The Phenotype of Patients with a Recent Fracture: A Literature Survey of the Fracture Liaison Service

The aetiology of fractures in patients aged 50 years and older is multifactorial, and includes bone- and fall-related risks. The Fracture Liaison Service (FLS) is recommended to identify patients with a recent fracture and to evaluate their subsequent fracture risk, in order to take measures to decrease the risk of subsequent fractures in patients with a high risk phenotype. A literature survey was conducted to describe components of the bone- and fall-related phenotype of patients attending the FLS. Components of the patient phenotype at the FLS have been reported in 33 studies. Patient selection varied widely in terms of patient identification, selection, and FLS attendance. Consequently, there was a high variability in FLS patient characteristics, such as mean age (64–80 years), proportion of men (13–30%), and fracture locations (2–51% hip, <1–41% vertebral, and 49–95% non-hip, non-vertebral fractures). The studies also varied in the risk evaluation performed. When reported, there was a highly variability in the percentage of patients with osteoporosis (12–54%), prevalent vertebral fractures (20–57%), newly diagnosed contributors to secondary osteoporosis and metabolic bone disorders (3–70%), and fall-related risk factors (60–84%). In FLS literature, we found a high variability in patient selection and risk evaluation, resulting in a highly variable phenotype. In order to specify the bone- and fall related phenotypes at the FLS, systematic studies on the presence and combinations of these risks are needed.


Introduction
Fractures constitute a major health care concern worldwide, as 50% of women and 20% of men at the age of 50 years will sustain a fracture during their remaining lifetime [1,2]. Since the world population is ageing, the annual number of fractures is expected to increase from 3.5 million in 2010 to 4.5 million in 2025, corresponding to an increase of 28% [3].
Fractures indicate an increased risk of subsequent fractures and premature mortality [4][5][6][7]. Current guidelines recommend secondary fracture risk evaluation in all men and women aged 50 years and older with a recent clinical fracture [8][9][10][11]. However, many fracture patients were not offered appropriate secondary fracture prevention, resulting in a care gap throughout the world [12].
Fracture Liaison Services (FLS) have been designed and implemented to diminish the care gap [13]. The key components and objectives of a FLS are multiple. Firstly, case finding by systematic identification and selection of fracture patients. Second, to adequately evaluate subsequent fracture risk using clinical risk factors for fractures and falls, dual-energy X-ray absorptiometry (DXA) and imaging of the spine for detection of previously unknown vertebral fractures. Third, analysis for eventual underlying secondary osteoporosis and metabolic bone disorders. Fourth, adequate treatment in patients at high risk, and fifth, development of a follow-up program [14].
Unfortunately, FLS are currently established in a small proportion of facilities that receive fracture patients worldwide [15]. The International Osteoporosis Foundation (IOF), American Society for Bone and Mineral Research (ASBMR), European League Against Rheumatism (EULAR), and European Federation of National Associations of Orthopaedics and Traumatology (EFORT) support the implementation of FLS as they identify this as the most successful approach for secondary fracture prevention [11,[15][16][17][18]. In this literature survey, we investigate what has been published on components of the bone-and fallrelated risk factor phenotype in patient attending the FLS.

Methods
A literature search was conducted in PubMed/Medline, EMBASE and CINAHL to identify relevant publications up to and including October 2016 using the following search terms: Fracture Liaison Service, fracture prevention service, fracture prevention clinic, fracture prevention program, osteoporosis clinic, and secondary fracture prevention. The search was limited to human studies in adults (18-64 years) and aged (C65 years) written in English. We specifically selected articles which reported components of the phenotype of patients at the FLS. Finally, additional relevant publications known to us were added.

Search Results
After removing duplicates, our search resulted in 373 potentially relevant publications. Based on title and abstract screening, 270 publications were excluded. Based on full-text eligibility assessment, 80 publications were excluded, resulting in 23 being selected. The reasons for exclusion were no FLS population (n = 40), and no components of the phenotype reported (n = 40). In addition, manual searches through the reference lists were performed, resulting in 10 additional publications. In total, 33 publications were included in this literature review (Table 1).

Patient Selection Procedure
The patient selection procedure can comprise up to three steps: (1) the identification and selection of patients with a recent clinical fracture for evaluation at the FLS, (2) the patients' response to the FLS invitation (i.e. the proportion of patients willing and able to attend the FLS), and optionally (3) the selection of a subgroup of FLS attenders to be included in the publication.
Identification and Selection of Patients for Evaluation at the FLS Patient identification and selection differed markedly across studies (Table 1 and Fig. 1). Twenty-nine studies identified and selected in-and outpatients [13,, two studies selected only inpatients [47,48], and two did not report this aspect of patient identification and selection [49,50]. With respect to age, 26 studies identified and selected patients age 50 years or older [13,48]. Five studies used other age criteria, namely patients aged 45 years and older [43,49], patients aged 75 years and older [44], or those who were postmenopausal [45,46]. In two studies, no age criterion was used [47,50]. Thirty-one studies identified and selected both men and women [13,[47][48][49][50], whereas two studies selected only postmenopausal women [45,46]. Patients with any fracture were identified and selected in 31 studies [13, 19-41, 43-47, 49, 50], whereas only patients with a nonvertebral fracture were selected in two studies [42,48].

Attendance
Selected patients were informed personally or through an information letter, except for the study by Fraser et al. [21], in which a letter was sent to the general practitioner informing them of the fragility fracture and invited referral to the fracture prevention clinic. In 17 (52%) of the 33 studies ( Fig. 1), 20-89% of the patients selected for evaluation at the FLS actually attended the FLS (Table 1 and Fig. 2) [19-25, 30, 33-36, 38, 43, 44, 47, 49].

