Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear authors, the topic of the paper is interesting, but some information are missing:

1. The ethical committee approval is mandatory regardless the type of study, when it involves humans or animals!
2. Some references are older than 10 years, please replace them!
3. The complication rate within 24 months following the procedure is in my opinion correlated with the comorbidities of the patients, give the major comorbidities of the patients for each group.
4. Give the mean age for each group, it can be correlated with the complication rate and mortality
5. Also the mortality can be correlated with the comorbidities 

Author Response

Dear authors, the topic of the paper is interesting, but some information are missing:

• The ethical committee approval is mandatory regardless the type of study, when it involves humans or animals!

We thank you for your valuable feedback. The approval code issued by the Ethics Committee has now been included. Line 91

• Some references are older than 10 years, please replace them!

Thank you very much. We have included more recent references

• The complication rate within 24 months following the procedure is in my opinion correlated with the comorbidities of the patients, give the major comorbidities of the patients for each group.

We listed the most common comorbidities identified, with a similar frequency across the three groups.

• Give the mean age for each group, it can be correlated with the complication rate and mortality
• Following this reviewer’s suggestion, we implemented multilevel modeling, nesting patients within individual surgeons, to assess the influence of patients’ age on the study outcome. Please see revised methods and results. The impact of patients’age on the outcome was negligible.

Table 2. Surgical features and peri-operative parameters.									adjusted for patients' age and FNF
Characteristics	Group A	Group B	Group C	Total	p-value	A vs B	A vs C	B vs C	p-value	A vs B	A vs C	B vs C
Surgical Time, min (SD)	45.8 (8.20)	58.15 (15.50)	70 (19.66)	63.53 (19.25)	0.0002	0.0061	< 0.001	0.0026	0.0002	0.0062	< 0.0001	0.0026
Skin Incision (cm)	10.08 (0.83)	12.37 (2.24)	15.03 (3.27)	13.63 (3.28)	< 0.0001	0.005	< 0.0001	0.0008	< 0.0001	0.005	< 0.0001	0.0008
Pre-Hb (g/dL)	12.62 (0.94)	12.57 (0.90)	12.60 (0.91)	12.59 (0.91)	0.8913				0.811
Post-Hb (g/dL)	9.61 (1.08)	9.19 (1.12)	8.66 (1.11)	8.94 (1.16)	0.0009	0.0377	0.0004	0.0033	0.0006	0.0278	0.0003	0.0026
ΔHb (g/dL)	3.01 (1.01)	3.38 (0.80)	3.93 (0.86)	3.65 (0.92)	0.0006	0.0441	0.0003	0.0021	0.0006	0.0412	0.0003	0.0021
LoS (day)	3.94 (1.00)	4.26 (1.21)	4.63 (1.34)	4.43 (1.29)	0.0321	0.1959	0.0135	0.0697	0.0334	0.1806	0.0135	0.0792

 

• Also the mortality can be correlated with the comorbidities

In accordance with the reviewers' suggestions, we have removed the paragraph on mortality as requested.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The authors present a large non-randomized series of surgical outcome data for hip artroplasty using a new and functionally better method. The data clearly show that skilled surgeons learn faster and have better results than juniors.

The only thing I find in the paper is that group B doctors characteristics are not described in the methods, otherwise than implicitly. 

Author Response

• The authors present a large non-randomized series of surgical outcome data for hip artroplasty using a new and functionally better method. The data clearly show that skilled surgeons learn faster and have better results than juniors.
• The only thing I find in the paper is that group B doctors characteristics are not described in the methods, otherwise than implicitly

Thank you very much for the suggestion. We have described the composition of Group B.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This study aims to assess the learning curve of different surgeons performing the direct anterior approach using the AMIS technique for femoral neck fractures.

1. The study includes two types of subjects: (1) patients and (2) surgeons. It would be helpful to also report the surgeons’ characteristics, such as gender and age.
2. In Section 3.2, please specify the number of patients treated by Group B.
3. There is some confusion regarding the classification of surgeons. For example, for Levels 2, 3, and 4 surgeons, are they equivalent to Groups A, B, and C?
4. The current statistical analysis is relatively basic. A multilevel modeling approach—such as nesting patients within individual surgeons—would provide a more appropriate and rigorous analysis.

5. When comparing more than two groups, ANOVA should be used instead of multiple t-tests.

Author Response

Reviewer III :

This study aims to assess the learning curve of different surgeons performing the direct anterior approach using the AMIS technique for femoral neck fractures.

