Evidence library

Recommends diagnosing soft-tissue infection clinically and grading its severity with the IWGDF/IDSA classification, a structured work-up for suspected osteomyelitis (probe-to-bone, plain radiographs, inflammatory markers and MRI where uncertain), deep tissue or bone sampling for culture rather than superficial swabs, and shorter, targeted antibiotic courses.

Both groups improved substantially, with no statistically significant difference in the primary outcome at 52 weeks between replacement and arthrodesis.

Deep surgical site infection did not differ significantly between high-dose dual-antibiotic and standard single-antibiotic cement.

REBOA did not reduce mortality; the Bayesian analysis showed a high probability of harm, with more deaths, many occurring early from bleeding.

Aspirin failed to show non-inferiority and was associated with a significantly higher rate of symptomatic VTE than enoxaparin, mainly driven by below-knee DVT. There was no significant difference in death or major bleeding.

Pain at 3 days was equivalent between the bandage and rigid immobilisation groups in both intention-to-treat and per-protocol analyses.

There was no significant difference in PRWE at 12 months between K-wires and casting, though a minority of cast patients later needed surgery for loss of position.

There was no significant difference in WOMAC at 18 months; single-stage patients had better function early after surgery, and single-stage revision was less costly.

Reconstruction gave a statistically significantly better KOOS4 at 18 months than rehabilitation. Around two-fifths of the rehabilitation group went on to have reconstruction within 18 months, so most avoided surgery in that period.

Vitamin D3 did not significantly reduce total, non-vertebral or hip fractures compared with placebo.

There was no clinically important difference in PRWE at 12 months between plating and casting, and no clinically important differences in quality of life.

Progressive exercise was not superior to a single advice session at 12 months. Injection gave a short-term benefit at 8 weeks but not at 12 months.

Disability at 18 weeks did not differ significantly between injection and microdiscectomy, and injection was the cheaper first-line option on economic evaluation.

There was no statistically significant or clinically important difference in OMAS at 16 weeks between brace and cast, nor in secondary analyses.

Replaced PRISMA 2009 with an expanded 27-item checklist, an abstract checklist and revised flow diagrams for both new and updated reviews, reflecting advances in how studies are identified, selected, appraised and synthesised.

There was no significant difference in ankle function at 16 weeks between cast and brace, and no clinically important differences at other time points; casting was not superior to functional bracing.

Non-inferiority of 6 weeks was not shown: persistent infection was more frequent with 6 weeks than 12 weeks, with the difference most marked in patients treated by DAIR.

Plaster casting was not superior to a removable brace: ankle function at 16 weeks did not differ significantly between groups, and secondary analyses were consistent.

Tourniquet use was associated with a higher risk of serious adverse events and slightly more early postoperative pain, with no meaningful gains in function and only a modest shortening of operating time.

Accelerated surgery did not significantly reduce either mortality or the composite of major complications at 90 days. Some secondary outcomes, such as delirium and time to mobilisation, favoured accelerated surgery.

Tourniquet use was associated with a higher risk of serious adverse events (such as venous thromboembolism and infection) and slightly worse early postoperative pain, with little or no functional benefit.

There was no significant difference in ATRS at 9 months between the brace and cast groups, and re-rupture rates were similar. The brace was not shown to be clinically superior.

There was no clinically or statistically significant difference in PRWE at one year between early fixation and casting. Complications potentially linked to surgery were more frequent in the fixation group, whereas the cast group needed occasional later fixation for non-union.

There was no clinically important difference in DASH at 12 months; bracing had a higher non-union rate, with some patients later converted to surgery.

All three groups improved substantially; differences between them did not reach clinical importance. Capsular release carried more risk and cost, and manipulation under anaesthesia was the most cost-effective.

Atypical femoral fracture risk rose with longer bisphosphonate use, was substantially higher in Asian women, and fell rapidly after stopping; in white women the hip and other fractures prevented far exceeded the atypical fractures caused.

There was no significant difference in deep surgical site infection at 30 days between incisional negative pressure therapy and standard dressings.

Pooled registry data suggested that just under 60% of hip replacements last 25 years; case series reported higher survival than registries, which are considered less prone to bias.

Immediate MRI was associated with lower mean costs per patient; the saving was not statistically significant at three months but became significant at six months. MRI also picked up significantly more fractures other than the scaphoid and had higher diagnostic accuracy than standard care.

