In-hospital care, complications, and 4-month mortality following a hip or proximal femur fracture: the Spanish registry of osteoporotic femur fractures prospective cohort study

We conducted a multi-centric, prospective cohort study in a representative 45 hospitals from 15 autonomic regions from Spain.

All subjects were followed up for up to 4 months after their index admission date.

The main measurements of this study were socio-demographic features of the patient (age, sex, body mass index, civil status, ethnic and educational background, place of residence previous to fracture), comorbid conditions such as previous fractures, diagnose of osteopenia, osteoporosis, rheumatoid arthritis, and secondary osteoporosis (which was confirmed if the patient had any of the following disorders related to osteoporosis: type 1 diabetes, adult osteogenesis imperfecta, hyperthyroidism or premature menopause (< 45 years old), chronic malnutrition or malabsorption, or chronic liver disease). Falls (number of falls in the last year), date of menopause, medications (anti-osteoporosis medication), fracture risk factors (FRAX tool), characteristics of the fracture episode (circumstances, site, and type of fracture), in-patient care received (pre-operative assessment, type of anaesthesia, ASA (American Society of Anaesthesiologists) physical status classification system) grade, surgery, the post-operative management (early mobilisation, constipation prevention or need or urinary catheterisation) and rehabilitation, prophylaxis of post-surgery complications, and secondary fracture prevention measures, type of multi-disciplinary units involved (group of health care professionals including physiotherapy, occupational therapy, nursery and medicine) who are responsible of the assessment and treatment of hip fracture patients) as well as baseline and post-fracture clinical outcomes (inpatient complications and up to 4-month mortality) were also collected.

A descriptive analysis was conducted for all measurements; continuous variables were summarised as mean, median, standard deviation, and quartiles (inferior, superior, minimum, maximum). The number and percentage of patients in each group was reported for each categorical variable. Kaplan-Meier curves were computed to depict post-fracture mortality up to 4 months of follow-up and after stratification per age and sex. All the analyses were stratified by sex, age, previous fractures history, and type of fracture.

A total of 997 subjects admitted to 45 hospitals were included between June 2014 and June 2016, of whom 856 (85.9%) completed 4 months of follow-up, 99 (9.9%) died in the same period, and 42 (4. 2%) were lost to follow up. Baseline characteristics of the study population are reported in Table 1. In brief, participants were mostly widowed Caucasian old women, with primary education levels, who lived in the community until they fractured. Key fracture risk factors were common, including previous fractures (36.5%) and at least one fall in the previous year (43%). Ten-year absolute fracture risk assessed using the FRAX tool was estimated at a mean (standard deviation) of 15.2% (9.0%) and 8.5% (7.6%) for major osteoporotic and hip fracture respectively. Regarding the characteristics of contributing hospitals, median (inter-quartile range) volume was of 290 (200–370) hip fractures per annum, and dedicated staff included a median (inter-quartile range) of 12 (7–15) clinicians.

Fracture type and circumstances at the time of fracture are reported in Table 1. Hip fractures were predominantly either inter-trochanteric (44.6%) followed by intra-capsular displaced (28%), whilst 7 (0.7%) were classified as atypical femoral fractures. Most common circumstance/s leading to fracture were tripping (56.1%) or slipping (23%).

The in-patient pre-operative care and treatments as well as the characteristics of the surgery carried out are detailed in Table 2. Upon admission, 60.6% of the subjects were assessed using a multi-disciplinary protocol, and a medical doctor reviewed 61.9% of them previous to surgery. The three most common surgical procedures carried out were (in order) (1) internal fixation with a short intramedullary nail (38.6%), followed by (2) bipolar-cemented hemi-arthroplasty (15%), and (3_ internal fixation with long intramedullary nail (14.9%). In 75.5% of these surgeries, spinal anaesthesia was used, and almost all of the subjects received either antithrombotic or antibiotic prophylaxis (99.8% and 98.2% respectively).

Different multi-disciplinary models were observed in Spanish hospitals during the whole hospital stay of the patient. An 89.5% of the participating hospitals reported formally co-ordinated care with either internal medicine or geriatrics. A multi-disciplinary protocol was used to guide care in 76.3% of the cases, and the process was usually co-ordinated by specific staff members (either specialised nurse/s or medical doctors).

Regarding surgical delays and time to rehabilitation, these were on mean (standard deviation) of 59.1 (56.7) hours and 61.9 (55.1) hours respectively. Total length of stay was of mean (standard deviation) 11.5 (9.3) days.

Post-operative care/management and complications after the fracture are summarised in Table 2. The most common inpatient complications were delirium (36.1%) and kidney failure (14.1%). The total in-hospital mortality was 2.1%.

At discharge, bone health was reportedly not assessed for 23.6%, assessed but treatment deemed unnecessary/inappropriate for 20.5% of the participants. In addition, 14.9% were awaiting a bone health clinical assessment, and 3.2% were discharged pending a DXA scan before treatment. Regarding anti-osteoporosis medication/s, a 21.4% of the participants were prescribed anti-osteoporosis drug therapy at discharge (8.3% continuing previous therapy and the other 12.1% newly started treatments during hospital admission), and the percentage of such treatment increased but remained suboptimal at 32.1% and 37.8% at 1 and 4 months respectively.

Mortality in the first 4 months, overall, and stratified by sex is reported in Tables 3 and 4 and in Figs. 1 and 2. Kaplan-Meier survival estimates show an overall decreasing trend of mortality from the admission date to the end of the 4th month after the fracture, reaching 11% at the end of the 4th month. Results stratified by sex showed that men had an increased mortality after discharge compared to women (Table 4). When stratifying by age, we found that survival rates were similar in subjects up to 85 years old after which it drops drastically reaching 67% of survival rates among the oldest subjects (> 90 years old) at the end of this period (figure 3 in the supplementary material).