The study includes two types of subjects: (1) patients and (2) surgeons. It would be helpful to also report the surgeons’ characteristics, such as gender and age.

Thank you for the suggestion. We have revised paragraph 3.2 to align more closely with your recommendations.

In Section 3.2, please specify the number of patients treated by Group B.

Done

There is some confusion regarding the classification of surgeons. For example, for Levels 2, 3, and 4 surgeons, are they equivalent to Groups A, B, and C?

We believe that by fully rewriting the paragraph, we may have resolved the confusion between the surgeons’ level of experience and the grouping of patients

The current statistical analysis is relatively basic. A multilevel modeling approach—such as nesting patients within individual surgeons—would provide a more appropriate and rigorous analysis.

Thanks to this Reviewer’s for this useful suggestion. We re-run the analyses using a linear multilevel modeling approach nesting  patients within individual surgeons using a compound symmetry matrix. Therefore, Table 2 has been revised accordingly. We derived the pairwise comparisons within the same modeling. Moreover, we used this models also to test the potential confounding effect of patients’ age and FNF.

When comparing more than two groups, ANOVA should be used instead of multiple t-tests.

We apologize for the mistake. We effectively used ANOVA for group comparisons. We revised the statistical methods section accordingly.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

PLEASE SEE THE ATTACHED FILE, "REVIEW OF PROSTHESIS MANUSCRIPT #3708803".

Comments for author File: Comments.pdf

Comments on the Quality of English Language

PLEASE SEE THE ATTACHED FILE, "REVIEW OF PROSTHESIS MANUSCRIPT #3708803".

Author Response

 

REVIEWER IV

 

REVIEW OF PROSTHESIS MANUSCRIPT ID 3708803 (Review submitted on June 30, 2025)

 

 

PART 1: MANDATORY EDITORIAL CORRECTIONS/REVISIONS

 

 

 

Line number(s)                                                    Revise to read:

 

2-3                                  “Surgeon learning curve for minimally invasive hemiarthroplasty using the direct anterior approach for treatment of femoral neck fractures in elderly patients”

Done

 

17-39                                             “Abstract. Background: Femoral neck fractures (FNFs) are

………reduced functional demands. The direct anterior approach (DAA) to perform HA is widely used because, among other attractive features, it facilitates recovery of functional outcomes by limiting iatrogenic muscle damage. The learning curve for surgeons who perform minimally invasive HA using the DAA approach is unknown. The purpose of the study is to perform this determination. Methods: 850 patients (age: 82 ± 6 years) who had suffered FNFs were enrolled for the study between January 2017 and September 2022. The patients underwent minimally invasive hemiarthroplasty (using the anterior minimally invasive surgical (AMIS) technique) and DAA (AMIS+DAA). The patients were divided into three groups, one operated on by surgeons who substantial experience in HA using DAA (Group A), another operated on by surgeons who have experience in adult hip arthroplasty (≥ 5 y) but had not specialized in using DAA (Group B), and surgeons who are being trained to perform adult hip arthroplasty (Group C). The metrics determined were duration of surgery, skin incision length, drop in hemoglobin level, length of hospital stay, complications experienced within 2 y of the procedure, Harris Hip Score, and mortality 6 mo, 1 y, and 2 y post-operation. For each of these metrics, the results were used to determine the learning curve for Groups B and C surgeons. Results: Using the learning curve

 

profiles obtained, it was calculated that in order to achieve the competence of Group A surgeons, Group B and Group C surgeons need to perform 46 and 102 consecutive procedures, respectively. Conclusion: For a HA patient to achieve    outcomes when treated for FNF using AMIS+DAA

requires that the surgeon should have performed a large number of this procedure. In other words, the surgeon learning curve is modest for Group B surgeons but substantial for Group C surgeons.”

 

Done

 

 

40-41                              “femoral neck fracture; hemiarthroplasty; direct anterior approach; anterior minimally invasive surgery.”

   

Done

 

 

 

45-53                              “Femoral neck fractures (FNFs)……..[1,2]. Usually, non- displaced fractures (Garden types I and II) are managed with closed reduction……………….[3] whereas fractures (Garden types III and IV) require hip replacement [4]. In most cases, hemiarthroplasty   (HA)   is........ used........... for older…………………..arthroplasty is preferred for high- demand…........... [7].”