There was no significant difference in the rate of secondary procedures at 24 months. THA gave a small functional and quality-of-life advantage that was unlikely to be clinically important, while dislocation was more common after THA.

Serious adverse events were associated with older age, comorbidity and male sex, while lifetime revision risk was greatest in younger patients, particularly younger men, with most revisions occurring in the early years after surgery.

PRP did not improve heel-rise endurance or patient-reported outcomes compared with placebo at 24 weeks.

Arthroscopy produced significantly better HOS-ADL scores at 8 months than physiotherapy and activity modification, with a difference around the prespecified minimum clinically important difference.

There was no significant difference in Oxford Knee Score at 5 years, with similar reoperation and complication rates; partial knee replacement was more cost-effective and showed small advantages in some secondary measures.

After network implementation there were improvements in risk-adjusted odds of survival, faster senior care and CT, and more patients treated in major trauma centres.

Major criteria (two positive cultures of the same organism, or a sinus tract communicating with the joint) confirm infection, while a weighted minor-criteria score classifies cases as infected, inconclusive or not infected. The new definition outperformed the 2011 MSIS criteria.

Robotic TKA was associated with significantly less early postoperative pain and analgesia requirement, faster attainment of straight leg raise and range of motion, and shorter hospital stay.

Physical therapy was non-inferior to arthroscopic partial meniscectomy for patient-reported knee function over 24 months. Some physiotherapy patients crossed over to surgery for persistent symptoms.

Aspirin was non-inferior to rivaroxaban for symptomatic VTE, with low event rates in both groups. There was no significant difference in major or clinically relevant bleeding.

Both groups improved, but hip arthroscopy gave significantly better iHOT-33 scores at 12 months than Personalised Hip Therapy, a difference judged to be clinically meaningful.

Major criteria (sinus tract communicating with the joint, or two positive cultures of the same organism) confirm infection; otherwise weighted minor criteria (serum CRP, D-dimer, ESR; synovial WCC, leucocyte esterase, alpha-defensin, neutrophil percentage, synovial CRP) give a score where 6 or more indicates infection and 2-5 is inconclusive. The new definition showed higher sensitivity than the 2011 MSIS criteria with very high specificity.

Screening increased uptake of osteoporosis medication but did not significantly reduce the primary outcome of all osteoporosis-related fractures; hip fractures, a prespecified secondary outcome, were reduced.

Both groups improved, but there was no significant difference between surgery and physical therapy at 2 years; many patients allocated to physical therapy crossed over to surgery, limiting interpretation.

There was no clinically relevant difference in pain between decompression and diagnostic arthroscopy at two years.

Both surgical arms improved slightly more than no treatment, but decompression gave no clinically important benefit over arthroscopy alone. The difference between the two surgical groups was not significant.

There was no significant difference in disability at 12 months, and no significant reduction in deep infection with negative pressure wound therapy.

Defined confirmatory criteria (fistula, sinus or wound breakdown communicating with bone or implant; purulent drainage or pus; the same pathogen in at least two deep tissue samples; microorganisms on histology) and suggestive criteria (clinical signs, raised markers, imaging, single positive culture) that should prompt further investigation.

Quality of life at four months did not differ significantly between implants, and any difference was below the minimum clinically important threshold. Mortality and mobility were also similar.

Romosozumab-then-alendronate significantly reduced new vertebral, clinical and hip fractures compared with alendronate alone.

Reoperation at 24 months did not differ significantly between implants, but avascular necrosis was significantly more common with the sliding hip screw. Exploratory subgroups suggested possible benefit from the sliding hip screw in smokers and in displaced or basicervical fractures.

The panel made a strong recommendation against arthroscopic surgery for almost all patients with degenerative knee disease, as any benefit was very small and short-lived.

No significant difference in DASH or other functional scores at any point; the operative group had an unacceptably high complication rate and the trial was stopped early for this reason.

Pregabalin did not significantly reduce leg pain compared with placebo at 8 weeks or at 1 year, and caused more adverse events such as dizziness.

Disability at 6 months did not differ significantly between plate and nail. Nailing gave significantly better function at 3 months.

There was no significant difference in Oxford Shoulder Score, re-tear rate or complications between arthroscopic and open repair.

Romosozumab significantly reduced new vertebral fractures at 12 months, with the benefit maintained after switching to denosumab; non-vertebral fracture reduction was not significant.

