Trochanteric Bursa Repair and Lateral Hip Pain After Hemiarthroplasty for Femoral Neck Fracture

This is a prospective, single-center, parallel-group, randomized controlled trial designed to evaluate the effect of trochanteric bursa repair on early postoperative lateral hip pain in patients aged 65 years and older undergoing primary hemiarthroplasty for acute intracapsular femoral neck fractures.

Background and Rationale

The trochanteric bursa facilitates gliding between the greater trochanter and the surrounding soft tissues. During posterior approach hip arthroplasty, the bursa may be longitudinally incised to obtain surgical exposure. Whether repairing this tissue at the end of surgery reduces postoperative lateral trochanteric pain remains uncertain.

Previous studies have primarily investigated trochanteric bursa management during total hip arthroplasty. Evidence specifically addressing bursa repair versus leaving the bursa unrepaired in older patients undergoing hemiarthroplasty for femoral neck fractures is limited.

Study Design and Randomization

A total of 102 participants will be randomized in a 1:1 ratio to either trochanteric bursa repair (n=51) or no bursa repair (n=51).

Randomization will use a computer-generated variable block sequence prepared by an independent investigator. Group allocation will be concealed using sequentially numbered, opaque, sealed envelopes.

During surgery, the operating surgeon will assess the integrity and suitability of the trochanteric bursa for primary anatomical repair before group allocation is revealed. Only patients with a bursa suitable for repair will undergo randomization.

Surgical Interventions

All patients will undergo primary hip hemiarthroplasty through a posterior surgical approach.

In the intervention group, the longitudinally opened trochanteric bursa will be anatomically repaired using three to four interrupted sutures with No. 1 Vicryl.

In the control group, the trochanteric bursa will be left unrepaired following the same longitudinal opening.

The trochanteric bursa will not be excised in either group. Capsular and short external rotator repair, implant system, postoperative analgesia, mobilization, and rehabilitation will follow the same routine clinical protocols in both groups.

Blinding

The operating surgical team cannot remain blinded after group allocation because of the nature of the intervention. Postoperative physiotherapists, outpatient outcome assessors, and radiographic assessors will remain blinded to treatment allocation. Complete participant blinding cannot be guaranteed.

Outcome Assessment

The primary outcome is lateral trochanteric pain experienced while lying on the operated side at six weeks postoperatively, assessed using an 11-point Numerical Rating Scale (NRS; 0-10), with higher scores indicating greater pain.

Secondary outcomes include lateral hip pain during daily activities, pain on trochanteric palpation, ability to lie on the operated side, mobility assessed using the Turkish version of the Cumulated Ambulation Score (CAS-TR), functional independence assessed using the Modified Barthel Index, and intraoperative and postoperative complications.

Preoperative assessments include cognitive and delirium screening using the Standardized Mini-Mental Test and 4AT, together with pre-fracture functional assessments.

Postoperative day 1 mobilization status and postoperative day 2 CAS-TR, 4AT, and early complications will be recorded. Follow-up assessments will be performed at 2, 4, and 6 weeks and at 3 months after surgery.

Radiographic assessments will use existing routine postoperative anteroposterior pelvic radiographs. No additional radiographs will be obtained solely for research purposes.

Statistical Analysis

The primary analysis will compare six-week side-lying lateral hip pain NRS scores between randomized groups among participants with an evaluable primary outcome, according to their assigned treatment. Between-group mean differences, 95% confidence intervals, and two-sided p-values will be reported. Repeated pain and functional measurements will be analyzed using mixed-effects models.

Missing primary outcome measurements, including those due to death, inability to lie on the operated side, or inability to provide a reliable pain rating, will be documented without assigning artificial pain scores.

Protocol and Statistical Analysis Plan Details

Neither arm will receive trochanteric bursa excision or a study-related steroid injection. The independent assistant physician preparing the randomization sequence will not participate in participant selection, surgery, or outcome assessment. The 001-102 allocation envelopes will be held in one ordered box; operating-room personnel will open the next envelope only after intraoperative bursa repairability has been confirmed. Envelope opening defines randomization. Complications or intervention deviations after randomization will not lead to removal of the randomization record or reassignment.

Follow-up targets are calculated from the date of surgery. Visit windows are plus or minus 3 days at weeks 2, 4, and 6, and plus or minus 7 days at month 3; actual visit dates will be recorded. Pain NRS and Modified Barthel assessments are not scheduled on postoperative day 2. For radiographic reconstruction measurements, one best suitable routine AP pelvis radiograph from the early postoperative period through month 3 will be selected and assessed by one blinded reader. Radiographic differences will not be used to exclude randomized participants.

Sample Size

Assuming a between-group difference of 1.0 NRS point and a common standard deviation of 1.4, a two-sided alpha of 0.05 and 90% power require 43 evaluable participants per group. Allowing approximately 15% loss or non-evaluability, 51 participants per group (102 total) will be randomized. These planning assumptions are extrapolated from total hip arthroplasty studies using VAS and are not directly observed six-week NRS data from this hemiarthroplasty population.

Prespecified Analysis

The primary evaluable population consists of randomized participants alive at week 6 with a reliably obtained operated-side lying lateral pain NRS. The primary effect will be estimated by unadjusted linear regression as the repair-minus-control mean difference, with a 95% confidence interval, effect size, and two-sided p-value (alpha 0.05), according to randomized assignment regardless of the intervention received. Robust standard errors will be used if model residual assumptions are substantially inappropriate; Mann-Whitney U testing and an appropriate confidence interval for the median difference will provide a sensitivity analysis.

A supporting adjusted regression will include prespecified age, sex, pre-fracture CAS-TR, classical Charlson Comorbidity Index (without age points), and cemented versus uncemented stem status. Covariates will not be selected using observed outcomes or univariate p-values. Repeated NRS, CAS-TR, and Modified Barthel measurements will use mixed-effects models with participant-level random intercepts and group, time, and group-by-time fixed effects; baseline values will be included where appropriate. Categorical outcomes will use chi-square or Fisher exact tests; day 1 mobilization will be summarized or analyzed using an appropriate ordinal method. Secondary analyses are supportive and exploratory.

Missingness due to death, loss to follow-up, not lying on the operated side, unreliable NRS reporting, or other reasons will be reported separately by group. No hypothetical NRS will be assigned. Prespecified sensitivity analyses will examine major complications affecting lateral pain, assessments outside visit windows, and nonparametric assumptions. Multiple imputation or mixed-model supporting analyses may be used when the amount and pattern of missingness justify them; outcomes after death will remain unassigned. Surgeon and cement-type interactions will be exploratory only if sufficient data are available. No interim efficacy/futility analysis, internal pilot, or sample-size re-estimation is planned.