 

 

Done

 

 

54-69                              Revise to read, “Different approaches are taken in HA surgery, among which are direct anterior, direct lateral, direct posterior, and anterolateral approaches [6, 8-19]. The direct anterior approach (DAA), introduced by Smith-Peterson and modified by Heuter, utilizes intermuscular and internervous planes (sartorius-tensor fascia lata (TFL) superficially and rectus femoris-gluteus deeply) to access the anterior capsule [16]. The direct lateral approach starts with a longitudinal incision over the greater trochanter and partial splitting of the discal gluteus and minimus to expose the anterior capsule [12,13]. The direct posterior approach involves tenotomy of the short

 

external rotators and piriformis at the greater trochanter to access the posterior joint capsule [11]. The anterolateral approach, introduced by Watson-Jones and modified by

???????? [AUTHORS: PROVIDE THIS INFORMATION] involves

retracting the gluteus medius and minimus through the interval with the TFL to reveal the superior capsule. DAA is widely used because, among other aspects, it is thought to facilitate functional recovery due to the preservation of external rotators and use of natural planes [18]. At our institution, we routinely use DAA and a specialized table and fluoroscopy [17].

In the literature, there are randomized trials and cohort studies in which DAA has been compared to other approaches [8,9,19] and studies in which DAA and a minimally-invasive technique (MIS + DAA) have been used [AUTHORS: PROVIDE REFERENCES

TO SUPPORT THIS STATEMENT]. However, a study in which the learning curve for surgeons who use MIS+DAA was determined is lacking. This is the purpose of the present study. In our study, HA was performed by three sets of surgeons (those who are very experienced in DAA, those who are experienced in HA but not specifically in DAA, and surgeons who are undergoing training in adult joint arthroplasty). The metrics determined were duration of surgery, skin incision length, decrease in patient hemoglobin level from pre-operation to post-operation, length of stay in hospital, complications, Harris Hip Score, and mortality.”

 

DONE

 

 

77-89                               “2.1. Study Design

“This study is……………..bipolar HA for FNFs classified as Garden type III or IV. The patients   September 2022. In each

procedure, anterior minimally invasive surgery (AMIS) technique (Medacta International, CITY?????, [AUTHORS: STATE CITY] Switzerland) and the DAA approach were used (AMIS+DAA method). Collected variables were patient demographic data, experience level of surgeon with regard to

 

DAA, skin incision length, duration of surgery, decrease in hemoglobin levels pre- versus post-operation, postoperative complications, Harris Hip Score (HHS) pre-operation and at 6 mo, 1 y, and 2 y mo post-operation.”

Done

 

99                              “American………score ≤ 3 [23]. Exclusion…             included

 

Done

 

102-106                           Delete the entire sub-section 2.4 because the information in it is now given in the revised sub-section 2.1.

 

DONE

 

 

107                                                “2.4. Surgeon Experience Classification”

DONE

 

 

 

 

108-114                            “Surgeons were                                        et al. [24]. Level 4 surgeons were senior

surgeons who had substantial experience with DAA, having performed a very number of HAs using this method or have participated in academic studies focused on this method (herein, designated Group A surgeons). Level 3 surgeons are those who have extensive experience (≥ 5 years) with HA but had not specialized in DAA (herein, designated Group B surgeons). Level 2 surgeons are newly certified surgeons who were being trained to perform adult hip arthroplasty (herein, designated Study Group 3 surgeons)

 

DONE

 

116                                                “2.5. Surgical Technique”

DONE

 

 

117-138                            “Each patient underwent…                           lateral to the anterior

superior iliac spine along the TFL. Blunt dissection……………..

 

….................... the risk of lateral femoral gutaneous nerve injury,

a known…………………the THL and satorius muscles was……………………………………………………….between the vasus lateralis and rectus femoris, and the ascending branch of the lateral femoral circumflex artery was isolated…………………………………………….day two. The patient was given acetaminophen (3 g/day)……………………….day one followed by a structured…............... for ~2 weeks.”

DONE

 

139                                                “2.c. Statistical Analysis”

DONE

 

140-152 “Demographic parameters……………………………carried out using the unpaired t-test and the Pearson………………respectively. For a surgeon experience level (Group A, B, or C), the variation of the each of the following variables with number of consecutive procedures performed by each surgeon was plotted: duration of surgery, skin incision length, DHb, and length of stay in hospital. The derived cubic interpolation fit to this plot was used to determine the learning curve for a surgeon experience group. The 75th, 80th, 85th, and 90th percentiles for this curve were determined using a root- finding numerical optimization for non-linear function. Mortality............... R.”

