Study design and participants

In September 2004, based on the guideline of the Dutch Institute for Health Care Improvement (CBO) on osteoporosis and the guideline on fall prevention, a large European hospital initiated a fracture and osteoporosis outpatient clinic in which fracture patients aged 50 years and above are investigated for osteoporosis [8, 10]. The aim of the outpatient clinic is to improve the care of patients with a clinical fracture, to diagnose osteoporosis, and to determine fall risk in fracture patients. Patients who present with a fracture either at the emergency department, at the outpatients clinic or who are hospitalized because of a facture were invited to the fracture and osteoporosis outpatient clinic.

For the present study, all patients with a clinical fracture that visited the hospital for fracture treatment between April and September 2005 were invited to participate. During the first consultation, the nurse specialized in osteoporosis and fall risk assessment informed and invited every patient individually. If the patient agreed to participate in the study, an informed consent form was signed and handed over to the osteoporosis nurse. The second consultation was combined with a dual-energy x-ray absorptiometry (DXA) measurement. During the second visit, patients were informed about the results of the DXA scan and anamnesis for fracture and fall risk assessment were performed. Patients were included if they were willing to undergo 1) a fall risk assessment and 2) a bone mineral density (BMD) measurement by DXA. Patients were excluded for the following reasons: not completing one of both assessments (fall risk and DXA), deceased, established dementia written down in medical history, living in another region, having a pathological (non – osteoporotic) fracture, no informed consent or no show on DXA appointment. Patients already receiving adequate treatment for osteoporosis were not invited to attend the fracture and osteoporosis clinic where the recruitment for the study occurred. The study was approved by the medical ethical committee of the hospital.

Follow-up and outcome assessment

Follow-up information on falls and fractures was collected by telephone interview, performed monthly for 3 months after the fracture. This three-month period was chosen to evaluate fall risk in the recovery period of the fracture. The osteoporosis nurse asked whether the patient had a fall, and if so, the number and timing of falls and whether the fall resulted in a fracture.

The primary outcome of this study is the incidence of patients who had a fall within the three months of follow-up. Falls were defined as unintentional events which result in a person coming to rest on the floor or a lower level [8, 9, 22, 24]. People who had fallen were classified as a faller or a recurrent faller. A faller was defined as someone who had fallen at least once within 3 months of follow-up, and a recurrent faller as someone who had fallen twice or more within 3 months of follow-up.

Measurements of risk factors

Fall risk was assessed by measuring balance, mobility, lower limb muscle strength, handgrip strength, cognitive status, activities of daily living, visual impairment and general measurement such as blood pressure. These risk factors were chosen based on their description in the Dutch guideline for prevention of falls in the elderly [8].

Balance was evaluated by the Four-Test Balance Scale, in which the patient was asked to perform feet together stand, semi-tandem stand, tandem stand and one leg stand [25]. If the patient was not able to hold at least one of these positions for 10 seconds, this counted as one fall risk.

Mobility was assessed by the Timed Get Up and Go Test, in which the patient is asked to rise from a chair, walk 3 meters, turn, walk back and sit down in the chair. If the patient was not able to perform the test within 12 seconds this was regarded as a reduced mobility and increased fall risk [26, 27].

Lower limb muscle strength was measured by the Chair Stand Test. In this test, the patient was asked to rise up and sit down from a chair as quickly as possible five times, while not using their arms if possible. If the patient was not able to complete the test within two minutes this was regarded as a fall risk [28].

Handgrip strength was measured by the Jamar dynamometer (Jamar, Irvington, NY). The patient was asked to squeeze two times on a handgrip strength indicator with both hands separately. For each hand the maximum score (in kg) was added up. If people squeezed with only one hand, the score of the other hand was replaced by the mean of the group that squeezed with both hands taking sex and dominant hand into account. For women, the cut-off point was < 30 kg, for men the cut-off point was < 50 kg [29].

The Abbreviated Mental Test, a questionnaire to test the cognitive status, was used to assess if patients were cognitive impaired. The cut-off point of this test is 8 (range: 0–10), with a score less than 8 suggesting abnormal cognitive functioning [30, 31].

The Groningen Activity Restriction Scale (GARS) was used to test disability in activities of daily living (ADL) [32]. Patients were asked to answer the questions regarding their abilities just before the fracture. The patient was considered as having a fall risk if he or she had difficulties with a least two out of the three following questions of GARS: a) Can you, fully independently, go up and down the stairs; b) Can you, fully independently, walk outdoors (if necessary with a cane); and c) Can you, fully independently, take care of your feet and toenails? Visual impairment was measured by the Snellen eye chart. Patients viewed the eye chart at a distance of 3 meters. If the visual acuity was less than 0.4, the patient was regarded as visually impaired and having a fall risk [33, 34].

Blood pressure was measured to determine if orthostatic hypotension was present [35]. It was measured both in lying and standing position (after one minute). Further, patients were asked for previous falls in the past 12 months, the ability to keep their balance, problems with walking, difficulties with rising from a chair, difficulties with dressing and undressing, the use of psychofarmaca, polypharmacy (taking 5 or more pills per day), osteoarthritis (patients were asked for previous medical attention for arthrosis of especially the joint in the lower limb), urinary incontinence (patient was asked for involuntary loss of urine), difficulties with reading the newspaper, and depression.

Statistical analysis

Statistical analyses were performed using SPSS for Windows version 12.0.1. Individuals without falls were compared with those with falls. First, univariate logistic regression models were fitted with all individual fall risk factors for the total group and women. If a risk factor had an Odds Ratio (OR) ≥ 2.0 it was retained for subsequent multivariate analysis.

The logistic regression analysis was fitted by the "Forward Likelihood Ratio" method. To identify the relationship between the fall risk factors and falling, OR's and 95% confidence intervals were calculated. Interaction was tested for all significant variables resulting from the univariate analysis.