To derotate or not ? The impact of a permanent derotation screw on the revision rate of dynamic hip screw fixation for intracapsular neck of femur fractures

In this retrospective study, we examine the impact that Background employing a permanent derotation screw (DRS) has on the rate of revision for 2-hole dynamic hip screws (DHS, a.k.a. sliding hip screws), used for internal fixation of intracapsular neck of femur (NOF) fractures. To the best of our knowledge, we are the first to examine the impact of using a derotation screw on DHS revision rate. : We obtained a list of 64 patients suffering intracapsular NOF fracture Methods treated with 2-hole DHS over a 5-year period, 28 of these were also treated with a DRS, forming our DRS group, 36 were not (non-DRS group). Fracture severity and patient demographics between the groups were compared to ensure homogeneity. The rate of revision to arthroplasty (total or hemi) of the two groups were compared. : The mean age in the DRS group was 70.79 years, 1.77 years lower Results than the non-DRS group (p=0.570). The DRS group had a rate of revision of 14%, in comparison with 39% in the non-DRS group (p=0.0299), corresponding with a number needed to treat of 4.06 derotation screws to prevent a single failure. : In this study, employing a permanent derotation screw alongside Conclusions a 2-hole DHS was associated with a significantly lower rate of revision to arthroplasty than using a 2-hole DHS alone. We would recommend this be further investigated with prospective randomized trials, to provide robust evidence and make clinical recommendations. 1 1 2 3


Introduction
Intracapsular neck of femur (NOF) fractures comprise one of the most common orthopaedic injuries 1 .The majority are treated with arthroplasty, as the femoral neck biomechanics and vulnerability of the blood supply lead to a high incidence of non-union and avascular necrosis following internal fixation.Some may be treated with fixation rather than replacement depending on patient factors and fracture configuration, however, the optimal fixation method is controversial 2 .
The choice of fixation method for intracapsular fractures is either cannulated hip screws (CHS), or dynamic hip screw (DHS) with or without a derotation screw (DRS).These devices provide stability in the plane of the femoral neck, whilst enabling compression at the fracture site to facilitate direct healing.
The biomechanics of basicervical fractures are influenced by fracture character and fixation method.Stankewich et al. (1996) investigated the biomechanical impact afforded to fracture configuration under cyclical and failure loading.They determined that force at the fracture site correlates with fracture angle.The more vertical the fracture angle, the greater the force resisted by the implant alone, and ultimate failure force correlated with the moment arm 3 .
The load through the hip when walking at 4km/h is approximately 238% body weight (BW), increasing to 250% when ascending stairs and 260% when descending 4 .Therefore a 70kg person loads the hip with approximately 1400-1500N when walking.The torsional force through the femur is also 23-83% larger when climbing stairs than when walking 4 .Using synthetic femurs Freitas et al. (2014) compared the load to failure for a pauwels III fracture fixed with a DHS plus DRS against a control group without fracture and found the mean load to failure in the DHS group was 1742N, compared to 1329N in the control group 5 .Blair (1994) states that a DRS provides rotational control during insertion of the lag screw, but no additional fixation thereafter 6 .This opinion is echoed by both clinical experience of others 7,8 and biomechanical testing by Swiontkowski et al. (1987)  9 .However, biomechanical studies of cadaveric fractures have shown that DHS with DRS gives superior stabilisation, theoretically reducing AVN and non-union 10,11 .
In this study we examine the impact employing a permanent derotation screw concomitantly with a 2-hole DHS has on rate of revision to arthroplasty in the treatment of intracapsular NOF fractures.

Patient selection
A list of 161 patients were identified as sustaining an intracapsular NOF fracture treated with internal fixation between April 2009 and April 2014.The patient follow-up notes and imaging were reviewed, excluding those treated with CHS, and ensuring follow up for at least one year.This left 64 patients treated with DHS, 28 of those treated with a derotation screw (the DRS group), and 36 without (the non-DRS group).
X-rays, operation notes, discharge summaries and clinic letters were reviewed to assess the outcomes associated with the treatments.Each fracture was assessed and scored using Pauwels and Gardens classifications 1 to ensure homogeneity between the DRS group and the non-DRS group with regards to fracture severity.The follow-up was reviewed and a negative outcome was defined as the need for revision surgery to hip arthroplasty.
The following inclusion criteria were applied: 1. Patient sustained intracapsular NOF fracture between the 1 st April 2009 and 31 st March 2014.
2. The fracture was treated with 2-hole dynamic hip screw.
3. Patient has been followed up for a minimum of 1-year following surgery.

