Elderly hip fracture patients are often referred for 24-hour electrocardiographic (ECG) monitoring to investigate falls of uncertain aetiology, yet there is limited evidence to support its use within these patients. Moreover, current guidelines for secondary prevention of hip fracture do not emphasise the usefulness of 24-hour ECG monitoring in determining fall aetiology,1 possibly because they have not used hip fracture specific data when making recommendations.

Although any serious injury or fracture, resulting from a fall in the elderly population, is associated with a high degree of morbidity, hip fracture is undoubtedly one of the most devastating. The probability of a further fall is significantly increased in the presence of an undiagnosed arrhythmia.2 While it is suggested that only 3% of hip fracture falls are due to arrhythmias,3 they represent an important potentially reversible aetiology. Moreover, the scope of this problem is increasing, with the UK annual hip fracture incidence forecast to reach 91,500 by 2015, costing between £5,000 and £12,500 per hospital case.1

Previous studies of 24-hour ECG monitoring in patients presenting with syncope have shown poor diagnostic yield.4,5 Similarly, research comparing monitoring results between ‘recurrent-fallers’ with ‘non-fallers’ has shown similar arrhythmia rates, and argues that 24-hour ECG monitoring is not a useful investigation for determining the aetiology.6 Thus, while the cost of each investigation is low, it becomes expensive in terms of overall diagnostic yield per test.3,6 We suspect this makes clinicians reluctant to request this investigation.
Twenty four hour ECG monitoring is a straightforward and non-invasive method of investigating arrhythmias as the cause of falls. In patients with poor re-call and with symptoms suggestive of an arrhythmia, both frequently encountered in patients with hip fracture, it appears a logical first line investigative choice. We set out to determine whether the 24-hour ECG is of use in detecting significant arrhythmia and whether it influences therapeutic outcomes.

Methods
We retrospectively collected data on patients admitted following a fractured neck of femur between April 2013 and January 2014. On admission, each patient underwent a comprehensive falls assessment. This included a complete medical and drug history, physical examination, 12-lead ECG, appropriate bloods tests and urinalysis. 24-hour ECG monitoring was requested in those patients with an uncertain nature of the fall, or an aspect of the history indicating a likely cardiac cause. Patients were fitted with a dual channel 24-hour ambulatory ECG (Space Labs Lifecard™) recorder as an in-patient, or during out-patient follow-up. Printed reports were reviewed by either a care of the elderly consultant or cardiologist, who determined the need for further management. For patients who underwent 24-hour ECG monitoring, the medical records and discharge summaries were examined for actions taken as a result of the monitoring.

Arrhythmia classification
Existing arrhythmia classification systems are largely based on assuming a diagnosis of syncope after a fall.3,4,7 However, although not all patients who fall experience a syncopal episode, many have transient dizzy spells or presyncopal symptoms prior to falling.3 We elected to divide patients into normal and abnormal 24-hour ECG recordings based on arrhythmias commonly recognised within syncopal and fall classification systems.4,6,7,8

Normal 24-hour ECG: 24 hour study or arrhythmias of no clinical significance eg. known rate controlled atrial fibrillation, ventricular ectopics, premature normal and aberrant beats, couplets, bigeminy and trigeminy. 

Abnormal 24-hour ECG: an arrhythmia with the potential to cause a fall, presyncope, or syncopal episode eg. supraventricular tachycardia of 10 or more beats at >150 beats per minute, bradycardia of <40 beats/min, bundle branch block, ventricular tachycardia, new persistent or paroxysmal atrial fibrillation and sinus pauses of >2 seconds whilst awake.

Results
One hundred and fifty nine hip fracture patients were admitted during the time period previously specified. Of this group 44 (28%) underwent 24-hour ECG investigation.

