The Fracture Liaison Service of the Virgen Macarena University Hospital Reduces the Gap in the Management of Osteoporosis, Particularly in Men. It Meets the International Osteoporosis Foundation Quality Standards

Objectives: To describe the Fracture Liaison Service (FLS), to know the characteristics of the patients attended with emphasis on sex differences, and to know the compliance of International Osteoporosis Foundation (IOF) quality standards. Methods: Observational, prospective research. All the consecutive patients that attended in usual clinical practice from May 2018 to October 2019, were over 50 years, and with a fragility fracture (FF), were included. Results: Our FLS is a type A multidisciplinary unit. We included 410 patients, 80% women. FF recorded in 328 women were: Hip (132, 40%), Clinical Vertebral (81, 25%) and No hip No vertebral (115, 35%). Those in 82 men were: Hip (53, 66%), Clinical Vertebral (20, 24%) and No hip No vertebral (9, 10%), p = 0.0001. Men had more secondary osteoporosis (OP). The most remarkable result was the low percentage of patients with OP receiving treatment and the differences between sex. Forty-nine (16%) women versus nine (7%) men had received it at some point in their lives, p = 0.04. The probability of a man not receiving prior treatment was 2.5 (95%CI 1.01–6.51); p = 0.04, and after the FF was 0.64 (0.38–1.09). Treatment adherence in the first year after the FLS was 96% in both sexes. The completion of IOF quality standards was bad for patient identification and reference time. It was poor for initial OP screening standard and good for the remaining ten indicators. Conclusions: the FLS narrowed the gap in diagnosis, treatment, and follow-up of fragility fracture patients, especially men. The FLS meets the IOF quality standards.


Introduction
Osteoporosis (OP) is a global public health problem [1][2][3]. It causes 8.9 million fractures annually worldwide. It is estimated that 1000 bone fragility fractures (FF) occur every hour. The probability of having a major FF from the age of 50 for the rest of life is 46% in women and 22% in men [1]. Spain is a country with a mean incidence of OP, with 150 to 250 new cases a year per 100,000 adults over 50 years [1,2].
In recent years, there have been advances in the management of patients with OP. Highlights include tools for estimating fracture risk, effective and safe treatments [1][2][3], being treated, with an emphasis on gender differences, and (iii) To know the degree of completion of The Best Practice Framework quality standards according to IOF.

Study Design
This study was a a prospective, observational, and analytical one in regular clinical practice. It took place from 1st May 2018, to 31st October 2019. Setting: VMUH is an 850 bed, teaching hospital, serving a population of approximately 481,296 inhabitants (157,428 over 50 years old) in Seville, Spain. This population can be treated in 34 PC health areas [18]. The VMUH belongs to the Public Health System of Andalusia, with universal coverage for medical consultation, laboratory tests, X-Ray films, bone densitometry test (BMD). It also covers all specific treatments for OP. All candidate patients are valued for inclusion in VMUH-FLS. After the signing of the informed consent, they are consulted face to face. Their clinical data are included in a real time database of 106 variables in a specialized free software for electronic case report form (OpenClinica ® LLC Enterprises Services Group, USA) [19], as described in the results section.

Inclusion and Exclusion Criteria
We included patients of ≥50 years old, cared for in the VMUH-FLS for an FF occurring in the previous 24 months. They agreed to participate in the study and gave their informed consent. FF was defined as a confirmed X-ray bone fracture, which derives from trauma that under normal conditions does not cause a fracture, after a fall from the patient's height and without acceleration mechanism. In the case of having two or more FFs, the index was coded with the greatest impact on physical function, quality of life, and mortality (hip, vertebra, humerus, distal end of radius (DER), and any others in the same order).
Exclusion criteria: (i) Patients of under 50 years of age, (ii) Patients who did not agree to participate, (iii) Patients with a noticeably short life expectancy (defined by a Paliar Index ≥10 points with a mortality risk at six months >61%) [20], and/or iv. Patients with trauma or pathological fractures.

