Secondary prevention and estimation of fracture risk




Abstract


The key questions addressed in this chapter are:


• How can individual risk of fracture be best estimated?


• What is the best system to prevent a further fracture?


• How to implement systems for preventing further fractures?


Absolute fracture risk calculators (FRCs) provide a means to estimate an individual’s future fracture risk. FRCs are widely available and provide clinicians and patients a platform to discuss the need for intervention to prevent fragility fractures.


Despite availability of effective osteoporosis medicines for almost two decades, most patients presenting with new fragility fractures do not receive secondary preventive care. The Fracture Liaison Service (FLS) model has been shown in a number of countries to eliminate the care gap in a clinically and cost-effective manner.


Leading international and national organisations have developed comprehensive resources and/or national strategy documents to provide guidance on implementation of FLS in local, regional and national health-care systems.


How can individual risk of fracture be best estimated?


Fracture risk calculators


Fragility fractures are the clinically significant consequence of osteoporosis. As such, being able to assess the future fracture risk is an essential component of any strategy to reduce the burden of disease. During the 1990s, fracture risk assessment was primarily informed by measurement of bone mineral density (BMD). Since the turn of the century, introduction of absolute fracture risk calculators (FRCs) has enabled clinicians to integrate clinical risk factors (CRFs) and BMD into fracture risk assessment. Details of several FRCs are discussed subsequently, which have been the subject of reviews and comparative evaluation .


The World Health Organization FRC FRAX ®


In 2008, the FRAX ® FRC became available as an online tool to calculate the 10-year probability of a major fracture (hip, clinical spine, humerus or wrist fracture) and, individually, the 10-year probability of hip fracture . FRAX ® FRCs are currently available in 51 countries and can be accessed online at www.shef.ac.uk/FRAX . The FRAX ® calculation is based upon age, body mass index (BMI) and a number of dichotomised variables including:




  • Previous fracture



  • Parental history of hip fracture



  • Current tobacco smoking



  • ‘Ever use’ of long-term glucocorticoids



  • Rheumatoid arthritis



  • Secondary causes of osteoporosis (including type I (insulin dependent) diabetes, osteogenesis imperfecta in adults, untreated long-standing hyperthyroidism, hypogonadism or premature menopause (<45 years), chronic malnutrition or malabsorption and chronic liver disease)



  • Alcohol consumption of ≥3 units per day



Fracture risk prediction can be enhanced by inclusion of femoral neck BMD. Notably, the FRAX ® algorithm takes into account the competing risk of death. As with all risk assessment tools in medicine, outputs from FRAX ® should be considered to be complementary to clinical judgement. The key limitations of FRAX ® are that dose response for several risk factors cannot be used to inform calculation of risk. These include smoking, glucocorticoid dose and the number of previous fractures experienced by the individual. Furthermore, other risk factors that may have a significant impact on fracture risk are not optional inputs, history of the patient’s fall being the most notable omission.


FRAX ® is available on several technology platforms, including versions for desktop computers, tablets and smart phones, in addition to the online version. FRAX ® has proved to be popular with almost 6 million individuals having undergone fracture risk assessment between June 2011 and August 2013 . A growing number of national clinical practice guidelines have FRAX ® -based intervention thresholds, including guidelines from Argentina , Canada , France , Switzerland , the UK and the USA .


The Garvan FRC


The Garvan FRC was developed using data from the Dubbo Osteoporosis Epidemiology Study in Australia . The calculator is available online at http://www.garvan.org.au/bone-fracture-risk and has been subject to independent validation in Canada and in New Zealand . The Garvan FRC generates 5- and 10-year fracture risk estimates for hip fracture and a broadly defined group of fragility fractures, and the algorithm is based upon five variables (age, gender, femoral neck BMD, number of falls in the past year and number of fractures since the age of 50 years). BMI can be substituted if BMD is not available.


QFracture ®


First published in 2009, and subsequently updated in 2012, the QFracture ® algorithm was developed from data drawn from a large primary care database in the UK . Qfracture ® is available online at http://www.qfracture.org/ and as an App for iPad and iPhone. This FRC was developed specifically for use within the UK and is based upon 25 variables.


