Abstract
Introduction:
Osteoporotic hip fractures receiving surgery are common and early rehabilitation is needed in the acute hospital setting. In Singapore, many receive rehabilitation after hip surgery in a tertiary hospital, and various clinical variables may be predictors of functional outcomes.
Methods:
We retrospectively reviewed 68 patients who went through inpatient rehabilitation in Singapore General Hospital. The primary outcomes of this study were to identify predictors which affect Functional Independence Measure (FIM) efficiency and motor FIM gain at discharge. The secondary outcomes include predictors affecting ambulation distance at discharge and rehabilitation length of stay (RLOS).
Results:
Age, dementia and days from fracture to surgery are important predictors of FIM efficiency; age and FIM efficiency are important predictors of ambulation distance; and type of fracture is an important predictor of RLOS. Patients of age <75 (OR 2.419, p=0.002), absence of dementia (OR 2.570, p=0.045) and those who received surgery <3 days from fracture onset (OR 2.529, p=0.036) achieved greater FIM efficiency. Younger patients of age <75 (OR 23.177, p=0.030) and those with FIM efficiency of more than 7 points per week (OR 38.963, p=0.05) achieved greater ambulation distance at discharge. Type of hip fracture is an important predictor for RLOS, with neck of femur fracture patients having shorter RLOS (OR 7.186, p=0.005).
Conclusion:
Age, dementia, days from fracture to surgery and type of hip fractures are important predictors of early functional outcomes in inpatient rehabilitation setting.
Introduction
Osteoporotic hip fractures are common worldwide, with the incidence rising as a result of ageing populations. 1 Hip fractures are associated with decline in physical function. 2 In the acute setting after hip surgery, many patients require inpatient hospital rehabilitation to regain physical function. Some of the goals of acute inpatient rehabilitation are to optimise mobility and shorten length of stay (LOS). Previous research suggests intensive rehabilitation in some patients may reduce rehabilitation duration. 3 LOS is only one of the many outcome measures used. Other measures in the acute setting include the Barthel Index, ADL score, Functional Independence Measure (FIM) and hospitalisation costs.3–7 These outcomes are affected by a variety of factors such as preoperative general condition, walking ability and type of surgery. 8 In Singapore, there are some studies on functional outcomes after hip fractures, but studies reviewing the use of FIM, ambulation distance at discharge from hospital and inpatient rehabilitation LOS as acute outcome measures are scarce.9,10 We sought to identify demographic and clinical predictors affecting the functional outcome of post-operative hip fracture patients in the early phase. Outcomes include FIM, ambulation distance achieved at discharge and rehabilitation LOS.
Materials and methods
We undertook a retrospective review of patients who went through rehabilitation in the Department of Rehabilitation Medicine, Singapore General Hospital (SGH). Data from January 2013 to December 2014 were retrieved. In total, 68 patients were transferred from the Department of Orthopaedics to the Department of Rehabilitation for inpatient rehabilitation after hip surgery. All 68 patients stayed in SGH throughout their course of rehabilitation.
Demographics and premorbid ambulatory status
Demographic information such as age, gender and pre-existing comorbidities (hypertension, diabetes, cerebrovascular accident, ischaemic heart disease, renal impairment or diagnosed dementia prior to hip fracture) were identified. The premorbid ambulatory status was determined on entry to rehabilitation: (a) those who ambulate independently without aid; and (b) those who ambulate with walking aid or wheelchair.
Medical information
These patients suffered from either unilateral neck of femur fracture or intertrochanteric fracture. Various hip surgeries were performed including cancellous screw fixation, proximal femoral nail anti-rotation (PFNA), hemiarthroplasty (unipolar and bipolar), dynamic hip screw insertion or total hip replacement. The mean time in days from sustaining a hip fracture to the time of receiving hip surgery was recorded. Complications recorded after hip surgery include cardiovascular events such as arrythmia, cardiac related-hypotension or acute coronary syndrome, urinary tract infection, pneumonia and lower limb deep venous thrombosis. The mean drop in haemoglobin (g/dL) after surgery was defined as difference in haemoglobin level before the surgery and immediately after surgery. Those who received blood transfusion after surgery were recorded as “yes” and those without “no”. Those who were discharged directly home were recorded as “home”, while “CH” refers to those transferred to community hospitals. Caregiver availability was recorded as “yes” or “no”.