Statistical Analysis
In a first phase descriptive statistics were calculated with measures of central tendency and dispersion. After that, a graphic analysis was conducted, comparing the main variables by sex, age in decades, and by FF type. Finally, parametric and non-parametric tests were calculated, according to the type of variable. Both sexes were at this point compared, considering as null hypothesis of no differences between them. A two-tailed p < 0.05, with Bonferroni correction for multiple comparisons was considered significant. The statistical analysis was carried out with the STATA v 10.1 package [21].
Sample size calculations. Descriptive analysis does not require sample size calculations. All patients included in the chosen period were selected and analysed. Figure 1 shows both characteristics and composition. They are described below.

Staff and Patients
The staff is composed by a secretary, two specialists in internal medicine, one rheumatologist, one case management nurse, two densitometer technicians, and three PhDs from the research group of the Andalusian Research Plan (PAIDI CTS/211). The clinical coordinator is an internist physician (FJOM), who relies on the case management nurse (MDJM). The head of the group is an internist and professor at the Faculty of Medicine (MJMG).
The type of patients attended are hospitalised and outpatients.

Pathways of Referral
The pathways are multiple, and a direct derivation from PC and various referral services serving hospitalised and outpatients is important. In addition, index cases are identified by the reviewing of emergency lists looking for codes related to FF according to the Minimum Basic Data Set of Andalusia (CMBD) [22] and the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) [23].
Regular visits are carried out to the Plaster and Emergency Room with review of outpatient lists. Surgical demand records are searched for admitted patients. There is also direct derivation by consultation sheets of the reference units (Orthopaedic Surgery and Traumatology, Rheumatology, Internal Medicine, Rehabilitation, PC, and others such as Medical and Radiotherapy Oncology, Urology, Gynaecology, and Endocrinology).

Flow of Patients
There is a dual flow of patients to the unit: • Two-way process: The collaboration with PC is critical and involves sending patients from PC, where they are cared for with an FF. Once evaluated in VMUH-FLS, patients are referred to PC for follow-up, with the option to be re-referred to the VMUH-FLS, if needed. From the reference units, patients with an FF are referred to the VMUH-FLS and vice versa, based on an early care agreement; • One-way process: patients with an FF are derived from referral services, served in the VMUH-FLS, and referred to PC (Figure 1).

Referral Protocols
Previous working meetings were held with all those concerned. They agreed to centralise care in VMUH-FLS. The corresponding diagnostic and treatment protocols were agreed, based on the SEIOMM Clinical Practice Guidelines [24] and the Spanish Society of Rheumatology (SER) [25]. Signage was placed with referral instructions in Traumatology and Orthopaedics consultations, Emergency rooms, and Plaster rooms. In these areas preprinted inter consultations and lab test sheets were provided. Well-functioning pathways were established by telephone, e-mail, WhatsApp, and Tele-consultation.
The referral protocols are different depending on the type of FF and are outlined in Figure 1.

•
Hip fracture: The patient is admitted in the emergency room service and is transferred to the Traumatology and Orthopaedics Units. They contact the Perioperative Internal Medicine Unit that proceeds with the ortho geriatric evaluation. When discharged, the patient is referred to the FLS; • DRE fracture: Most are served in the Emergency room and move to the plaster room, where they are evaluated and derived to the FLS. Those requiring surgical treatment are captured by CMBD [22]