Patient demographics and fracture severity
Patient demographics between the groups are summarised in Table 1.The Pauwels and Gardens scores are shown in Table 2 and Table 3. Pauwels score cut-offs are 0-30° for 1, 30-50° for 2 and >50° for 3 12 .The mean fracture angle in the DRS group was 39.78 degrees (SD 11.11) compared to 35.18 degrees (SD 9.69) in the non-DRS group, and the proportion of each fracture character in the 2 groups shows a similar distribution.We therefore determined that the groups were sufficiently homogeneous to allow comparison.

Rate of Revision
The patients in the DRS group had a significantly lower rate of revision to arthroplasty than those in the non-DRS group (p=0.0203), as shown in Table 4. Without a derotation screw the revision rate was 39%, in comparison to 14% when a DRS was used.

Discussion
Employing a permanent derotation screw alongside a dynamic hip screw seems to offer protection against requirement for revision to arthroplasty, carrying a relative risk reduction of 66% and NNT of 4.06.This NNT suggests the clinical impact of routinely employing a DRS could be quite significant, and needs to be further investigated with robust, prospective clinical studies.To the best of the our knowledge, there are no previous studies analysing the impact a derotation screw has on the failure rate of 2-hole sliding hip screws when used for treating intracapsular hip fractures.
Of the 18 patients requiring revision, 16 underwent total hip arthroplasty, 1 underwent hemiarthroplasty and one (in the non-DRS group) was managed conservatively despite requiring revision.We included this patient as it was documented that they required revision to arthroplasty, but were not fit for surgery, and therefore met our definition of a negative outcome.

Limitations
When reviewing images it was not possible to ascertain whether an intraoperative derotation wire had been used as these images were rarely saved, and operative notes were unreliable in reporting this.Our non-DRS group therefore likely contained some patients that were covered with an intraoperative derotation wire and some that were not.

Conclusions
This study shows a reduced rate of revision to arthroplasty when a permanent DRS was used alongside a 2-hole DHS for fixation of intracapsular neck of femur fractures when compared to DHS alone.Given effect size suggested in this study and potential improvements in patient care that could be achieved we recommend this is an area that should be investigated with a randomised controlled trial.

Consent
All data collected for this study was collected through Hull Royal Infirmary's hip fracture database, which is gathered for the national hip fracture database (NHFD).From the NHFD website: "the NHFD is approved by the NHS England HRA Confidentiality Advisory Group (CAG) to collect patient data without consent under Section 251 exemption.(This approval was formerly administered under the NIGB-ECC/PIAG)."and "patients do not need to give formal consent" for data to be collected, but "may opt out if they wish".Essentially the article is concise and well written.The problem is the small number of patients studied, there really have not been a sufficient number to be able to justify the conclusions stated in this article.
Other comments are -There must have been many more patients treated by internal fixation over the 5 year study period.This is a selected group of patients.How were they selected?Numbers not just percentages must be given in the abstract.
There are only very limited presentation of the patient demographics.
There is only need for one decimal place.
Were there more displaced fractures in the de-rotation screw group?
The treatment of the complications should be in the results not the discussion.

If applicable, is the statistical analysis and its interpretation appropriate? No
Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Partly
No competing interests were disclosed.

Competing Interests:
I have read this submission.I believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.reviewed patient imaging and eliminated those that had been miscoded (2481 patients: mostly patients that had actually been treated with arthoplasty, or extracapsular fractures).A further 39 were eliminated if they did not have clinical follow up for 1 year post op.This left 122 patients, 58 treated with cannulated screws and not included in this study, 64 teated with 2-hole DHS.
We don't doubt there were also patients with IC NOFs treated with fixation that were incorrectly coded and so lost to this study, but there was no intentional patient selection.This would have been included in the limitations section given more words to use.

There are only very limited presentation of the patient demographics.
There is a 1000 word limit and so we only reported age, sex, and fracture angle.
Were there more displaced fractures in the de-rotation screw group?The distribution of Garden classification and Pauwel scores can be found in tables 2 and 3 The treatment of the complications should be in the results not the discussion True.There is a lot more we would have liked to included in the discussion but were limited by the 1000 word limit No competing interests were disclosed.Competing Interests:

Dataset 1. Source data used as a basis for the findings in this study http://dx.doi.org/10.5256/f1000research.11433.d161270
Data collected for this study was collected through Hull Royal Infirmary's hip fracture database, which is gathered for the national hip fracture database (NHFD).

Table 2 . Distribution of Pauwels classification in the DRS and non-DRS group.
Department of Orthopaedics, Peterborough and Stamford Hospitals NHS Foundation Trust, Peterborough City Hospital, Bretton Gate, Peterborough, Cambridgeshire, UK