Inclusion and exclusion criteria
Inclusion criteria include: (a) medically stable patients after surgery and (b) patients who can bear at least partial weight on the surgical side. Exclusion criteria: (a) patients who are non-weight bearing; (b) those who chose to select rehabilitation in step-down services in a community hospital after surgery; (c) patients who chose not to participate in the therapy programme offered by the inpatient rehabilitation department. The weight-bearing status of all patients was determined by the individual surgeons who operated on the patient. Patients who chose not to participate in the rehabilitation programme were discharged directly home from the Department of Orthopaedics. Those who chose step-down services were transferred directly from the Department of Orthopaedics. Retrieval of data in this study was approved by Singhealth Centralised Review Board.
Outcome measures
Admission FIM was assessed on the day of transfer to the Department of Rehabilitation Medicine. FIM is a widely used scale to assess physical and cognitive disability and focuses on the burden of care. Items are scored based on the level of assistance required by an individual to perform his or her activities of daily living. The scale consists of 13 items of physical domains and five items of cognitive domains, where each item is scored from a scale of 1–7; 1 represents total dependence and 7 indicates complete independence. During the course of rehabilitation, FIM scores were scored weekly by the rehabilitation team which comprises the rehabilitation physicians, the physiotherapists, the occupational therapists and the speech therapists. A discharge FIM score for each of these 68 patients was collected on the week of discharge. Motor FIM gain for each patient is defined as the difference between discharge motor FIM and admission motor FIM. FIM efficiency per week is defined as motor FIM gain divided by rehabilitation LOS.
In terms of LOS, rehabilitation LOS refers to length of stay in the Department of Rehabilitation Medicine. We sub-divide patients into those who need 14 days and above for rehabilitation and those who need less than 14 days. Ambulatory distance at the point of discharge was defined as the minimal distance (in metres) a hip fracture patient can ambulate after surgery with or without the use of walking aid. The ambulatory status at discharge was defined as those who could ambulate without aid and those who need aid to walk.
Statistical analysis
Descriptive statistics for quantitative variables are presented as mean, standard deviation (SD) and n (%) for categorical variables. The variables included in this model were: age, gender, premorbid ambulation status, comorbidities, clinical factors such as type of fractures, rehabilitation LOS, medical complications after surgery (pneumonia, urinary tract infection, cardiac complications and deep venous thrombosis), drop in haemoglobin after surgery, administration of blood transfusion, and other variables such as caregiver availability, discharge destination, ambulation distance and status at the point of discharge. Independent t-test was used in univariate analysis to compare quantitative variables, while multiple linear regression analysis was used to determine the variables associated with FIM efficiency, motor FIM gain, ambulation distance and rehabilitation LOS. Significance was determined at p⩽0.05, and statistical analysis was done using SPSS version 20.
Results
Baseline characteristics
The mean age of the patients was 75.9 years ± SD 9.2. Of these, 31 (45.6%) were of age <75. The majority (77.9%, n=53) were female. Six (9.1%) had pre-existing dementia diagnosed prior to sustaining hip fracture. The two most common comorbidities were hypertension (69.1%, n=47) and diabetes (29.4%, n=20). Some 57 patients (83.8%) were premorbid independent without aid. Treatment in 67.6% (n=46) was for neck of femur fractures, and the remaining were intertrochanteric fractures. Hemiarthroplasty was the most common surgery performed (57.4%, n=39), followed by PFNA (25.0%, n=17). The mean total LOS was 31.8 days ± SD 12.6. The average number of days from sustaining a hip fracture to surgery was 4.75 ± 4.23 days. The mean acute LOS was 11.4 days ± SD 7.6. The mean rehabilitation LOS was 20.4 days ± SD 10.1. The mean FIM score on admission to rehabilitation was 85.5±SD 13.5 and the mean discharge FIM score was 98.1±SD 15.3. Mean motor FIM gain was 12.7± 7.5. Mean FIM efficiency per week was 4.84 ± SD 3.55. All 68 patients had full cognitive FIM scores of 35 points. Fifteen patients (22.1%) developed urinary tract infection during their hospital stay and were treated with either oral or intravenous antibiotics. Eleven (16.2%) developed cardiac complications after surgery but were responsive to medical treatment. None developed pneumonia or deep vein thrombosis during their hospital stay. The mean haemoglobin after surgery was 9.8 ±SD 1.4 g/dL. Mean drop in haemoglobin after hip surgery was 1.8 ±SD 0.9 g/dL. Some 59 patients (n=86.8%) were discharged home and the remaining to community hospital; 58 (n=85.3%) were ambulating with aid on discharge, while 10 (n=14.7%) did not require aid. The mean ambulation distance in metres was 57.9 ±SD 46.3. No statistical significance was found between neck of femur fracture and intertrochanteric fracture cohorts with regards to demographics, comorbidities and premorbid ambulatory status. The baseline characteristics are shown in Table 1.