 

 

 

 

155-158                           “850 patients who experienced FNFs and who received a HA and met all the inclusion for the study were enrolled. On each of the characteristics of the patients, the patient groups were matched (Table 1).”

 

 

DONE

 

160-164                           Delete what is written in this sub-section because it is very confusing. Replace it with a simple statement such as, “3 Group A surgeons operated on 90 patients, 4 ???? [AUTHORS: CONFRM THIS NUMBER] Group B surgeons operated on 280 patients, and 4 [AUTHORS: CONFIRM THIS NUMBER] Group C surgeons operated on 480 patients. For each surgeon group, the procedures were equally divided among the surgeons.”

We have completely rewritten this paragraph in accordance with the suggestions provided by all the reviewers

 

 

167                                  “SD: standard deviation”.

 

DONE

 

 

169                                                “Duration of surgery”

DONE

 

 

170-179                           “The duration of surgery by Group A surgeons (mean: 45.8 min; range: 31-86 min) was significantly shorter than that for Group B surgeons (mean: 58.1 min; range: 37-121) and for Group C surgeons (mean: 70.0 min; range: 38-138 min) (Table 2). In order to achieve a duration of surgery corresponding to the 75th percentile time achieved by Group A surgeons (50 min), Group B and C surgeons would need to perform 36 and 96 consecutive procedures, respectively (Figure 1 and Table 3).”

DONE

 

180                                                “Skin incision length”

DONE

 

 

181-188                           “The skin incision length made by Group A surgeons (mean: 10.1 cm; range: 8.6-12.1 cm) was significantly shorter than that made by Group B surgeons (mean: 12.4 cm; range: 8.9-20.4 cm) or Group C surgeons (mean: 15.0 cm; range: 8.9-25.2 cm) (Table 2). In order to achieve a skin incision length

 

corresponding to the 75th percentile length achieved by Group A surgeons (10.7 cm), Group B and C surgeons would need to perform 46 and 102 consecutive procedures, respectively (Figure 2 and Table 3).”

 

DONE

216                                                “Blood Loss”

DONE

 

217-223                           “For patients operated on by Group A surgeons, DHb (mean: 3.0 g/dL; range: 1.2-7.8 g/dL) was significantly smaller than for those operated on by Group B surgeons (mean: 3.4 g/dL; range: ???? [AUTHORS: STATE AMOUNT] g/dL) or Group C surgeons (mean: 3.9 g/dL; range: 2.1-7.2 g/dL) (Table 2). In order to achieve DHb corresponding to the 75th percentile loss experienced by patients operated on by Group A surgeons (3.5 g/dL), Group B and C surgeons would need to perform 13 and 75 consecutive procedures, respectively (Figure 3 and

Table 3).”

 

DONE

 

225                                                “3.c. Length of stay in hospital”

DONE

 

 

226-229                           “In order for to achieve patient length of stay in hospital to correspond to the 75th percentile value when Group A surgeons performed the surgery (5 days), Group B and C surgeons would need to perform 12 and 16 consecutive surgeries, respectively (Figure 4 and Table 3).”

DONE

 

 

231-236                           “Types and frequencies of complications that presented within 24 mo postoperatively are given in Table 4. The most frequent complication was hematoma but none of the patients required

drainage. For a given complication, the fewest was experienced by patients who were operated by Group A surgeons.”

 

DONE

 

238                                                “3.8. Harris Hip Score”

DONE

 

 

239-244                           “Harris Hip Scores (HHSs) are presented in Table 5. At 6 mo following the procedure, HHS for patients who were operated on Group A surgeons was significantly higher than that for those operated on by either Group B or Group C surgeons. The same trend was found at 1 y following the procedure. However, at 2 y following the procedure, the difference in HHS for patients in the three study groups was not significant.”

 

DONE

 

255-260                           “The mortality data are given in Table 6. At a given time following the procedure, the mortality number and percentage are each significantly lower for patients who were operated on by Group A surgeons than for those operated on by either Group B or Group C surgeons.”

 

In line with the recommendations provided by the other reviewers, we have decided to remove the paragraph addressing mortality

 

 

266-346                           “In recent years, DAA has gained increasing attention………..curve, especially when compared to other approaches, such as.............. [25-29]. The learning curve for DAA

when used in total hip arthroplasty has been reported in many studies [25-28]. In contrast, in the case of HA, only a few studies have focused on learning curve [AUTHORS: PROVIDE REFERENCES AND, THEN, STATE THE SHORTCOMINGS OF

THESE STUDIES]. The purpose of the present study was to determine the learning curves of surgeons who have experience in HA (≥ 5 y) but not specifically in DAA (Group B) and those who are in training to become adult hip arthroplasty surgeons (Group C).

