertiary hospital.Clin Orthop Relat Res. 2011; 469: 319-329https://doi.org/10.1007/s11999-010-1487-1
- Assuring the long-term total joint arthroplasty: a triad of variables.Bone Joint J. 2019; 101-B: 11-18https://doi.org/10.1302/0301-620X.101B1.BJJ-2018-0377.R1
- Robotics versus navigation versus conventional total hip arthroplasty: does the use of technology yield superior outcomes?.J Arthroplasty. 2021; 36: 2801-2807https://doi.org/10.1016/j.arth.2021.02.074
- A comparison of risk of dislocation and cause-specific revision between direct anterior and posterior approach following elective cementless total hip arthroplasty.J Arthroplasty. 2020; 35: 1651-1657https://doi.org/10.1016/j.arth.2020.01.033
- The Ranawat Sign. A specific maneuver to assess component positioning in total hip arthroplasty.J Orthop Tech. 1994; 2: 59-61
Article info
Publication history
Published online: February 07, 2022
Accepted:
February 1,
2022
Received in revised form:
January 16,
2022
Received:
December 1,
2021
Footnotes
Funding Source: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
No author associated with this paper has disclosed any potential or pertinent conflicts which may be perceived to have impending conflict with this work. For full disclosure statements refer to https://doi.org/10.1016/j.arth.2022.02.002.
Identification
Copyright
© 2022 Elsevier Inc. All rights reserved.