A systematic approach to fracture risk assessment


Osteoporosis is the most common bone disease and fragility fractures affect a substantial proportion of the population; one in two women aged over 50 years will suffer a fracture, as will one in five men . Accordingly, a key challenge for policymakers and health-care professionals is where to start with a chronic disease of this scale? Considerable effort has been put into answering this question over the last decade. A crucial observation is that fracture begets fracture ; individuals who have already suffered a fragility fracture are twice as likely to suffer second and subsequent fractures in the future compared to their peers who are yet to break a bone . Furthermore, consideration of the previous fracture history of individuals who have suffered the most serious fragility fracture, a hip fracture, reveals that around a half of the patients who suffered hip fractures had previously broken another bone . Epidemiological estimates from several countries suggest that approximately 16% of post-menopausal women have suffered a fragility fracture . Thus, implementation of secondary fracture prevention strategies first, followed by primary prevention strategies – as time and resources permit – provides a rational stepwise approach to maximise health gains in the shortest time frame, as illustrated in Fig. 1 .




Fig. 1


A systematic approach to fragility fracture prevention.


The secondary fracture prevention care gap


The range of therapies available to treat osteoporosis is broad and the variety of dosing regimens is quite unique in medicine. These include daily , weekly and monthly oral tablets and daily , quarterly , semi-annual and annual injections . Several of these therapies have been shown to reduce the incidence of fractures by 30–50%, even amongst individuals who have suffered multiple prior fractures . A growing body of evidence also suggests that osteoporosis treatment is associated with reduced mortality .


So, it is clear that patients who present to urgent care services with fragility fractures are at a considerably increased risk of suffering subsequent fractures in the future, and a broad range of effective treatments are available to reduce fracture risk in this population. These facts make it all the more surprising, and disappointing, that almost 20 years after bisphosphonates became available throughout the world, the vast majority of fragility fracture sufferers receive neither assessment nor treatment to reduce their future fracture risk .


Systematic literature reviews have sought to understand why secondary fracture prevention is not delivered in routine clinical practice. Elliot-Gibson and colleagues identified the following issues :




  • Cost concerns with respect to diagnosis and treatment



  • Time required for diagnosis and case-finding



  • Concerns with respect to polypharmacy



  • Lack of clarity regarding where clinical responsibility resides



The final point is consistent with experience reported throughout the world and plays a central role in the existence, and persistence, of the secondary fracture prevention care gap. Harrington from the USA characterised osteoporosis care of fragility fracture sufferers as a Bermuda Triangle comprised of orthopaedic surgeons, primary care physicians and osteoporosis specialists into which the fracture patient disappears . Chami and colleagues from the UK investigated the phenomenon in detail identifying a ‘Catch-22’ situation . Orthopaedic surgeons deliver the best acute fracture care but generally do not initiate investigation for the underlying cause of the fracture, relying on the primary care team to investigate for osteoporosis. Primary care physicians generally do not investigate patients who have recently suffered fragility fractures unless recommended to do so by a hospital specialist. In the absence of a pro-active case-finding system, hospital-based osteoporosis specialists (often endocrinologists or rheumatologists) are unlikely to interact with the fracture patient who is admitted to hospital, and very unlikely to interact with those managed exclusively in the outpatient setting. The consequence of patients ‘slipping through the cracks’ of guidelines in this way has been described as a fragility fracture cycle , which clearly needs to be broken if the global burden of the disease is to be reduced .




What is the best system to prevent a further fracture?


What an effective secondary fracture prevention system does


The key elements in effective provision of secondary preventive care have been described as the 3i’s: identification, investigation and initiation (of intervention): .




  • Identification: A robust system must be established to ensure that when patients present to urgent care services with a fragility fracture, it is recognised as such, documented and the patient is scheduled for investigation and, if appropriate, intervention.