FLS Care
It is carried out on a one-time visit that lasts approximately 3.5 h. In a day of consultation, a physician attends to 12 first visits and four reviews. There are two consultations on Tuesdays and two on Fridays. The rest is by telemedicine.
Once a possible case has been identified, the nurse checks the medical history to verify that the patient meets the inclusion criteria. In this case, she contacts the patient by phone and fixes an appointment, during which, and after explaining the project and signing the consent, a physician and nurse carry out the comprehensive evaluation.
The evaluation includes (i) Anamnesis, (ii) Physical examination with anthropometric measurements (weight, height, and body mass index (BMI)), (iii) Collecting sociodemographic variables, iv. Identifying the mechanism that caused the fracture, smoking, alcoholism, and personal history of fractures and history of falls and treatments, v. Estimating the risk of fracture according to the instrument Fracture Risk Assessment Tool (FRAX ® ) [26].
In addition, some validated Spanish versions of the following specific tests have been used: Generic quality of life measurement (EuroQol5D) [32].
Physicians review the results of a specific analysis: full blood count, Erythrosedimentation rate (ESR), serum levels of calcium, phosphorus, alkaline phosphatase, parathormone, 25(OH) vitamin D, total proteins, albumin, creatinine, glomerular filtration rate, thyrotropin, and tyrosine, basic urine analysis, and calcium/creatinine ratio in urine.
Other tests are added according to clinical guidance (antibodies to transglutaminase and gliadin, protein electrophoresis, testosterone levels, and 24-h urine analysis or autoantibody tests).
The search for morphometric vertebral FF is performed by reviewing lateral X-rays of the dorsal and lumbar spine, with a specific protocol agreed with the Radiology department. The femur neck and lumbar BDM with a dual energy X-Ray absorptiometry (Dexa Hologic Discovery densitometer ® ) is performed and assessed.
Comorbidity and the type and number of drugs at the time of FF are evaluated. Looking for drugs (glucocorticoids, aromatase inhibitors, therapy of prostate cancer, psychotropics, etc) related to OP and falls is important.
Clinical evaluation is recorded in a real time database (OpenClinica ® ) containing 106 variables [20]. A report is obtained from the database and is computerised in the Andalusian Health Service (Diraya) [33] . With all the information the patient is re-evaluated in a single act, during which on suspicion of sarcopenia, femoral rectum muscle ultrasound is performed (ultrasound General Electric Logiq v2 ® ).
Finally, a diagnosis is made regarding OP, risk of falls, FF and a comprehensive targeted treatment plan is done for each patient. The plan includes: 1.
Verbal and written explanation of OP; 2.
Measures of primary prevention of falls and OP; 3.
Exercise plan to improve physical function; 5.
A plan to reduce the risk of falls.
A written report is issued, including a table of exercises aimed at improving muscle balance and function, and a specific drug treatment for OP is instructed. A copy of the report is forwarded to the PC physician.
Follow-up. It is done based on risk. Patients at low and medium risk of FF and falls are referred to PC and monitoring is conducted by phone in the following 6, 12, and 24 months.
Patients at high risk of FF and falls, and those with suspected poor adherence to treatment are followed in FLS with personal appointments every 3, 6, 12, and 24 months, depending on the judgment of the physician.
Adherence is measured by phone or face-to-face interview; and confirmed by reviewing the dispensing of the drug through electronic prescription of the Andalusian System of Health [34].

Patient Recruitment
A total of 450 patients were included in an 18-month study. Of which 351 (78%) were attended to in single-act consultation. The patient flowchart is shown in Figure 2. Forty patients (9%) were excluded. The causes of exclusion were a noticeably short life expectancy (Paliar index ≥10) in 12 (3%) patients, six patients (1%) with age <50 years, three patients (0.6%) with high-impact fractures, four patients (0.9%) with refusal to participate. In total, 15 duplicated records were removed after a database clean-up.

Clinical Characteristics and Treatment of the Patients
Data from 410 patients were analysed and are shown in Table 1, aged 73.5 (±10.2) with a lower limit of 51 and upper limit of 94 years. Most were women (n = 328, 80%). There were no differences between women and men in terms of age, referral time, or BMI, but in weight and height. Men were taller and heavier, as expected. There were clinical and statistical differences in the type of FF (Table 1 and Figure 3). Hip FF was more common in men, whereas DRE and humerus was more common in women. Vertebral FF and the Other FF group had equal gender distribution. In 11% of cases two or more FF were found. This percentage ranged from 5% to 12% in patients with first hip, vertebral, and DRE FF and rose to 67% in the group with other FFs. The anatomic sites of the second FF were vertebral n = 9 (21%), DRE n = 6 (14%), tibia n = 6 (14%), and+ fibula n = 4 (9%).
Two-thirds of patients were evaluated in the FLS within 6 months after the index FF. Most of the FF occurred in patients between 70 to 79 years of age, followed by patients of age between 80 to 89 years, with no gender differences. The prevalence of smoking was three times higher in men. The prevalence of alcoholism was five times higher. Additionally, secondary causes of OP were commonly found in men. In addition, they had lower physical activity. There were differences in the referral units to FLS, related to type of FF.