In terms of duration of surgery, we used our results to determine that Group B and Group C surgeons need to perform 36 and 96 consecutive cases, respectively, for their duration of surgery to be at the 75th percentile of Group A surgeons (those who have extensive experience in using DAA to perform). Our learning curve findings are comparable to those reported by Burnham et al. [26], who..... training.

Precision in placement and length of skin incision play crucial roles in…   when needed. We found that the incision length

created by Group A surgeons was significantly shorter than those created by Group B or Group C surgeons; thus, we determined that Group B and Group C surgeons need to perform 46 and 102 consecutive cases, respectively, for the length of the incision they made to be at the 75th percentile of Group A surgeons. These results are consistent with those reported by Jin et al. [31],........... time.

Intraoperative blood loss,………….……level, was significantly less in cases performed by Group A surgeons compared to the levels in cases performed by Group B and Group C surgeons; thus, we determined that Group B and Group C surgeons need to perform 13 and 75 consecutive cases, respectively, for the

 

blood loss by patients to be at the 75th percentile of Group A surgeons.

At both 6 mo post-procedure, HHS score for patients operated on by Group A surgeons was significantly higher than that for patients operated by Group B or Group C surgeons. This trend was also seen at 1 y post-procedure. Although this trend was not seen at 2 y post-procedure, the improvement in the earlier periods (6 mo and 1 y) is clinically significant, especially for patients in this age group.

Our study has many strengths, namely, relatively large sample size (especially, in Group C cases (480), patients enrolled from many centers [AUTHORS: IS THIS STATEMENT CORRECT?

IF SO, IT CONTRADICTS THE STATEMENT BELOW UNDER THE

SECOND LIMITATION], and stratification of surgeons by experience level with using DAA to perform HA. We recognize that our study has many limitations. First, it was a retrospective case study [AUTHORS: EXPAND THE LIMITATIONS OF THIS

TYPE OF STUDY]. Second, the study was conducted at one center [AUTHORS: ON LINE 78, IT IS STATED THAT THIS WAS A    “…….MULTICENTER……..”.    RESOLVE    THIS

CONTRADICTION]…………………….preoperative protocols. Third,…………………three groups, some heterogeneity in prior…………….experience level with DAA. Fourth, there is the issue of lack of randomization in the assignment of surgeons to cases; that is, the surgeons were not assigned randomly to cases, which could have introduced selection bias. Fifth, there many relevant quantitative outcome metrics (such as revision rate within 2 years following the procedure, EQ-5d Score (part of the standardized Health Related Quality-of-Life (HRQoL) instrument), and leg length discrepancy) and many relevant subjective outcome measures, notably patient reported outcome measure and pain score, that were not obtained.”     This is a multicenter  study, we resolve this conctradiction

 

363-369                           “In a retrospective case study, we determined the variation of each of 7 parameters with the number of HA procedures using a minimally invasive technique and DAA (AMIS+DAA) performed by 3 groups of surgeons (those who had very extensive experience in performing HA using DAA (Group A), those who had experience in HA (at least 5 years) but had no specialized training in DAA (Group B), and those who were undergoing training to perform adult hip arthroplasty). We used these results and determined the learning curve for Group B surgeons to be modest (conservative estimate: 46 consecutive cases) and for Group C surgeons to be substantial (conservative estimate: 102 consecutive cases). Our results underscore that AMIS+DAA is technically challenging and, as such, HA surgeons

need to be offered structured pathways for them to become proficient on this procedure, which, in turn, will contribute to better outcomes for patients who elect for HA to treat their femoral neck fracture(s).”