Immediate total-body CT did not significantly reduce in-hospital mortality compared with a standard imaging approach.

LIPUS did not improve functional recovery or speed radiographic healing compared with sham; the trial was stopped early for futility.

Simple decompression and transposition gave similar clinical improvement, with more wound complications after transposition; evidence comparing endoscopic with open release and for conservative measures was of low quality, although advice on avoiding prolonged elbow flexion and pressure may help mild cases.

Defined FAI syndrome as a motion-related clinical disorder requiring a triad of symptoms, clinical signs and imaging findings (cam or pincer morphology); morphology alone without symptoms is not FAI syndrome.

There was no clinically relevant or significant difference in KOOS4 at 2 years, and the exercise group had better thigh muscle strength in the short term.

Close contact casting gave ankle function equivalent to surgery on OMAS at 6 months. Casting had fewer infection and wound problems but more radiological malunion, and a minority of casting patients converted to surgery after early loss of reduction.

Return to work did not differ between groups, but early weight bearing gave better ankle range of motion and early functional outcome at 6 weeks, without an increase in complications.

No procedure showed clear superiority over another, including over trapeziectomy alone, for pain, function, quality of life or adverse events; most trials had unclear risk of bias.

Joint-preserving procedures are indicated before femoral head collapse, with good mid- to long-term results in several series; once collapse has occurred, total hip arthroplasty gives excellent long-term outcomes, whereas resurfacing and hemiresurfacing have performed poorly.

There was no significant difference in 24-hour or 30-day mortality, but the 1:1:1 group achieved haemostasis more often and had fewer deaths from exsanguination in the first 24 hours, without more complications.

Surgery was associated with a lower risk of recurrent dislocation, but the evidence was of low or very low quality and any difference in patient-reported function was uncertain.

The posterior approach significantly reduced Trendelenburg gait and stem malposition compared with the lateral approach, with non-significant trends towards less dislocation and heterotopic ossification; neither approach gave better function.

There was no significant difference in Oxford Shoulder Score between surgical and non-surgical treatment over two years.

TKR plus non-surgical treatment produced significantly greater improvement in pain, function and quality of life at 12 months, but with more serious adverse events; most in the non-surgical group improved and many had not undergone TKR by 1 year.

Group-level minimal important change was roughly 11 points for the OHS and 9 for the OKS (about 8 and 7 points for individual patients), whereas the between-group minimal important difference for trials was about 5 points for both scores.

The cast excluding the thumb produced significantly greater extent of union on CT at 10 weeks; overall union was almost universal and there were no significant differences in motion, grip, wrist scores or pain.

Concluded that atypical femoral fractures behave as stress or insufficiency fractures and revised the case definition: transverse orientation was emphasised, minimal comminution was now allowed, localised periosteal stress reaction became a major feature, and drug and disease associations were removed from the criteria. The relative risk with bisphosphonates is high but the absolute risk is low, rising with long-term use and appearing to fall after the drug is stopped.

There was no significant difference in Kerr-Atkins score at 2 years, and the operative group had more complications and reoperations.

Tranexamic acid was associated with significantly lower transfusion rates, with no increase in thromboembolic events, acute renal failure or other complications.

The redesigned pathway, with leaflet-supported direct discharge of minor stable injuries, did not increase unplanned ED reattendance, shortened consultations where backslabs were replaced by removable splints, and reduced unnecessary face-to-face fracture clinic review.

Kinematically aligned knees had significantly better pain relief, Oxford, WOMAC and Knee Society scores and greater flexion at 2 years than mechanically aligned knees.

UKA had significantly lower early mortality, fewer thromboembolic events, myocardial infarctions and strokes, shorter stay and fewer readmissions, but significantly higher revision and reoperation rates.

There was no clinically relevant or statistically significant difference in PRWE at 3, 6 or 12 months, nor in quality of life or complication numbers.

Fragility pelvic fractures differ from high-energy pelvic ring injuries and are poorly described by existing systems. Most are minimally displaced and settle without surgery, but some progress to displacement, nonunion and persistent instability; the proposed graded classification links instability to treatment, from minimally invasive fixation to complex reconstruction.

Plate fixation markedly reduced non-union and gave slightly better DASH and Constant scores at 1 year, but these functional differences disappeared once patients with non-union were excluded.