Baseline characteristics
The average age of the 44 patients referred for investigation was 85 years. The gender split was approximately 2:1 with 68% (n=30) female and 32% (n=14) male. Twenty one patients (47%) had a documented cardiac history. Within this group, 27% (n=12) had a history of atrial fibrillation or an episode of paroxysmal atrial fibrillation on a previous hospital admission, one patient had a diagnosis of congestive cardiac failure and two patients had undergone prior pacemaker implantation. Additionally, eight (18%) had ischaemic heart disease, 29 were hypertensive (68%) and six were (14%) were diabetic.

Results of 24-hour ECG monitoring
Of the 44 24-hour ECG recordings, 18 (41%) were considered normal and 26 (59%) abnormal. Of the 26 patients with an abnormal 24-hour ECG, nine (35%) required treatment. Three (12%) underwent PPM insertion: two for sinus pauses of >2 seconds and one for atrioventricular block. A further six patients (23%) required beta blockade for episodes of supra ventricular tachycardia, thought responsible for their fall. The remaining 17 patients were followed up in the outpatient setting prior to commencing anti-arrhythmic medication. 

Discussion
Of 159 consecutive admissions for fractured neck of femur, around a third (n=44, 28%) were deemed to require 24-hour ECG monitoring and abnormalities were seen in 59% (n=26) of these patients. These figures show that a significant percentage of monitored patients were diagnosed with an arrhythmia requiring therapeutic intervention. Holter monitoring in this group of patients appears a useful investigation and consistent with previous reports that continuous 24-hour ECG monitoring is beneficial when the pre-test probability of an arrhythmia is high.4,8

To our knowledge this is the only paper to examine therapeutic outcomes of 24-hour cardiac monitoring in a select cohort of patients with fractured neck of femur. A search of the literature using the search terms ‘hip fractures and ‘Holter ECG recording’ failed to produce any recent publications specifically addressing this subject. Three relevant studies conducted in the 1980s looked at arrhythmia rates following neck of femur fractures.9,10,11 Two of these studies, were small, each reviewing 19 hip fracture patients and showed no significant difference in arrhythmia rates existed when compared to controls.10,11 The third study found arrhythmia rates in 40% of consecutive hip fracture patients versus 12% of controls.11 Findings and methodology used in our study would be most in keeping with those of Abdon & Neilson, as one of these studies excluded patients in institutional care,10 and the second failed to discuss selection criteria and excluded patients with dementia.11 Furthermore, all three studies included patients that current arrhythmia classifications would consider
as ‘normal’.

  To compare our findings with recent research we reviewed studies of syncope and falls in the elderly.4,5,12,13,14 These studies suggest that arrhythmia related syncopal episodes occur in up to 20% of elderly patients, with arrhythmia rates being detected in up to 49% of elderly fallers.4,6,9,10,15 When we compared arrhythmia rates we found significantly higher instances of treatable arrhythmias, namely supraventricular tachycardia (including atrial fibrillation), and sinus or ventricular pauses existed in our study.

A recent study investigating syncope in octogenarians highlighted that Holter monitoring identified up to 8% of octogenarians who required pacemaker implantation and supports the validity of 24-hour ECG as an investigation in these patients. Given that the average age of a male hip fracture patient is 83 and a woman is 84,16 our study would support the use of 24-hour ECG in hip fracture patients. 

The benefit of longer recording periods, in terms of greater arrhythmia yield, remains unanswered in hip fracture patients. However, detecting paroxysmal atrial fibrillation and more specifically, ECG Holter recording in stroke survivors, suggest that a seven day recording will document the arrhythmia in around 70% of patients.17 Secondary prevention forms a crucial part of the care pathway for patients presenting with a neck of femur fracture.1 Much of the current guidance is based on studies in recurrent fallers. 

The consensus is that the cause of a fall is often multi-factorial.14,18 Considering morbidity and mortality in hip fracture patients, ‘casting a wide net’ and investigating common causes of falls is the most effective approach to secondary prevention. A significant number of hip fracture patients fail to recall the events surrounding a fall.6,8,12,15 In this group, given a poor history, average age, and high instance of cardiovascular disease and cognitive impairment, the pre-test probability of having an arrhythmia is high.

