Trauma

Periprosthetic femoral fracture versus native hip fractures

Thankfully, periprosthetic fractures following total hip arthroplasty (THA) are relatively uncommon, being reported to complicate between 0.1% and 4% of patients during the lifetime of their prosthesis. However, with increasing numbers of arthroplasties being performed, and an ageing population, the number of these fractures can be expected to increase. Much has been written on the outcomes following native hip fractures, particularly in terms of mortality, but much less is known about the mortality following periprosthetic fractures of the hip. The authors of this study from New York, New York (USA) set out to evaluate the mortality of periprosthetic hip fractures, comparing them to native femoral neck and intertrochanteric hip fractures over a nine-year period.¹ This report is based on a database query of 1655 patients, all with a periprosthetic fracture of the proximal femur. Patients had a mean age of 78.9 years. A comparative group of 97 231 patients with a native hip fracture had a mean age of 82.6 years. Patients with a periprosthetic fracture had a lower mean comorbidity score and also had longer surgical delays. At the traditional 30 days following injury, the mortality rate for patients with a periprosthetic fracture was 3.2%, compared with 4.6% for those with a native hip fracture. Adjusting for possible confounding factors, there was no difference between the two groups for risk of death at one month. At six months, the mortality rate for those patients in the periprosthetic group was 3.8%, compared with 6.5% for those native hip fracture group. Adjusting for confounding factors, the risk of death remained lower in the periprosthetic group compared with the native hip fracture group. These differences in mortality became even more marked at one year (9.7% vs 15.9%) in the native hip fracture group. The risk of death was 29% lower for the periprosthetic fracture group compared with the native hip fracture group. The authors also identified factors that increased the mortality risk. In the native hip fracture group, advanced age, male gender, being Caucasian, higher comorbidity scores, and surgical delay were all associated with increased mortality risk. Those patients in the periprosthetic fracture group saw an increased risk in mortality if they were aged 80 years and over, were of male gender, and had higher comorbidity scores. Interestingly, there were no significant associations with surgical delay in this patient group. The authors did find that there was an increased risk of mortality associated with revision arthroplasty for periprosthetic fracture, as opposed to open reduction and internal fixation. This study was interesting for several reasons. First, it identified that the mortality risk is similar for both groups in the acute period following the injury, but, over the longer term, those in the periprosthetic group fared better. This is one of the largest studies of this kind published to date and involves patients from multiple institutions, enabling the authors to exclude a number of important potential biases. The difference in mortality risk for those patients who had fixation as opposed to replacement for their prosthetic fracture is not too surprising. Revision hip arthroplasty involves a greater physiological insult, with longer operative time and higher blood loss. This study also suggests that surgical delay was not a risk factor for increased mortality in the periprosthetic group, unlike the widely accepted (but yet to be proven) view in those patients with native hip fractures. Previous studies have confirmed that surgical delay in patients with native hip fractures is associated with a worse prognosis. Perhaps the periprosthetic hip fractures group have fewer significant comorbidities? What is clear is that the incidence of periprosthetic hip fractures is likely to increase, and that all institutions need a standardized pathway to manage these high-risk patients with multidisciplinary involvement, including physicians and surgeons with the necessary expertise.

Factors associated with revision surgery after internal fixation of hip fractures

Surgical timing 30-day mortality in adults undergoing hip fracture surgery

On the face of it, patients who are delayed for their hip fracture surgery are likely to be at a greater risk of complications and are more likely to suffer those complications through delay to surgery. Any prolonged period of bed rest, supine, would appear to increase frailty, and the subsequent increased risk of anaesthesia would clearly increase the risk of death. However, despite many studies examining the potential association, there is still no clear-cut answer one way or the other. The investigators in the current study aimed to perform a population-based, retrospective cohort study of 42 230 adults (mean age, 80.1 years) undergoing hip fracture surgery over a five-year period at 72 hospitals in Toronto (Canada).³ The authors undertook a relatively advanced analysis using risk-adjusted restricted cubic splines to model the probability of each complication according to wait-time. The inflection point (in hours) when complications began to increase was used to define early and delayed surgery. Their main outcomes measure was mortality within 30 days. Secondary outcomes included a composite of mortality or other medical complications (myocardial infarction, deep vein thrombosis, pulmonary embolism, and pneumonia). The authors established that the risk of complications increased significantly when wait times were greater than 24 hours, irrespective of the complication considered. Compared with 13 731 propensity-score-matched patients who received earlier surgery, the 13 731 patients who received surgery after 24 hours had a significantly higher risk of 30-day mortality (6.5% vs 5.8%) and the composite complications outcome (12.2% vs 10.1%). The authors concluded that a wait time of 24 hours for hip fracture surgery may represent a threshold-defining higher risk. Whilst this is a straightforward database trawl study, it is powerful enough to quantify what has always been suspected: an excess complication rate following delay to surgery. The authors’ use of a composite complication outcome and a cubic splines method allows complex non-linear associations to be defined. Rather than simply selecting an arbitrary cut-off for comparing early versus late complications, the authors were able to accurately establish the best threshold by determining the point at which the complications rates are likely to rise from surgical delay.