Characteristics of the 68 patients after hip surgery.
Predictors of FIM efficiency, motor FIM gain, ambulation distance and status
Univariate analysis between clinical characteristics and outcomes was correlated using independent sample t-test. Age >75 (mean=3.6±SD 2.5, p=0.002), dementia (mean=1.7±SD 1.3, p=0.023), intertrochanteric fracture (mean=3.4±SD 2.1, p=0.005), blood transfusion (mean=3.7±SD 2.3, p=0.029), cardiac complications after hip surgery (mean=3.0±SD 4.6, p=0.04) and >3 days from fracture to surgery (mean=6.7±SD 4.2, p=0.000) were more likely variables leading to lower FIM efficiency per week. Cardiac complications (mean=9.3±SD 4.6, p=0.027) and patients with dementia (mean=6.7±SD 3.7, p=0.005) were variables associated with lower motor FIM gain. Likewise, age >75 (mean=23.4±SD 11.8, p=0.007), dementia (mean=29.2±SD 19.4, p=0.026), intertrochanteric fracture (mean=26.7±SD 12.2, p=0.002) and those with FIM efficiency of <7 (mean=12.8±SD 4.5, p=0.001) were associated with longer rehabilitation LOS. Age <75(mean 75.7±SD 57.6, p=0.006) and those with higher FIM efficiency (mean 15.1±SD 5.4, p=0.019) achieved greater ambulatory distance (See Figures 1(a, b) and Table 2).
Univariate analysis associating FIM efficiency with days from fracture to surgery, comparing cohort with <3 days to surgery to ⩾3 days to surgery.
Univariate analysis showing higher FIM efficiency in cohort with age <75.
Univariate analysis on association between FIM efficiency, motor FIM gain, ambulation distance and rehabilitation LOS.
Using multiple linear regression analysis, age <75, absence of dementia and <3 days to surgery remained as predictors for greater FIM efficiency. Neck of femur fracture was a significant predictor for shorter rehabilitation LOS, whereas age <75 and FIM efficiency >7 points per week were significant predictors for higher ambulation distance at discharge. FIM efficiency was higher in those <75 years old (OR=2.42, 95% CI: 0.90–3.94, p=0.02). Patients without dementia gained greater FIM efficiency (OR 2.57, 95% CI: 0.06–5.08, p=0.045). Receiving earlier surgery <3 days from fracture was more likely to achieve higher FIM efficiency (OR 2.53, 95%CI: 1.03–4.03, p=0.001). Type of hip fracture was also a significant predictor, where neck of femur fracture patients were more likely to achieve shorter rehabilitation LOS (OR 7.19, 95% CI: 2.22–12.15, p=0.005). Age of <75 years (OR 23.12, CI: 2.30–44.05, p=0.03) and FIM efficiency per week of >7 (OR 38.96, CI: 13.89–64.04, p=0.003) were strong predictors of ambulation distance at discharge. Absence of cardiac complications and dementia were not statistically significant for motor FIM gain (Table 3).
Multivariate analysis on association between FIM efficiency, ambulation distance and rehabilitation LOS.