Sex Differences
The most important gender difference was the low percentage of patients receiving OP specific treatment prior to FF; 49 (16%) women versus 9 (7%) men had been treated at some point in their life before the FF, p = 0.04. After the FF, specific treatment for OP was started in the reference unit in 271 (86%) women versus 48 (63%) men (Table 1 and Figure 4). All patients, both women and men were treated after the VMUH-FLS intervention. The probability for a man not to receive treatment at some point prior to FF was 2.5 (CI95% 1.01-6.51); this difference lowered to 0.64 (0.28-1.09) in the referral unit after the FF; and disappeared in the FLS, as 100% of patients of both sexes received specific treatment.
The drugs used are shown in Figure 5. In total, 65% started treatment with bisphosphonates; 24% with denosumab; 8% with teriparatide and the rest selective oestrogen receptor modulator receptors or hormone replacement therapy. In 188 (47%) patients the specific treatment for OP started in the referral unit was changed in the VMUH-FLS, after identifying barriers in therapeutic adherence. The most common change was from oral bisphosphonate to zoledronic acid IV in 20% of the cases. Figure 6 summarises the data with compliance with the quality standards. The completion of IOF quality standards was poor (red traffic light) for patient identification and FLS reference time items. It was average (amber traffic light) for initial OP screening standard and was good (green light) for the remaining 10 indicators.  • Quality standard 3. Time lapse from FF to VMUH-FLS assessment. The evaluation in the FLS was done after 5.9 (±4.7) months, with a lower limit of 3 days and an upper limit of 24 months. Nearly one-third (27%) were evaluated in the following three months after the FF; and 71% within 6 months after the FF. The indicator received a red traffic light. • Quality standard 4. Adherence to guidelines and number of patients with DXA. Each month the members of the FLS meet and discuss adherence to the guidelines, especially in those difficult or refractory cases. Regular meetings with the reference units and PC were also held. Thus, treatment guidelines are known and applied in a 98% of cases. As a sample, a DMO was prescribed and performed to 365 (89%) patients in the femoral neck and to 366 (90%) in the lumbar spine. The indicator received a green traffic light. • Quality standard 5. Fall risk estimation and prevention. The risk of falls was valued with the SPPB test to 298 (73%) and the Downton scale to 339 (83%) of the patients. In all cases where the risk of falls was high, the nurse initiated an oral and written exercise programming, that was supported by the physician. The indicator received a green light. • Quality standard 6. Secondary causes of OP screening. A cause of secondary OP was sought in all patients by anamnesis and physical examination. The results of the blood and urine tests aimed at investigating secondary OP causes were collected in the database in 61% to 77% of cases, as shown in Figure 6. The indicator received amber light. • Quality standard 7. Multidimensional assessment of potentially modifiable health and lifestyle. Estimations of level of regular exercise and of the degree of physical dependence; the use of OP related drugs, the use of fall related drugs and dairy consumption were made in 92% of the cases. The indicator received a green light. • Quality standard 8. Specific treatment for OP. Following the evaluation in FLS, specific treatment for OP was initiated in 100% of cases. The indicator received a green light. • Quality standard 9. Adherence of specific treatment for OP at 12 months. The nurse contacted all patients by phone at 3, 6, and 12 months. The adherence at 12 months was 95%. This information was compared with the review of the dispensing of the drug in the electronic prescription and the fulfilment at 12 months was 94%. The indicator received a green light.