WOLLF for lower-limb open fractures

The Wound Management of Open Lower Limb Fractures (WOLLF) study, one of the most hotly anticipated trials in trauma, has recently been published in JAMA.⁴ The authors designed a large pragmatic randomized controlled trial comparing dressing types in open lower-limb (chiefly tibial and femoral) fractures that were not suitable for primary closure. A total of 460 patients with grade III open lower-limb fractures, aged over 16 years, were recruited across the UK Major Trauma network over a three-year period. Patients were all treated within 72 hours of fracture and did not undergo primary closure, so, as such, they had a dressing applied prior to delayed flap closure or delayed primary closure. The pragmatic nature of the trial was such that patients randomized to the intervention group were treated with a vacuum-assisted dressing (VAC) of any variety, and the control group had a standard dressing (including bead pouch and all other types of non-vacuum dressings). Outcomes were assessed in terms of limb function (disability rating index) and health-related quality of life (using the Euro-Qol 5D (EQ-5D) and health-economic analysis). The study team recruited 226 patients to the VAC group and 234 patients to the standard dressing group. The bottom line result was that there were no differences in any of the primary outcome measures at 12 months. However, perhaps more surprising was the finding that there were no differences in the numbers of deep surgical site infections: 7.1% (n = 16) in the VAC versus 8.1% (n = 19) in the standard dressing group. From the quality-of-life perspective, which would also be expected to be different if there were a difference in secondary complications, there was a barely detectable difference of 0.02 points on the EQ-5D between the two groups. This is the largest randomized trial reported in open lower-limb fractures to test different dressings. It looks, somewhat counterintuitively, like the VAC dressings do not improve self-rated disability at 12 months.

Dosage of prophylactic antibiotics and open fractures

There is, unusually, almost universal agreement within the orthopaedic trauma fraternity that the best and most important treatment for open fractures is antibiotic prophylaxis on admission. Most systems are centred on early administration of antibiotics to minimize later infection. These protocols, however, focus on a fixed dose of antibiotics, whatever the size and weight of the patient. We were interested, here at 360, to read this simple but insightful report from Memphis, Tennessee (USA), in which the authors aimed to establish how often use of standard protocol results in the administration of a dose of antibiotics that is too low, and to try to establish if this has an effect on eventual infection rates.⁵ The authors evaluated infection rates at one year in 63 patients with open tibial fractures (Gustilo–Anderson type IIIA or IIIB) presenting open extra-articular tibial fractures to a single centre over a five-year period. The infection rate was high, as expected, with 33% (n = 20) of patients developing infection by the one-year follow-up point. The most startling observation was that, of the 20 patients who went on to develop a deep infection, around half of the patients had an appropriate dose of antibiotics. Of those who did not develop a deep infection, around three quarters had received the correct dose of antibiotics. This difference was significant and clearly suggests, although it is not entirely causally proven, that underdosing of intravenous antibiotics may be in part to blame. This suggestion is further underlined by the observation that, in this patient series, there was a higher incidence of cefazolin-sensitive infections in those that were underdosed. It is an easy win to introduce a weight-based dosing regimen in those antibiotics that require it, and the suggestion here is that this will reduce the incidence of antibiotic resistance. The caveat to this, of course, is that there may well be an association with being overweight, comorbidity, and susceptibility to infection.

Competency in hip hemiarthroplasty

Neglected pilon fractures and the Ilizarov technique

The next big thing in compartment syndrome?

Our understanding of the pathophysiology of compartment syndrome has moved on tremendously. What was once perceived as a purely pressure-related phenomenon has been attacked from all sides, with studies being undertaken into the local metabolism, tissue PaO₂ levels, and a dizzying array of other measurables. Just about anything you can measure in a limb is being measured in a limb. However, whilst the majority of these new and exciting technologies haven’t actually made it off the drawing board, spectroscopy has done. The initial papers describing the technique, which are now a few years old, have been surpassed by this latest article from The Bone & Joint Journal, which originates in Athens, Georgia (USA). The study evaluated near-infrared spectroscopy (NIRS) with the aim of validating as a clinical measure in continuous, non-invasive setting. The authors report a diagnostic study using 86 patients, all presenting with a severe leg injury, and 23 normal controls. Monitoring was undertaken for up to 48 hours and NIRS values were recorded. A fairly thorough but exploratory approach was taken with the statistics, and the longitudinal data sets were evaluated with graphical methods, bivariate comparisons, and multivariable multilevel modelling. The series included seven patients with compartment syndrome confirmed decompression. The continuous NIRS monitoring was able to pick out those with compartment syndrome in this cohort. On average, there was a 3% drop in NIRS in at least one injured compartment below the uninjured contralateral compartment in all seven patients with known acute compartment syndrome. We are in desperate need for a better diagnostic test for compartment syndrome. In the awake patient, clinical examination remains probably the most sensitive and specific test, and certainly represents the gold standard. However, in the unconscious or anaesthetized patient, compartment pressure monitoring is often used, which has significant drawbacks. A non-invasive continuous monitoring test would be a very welcome addition to the diagnostic armoury.