Both groups improved, with no significant difference between partial meniscectomy and sham surgery for any primary outcome at 12 months, and no significant difference in subsequent knee surgery or serious adverse events.

Methylprednisolone significantly improved symptoms at 10 weeks compared with placebo and modestly reduced the rate of surgery at one year, but most patients in all groups still went on to have surgery.

At 5 years there was still no significant difference in KOOS4, other patient-reported outcomes or radiographic osteoarthritis between groups. About half of the rehabilitation-first group had still not needed reconstruction.

Patients given corticosteroid had significantly worse 1-year recovery and higher recurrence than those given placebo injection. Adding physiotherapy to steroid did not offset this harm.

All groups improved, with no significant difference between PRP, glucocorticoid and saline at 3 months. Glucocorticoid gave better pain relief at 1 month and reduced tendon thickness and Doppler activity.

Roughly 10-34% of patients after TKR reported unfavourable long-term pain, compared with about 7-23% after THR, a higher proportion after knee than hip replacement.

Stemmed metal-on-metal hips had significantly poorer survival than other bearings, with larger heads failing earlier, whereas larger heads improved survival for ceramic-on-ceramic bearings; younger women with large metal-on-metal heads had particularly high revision rates.

Percutaneous and open release both resolved triggering in all cases and were similarly effective, while steroid injection had lower cure rates even when a second injection was allowed.

There were no significant differences in pain, grip or pinch strength or movement between the three techniques; good pain relief was maintained over time, although pinch strength declined regardless of procedure.

Male sex, sports injury, snuffbox pain on ulnar deviation within 72 hours and scaphoid tubercle tenderness at two weeks independently predicted fracture. Absence of snuffbox pain on ulnar deviation early on was associated with no fracture, and when all four factors were present the fracture risk was very high.

No significant difference in patient-reported outcome between techniques; the double-incision group had a small gain in flexion strength. Lateral antebrachial cutaneous nerve neuropraxia was more frequent with the single anterior incision.

Most individual shoulder tests had limited diagnostic accuracy on their own, and many labral and impingement tests performed poorly.

Recurrence was significantly and substantially more frequent and earlier after needle fasciotomy than after limited fasciectomy, although PNF had faster recovery and fewer complications.

Plating gave better early PRWE and DASH scores and better grip strength throughout, but by 12 months there was no significant difference in range of motion, pain, PRWE or DASH; anatomical reconstruction did not translate into better motion or daily function.

There was no significant difference in Oxford Knee Score or any other outcome at 5 years, including reoperation and later patella-related surgery, and total healthcare costs did not differ between resurfacing and non-resurfacing.

Only around half of binders sat at trochanteric level, and placement too high was the commonest error. In open-book injuries, high binders left a significantly wider symphyseal gap than correctly placed ones.

Treatment within 1 hour, and between 1 and 3 hours, significantly reduced bleeding deaths, whereas treatment after 3 hours appeared to increase the risk of death from bleeding.

CT and MRI performed comparably with no significant difference in sensitivity, specificity or accuracy; both were better at excluding than confirming a fracture, and MRI produced more false positives.

Laminar flow and space suits were not associated with lower early revision for deep infection; early infection revision was significantly higher with their use in both hip and knee replacement, independent of age, diagnosis and operating time.

The share of patients over 60 rose significantly over the study period. Older patients significantly more often had fractures involving displacement of the anterior column, frequently with a separate quadrilateral plate fragment or roof impaction, while their posterior wall fractures often showed comminution and marginal impaction, features linked to poorer outcomes.

Uncemented THA had significantly lower revision-free survival than cemented THA, mainly because of poorer uncemented cup performance; uncemented stems did better than cemented stems, though unrecognised intraoperative femoral fractures may cause early stem failure, and the commonest modern uncemented cups performed similarly to cemented cups.

Combines six clinical and radiological components into a weighted score that classes lesions as stable, potentially unstable or unstable, with the higher categories prompting a spinal surgical opinion.

There was no significant difference in 15-year survival between knees aligned within the neutral range and outliers.

Re-rupture rates were low and not significantly different between groups, and there was no significant difference in strength, function or range of movement. Surgery carried more complications, mainly wound-related.

Primary repair gave significantly better Constant and ASES scores, pain-free abduction and pain relief at one year. Roughly a fifth of the physiotherapy group went on to secondary repair because of insufficient improvement.