Our study has shown 59% of investigated patients had an abnormality recorded during monitoring and, importantly, 35% of this subgroup required therapeutic intervention.
Other studies have shown similar arrhythmia rates between hip fracture patients and controls.9,10,11 although the primary cause of the hip fracture may still have been an arrhythmia. Comparing arrhythmia rates between case and controls is of limited value, whereas the diagnosis of an arrhythmia requiring therapeutic intervention is highly relevant. Treatment based on the outcomes of 24-hour ECG monitoring clearly has the potential to prevent readmission and further fragility fractures.

Conclusion
On the basis of our study, we feel that 24-hour ECG monitoring is a highly effective method of investigating arrhythmia as a cause of a hip fracture. Given the pre-test probability of a positive finding in hip fracture patients, this investigation should be routinely requested in patients with neck of
femur fractures and integrated into secondary prevention guidelines.

Conflict of interest: none declared

References
1. British Orthopaedoc Association. The Care of Patients with Fragility Fractures 2007; 1–80. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2060261/ Accessed 10/07/15
2. Cummings SR, Nevitt MC, Browner WS, et al. Risk factors for hip fracture in white women. N Engl J Med 1995; 332(12): 767–73
3. Brignole M. Distinguishing syncopal from non-syncopal causes of fall in older people. Age Ageing 2006; 35 Suppl 2:ii46–ii50
4. Kühne M, Schaer B, Moulay N, et al. Holter monitoring for syncope: diagnostic yield in different patient groups and impact on device implantation. QJM 2007; 100(12): 771–77
5. Kühne M, Schaer B, Sticherling C, Osswald S. Holter monitoring in syncope: diagnostic yield in octogenarians. J Am Geriatr Soc 2011; 59(7): 1293–98
6. Davison J, Brady S, Kenny RA. 24-Hour ambulatory electrocardiographic monitoring is unhelpful in the investigation of older persons with recurrent falls. Age Ageing 2005; 34(4): 382–86
7. Bass EB, Curtiss EI, Arena VC, et al. The duration of Holter monitoring in patients with syncope: Is 24 hours enough? Arch Internam Med 1990; 150(5): 1073–78
8. Brignole M, Alboni P, Benditt DG, et al. Guidelines on management (diagnosis and treatment) of syncope--update. Europace 2004; 6(6): 467–537
9. Abdon NJ, Nilsson BE. Episodic cardiac arrhythmia and femoral neck fracture. Acta Med Scand 1980; 208: 73–76
10. Taylor IC, Medicine G, Stout RW, et al. Cardiac arrhythmias and femoral neck fracture. Age Ageing 1983; 12: 118–23
11. Mills RM, Muscente D. Cardiac arrhythmia and hip fracture. Int J Cardiol 1984; 6(3): 382–85
12. Carey BJ, Potter JF. Cardiovascular causes of falls. Age Ageing 2001; 30(S4): 19–24.
13. Aharonoff GB, Dennis MG, Elshinawy A, et al. Circumstances of falls causing hip fractures in the elderly. Clinical Orthopaedics and Related Research 1998; 348: 10–14
14. Rubenstein LZ, Josephson KR, Osterweil D. Falls and fall prevention in the nursing home. Clin Geriatr Med 1996; 12(4): 881–902
15. Tan MP, Kenny RA. Cardiovascular assessment of falls in older people. Clin Interv Aging 2006; 1(1): 57–66
16. Healthcare Quality Improvement Partnership. The National Hip Fracture Database National Report. 2010 1–72. https://www.rcseng.ac.uk/news/docs/NHFD%20(final).pdf Accessed 10/07/15
17. Camm AJ, Kirchoff P, Lip GYH (2010) Guidelines for the management of atrial fibrillation. Eur Heart J 2010; 31; 2369–29
18. Rubenstein LZ. Falls in older people: epidemiology, risk factors and strategies for prevention. Age Ageing 2006; 35 Suppl 2: ii37–ii41