Components of the Phenotype Age and Gender
In 29 of the 31 studies in which both men and women were included, the proportion of men ranged from 13 to 30% (  [19,34]. In both men and women, mean age was highest in hip fracture patients [40].

Body Mass Index
Mean body mass index (BMI) was reported in nine studies, ranging from 24 to 29 kg/m 2 (

Bone Mineral Density
In all 33 studies, bone mineral density (BMD) measurement at the lumbar spine and hip was performed (Table 3 and Fig. 4) [13,, with additional measurements at the distal radius in one study [33]. Based on the lowest T-score, osteoporosis was diagnosed in 12-54% of patients   [13,19,29,48], and vertebral fracture [19], and least in patients with a foot, and clavicle fracture [19]. Classified according to Center et al. [6], osteoporosis was found in 31% of patients with a minor, in 49% of patients with a major, and in 58% of patients with a hip fracture [30]. Osteopenia was found in 49% of patients with a minor, in 39% of patients with a major, and in 42% of patients with a hip fracture [30].

Trabecular Bone Score
Only Nassar et al. [48] reported the trabecular bone score (TBS) in non-vertebral fracture patients at the FLS. Mean TBS was 1.201 ± 0.113 and mean TBS was lower in patients with VFs than in those without VFs in VFA (1.156 ± 0.108 vs. 1.227 ± 0.107, p \ 0.0001).

Daily Calcium Intake
Only three studies reported mean daily calcium intake [19,33,38], ranging from 759 to 912 mg/day, and two studies reported daily calcium intake \1200 mg/day, ranging from 86 to 91% of patients [30,38]. Daily calcium intake \1200 mg/day was similar for men and women, age decades, fracture location according to Center et al. [6], and patients with a normal BMD, osteopenia, and osteoporosis [30].

Fracture Risk Assessment Tools
FRAX score for major fractures was 8-13% in four studies, and for hip fractures 3-7% in four studies [23][24][25]28]. In  Fig. 4 Percentage of studies reporting assessments for fracture risk evaluation Table 3 Performance of assessments (DXA, VFA, laboratory tests, and fall risk assessment), and when reported, the results  ? Excl. Excl.

Fall-risk Assessment
Fall-risk assessment was reported to be performed in eight studies (Fig. 4) [29, 35-38, 40, 42, 44]. Only four studies [35,36,38,42] reported prevalence rates of fall-risk factors, with at least one fall-risk factor in 60-84% of patients (Table 3). All fall-risk factors were more frequently reported in women, with the exception of impaired vision, which was found in 25% of women and 31% of men [35].

Discussion
This survey aimed to describe the bone-and fall-related components of the phenotype of patients attending the FLS based on 33 FLS related papers. The reported phenotypic characteristics varied widely among the various publications with regard to the mean age, proportion of men, and fracture location. In addition, the proportion of patients with osteoporosis, prevalent vertebral fractures, newly diagnosed contributors to secondary osteoporosis and metabolic bone disease, and proportion of patients with fall-related risk factors varied substantially across studies. Although, there is a great heterogeneity in components of the phenotype, the prevalence rates of these components were high. The heterogeneity of reported phenotypes of FLS patients can be explained by several aspects. Firstly, the variability in the FLS patients phenotype can be explained by differences in patient selection and FLS attendance. Positioning papers on secondary fracture prevention by the ASBMR, IOF, and EULAR/EFORT [11,15,18], recommended that all patients aged 50 years or older with a recent fracture should have their risk for subsequent fractures evaluated at the FLS. In three out of four studies, this recommendation was implemented successfully. Nine studies selected another group of patient for evaluation at the FLS based on different selection criteria (only inpatients, only women, only patients aged 75 years or older, only NVF patients). Additionally, various combinations of selection criteria were used, such as only low-trauma or fragility fracture patients, or excluded patients with pathological fractures. Further, FLS attendance rates ranged from 20 to 89%. This indicates that achieving adequate FLS patient selection and attendance is a major challenge and often hampered by logistic obstacles. It has been shown that FLS care with a central coordinator (often a specialised nurse) is the most appropriate clinical organization model for secondary fracture prevention [11,15,18]. Although capturing all fracture patients is the ultimate goal, it has been suggested that an FLS may Table 3 continued Author DXA Normal BMD initially focus on a subgroup [15]. Once secondary fracture prevention for these patients has been well-established, the scope of the FLS should be expended to eventually include all fracture patients. In addition, other approaches, such as an orthogeriatric service, may have been established in hospitals to systematically optimise care of hip fracture patients, including components covered by a FLS [52]. This type of service of course alters the phenotype of the patients attending the FLS. In our literature survey, all but six studies focussed on all patients regardless of their fracture location. Of these six studies, one study [50] excluded hip fractures. Second, as recommended in the positioning papers, risk evaluation should include dual-energy X-ray absorptiometry (DXA), and vertebral fracture assessment (VFA), and on indication, laboratory tests, and fall risk assessments [18]. DXA evaluation was performed in all studies, imaging of the spine in nine studies, laboratory tests in 21 studies, and fall risk evaluation in eight studies. Since these assessments often have to be justified through local business cases supported by solid health economic analysis, which are currently lacking, implementation of these assessments is not always feasible. Hence, the reported outcomes of the various bone-and fall-related components of FLS patients may be influenced not only by patients selection and attendance rates, but also by the possibility to perform additional assessments in all FLS patients [52].
Based on these results in literature, it is difficult to describe the full spectrum of bone and fall risks in patients attending the FLS. In the context of fracture prevention, knowledge of the presence and combinations of the risk factors will guide the need for evaluation and treatment. In this literature survey of FLS, we found a high variability in patient selection and fracture risk evaluation. In order to specify the bone-and fall-related phenotypes at the FLS, systematic studies on the presence and combinations of these risks are needed.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://crea tivecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.