Both groups improved, with no significant difference in VISA-A between PRP and saline at 24 weeks.

Staples were associated with a significantly higher risk of superficial wound infection than sutures, particularly after hip surgery.

Decolonisation more than halved hospital-acquired S. aureus infection, with the largest effect on deep surgical site infection; mortality did not differ.

Overall surgical site infection was significantly lower with chlorhexidine-alcohol (about 10% vs 16%), driven by fewer superficial and deep incisional infections; organ-space infection did not differ.

The PRP group had significantly more patients with successful pain and function outcomes at 1 year; the steroid group improved early and then declined.

Corticosteroid gave better early relief but its effect faded, whereas the PRP group had significantly more successful outcomes at 1 year.

86% of evaluable patients responded, including every patient assessed histologically (near-elimination of giant cells); serious adverse events were uncommon and none was judged treatment related.

There was no significant difference in KOOS4 change at 2 years between the two strategies; around 40% of the rehabilitation-first group had delayed reconstruction, so most avoided surgery without a worse outcome.

Collagenase achieved the primary end point in significantly more joints than placebo, with better results at the MCP than the PIP joint; adverse events were mostly local (swelling, bruising, skin tears), with rare tendon rupture.

Surgery gave significantly greater improvement in function and symptoms at 12 months, although both groups improved and the size of the difference was modest.

TEA gave significantly better Mayo scores and shorter operations, and several ORIF patients were converted intraoperatively to TEA because fixation was not achievable.

A short period in a below-knee cast gave a faster recovery than tubular bandage, with the Aircast brace also beneficial; the Bledsoe boot did not show benefit. Differences narrowed by 9 months.

Denosumab significantly reduced new vertebral fractures (by about two thirds), with significant reductions in hip and non-vertebral fractures.

Pain relief, grip strength and satisfaction were broadly similar between the two salvage procedures; four-corner fusion had more complications, while proximal row carpectomy carried a higher risk of later radiocapitate arthritis.

Femoral head involvement and age at diagnosis were the strongest prognostic factors; in children over 6 years with more than half of the head involved, osteotomy gave significantly better femoral head sphericity than non-operative treatment.

Early arthroscopic Bankart repair significantly reduced recurrent instability and improved shoulder-specific function compared with lavage.

Recommended a well-centred AP pelvis plus lateral views (e.g. cross-table lateral, frog-leg, Dunn) and set out key measures such as lateral centre-edge angle, Tonnis angle, crossover sign and head-neck offset.

Roughly 60% of hips were preserved at 20 years, and older age, poorer preoperative function and pre-existing arthritis predicted failure.

In as-treated analysis surgery gave significantly greater improvement in pain and function at 2 years; intention-to-treat analysis showed a significant benefit only for bodily pain because of crossover.

Reamed nailing reduced events in closed fractures, an effect largely due to fewer autodynamisations, with no clear difference in open fractures. The trial also showed that delaying reoperation for nonunion for at least 6 months reduces unnecessary procedures.

Adding arthroscopic surgery gave no significant additional benefit in WOMAC score at 2 years over physical and medical therapy alone.

Drains gave no significant benefit for infection, haematoma or reoperation, but were associated with a significantly higher blood transfusion requirement; undrained wounds needed more dressing reinforcement.

Zoledronic acid significantly reduced new clinical fractures (by about a third) and was associated with a significant reduction in all-cause mortality.

Zoledronic acid markedly and significantly reduced vertebral fractures (around 70%) and significantly reduced hip fractures (around 40%) versus placebo.

Early surgery produced faster relief of leg pain and faster perceived recovery, but by 1 year outcomes were similar between groups; a substantial proportion of the conservative group eventually had surgery.

Describes how THR transformed care of elderly arthritic patients, that both cemented and uncemented fixation can be durable, and how newer bearings, less invasive surgery and computer navigation aim to meet younger patients' demands, with cost-effectiveness governing future developments.

Neither group had a major loss of reduction or an iatrogenic ulnar nerve injury; mild loss of reduction, radiographic angles, alignment and elbow motion did not differ significantly between lateral-entry and crossed pinning.

Plate fixation gave significantly better Constant and DASH scores, faster union and lower rates of non-union and symptomatic malunion than sling treatment.

The hook test was highly sensitive and specific for complete rupture, outperforming MRI in this series.

