The Difference of Acetabular Cup Orientation Between Two Surgical Sides After Primary Total Hip Arthroplasty by Right-handed Surgeons

Background: We performed the retrospective cohort study to compare the acetabular cup 19 orientation, including anteversion angle (AA) and inclination angle (IA), of dominant hand side and 20 non-dominant hand side after primary total hip arthroplasty (THA) by right-handed orthopedic 21 surgeons. The results are still needed to be verified in future.


Methods:
Between January 2018 and December 2018, 290 patients who aged below 60 years and 23 underwent primary THA were retrospective screened. Patients who had hemiarthroplasty, previous 24 hip surgery, ankylosing spondylitis, developmental dysplasia of hip (DDH, Crowe type-Ⅲ and type-25 Ⅳ), severe comorbidity, missing information, inferior quality radiographs were excluded. 26 According to the surgery side, all patients were divided into left group and right group. Postoperative 27 plain radiographs were analyzed to compare the AA and IA between left and right side. Univariate 28 and stepwise multivariable linear regression to control included confounding factors. Stratified 29 analysis was performed to identify whether the operation approach can affect the result, including 30 anterolateral (ALA) and posterolateral approach (PLA). 31 Results: The mean AA was 17.7° (range 6.0° to 30.0°) and 21.0° (range 9.5° to 35.0°) for the left 32 and right side respectively. The mean difference was 3.28° (95% CI: 1.92 -4.64; P<0.001). The 33 mean IA was 41.1° (range 24.0° to 59.0°) and 40.1° (range 20.5° to 56.0°) for the left and right side 34 respectively (P=0.314). 113 patients' AA within the "safe zone" in the left (93.4 %), while the right 35 was 93 patients (82.3 %) (P=0.009). 95 patients' IA within the "safe zone" in the left (78.5 %), while 36 the right was 97 patients (85.8 %) (P=0.144). The IA of ALA group was smaller than PLA group in 37 both sides. The mean difference was 3.98° (95% CI: 1.22 -6.74; P=0.005). 38

Conclusions:
We concluded that AA in left side may be more accurate than right side after primary 39 THA by right-handed surgeons. The IA was no difference between the two sides, while it was 40 smaller in ALA than in PLA. The results are still needed to be verified in future. THA, as misalignment leads to a higher incidence of dislocation, polyethylene wear, osteolysis, or 50 irritation of the iliopsoas tendon [3][4][5][6]. 51 Handedness is defined as the more skillful and flexible of using one hand than the other [7]. The 52 influence of surgeon handedness on surgical procedure and clinical outcomes has been reported 53 previously, especially in general surgery [8], dentistry [9] and urology [10]. Bones are 54 symmetrically distributed in humans, so the surgeons' handedness may have a greater impact on 55 orthopedic surgery than on non-orthopedic surgery [11]. Standing on the right side of the patients 56 during right THA will allow the right-handed surgeon to ream and implant acetabular component 57 comfortably and conveniently. However, it becomes inconvenient when performing the left THA. 58 Up to now, there are extremely limited articles reported the surgeon handedness in orthopedic 59 surgery, especially in primary THA. So, the influence of surgeon handedness on acetabular cup 60 orientation remains unclear. 61 Therefore, we performed this retrospective cohort study to compare the difference of left and 62 right side primary THA by right-handed surgeon from two aspects: (1) is there any difference 63 between the two sides? (2) which is more accurate? 64

Variables 89
We conducted the retrospective analysis of all patients' medical records. The data were extracted 90 independently by two researchers. Any difference would be negotiated by a third party. We 91 identified outcomes reported in the previous studies. Radiographic outcomes including AA, IA. 92 Basic demographic information including gender, age, height, weight, body mass index (BMI), 93 diagnosis, operation approach and American Society of Anesthesiologists (ASA). 94 Postoperative plain radiographs were conducted in anterior-posterior position (AP) of the 95 operation side routinely, which were recorded and evaluated in Picture Archiving and 96 Communication System (PACS). The measurements were made by two researchers independently 97 using a standardized protocol. Only bilateral obturator foramina equally sized would be accepted. 98 The intraclass correlation coefficient (ICC) was calculated to measure the inter-observer reliability 99 for AA and IA. IA is defined as the angle between the long acetabular axis and the horizontal plane  Smirnov test was applied for normal distribution while Mann-Whitney U test for non-normal 116 distribution. Chi-square test and Fisher's exact test were applied to test categorical variables. 117 Baseline variables considered clinically relevant or univariate in relation to outcomes were included 118 in the multiple linear regression model. Considering the number of events available, the included 119 variables were carefully selected to ensure parsimony of the final model. All statistics were 2-tailed, 120 p < 0.05 were considered statistical significance, p < 0.001 were considered dramatically statistical 121 significance [14]. Scatter plots of the variables were graphically presented. Stratified analysis was 122 performed according to different operation approach, including anterolateral (ALA) and 123 posterolateral approach (PLA). ICC was measured and estimated using the grouping recommended 124 by Landis and Koch [15]. A score between 0.61 and 0.8 indicated substantial content, while a score 125 higher than 0.81 indicated near-perfect agreements content. 126