Discussion
This observational and prospective study analyses the first 18 months of operation of an FLS at a teaching hospital in the Public Health system in southern Spain. The VMUH-FLS treats external and in-patients after an FF. The aim of the study was to describe the characteristics and the operation of the VMUH-FLS, to know the type of patients seen after 18 months of operation, and to know the level of completion of the quality standards according to the IOF and NOS.
It is a Type A and multidisciplinary FLS. In this we proactively intend to: (i) Identify patients accurately with a new FF, including vertebral ones. (ii) Assess and stratify the FF risk, fall risk and severity of OP. (iii) Treat each case with no pharmacological measures and of course with specific drugs for OP in the long term. iv. The follow-up of the patients. [8,[13][14][15][16].
The figure of the clinical leader to reach agreements and collaboration with the reference units and PC is key since these referral units are different from each other. For example, surgical versus medical, emergence units and elective care, and PC [8][9][10][11][12][13][14][15]. This clinical leader seeks common protocols for identifying, diagnosing, treating, and monitoring patients agreed with all these referral units. In VMUH-FLS this is achieved through sessions and narrow communication channels, commented in the FLS description.
The second key figure is that of the nurse, who coordinates the functioning of the unit and the patient care. She is essential in risk estimations, patient follow-up, and treatment adherence. She is the connection among the reference units, PC, and the patient [8][9][10][11][12][13][14][15].
The third pillar of the FLS is the research group with its leader. They have made it easy to connect between the healthcare and the research. Sustainability is guaranteed as our unit gets its funds from resources of the Internal Medicine and Rheumatology Departments, use their own resources from the Andalusian Health System. All the staff spend a day a week within their usual clinical practice. However, ensuring the funding of the nurse is an outstanding issue, as it should not depend on scholarships, but should also be staff of the Health Service [8][9][10][11]. We hope that when we have results of health indicators, such as reduction in re-fracture and mortality rates, the nurse will become a full-time worker.
When analysing the characteristics of all the 408 patients, these patients are like those with OP in Europe [1][2][3]8,9,[13][14][15][16][17][35][36][37][38][39] and Spain [40]. In Europe, one in two women and one in three men of over 50 years will have an FF, and Spain is one of the European countries with one of the most ageing populations in the world . More and more fractures are being addressed in patients >80 years, and the differences between gender decrease in them [1][2][3]7]. A recent systematic review of 33 FLSs shows great heterogeneity in the type of patients and FF, depending on the individual characteristics of the FLSs and the health system where they are based [43]. This review showed patients between the age of 64 and 80, with a man rate of 13% to 30%, and only two FLSs included women. Our study involved four women (80%) for each man (20%), in the upper range of the distribution.
In the review as well as in our study, differences in the type of FF between genders were found. Hip FF occurred most often in men, spine FF alike in both men and women, and DER in women. Men most often had secondary causes of OP, smoking and alcoholism, as well as lower physical activity. These data are described in the literature [1][2][3][7][8][9][13][14][15][16][17][36][37][38][39][40][41]. As far as 2013, Ganda and colleagues had noticed gender disparities in FLS, in the identification, research and treatment of OP [15].
Our data show that men with FF exercise less than women, perhaps because they have more smoking and drinking habits. In our study the most significant difference was the low proportion of men (7%) in contrast with women (16%), who received specific treatment for OP prior to FF and resembling those found in a previous report [7]. After the attention by the FF index in the reference unit, it increased to a 37% and a 47%, respectively. And in the VMUH-FLS it was of 100%, for both sexes. Undoubtedly the implementation of the FLS has helped to improve treatment rates, not only in the FLS per se, but in the reference units, as described [1][2][3][8][9][10][11][12][13][14][15][16][17].
A second systematic review confirms these differences and disparities between genders in the health care of patients with a first FF. This shows how the disparity tends to improve after the attention at FLS [44]. Not only do men receive specific treatment for OP less frequently. They also see doctors less frequently, and therefore the FLS. Furthermore, they are given fewer diagnostic studies, including BMDs [44][45][46]. Their adherence and therapeutic completion are lower, and they are subjected with higher mortality after a hip FF [1][2][3]7,15,[44][45][46]. In this way, the inclusion in FLS decreases significantly gender disparities.
