Falls and fractures in the acute care setting: A retrospective cohort study

Falls with fracture in hospitalised patients remain a common occurrence with significant morbidity and mortality. Our objectives were to determine the characteristics of patients who suffer falls with fractures in hospital, and to examine whether outcomes in this cohort differ from those of patients who fall without sustaining a fracture.


| INTRODUCTION
Falls in the hospital environment remain common, with seven falls per 1000 hospitalisations that result in significant harm, as per current 1 and previous studies. 2,3Injuries are seen in one-third of cases and fractures in 1%-2% of cases, 2,4 although it is acknowledged that no fall is entirely without harm 5 and the psychological sequelae of falls are considerable. 6Reported incidence of falls varies between health-care settings, with data from some jurisdictions suggesting that incidence of in-hospital falls may be increasing in terms of both absolute of falls and the proportion of admissions with falls. 1 In addition to the physical and psychological consequences for patients, in-hospital falls have a sizable economic impact.Whether or not they are associated with injury, falls are associated with increased hospital costs and increased length of stay (LoS). 7This is further prolonged when injury is incurred, 7 and falls with severe injury have been demonstrated to increase the cost of the index hospital stay by 30%. 8Whilst multicomponent interventions may reduce the rate of falls in health-care settings, studies examining their implementation in the acute setting are yet to demonstrate this impact. 93][4] Characteristics of the patient cohort who sustain in-hospital falls with fractures have not been described in detail, and outcomes among this group are unclear.Characterisation of this cohort may facilitate the development of targeted interventions to prevent inhospital harm and will serve to highlight the potential deleterious impacts of such events.We hypothesised that falls with fracture would be associated with increased LoS, readmission and mortality rates.

| METHODS
This retrospective cohort study was compiled using coding data.Ethical approval was granted by the Central Adelaide Local Health Network Research Services, Royal Adelaide Hospital Clinical Trial Centre, Adelaide, South Australia, Australia, on 12 April 2021, ref: CALHN/ HREC/14284.Electronic medical records pertaining to all inpatient admissions in a large tertiary public hospital over a six-year period (2012-2017) were interrogated.All patients coded as having suffered a fall in hospital during this period were identified.These patients were divided into two cohorts: those who did and those who did not suffer fractures as a result of their fall.
Patient's age and sex were recorded.The presence of selected comorbidities, where available, (dementia, delirium, malignancy, heart failure, chronic obstructive pulmonary disease [COPD] and osteoporosis) was noted using coding data.These comorbidities were selected for their relevance to bone health, falls risk and increased prevalence in the older inpatient.Where available, the fracture site was documented.Outcome measures collated included length of stay, inpatient mortality and 6-month readmission rates based on the hypothesis that a fall with fracture portends adverse outcomes.Data were analysed using SPSS version 26. 10 Descriptive statistics were performed.Risk factors for falls with fracture were assessed using binary logistic regression.Outcome measures were compared between those who did and did not sustain fractures.Linear regression was used to estimate change in LoS due to various exposures, including inpatient fracture.Findings were reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for observational studies. 11

| RESULTS
There were 236,720 inpatient admissions and 721 (<1%) inpatient falls recorded over the six-year period, 128 (18%) of which were associated with a fracture.Neck of femur fractures represented the most common fracture site (29%), followed by humerus (9%) and radius (7%) fractures.Within the group of patients who fell and did not sustain a fracture, 41 (7%) died, whilst 17 (13%) patients who fell and sustained a fracture died.Patients who suffered a fracture as a result of their fall were older than those who did not suffer a fracture (76.8 vs. 72.8years, p < 0.010) and more likely to be female (53% vs. 44%, p < 0.020).A higher incidence of delirium (30% vs. 21%, p < 0.040) and dementia (23% vs. 13%, p < 0.004) were noted in the fall with fracture cohort.The prevalence of other comorbidities studied did not differ significantly between cohorts (Table 1 below).Only 4% of the patients who suffered a fall with fracture were listed as having osteoporosis in the coding data.

Practice Impact
This study helps assist the clinician in prognostication for those who fall and sustain fractures, suggesting increased mortality and length of stay.Given the poor outcomes following a fall with fracture, there may be a benefit from preventing harm from falls as well as their occurrence.
Multiple linear regression analysis was utilised to determine factors associated with length of stay in all patients who sustained a fall.Linear regression demonstrated that a fall with fracture added 9.2 days to length of stay (95% CI 5.5-12.9,p < 0.001).Two other factors significantly influenced length of stay, namely delirium (10.6 days, 95% CI 7.2-14, p < 0.001) and a diagnosis of malignancy (4.7 days, 95% CI 0.8-8.7,p < 0.03).

| DISCUSSION
In this retrospective single-centre study, 18% of coded falls recorded were associated with a fracture.The fall with fracture cohort was older and more likely to have a diagnosis of delirium or dementia.Advanced age, cognitive impairment and delirium are well-known to represent risk factors for falls in hospital inpatients.This study highlights that these characteristics are also associated with fractures among those who fall in hospital.
Older age represents an independent risk factor for fracture 12 in comparison with the general population, with patient-related adverse events in older people admitted to hospital being significantly associated with both dementia and delirium. 13his study serves to highlight the potential gravity of falls associated with fractures in the hospital setting.The odds of dying in hospital were almost twofold greater in the fall with fracture cohort.Fracture was confirmed as an independent risk factor for inpatient death among people who fall in hospital, surpassed in significance only by a documented diagnosis of malignancy.Length of stay was significantly prolonged in the fall with fracture group.As well as increasing costs, prolonged LoS is associated with a greater risk of hospital-acquired infection and functional decline. 14he deleterious effects of falls with fractures in hospital are clearly illustrated and underscore the need for effective interventions to prevent such events.An increasing tendency for governments and insurers to financially penalise hospitals for health-care errors ensures that effective strategies to prevent injurious falls are of interest not only to patients and clinicians but also to health system administrators.Whilst multifactorial interventions may reduce the rate of falls in hospital, 9 multiple studies demonstrating reduced rates of falls have failed to impact rates of injury. 15Better delineating the patient cohort affected by injurious falls could help inform targeted interventional strategies.
The retrospective nature of this work and its dependence on electronic coding data must be highlighted.The reported rate of falls in the coding data is significantly below previous estimates. 1This may reflect a failure to code (and perhaps even document) falls deemed to be less serious.The recorded prevalence of osteoporosis in this study was notably low at 4% in the falls with fracture group.This is almost certainly a gross underestimation 1 particularly if one assumes that most fractures will have been sustained in association with a fall from standing height or less.It is likely other comorbidities that might have contributed to fracture risk were similarly uncoded.Given the use of coding data, further details such as fall location in hospital and mitigation strategies in place at time of fall were not available.Data on other injuries other than fractures were not collected and, as such, limited conclusions are able to be drawn on the contribution of other injuries to patient mortality.

| CONCLUSIONS
Despite the aforementioned flaws, this study plays a valuable role in better characterising the patient cohort at risk of falls with fractures in the acute hospital setting.We highlight the grave consequences of these events and the need for prospective studies of interventions designed to prevent their occurrence.Whilst falls prevention should remain a priority, interventional studies aimed specifically at preventing fractures associated with these events may improve these patients' outcomes in terms of mortality and length of stay.
Comparison of cohort characteristics and comorbidities.
T A B L E 1