Recommended a simplified scoring system in which higher scores indicate better outcome, and gave guidance on handling missing items and interpreting scores.

Packing was associated with rapid haemorrhage control, reduced post-procedure transfusion and low mortality from acute haemorrhage, with only a minority needing subsequent angioembolisation.

Proposes that successful healing needs all four elements together, so nonunion treatment should identify and address each deficient element rather than relying on a single adjunct.

There was no significant difference in clinical scores or nerve conduction between transposition and simple decompression.

Recurrence was common, mostly within the first two years, and was most strongly associated with young age and male sex.

The Lachman test had the best overall sensitivity for ACL rupture, while the pivot shift test was highly specific and useful for ruling the diagnosis in. The anterior drawer test performed better in chronic than acute injuries.

Splinted children had better physical function at 2 weeks and less difficulty with daily activities such as bathing than those in casts, with similar pain and no refractures in either group.

Injection gave the best short-term results at 6 weeks, but at 1 year it was significantly worse than physiotherapy and wait and see, with high recurrence. Physiotherapy was better than wait and see early on, with little difference by 1 year.

The weighted clinical criteria (CTS-6) correlated well with the judgement of an independent expert panel, providing a consistent clinical definition of CTS.

Fever was the strongest single predictor and raised CRP was an independent predictor; combining CRP with the Kocher criteria increased the probability of septic arthritis as more predictors were present.

Fusion produced only a small difference in Oswestry score of borderline significance, with no meaningful advantage in walking, while carrying surgical risk and greater cost.

Scores of 0-8 predicted survival under 6 months, 9-11 at least 6 months and 12-15 at least 1 year, and predicted prognosis agreed well with actual survival.

Significantly more surgical patients were ambulatory after treatment (84% versus 57%) and they retained walking ability for longer, with reduced steroid and opioid requirements.

Screw fixation gave substantially shorter times to union and return to sport, while casting had a high failure rate from non-union, delayed union and refracture.

Outcomes were similar, but transposition had significantly more complications.

Early fixation offered no lasting functional advantage over casting by one year, although some early benefits in function were seen. Non-unions in the cast group united after subsequent fixation, and surgery carried its own complications.

Surgery plus radiotherapy significantly improved the proportion able to walk and the duration of walking, and more patients who were non-ambulant at presentation regained walking; corticosteroid and opioid needs were also lower.

Older age, ulnar (rather than median) nerve injury, more proximal injury and longer delay to repair were associated with worse motor recovery; age and delay also predicted sensory recovery.

Most nonunions united, and those that failed to heal had received significantly fewer and less concentrated progenitor cells, suggesting a dose threshold.

Radial nerve palsy complicated about 12% of humeral shaft fractures, most often middle and middle-distal transverse or spiral fractures; most recovered spontaneously with conservative care, and final outcomes were similar whether nerves were explored early or managed expectantly.

Clinical outcomes did not differ significantly between the two operations, but complications were significantly more frequent after transposition.

Ondansetron, dexamethasone and droperidol each reduced PONV risk by roughly a quarter, and propofol and avoiding nitrous oxide also helped; the effects were independent and additive, so combining interventions reduced risk further.

Most elbows regained concentric stability and a functional arc of motion, though reoperation for stiffness or complications was not rare.

Both groups improved at 2 years with no clinically important difference in clinical outcome; microfracture showed better SF-36 physical component improvement, and histology did not show clear superiority of ACI.

Children over 8 years at onset with lateral pillar B or B/C border hips had significantly better outcomes with surgery, whereas younger children with these groups did well regardless of treatment, and lateral pillar C hips did poorly whatever the treatment.

Very few feet required extensive corrective surgery; relapses were strongly associated with poor adherence to foot abduction bracing.

Sensitivity was almost 100% with low negative likelihood ratios for ankle and midfoot in adults and children, so a negative rule makes fracture very unlikely; specificity was only modest.

Introduced a five-level hierarchy, from high-quality RCTs (level I) to expert opinion (level V), applied separately to therapeutic, prognostic, diagnostic and economic studies.

ACI gave better functional results and a higher proportion of excellent or good arthroscopic repairs than mosaicplasty at 1 year; mosaicplasty performed particularly poorly in patellar defects.

Combined systemic and cement antibiotic prophylaxis carried the lowest revision risk; systemic prophylaxis alone was associated with higher revision for any cause, aseptic loosening and infection. Within the combined regime, four systemic doses on the day of surgery performed best.