Results 127
There were 234 patients (109 female and 125male) fulfilled the inclusion criteria and were 128 recruited for the final analysis. According to the operation side, they were grouped into the left 129 (N=121) and the right (N=113). Among all of the enrolled participations, the mean age was  Table 1). The mean AA was 17.7° (range 6.0° to 30.0°) and 135 21.0° (range 9.5° to 35.0°) for the left and right side respectively. The mean difference was 3.28° 136 (95% CI: 1.92 -4.64; t=-4.741, P<0.001) ( Table 2). The mean IA was 41.1° (range 24.0° to 59.0°) 137 and 40.1° (range 20.5° to 56.0°) for the left and right side respectively. There was no statistical 138 significance in AA between the two sides (P=0.314). There were two outliers in the right side 139 (137,138; Fig.3A) After removing the two outliers, there were still no statistical significance in AA 140 between the two sides (P=0.502) (Fig.3B). The IA of ALA group was smaller than PLA group in 141 both sides (P=0.005). The mean difference was 3.98° (95% CI: 1.22 -6.74; t=-2.840, P=0.005) 142 (Table 3). Scatter plots of AA and IA between the two sides were graphically presented in Fig. 4. 143 There were 113 patients' AA within the "safe zone" in the left group (93.4 %), while the right was 144 93 patients (82.3 %). It was statistical significance (P=0.009). There were 95 patients' IA within the 145 "safe zone" in the left group (78.5 %), while the right was 97 patients (85.8 %). It was no statistical 146 significance (P=0.144) ( Table 4). There was no difference in AA (P=0.165) and IA (P=0.342) 147 between the two sides after primary THA by ALA (Table 5). The AA was smaller in the left side 148 than in the right side after primary THA by PLA (mean difference=3.25°, 95% CI: 1.85 -4.65; t=-149 4.572, P<0.001), while the IA was no difference (P=0.393) ( Table 6). Analysis of inter-observer 150 correlation coefficients for AA (ICC 0.814, P<0.001), and IA (ICC 0.906, P<0.001) were excellent. 151 71% of AA and 88% of IA of right side was within the "safe zone". Both of the percentage of AA 171 and IA within "safe zone" were smaller than us. They thought that the placement of cup in dominant 172 hand side was more accurate than in the non-dominant hand side, which was contrary to us. In 173 Kong's study [17], the AA in left side was 22.44°, which was smaller than right side 24.77°. In 174 Pennington et al. 's study [18], they found that the difference of AA was 3° between the dominant 175 and non-dominant side of surgeons, which was a little smaller than our study. 176 Combined with previous studies, we could recognize that surgeon handedness may be have an 177 effect on AA. And we could assume that the AA in non-dominant hand side may be smaller than 178 dominant hand side, but it remains unclear which is much better. Several studies [19][20][21] had 179 reported that AA in THA was a key factor which was related to hip component dislocation. The

Conclusion 226
From the study, we concluded that AA in left side may be more accurate than right side after 227 primary THA by right-handed surgeons. The IA was no difference between the two sides, while it 228 was smaller in ALA than in PLA. The results are still needed to be verified in future.

Ethics approval and consent to participate 237
We have obtained oral consent from the participants and ethical approval from the ethics 238 committee. The ethical review was approved by the Human Subjects Committee of the Ethics 239 Committee of Peking University Shenzhen Hospital (2019). 240

Consent for publication 241
We have got an agreement on publication. 242

Availability of data and materials 243
All data collected or analyzed in the study were included in this published article. 244

Competing interests 245
There were no competing interests.