Mortality following combined fractures of the hip and proximal humerus

Aziz Haque, Harvinder Pal Singh

Aziz Haque, Harvinder Pal Singh, Department of Orthopaedic Surgery, University Hospitals of Leicester, Leicester B170RH, United Kingdom

Abstract

Key Words: Hip fracture; Proximal humerus fracture; Mortality; Return to home; Orthopaedics; Trauma

INTRODUCTION

Hip fractures are common with over 66000 patients sustaining the injury in England and Wales in 2019[1]. We also know that over a third of all fractures occur in patients over the age of 65[2]. With an aging population the number of hip fractures presenting to our hospitals is likely to increase. Hip fractures are associated with a higher mortality in both the immediate and late post-operative periods. Mortality is usually quoted as around 10% at 30 d and 30% at 1 year[1]. There are well established guidelines in place for the management of hip fractures in the elderly[3].

Fractures of the proximal humerus are the third most common fragility fracture after hip and distal radius accounting for around 6% of all fractures[2]. Longer term community studies have shown a higher mortality in those patients that have suffered a fracture of the proximal humerus with Wilsonet al[4]showing a two fold increase at 1 year (9.8%) and 5 years (28.2%)[5]. Mortality is likely to be higher with advancing age and increasing number of comorbidities[4,5].

Mortality following combined hip and proximal humerus fractures is likely to be higher and the aim of our study was to describe 30 d and 1-year mortality for these patients and compare them to patients who have suffered a hip fracture alone.

MATERIALS AND METHODS

All hip fractures over the age of 60 presenting to a single large trauma unit were identified using a prospective national database of hip fractures (National Hip Fracture Database) from January 2010 to December 2016. Patients sustaining a proximal humerus fracture in addition to their hip fracture were identified using hospital coding data which was then confirmed with radiographic review. All pathological hip and proximal humeral fractures were excluded in addition to those sustaining other significant orthopaedic injuries. Office of national statistics data was used to verify mortality in our patient group. We calculated 30 d and 1-year mortality for those patients that had sustained a hip fracture alone and those that had sustained a combined hip and proximal humeral fracture. Other variables recorded included age, gender, inpatient stay and discharge destination. A radiographic review of all proximal humeral fractures was performed to classify them and record their treatment.

Statistical analysis was carried out using the SPSS software (IBM, Armonk, NY, United States). Mortality and other secondary outcome measures were compared using Fisher’s exact test due to a large difference in sample size between the two groups. Difference was considered significant if aPvalue of < 0.05 was reached.

RESULTS

In total, 4131 patients were identified with a hip fracture in the study period. Out of these 40 had sustained both a hip fracture and a proximal humerus fracture. Mean age in the hip fracture cohort was 80.9 years and in the combined fracture group 80.3 years. The male to female ratio in the hip fracture group was 1:2.4 and in the combined fracture group 1:3.4. In terms of our primary outcome measure the 30-d mortality in the hip fracture cohort was 7.2% compared to 12.5% in the combined cohort (P= 0.163). The 1-year mortality for our hip fracture cohort was 26.4% compared to 40% for the combined fracture cohort (P= 0.038) (Table 1).

In terms of secondary outcome measures, mean hospital stay was 14 d for hip fracture patients compared to 16.3 d for patients with the combined injury (P= 0.163). Only 29% of patients with the combined injury were discharged back to their own home compared to 47% in the hip fracture group (P= 0.022). Four of the 40 proximal humerus fractures had operative fixation. No difference in mortality was seen between different fracture types or methods of treatment.

DISCUSSION

Sustaining a hip fracture is amongst the commonest reasons for elderly patients needing emergency surgery. This comes with increased risk both in terms of morbidity and mortality[1]. Few patients will go on to achieve their full pre-injury abilities, which means that a significant proportion of patients would require increased level of care[6]. This poses a significant socioeconomic burden on the National Health Service. The nation institute for health and care excellence estimates 30 d mortality of upto 10% and 1-year mortality of upto 30% in those patients that have sustained a hip fracture[3]. National guidelines and the formation of the national hip fracture database have gone a long way in improving and standardising the care provided to these patients in England[1].