 

 

395                                                “Acronyms” DONE

 

 

396                                  Delete this line.

In this list of acronyms, (a) revise the definition of LoS to be, “Length of stay in hospital”; (b) correct “HHP” to read, “HHS”;

(c) delete, “ASIS”, LFCN” and “LFCA”; and (d) add, “BMI          Body Mass Index”.DONE

 

 

 

Table 1                           *Revise the heading to read, “Table 1. Characteristics of the patients in the three study groups” DONE

*In column 1, instead of “Patients (n)”, use, “Number (n)” DONE

*In column 1, instead of “BMI (kg/cm2)”, use, “Body mass index (kg/cm2)” DONE

*In column 1, instead of “Female , (%)”, use, “Women, n (%)” DONE

*In column 1, instead of “Right side”, use, “FNF on right side, n (%)” DONE

*Delete the whole of column 5 (“Total”) DONE

*Delete the whole of column 6 (“p-value”) and replace it the p-vales for A vs B, A vs C, and B vs C, as is done in Table 2. DONE

 

 

 

 

Table 2                           *Revise the heading to read, “Table 2. Summary of surgical and other parameters (mean, SD)”  DONE

*Supply the missing p-values for Pre-Hb. DONE

*Delete the whole of column 5 (“Total”) DONE

*Revise the footnote on lines 101-192 to read, “SD: standard deviation;…............. LoS: length of stay in hospital.”Done

 

 

Table 3                           *Revise the heading to read, “Table 3. Values of parameters when surgery was performed by Group A surgeons and determined number of consecutive procedures needed to be performed by Group B and Group C surgeons to achieve the stated parameter value for Group A surgeons.” DONE

*In column 2, revise all the entries to read, “75th”, “80th, “85th”, and “95th” DONE

*Revise, “Surgical time min (SD)” to read, “Duration of surgery (min)” DONE

*Revise, “Skin incision (cm)” to read, “Skin incision length (cm)” DONE

 

*Revise line 197 to read, “SD:……….LoS: length of stay in hospital.”

DONE

 

 

 

 

Table 4                           *Revise the heading to read, “Table 4. Types and frequencies of complications presented within 24 mo following the procedure”

                                      DONE

*Delete the whole of column 5 (“Total”)  DONE

*Create a new column 5 to contain the p-vales for A vs B, A vs C, and B vs C, as is done in Table 2. DONE

 

Table 5                           *Revise the heading to read, “Table 5. The Harris Hip Score values (mean, SD)” DONE

*Correct the heading of column 1 to read, “HHS” DONE

*In column 1, correct, “HHP 6 m, (SD)” to read, “HHS 6 mo” DONE

*In column 1, correct, “HHP 1 y” to read, “HHS 1 y” DONE

*In column 1, correct, “HHP 2 y” to read, “HHS 2 y” DONE

*Supply the missing p-values in row 3 DONE

*Revise the footnote on lines 250-252 to read, “HHS c mo: HHS at c mo following the procedure; “HHS 1 y: HHS at 1 y following the procedure; “HHS 2 y: HHS at 2 y following the procedure.

 

 

Table 6                                         *Revise the heading to read, “Table 6. Mortality data (number and %) ”

*Delete the whole of column 5 (“Total”)

*Delete the whole of column 6 (“p-value”)

*Create a new column 5 to contain the p-vales for A vs B, A vs C, and B vs C, as is done in Table 2.

* Delete the footnote on line 262.

 

In line with the recommendations provided by the other reviewers, we have decided to remove the paragraph addressing mortality

 

Figure 1.                         ∗Instead of labeling the surgeon groups as “Junior”, “Medium”, and “Senior”, revise them to be “Group A”, “Group B” and “Group C”. DONE

 

 

*Revise the          label of the vertical axis to read, “Duration of surgery (min)” DONE

 

 

*Revise the label of the horizontal axis to be, “Case number”. DONE

 

 

*Revise the Figure Caption to read, “Figure 1..... experience.

The black curve is the group-specific interpolation.”

DONE

 

 

 

Figure 2.                         ∗Instead of labeling the surgeon groups as “Junior”, “Medium”, and “Senior”, revise to them be “Group A”, “Group B” and “Group C”. DONE

 

 

*Revise the label of the vertical axis to read, “Skin incision length (cm)” DONE

 

 

*Revise the label of the vertical axis to be, “Case number”. DONE

 

 

*Revise the Figure Caption to read, “Figure 2..... experience.

The black curve is the group-specific interpolation.”