Described cam (aspherical femoral head-neck junction) and pincer (acetabular over-coverage) impingement and their characteristic patterns of labral and cartilage damage, proposing FAI as a cause of early osteoarthritis.

Most patients achieved good or excellent hip scores with a low revision rate for acetabular component loosening, supporting acute arthroplasty as a viable option in selected patients.

Functional outcomes at 2 years were similar after salvage and amputation, with substantial disability in both groups. Poor outcome was linked to complications needing readmission and to social and economic factors rather than treatment choice, and reconstruction carried a higher rehospitalisation rate.

Radiographic abnormality by age 2 was identical in the two arms, while the ultrasound arm needed significantly less abduction splinting; rates of surgical treatment did not differ significantly.

Preoperative radiotherapy roughly doubled wound complications (35% versus 17%, significant); tumour size and site were also risk factors, and overall survival was marginally better with preoperative treatment.

Symptomatic patients had significantly larger alpha angles than controls, establishing the alpha angle as a measure of cam morphology.

The higher dose of rhBMP-2 significantly reduced secondary interventions, accelerated healing and reduced infection rates, particularly in Gustilo type III injuries.

Surgery gave significantly higher success rates than splinting at all follow-up points, and many patients in the splint group went on to request surgery.

Simple partial head fractures did well, but comminuted fractures with more than three fragments had a high rate of early failure, nonunion and poor results.

The derived rule had very high sensitivity for clinically important cervical spine injury while potentially reducing the number of radiographs ordered.

Proactive geriatric consultation significantly reduced the incidence of delirium (about one third vs one half) and reduced severe delirium by a greater margin.

Surgery gave significantly less pain and better function and satisfaction at 12 months than orthoses or waiting. Orthoses gave only short-term symptom relief.

Electrodiagnosis helps localise the lesion and quantify axon loss, but distal conduction persists for a period after transection, so studies are most informative once Wallerian degeneration has occurred; EMG evidence of reinnervation precedes clinical recovery.

The large majority of fractures united with bracing, with non-union uncommon and residual angulation usually functionally and cosmetically acceptable.

Decompression within 48 hours was associated with significantly better recovery of sensory and motor deficits and urinary and rectal function than later surgery, with no extra gain from operating within 24 hours versus 24-48 hours; pre-existing chronic back pain and rectal dysfunction were associated with worse sphincter outcomes.

Patients managed with external fixation were more severely injured, yet the staged approach was safe with low conversion complication rates, supporting its use in unstable patients.

Overall sensory recovery was similar between groups; conduits gave better two-point discrimination than end-to-end repair for small gaps and outperformed autograft for gaps of 8 mm or more.

The probability of septic arthritis rose steeply with the number of predictors present, being very high when all four were present and very low when none were.

The probability of septic arthritis rose steeply with the number of predictors present, being very high when all four were present and very low when none were.

Infants whose biceps recovered within the first three months achieved normal function, whereas recovery in the fourth to sixth months was followed by significantly worse Mallet scores; this supported microsurgical repair for infants with no biceps recovery by six months.

Among patients completing the study, cure was significantly more frequent with the rifampicin combination than with ciprofloxacin alone.

The questionnaire showed good internal consistency, reproducibility and validity, and was sensitive to change after knee replacement.

Alendronate roughly halved new radiographic vertebral fractures and also significantly reduced hip and wrist fractures compared with placebo.

The 12-item questionnaire was reliable, correlated with established measures and was responsive to change after surgery, supporting its use as a joint-specific patient-reported outcome.

Most hips achieved an anatomical reduction, which was harder to obtain in complex patterns, older patients and with longer delay to surgery. Clinical outcome tracked the radiographic result: anatomical reduction and a congruent roof predicted good results, whereas femoral head injury, older age and operative complications predicted poorer ones.

The rule identified all clinically important fractures in the validation cohort (sensitivity of 100%) and would have reduced the need for knee radiographs substantially.

Biopsy problems (diagnostic errors, complications, altered treatment and unnecessary amputation) were significantly more frequent when biopsy was done at a referring institution than at a treating centre.

Extended Ficat-type staging by incorporating MRI findings and quantifying the extent of head involvement, which helps predict progression and guide treatment.

Useful elbow flexion was restored in the reported cases without significant loss of ulnar nerve function in the hand.