The VMUH-FLS allows the identification of patients of 80 or more years, assuming one third of the total. These elderly people, who often have frail health, constitute a group of "very high" risk of refracture and increased mortality. Some special care must be taken for them, with the collaboration of the orthogeriatric units [47]. These elderly people should be treated in an intensive way, with a specific plan of no pharmacological and pharmacological treatment, and it should be equal between women and men, younger or older without differences.
A recent paper has improved this information and added a traffic light system for its evaluation, based on the percentages at which the indicator is reached [48]. After the analysis we find a poor assessment (red light) in two of the items. The first was patient identification, while the computer system, both CMBD [20] and ICD-10 [21], allows us to identify more than 99% of FFs. The problem arises when capturing these patients and including them in the VMUH-FLS. During the first 18 months we captured only 40% of the patients in the FLS. This low rate of patient uptake is multifactorial. We place it on the learning curve of all the members of the unit. Another possible cause is the lack of interaction with the referral units, mainly PC. PC sent us only a 10%of the patients. For this reason, we must improve communication and interaction flows with PC and other referral units, to improve patient recruitment. This communication and the use of the data provided to us by the statistics and computers lists, will improve the recruitment of both out and in-patients.
The other point with a red traffic light was the time elapsed between the FF index and the evaluation of the patients in the VMUH-FLS. 70% of patients were treated at FLS in the first 6 months after FF, and only one-third in the first 3 months, according to recommendations. This period can be improved. Again, the learning curve caused a delay in the FLS attention. The fact that hip FF is common also contributes to this. Hip FF leads to poor outcomes in mobility and physical function. Some patients, of 80 or more years and a hip FF, delayed the appointments to the FLS due to mobility problems. Both the ortho geriatric service and learning of the FLS members, as well as being aware of the failure, will minimise the delay.
The amber light indicator was the assessment indicator. This organizational problem was detected at the outset. The current coordination between secretariat, blood sampletaking service, densitometer rooms and radiology has collaborating in the improvement of this figure.
The advantages of the study are (i) This is a prospective patient cohort, (ii) It includes patients of usual clinical practice in a real-time database, (iii) It incorporates out and in-patients, and iv. The FLS is Type A. We also have resources for its implementation and maintenance. Being the resources mainly, except for the nurses, dependent of the health system, our FLS will be sustainable in the years to come. Another advantage is the multidisciplinary team with efficient leaders, and all the staff performing their work properly.
The limitations of the study are the low percentage of patients enrolled. We expect to increase it over time in the future. Elderly patients with comorbidities, treated with drugs related to OP, reduced physical capacity and cognitive decline, with high risk of falls and re-fractures require concentrated efforts in the fields of Internal Medicine, Rheumatology, Rehabilitation, Orthopaedic Surgery, and Traumatology, which are sometimes complicated.

Conclusions
HUVM-FLS is a type A multidisciplinary unit that in 18 months of operation has identified, evaluated, and treated 408 patients with OP and an incident FF. Its operation has narrowed the gap in diagnosis, treatment, and follow-up of FF patients, especially men. It is essential to improve patient recruitment, reduce referral times and increase the overall assessment of the patients.  Informed Consent Statement: All the patients received a printed information sheet. After reading it and explaining the protocol, they had to sign an Informed Consent form, according to the ICH Good Clinical Practice E6 guidelines.
Data Availability Statement: Per protocol, all the materials related with the study are stored in the FLS. The electronic clinical records of each patient are in their Diraya, a special software of the Andalusian Health System. The database of FLS is filled in real time and is stored in OpenClinica ® , a special module of the Diraya, on the web page of the Andalusian Health System. The data of the patients are confidential and protected by the current organic data protection law.