In our study we wanted to compare the outcomes of a small cohort of hip fracture patients that have also sustained a concurrent proximal humerus fracture. Similar to other previous studies we found that combined hip and proximal humerus fractures were more prevalent in females (male:female = 1:3.4) and sustained by a comparably elderly group of patients (mean age 80.3 years). In our analysis patients with combined hip and proximal humeral fractures had higher mortality at 30 d and 1 year when compared to those patients that had suffered a hip fracture alone. Although the difference was not statistically significant at 30 d, it did become significant at one year. Patients with these combined injuries were also less likely to return to their own home.

Combined proximal humerus and hip fractures are relatively rare and we found that only 1% of hip fractures had sustained this injury. A single unit would therefore only expect to see a few of these injuries a year. In our cohort only 4 patients out of 40 had their proximal humerus fractures treated operatively. Patients that have sustained a combined hip and an upper limb fracture are likely to face greater difficulty in terms of rehabilitation. With proximal humeral fractures this can be more of an issue as splints and adapted walking aids cannot be used to help improve mobility as it can for distal radius fracture for example. It is unclear whether this is the actual reason for increased mortality, but the socioeconomic burden of these combined injuries is clear.

Previous studies with smaller numbers have described a possible increased risk of mortality with proximal humeral fractures but their numbers have been smaller. Robinsonet al[7]found 21 proximal humeral fractures in 1971 consecutive hip fracture patients and suggested that mortality was increased but when hip fractures were associated with a distal radius fractures, their mortality appeared to be slightly lower. Mulhallet al[8]looked at 760 hip fractures and only found 5 fractures of the proximal humerus in their retrospective review, they found that with upper limb injuries there was an increase in inpatient stay with greater difficulties in mobilisation. Similar results were also reported by Kanget al[9]in 2019 when they found 35 upper limb fractures in 1018 hip fracture patients. Only 8 had sustained a combined proximal humerus fracture with increased mean inpatient stay. We have included 4131 hip fractures in our study with 40 also sustaining a proximal humerus fracture.

Treating all proximal humeral fractures operatively in those patients that have also sustained a hip fracture may not be possible due to the nature of their co-morbidities. Operative treatment may, in addition, not lead to a lower mortality and the benefits of surgery in terms of rehabilitation may be limited by post-operative restrictions. More work is needed in this area, however, due to the rarity of these injuries, large studies would be difficult to perform.

Table 1 Summary of mortality data for isolated hip fractures and combined hip and proximal humerus fractures

The main limitation of our study is its single centre retrospective design, use of hospital coding data and small number of patients particularly in the combined hip and proximal humeral fracture group. This was, however, unavoidable due to relative rarity of this combined injury. The large difference in sample size between the two groups may lead to difficulties in statistical analysis and this is acknowledged.

CONCLUSION

This retrospective review of a prospectively collected database over 7 years in a large trauma unit has looked at 4131 hip fractures with 40 out of those sustaining a combined proximal humerus fracture. These combined injuries are relatively rare but are likely to be associated with a higher mortality at 1 year. They also cause increased difficulty in terms of patients reaching their pre-injury abilities as they are also less likely to return to their own home following discharge. Our hope is that publication of this information would not only lead to a more informed discussion with patients and their families, but also generate discussion amongst trauma and shoulder surgeons about developing better strategies of treating these combined injuries.

ARTICLE HIGHLIGHTS

Research background

Combined hip and proximal humerus fractures are rate injuries that may be associated with increased mortality and morbidity.

Research motivation

The motivation for this study came from the idea that patients being admitted to our unit following these combined injuries may not have been getting good care for the proximal humerus fracture as hip fractures were the priority. So we wanted to see if mortality is higher for these patients and inpatient stay. Also the motivation was to increase awareness of this injury with other trauma surgeons so that more consideration can be given to these injuries.

Research objectives

To compare mortality and inpatient stay for patients with combined hip and proximal humeral fractures and hip fractures alone.

Research methods

Retropective single centre analysis of local data from a national database.

Research results

Increased mortality and inpatient stay with combined injuries.

Research conclusions

Combined fractures of the hip and proximal humerus are associated with increased morbidity and mortality when compared to isolated hip fractures. These combined injuries are relatively rare and more equal consideration should be given to both fractures when they do occur.

Research perspectives

This is important for those surgeons looking after patients with hip fractures and specialist upper limb surgeons.