Discussion
Previous studies have identified various important predictors for functional outcome after hip fracture surgery.4,11–14 Advanced age, dementia, low premorbid functional level, general medical conditions, living with a relative or having adequate social support, fracture type, time to surgery, pain muscle strength and low haemoglobin levels are common cited predictors. Age ranging from more than 65 to more than 80 years has been extensively studied in the past, and is associated with poorer early functional recovery.4,15,16 Our study chose 75 years as the cut-off, which is line with previous studies and the mean age of our 68 patients. FIM is an easy-to-use and commonly used standardised measure of general disability in most rehabilitation units. 17 Motor FIM gains and FIM efficiency have been used in hip fracture studies as part of short-term outcome measures.18–20 In SGH Department of Rehabilitation Medicine, the FIM is a quality measure for Rehabilitation Physicians and Allied Health to keep track of functional improvement. As expected, lower FIM efficiency was achieved in those >75 years old. We postulated that older patients may require more days of therapy to achieve functional gains, as they are more likely to have reduced muscle mass and muscle strength, with slower cognition functions. Lefaivre et al. suggested that early hip fracture surgery should be practised to reduced hospital LOS and medical complications. 14 Morever,existing guidelines suggest that earlier surgery is associated with a better functional outcome and shorter hospital stay. 21 These findings are consistent with results from our study where the cohort with less than 3 days from fracture to surgery had shorter rehabilitation LOS and greater FIM efficiency. The SGH rehabilitation team could possibly collaborate with Orthopaedic Surgeons on a future seamless hip fracture pathway to advocate logistic arrangement for early surgery and direct early transfer of post-surgical hip fracture patients to inpatient rehabilitation based on a clinically validated set of criteria. A set of cognitive screening tools such as the Mini-mental state examination would also be needed to identify patients with dementia.
In terms of rehabilitation LOS, SGH inpatient rehabilitation services provide acute intensive rehabilitation to patients, where the therapy programme is usually about 2 weeks, hence our study used 14 days as the cut-off period. The neck of femur fracture cohort in our study had shorter rehabilitation LOS. Fox et al. investigated the characteristics of neck of femur versus intertrochanteric fractures, and found that one of the factors leading to longer hospital stay and recovery in intertrochanteric fracture was lower trabecular bone density compared with the neck of femur counterpart. 22 Bone mineral density (BMD) was not used as one of the clinical variables in this study as not all of the 68 patients were screened. Future studies on hip fractures treated at SGH could include a hospital guideline on screening BMD in osteoporotic hip fractures and the impact of osteoporosis on functional recovery and LOS.
One of the goals in hip fracture rehabilitation is to encourage early mobility within the home and in the community. Minimum ambulation distance has been studied in Singapore, the United States and Australia for walking in the community and within the home environment.23 Older patients aged 75 years and above with hip fracture achieved a lower mean ambulation distance of 42.9 m compared with those younger than 75 years with a mean distance of 75.7 m. This review of current data is a pilot project towards better understanding of the ambulation status of patient at discharge. There is a need for the inpatient rehabilitation team to develop suitable exercise programme, to move beyond using FIM and develop a set of more robust outcome measures to better reflect the ambulation status of older patients. The primary vision is to help all inpatient hip fracture patients achieve community ambulation as soon as possible. We suggest the continuing use of FIM as one of the acute outcome measures in the inpatient hip programme. One possibility is to consider using motor FIM efficiency of 7 points or more as a weekly benchmark for individual hip fracture patients when setting rehabilitation goals.
Limitations
Several variables and previously validated outcomes were not included in our review, such as pain score, muscle strength, 10-m walk test and grip strength. The current data reviewed hip fracture patients who were previously diagnosed with dementia; there was lack of data on the cognitive status of these patients after surgery. Use of Mini-mental state examination as a screening tool during inpatient rehabilitation might better reflect the existing cognitive status. A small sample size in our study also explained the statistically insignificant outcome in some of our clinical predictors when correlated to functional outcome. A future prospective study of optimal sample size should be considered with inclusion of these important clinical predictors and outcome measures.
Conclusion
In evaluating short-term outcomes of the hip fracture population, age, presence of dementia, days from fracture to surgery and type of hip fracture are important predictors of FIM, ambulation distance and rehabilitation LOS in an inpatient hip fracture rehabilitation setting.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Declaration of Conflicting Interest
The authors declare that there is no conflict of interest.