Figure 3.                         ∗Instead of labeling the surgeon groups as “Junior”, “Medium”, and “Senior”, revise them to be “Group A”, “Group B” and “Group C”. DONE

 

 

*Revise the label of the vertical axis to read, “DHb (g/dL)” DONE

 

 

*Revise the label of the horizontal axis to be, “Case number”. DONE

 

 

*Revise the Figure Caption to read, “Figure 3..... trajectories

of loss of hemoglobin by patient (DHb) through…………experience. The black curve is the group- specific interpolation.” DONE

 

 

 

 

Figure 4.                         ∗Instead of labeling the surgeon groups as “Junior”, “Medium”, and “Senior”, revise them to be “Group A”, “Group B” and “Group C”. DONE

 

 

*Revise the label of the vertical axis to read, “Length of hospital stay (d)” DONE

 

 

*Revise the label of the horizontal axis to be, “Case number”. DONE

 

 

*Revise the Figure Caption to read, “Figure 4…………

…………experience. The black curve is the group- specific interpolation.” DONE

 

 

 

PART 2: SUBSTANTIVE ISSUES/QUESTIONS

• In sub-section 2.6 (the original sub-section 2.7), explain why, for a given outcome parameter, the 75th percentile value in Group A surgeons database was used as the benchmark for estimating the learning curve for Group B and Group C

 

 

• In the Discussion Section (perhaps, beginning at line 333), add comments on the attractive aspects and shortcomings of the method used to estimate the learning curve in this study versus methods used in literature studies, two widely used ones being the cumulative summation method (CUSUM) and the risk-adjusted cumulative summation method (RA-CUSUM).

 

 

• The study has two major design First, the number of cases treated by the surgeons is a confounding variable because of the uneven distribution of the number of cases; namely 90 in Group A , 280 in Group B, and 480 in Group C. Second, Group A surgeons may have used their seniority to cherry pick cases; for example, they preferred to operate on patients in whom the FNF was on the left femur (see Table 1; in Groups A, B, and C, the % of patients whose FNF was on the right femur were 40, 53, and 52, respectively).

Taken together, these flaws may undermine the plausibility of the study and, hence, the results. Thus, the authors must address these issues in a detailed manner, including suggestion of way(s) to get around/accommodate them.

 

We re-run linear multilevel analyses adjusting also for patients’ age and FNF and results were almost overlapping.

Table 2. Surgical features and peri-operative parameters.									adjusted for patients' age and FNF
Characteristics	Group A	Group B	Group C	Total	p-value	A vs B	A vs C	B vs C	p-value	A vs B	A vs C	B vs C
Surgical Time, min (SD)	45.8 (8.20)	58.15 (15.50)	70 (19.66)	63.53 (19.25)	0.0002	0.0061	< 0.001	0.0026	0.0002	0.0062	< 0.0001	0.0026
Skin Incision (cm)	10.08 (0.83)	12.37 (2.24)	15.03 (3.27)	13.63 (3.28)	< 0.0001	0.005	< 0.0001	0.0008	< 0.0001	0.005	< 0.0001	0.0008
Pre-Hb (g/dL)	12.62 (0.94)	12.57 (0.90)	12.60 (0.91)	12.59 (0.91)	0.8913				0.811
Post-Hb (g/dL)	9.61 (1.08)	9.19 (1.12)	8.66 (1.11)	8.94 (1.16)	0.0009	0.0377	0.0004	0.0033	0.0006	0.0278	0.0003	0.0026
ΔHb (g/dL)	3.01 (1.01)	3.38 (0.80)	3.93 (0.86)	3.65 (0.92)	0.0006	0.0441	0.0003	0.0021	0.0006	0.0412	0.0003	0.0021
LoS (day)	3.94 (1.00)	4.26 (1.21)	4.63 (1.34)	4.43 (1.29)	0.0321	0.1959	0.0135	0.0697	0.0334	0.1806	0.0135	0.0792

 

 

 

• Use of the unpaired Student’s t test to conduct statistical comparisons of group means in the case of continuous variables is not acceptable. Being a parametric test of comparison, the unpaired t-test should be used only if each of the populations being compared is normally Thus, prior to performing the unpaired t-test, a test of normality of each population (for example, the Anderson-Darling Test) should have been conducted. To avoid this two-step approach, the group comparison should have been carried out using a non-parametric method, such as the Kruskal-Wallis test. In light of the error committed by the authors on this matter, all of the group comparisons, in the case of continuous variables, must be re-done and the Results Section (including the p-values given in Tables 1, 2, 4, 5, and 6) must be corrected and the Discussion Section must be corrected and the Conclusions Section must be corrected.

 

As suggested by other Reviewer we re-run analyses following a linear multilevel modelling approach nesting patients within individual surgeons. Before doing this, we verified the normal distribution of continuous variables at issues.  Study on continuous variables’ normal distribution.

Since the large number of observations, we believe that the large statistical power could leading to obtain false positive results (rejecting the null hypothesis of normal distribution) which is known for some statistical tests.