Most femoral condyle lesions had good or excellent clinical results with repair tissue resembling hyaline-like cartilage on biopsy; results for patellar lesions were noticeably poorer.

Avascular necrosis developed in almost half of unstable slips and in none of the stable slips, and satisfactory results were far more frequent in stable hips; early reduction was not shown to be linked to AVN.

The refined rules detected essentially all clinically significant malleolar and midfoot fractures in validation, while allowing a substantial reduction in radiography.

Supplementation significantly reduced hip fractures and other non-vertebral fractures, with correction of secondary hyperparathyroidism.

Transfusion requirement and mortality varied by injury pattern, with anteroposterior compression and combined injuries carrying the highest blood loss and mortality; the authors concluded the classification-based protocols reduced pelvic-related morbidity and mortality.

Class 1 traumatic tears are subdivided by site (central, ulnar, distal and radial attachments), and Class 2 degenerative lesions are graded progressively from TFCC wear through perforation, chondromalacia and lunotriquetral ligament tear to ulnocarpal arthritis, reflecting ulnar impaction.

Lesions that fractured had a higher mean score than those that did not (10 versus 7), and fracture risk climbed as the score rose above 7; Mirels proposed that lesions scoring 7 or less can be irradiated safely, whereas higher scores merit prophylactic fixation before radiotherapy.

Defined the MCID as the smallest difference in score that patients perceive as beneficial and that would mandate a change in management, and showed it could be estimated using global ratings of change.

Argued that stable (type A) injuries do well without surgery, whereas unstable injuries, particularly vertically unstable posterior ring disruptions, are associated with persistent pain and deformity and benefit from anatomical reduction and stable fixation.

Showed that large acetabular corrections were achievable through a single approach with the posterior column left intact, allowing early mobilisation without external fixation.

Surgical repair gave no advantage over non-operative treatment.

Showed that correlation coefficients are misleading for judging agreement and proposed a simple graphical approach: plot differences against means and report the mean difference (bias) and limits of agreement derived from the spread of the differences.

Three progressive radiographic groups were seen - change confined to the scaphoid, radioscaphoid arthritis, then generalised wrist arthritis - with longer non-union duration in each successive group. Displacement and carpal instability correlated with more severe arthritis, and few non-unions remained arthritis-free after ten years.

Described the scapholunate advanced collapse (SLAC) pattern as the commonest form of wrist degenerative arthritis, with scaphotrapeziotrapezoid arthritis the next most frequent; in SLAC, arthritis progresses from the radioscaphoid to the capitolunate joint while the radiolunate joint is characteristically spared.

Deep joint sepsis was roughly halved with ultraclean air and reduced further when whole-body exhaust suits were also worn; prophylactic antibiotic use was associated with lower sepsis, though this was not tested in a controlled way.

Defined four progressive stages of perilunar instability: scapholunate disruption, capitolunate disruption, lunotriquetral disruption and finally volar lunate dislocation.

Showed that the cartilaginous infant hip can be imaged diagnostically with ultrasound, a harmless, cheap and non-invasive method suited to screening, laying the foundation for the later Graf classification.

Stage I (low grade), II (high grade) and III (metastatic), each subdivided into A (intracompartmental) and B (extracompartmental), stratified prognosis and linked to defined surgical margins (intralesional, marginal, wide, radical).

Each tissue tolerates a different level of strain: granulation tissue tolerates high strain, cartilage intermediate and lamellar bone very little, so bone can only bridge the gap once strain falls below its tolerance.

Cups within about 15 degrees (plus or minus 10) of anteversion and 40 degrees (plus or minus 10) of inclination dislocated less often than those outside this 'safe zone'; increased anteversion was linked to anterior dislocation, and risk was highest early after surgery and in previously operated hips.

Introduced the Gustilo-Anderson classification (types I to III) and showed infection risk rose with severity of soft tissue injury, supporting prophylactic antibiotics and leaving severe wounds open.

Described the two-column concept, the iliac and obturator oblique views and radiographic lines needed to read acetabular fractures, forming the basis of the later 10-pattern Letournel classification (5 elementary, 5 associated).

First degree equates to neurapraxia and fifth degree to complete transection; second to fourth degrees subdivide axonotmesis by progressive loss of endoneurial and perineurial continuity, with prognosis worsening as more layers are lost and fourth/fifth degree injuries generally needing surgery.