Materials and methods
Clinical study
We identified all patients presenting to our unit with concomitant hip and wrist fractures between July 2004 and April 2011. We recorded the length of stay and date of mortality. We collected and compared similar data on patients presenting to our institution with isolated hip fracture between January 2010 and December 2010. Demographic data and information relating to the injury was collected, including the laterality of fracture for both groups. Our analysis looked retrospectively at outcomes for a large cohort of patients treated.
Normality testing was performed with the Shapiro–Wilk test. Normally distributed data was compared with the two-sample t-test. Non-parametric data underwent analysis with the Mann–Whitney U test. Nominal data was compared using Fisher’s exact test. The 95 % confidence interval for proportions was calculated using standard methods. Cox’s proportional hazards ratio was used to determine the effect of concomitant wrist fracture in patients with hip fracture while adjusting for age and gender.
A PubMed search was performed to conduct the meta-analysis. Search terms were the MeSH term (neck of femur fracture) in combination with “wrist”, “upper limb” or “distal radius”. An iterative process was then used with the papers identified and their references. Meta-analysis was performed by pooling numerical data. When values were heterogeneous, in particular when combining data regarding early mortality, where some studies looked at in-hospital mortality and others recorded 30-day mortality, the random effects model was used.
Discussion
Wrist and hip fractures are two of the most common and clinically significant fractures treated by the orthopaedic surgeon. When they occur in combination they pose a peculiar challenge. Expeditious surgery is mandated in hip fracture patients as it is thought to improve survivorship [
6]. In England and Wales, the Department of Health provides financial incentives, in the form of the Best Practice Tariff, to National Health Service Trusts where patient care meets a minimum standard [
7]. Surgery within 36 h is one criterion to be satisfied if trusts are to enjoy the Tariff. Functionality and independence are dependent upon hip and wrist function. No previous studies have examined the effect of concomitant wrist fractures in hip fracture patients, correcting for potential confounders.
In the present study the combination and singular fracture cohorts were recruited from different time frames. This was necessary to include sufficient numbers of patients with hip and distal radius fractures. This may potentially impact upon our findings. Over the periods in question (2004–2011 for hip/wrist and 2010 for hip fracture) operative fixation remained the mainstay of treatment for hip fractures. In particular, there was no significant difference in the method of fracture fixation for the two cohorts (Table
3).
Those with combination hip and wrist fractures did not have a significantly different 30-day, 90-day and 1-year mortality compared to those with isolated hip fractures. There was a considerably higher proportion of women in the combination fracture group (89 vs 69 %, respectively p < 0.0001). Correcting for both age and gender with Cox’s proportional hazard regression analysis, we found that the presence of an associated wrist fracture did not significantly impact upon mortality in patients with hip fractures.
A review of the literature, involving smaller studies, suggests equally interesting findings (Table
4). In 2002 Mulhall [
2] performed an analysis of all patients presenting to his institution with simultaneous hip and upper limb fractures. He found that wrist fractures were the most common upper limb fracture associated with hip fractures. He also observed a significant female preponderance when compared to patients with isolated hip fractures. The combination fracture cohort had a longer in-hospital stay and lower in-hospital mortality.
Table 4
Meta-analysis of studies exploring effect of coincident hip fracture in patients with hip and wrist fracture
| 28 | 3.7 % | 77 vs 84* | 3:1 vs 8:1* | 10.3 vs 5.6 % (in-hospital mortality)* | | 15.6 vs 20.4 (mean) | |
| 33 | 2.6 % | 78 vs 79 | 2:1 vs 6:1 | | | 17 vs 23 | |
| 34 | 1.8 % | 82 vs 83 | 4:1 vs 7:1 | 6.4 vs 7.7 % | 28 vs 19 % | 13 vs 17.5 | |
| 46 | | | 7:1 (no data of isolated hip fracture) | | | 13 (no data of isolated hip fracture) | |
This study | 88 | 1.7 % | 80 vs 79 | 4:1 vs 9:1 | 9.6 vs 9.1 % | 30.6 vs 25 % | 13 vs 18 | 0.86 (95 % CI 0.57–1.28) |
Meta-analysis | 229 | 2.0 (95 % CI 1.7–2.4) | 79.8 vs 80.5 | 3:1 vs 7:1 (p < 0.0001) | Relative risk 0.93 (95 % CI 0.53–1.65) | 29 vs 24 % (p = 0.2) relative risk 0.81 (95 % CI 0.58–1.13) | | |
Tow et al. [
4] performed a matched case–control study. In this they compared 33 patients with coincident hip and wrist fractures with 33 patients suffering from isolated hip fractures. The comparators were matched for age and gender. They observed a similar female predilection. Tow et al. interestingly observed that the combination fracture group were slightly more osteoporotic than those in the isolated hip fracture group, but the difference was not statistically significant.
Robinson and co-workers, in 2012, analysed the features of patients with concomitant hip and upper limb fractures. Similar to Mulhall they observed distal radius fractures to be the most common associated injury [
5]. Consistent with our study and preceding works, Robinson noted that there was a high female:male ratio and longer length of hospital stay, in instances of hip and concomitant wrist fracture.
Pooling the available data from the literature, we observed that 2 % (95 % CI 1.7–2.4) of patients with hip fracture suffered a concurrent wrist fracture. The narrow confidence interval suggests the accuracy of the value. Both cohorts had a similar age. All previous studies found a much higher proportion of female patients in the group with combined wrist and hip fracture. We considered whether the similar survivorship observed in patients with simultaneous hip and wrist fractures, in spite of the presence of two fractures, was due to the female preponderance acting as a confounder. Male patients have a much higher mortality following hip fracture compared with women. The most recent meta-analysis, involving in excess of 64,000 patients, indicates that male sex engenders a 1.7-fold increase in mortality compared to female patients [
8]. We thus decided to adjust mortality for gender and age. In this present study, using Cox’s proportional hazard analysis adjusted for age and gender, there remained a non-significant difference in survivorship in patients with hip and wrist fractures compared to those with isolated wrist fractures. No adjustment was made for potential differences in co-morbidities between the two cohorts. However, both samples were sufficiently large to be representative and correction was made for the pre-eminent difference, namely gender. Further, differences in the mortality between male and female hip fracture patients are not related to co-morbid status [
9].
A minority of patients with hip fractures sustain concomitant wrist fractures. However, given the incidence of hip fractures, this number is not negligible. This is the largest study exploring the outcome of concomitant hip and wrist fractures. This is the first meta-analysis of studies examining the natural history of patients with synchronous hip and wrist fractures. The combination fracture occurs much more commonly in women and patients require a greater length of hospitalisation. The patients who sustained simultaneous hip and wrist fractures suffered no significant difference in survivorship when compared to those who suffer isolated hip fractures. It is tempting to assume that the combination fracture is indicative of a frailer patient and poses a greater risk to life. However, our findings and the meta-analysis suggest that the combination hip and wrist fracture does not portend increased mortality compared to patients with isolated hip fractures.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.