Therefore, we preferred to report here the normal qq-plot and density plot for each continuous variable which revealed, without any doubt, their normal distribution.

Age:

 

BMI:

 

Duration of Surgery:

 

Skin Incision lenght:

 

 

 

Pre-Hb:

 

Post-Hb:

 

ΔHb (hemoglobin loss):

 

Length of stay in hospital:

 

 

Mortality should not be used as one of performance metrics in this study because there is no evidence to show that receiving HA using the DAA + AMIS plays any role in mortality. After all, mortality is a consequence of many factors. Thus, mention of mortality should be deleted from every Section of this manuscript (Abstract, Introduction, sub-section 3.9, and Discussion). In line with the recommendations provided by the other reviewers, we have decided to remove the paragraph addressing mortality

 

 

 

• A large number of the articles cited (13 out of 34 (38%)) are studies on total hip arthroplasty alone [see #s 10, 11, 16, 17, 18, 20, 25, 26, 27, 28, 30, 31, and 33]. This is unacceptable given that the present study was on hip hemiarthroplasty. This deficiency undermines the quality of the Introduction Section and the Discussion Section of the manuscript. The authors must rectify this deficiency.

 

Articles 17, 18, 20, and 33 have been revised and updated as requested.

However, we kindly wish to clarify that articles 10, 11, 16, 25, 26, 27, 28, 30, and 31 could not be modified for the following reasons:

 

Articles up to number 16 indeed focus on total hip arthroplasty (THA), but they are referenced solely to describe the surgical approach and access route. In our view, whether the study discusses THA or hemiarthroplasty (HA) does not influence the validity of the information in this context.

 

Articles 25 to 28 are among the very few that specifically address the learning curve associated with the direct anterior approach (DAA). While these studies pertain exclusively to THA, there is currently no literature available that evaluates or compares the learning curve for DAA in HA procedures. As such, these references remain the most relevant and informative sources available on this specific technical aspect.

 

We hope this clarifies the rationale for maintaining these references.

 

 

 

PART 3: OVERALL ASSESSMENT

This study is an important contribution to the literature on the methodology and application of the learning curve concept for orthopaedic surgeons. The study has many attractive features, three of which are highlighted here. The first is the use of three sets of surgeons that represent the full spectrum of proficiency in hemiarthroplasty using the direct anterior approach (DAA). Second, the case study in which the surgeons performed HA using a combination  of  a  minimally-invasive  technique  and  DAA  is  innovative. Third, in the manuscript, the authors provide all the requisite details on the patients and the surgical procedure and present results on a large collection of outcome metrics which they then utilized to determine the learning curve. In short, with this study, the authors have both expanded and enhanced the literature on learning curve for orthopaedic surgeons.

The study and the manuscript do, however, have shortcomings/limitations, the details of which are presented in Parts 1 and 2 of this review. Nonetheless, it is worth reiterating two of these shortcomings/limitations. First, randomization was not used to assign cases to the surgeons. Second, use of a parametric test (unpaired Student’s test) for group comparisons when the normality of each of the populations being had not been established is unacceptable.

 

 

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The paper was improved! I have no further comments!

Author Response

We are very grateful for your valuable contribution to the improvement of our manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

Ok

Author Response

We are very grateful for your valuable contribution to the improvement of our manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

The authors have produced a much-improved Revised Manuscript. However, the authors must:

(1) make several editorial revisions to the Revised Manuscript, and

(2) address three substantive issues that were raised with respect to the Original Manuscript but were not addressed in the Revised Manuscript.

For details regarding items (1) and (2) above, please see the uploaded file titled, "REVIEW OF REVISED PROSTHESIS MANUSCRIPT #3708803".

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Please see Part 1 of the uploaded file titled, "REVIEW OF REVISED PROSTHESIS MANUSCRIPT #3708803".

Author Response

Lines 34-35 : done

68 and 75: The references have been modified in line with your suggestions

81-82: done

95-96 : done

107:  done

123 : done

142 : done

149-153:  done

154-161: done

165-166 : done

Table 1 : done

186 : done

Table 2 : done

Table 3 : done

211 : done

215 : done

219 : done

223 : done

227 : done

246-247 : done

Table 5 : done

269-270 :  done

298 : done

306-307 : done

318 : done

323-326 : done

326-332 : done

333-350 : done

361-363 : done

395 : done

396 : The sentence has been removed

ΔHb: The sentence has been modified

 

Part 2 

1)  Done

2) 296-299

3) 373-376