  • Investigation: An assessment of future fracture risk is likely to include BMD testing for a significant proportion of fracture patients, particularly those under the age of 75 years. For patients who have suffered non-vertebral fractures, spinal imaging can identify previously undiagnosed/unrecognised vertebral fractures, which have implications on future fracture risk . Investigations for secondary causes of low bone mass should also be undertaken.



  • Initiation: Fracture patients with osteoporosis/low bone mass should be treated with osteoporosis medicines in accordance with national/regional/local management guidelines. The need for non-pharmacological interventions and identification of lifestyle risk factors should be considered. Individuals who may be at an increased risk of falling should be counselled and/or referred to falls prevention services for mitigation of falls risk factors.



An additional key step is delivery of care on a long-term basis. As with many other chronic diseases, adherence with osteoporosis treatment has been shown to diminish rapidly for around half of patients who start taking anti-osteoporosis medicines . In this regard, ascertainment that a previously undiagnosed individual has osteoporosis on account of presentation with a fragility fracture provides a potentially life-altering opportunity; agreement on and implementation of a long-term management plan with the patient could prevent them from suffering a hip fracture in the future. Tailoring education about osteoporosis and future fracture risk to the individual patient’s needs is likely to influence adherence with the management plan . In health-care systems that rely upon primary care physicians (PCPs) to manage long-term conditions, effective communication between the system that delivers the 3i’s of secondary preventive care and the PCP is essential.


Effective systems for new fracture patients: Fracture Liaison Services


Several systematic reviews have analysed studies from systems that have been established to deliver secondary fracture prevention . The most recent of these sought to classify systems based upon the intensity of the service provided: .




  • Type A: A 3i model that assumes direct responsibility for identification, investigation and initiation of osteoporosis treatment for fragility fracture patients.



  • Type B: A 2i model that identifies and investigates fracture patients, but leaves initiation of the osteoporosis treatment to the PCP.



  • Type C: A 1i model that identifies fracture patients and sends an alert to the PCP to suggest that further assessment is needed. These models do not schedule BMD testing.



  • Type D: This could be termed ‘Zero i’ model because the service provides educational material to the patient, but neither informs nor educates the PCP on the need for fracture risk assessment or osteoporosis treatment.



The proportion of patients who underwent BMD testing and received osteoporosis treatment for each of the classifications is shown in Table 1 . Clearly, Type A and Type B models deliver considerably higher treatment rates than the less intensive Type C and Type D models do. The proportion of fragility fracture patients who receive treatment will be dependent upon intervention thresholds in national/regional/local clinical practice guidelines. The Technology Appraisal Guidance on osteoporosis treatments for post-menopausal women published by the National Institute for Health and Clinical Excellence (NICE) in the UK in 2005 provides an illustration . NICE assumed that 50% of fragility fractures occur in women aged over 75 years and 25% in women aged 65–74 years and 50–64 years. Adherence to the guideline would allow for up to 100% of the over 75-year-olds to receive treatment, 60% of the 65–74-year-olds and 20% of the 50–64-year-olds. Based on these assumptions, up to 70% of post-menopausal women who suffer fragility fractures would be eligible for osteoporosis treatment. Given that it is unlikely that all women over 75 years of age with fractures would receive treatment, on account of comorbidities and clinical appropriateness for individuals with a very short expected survival, the optimal evidence-based osteoporosis treatment rate is likely to be in the range of 50–70%. However, there is currently no consensus on a precise figure or range.



Table 1

Effectiveness of systems for secondary fracture prevention of various intensity





































Model Type BMD Testing Treatment initiation
Control (%) Intervention (%) Control (%) Intervention (%)
Type A (3i) 23.8 79.4 a 17.9 46.4 a
Type B (2i) 9.2 59.5 a 19.9 40.6 b
Type C (1i) 13.5 43.4 a 7.5 23.4 a
Type D (Zero i) 11.4 8 c

a P < 0.001 versus control.


b P = 0.01 versus control.


c P = 0.06 versus control, non-significant.



Systems to deliver secondary fracture prevention have been referred to by several terms including Fracture Liaison Services (FLSs), Osteoporosis Coordinator Programs and Direct Care Manager Programs. In accordance with the International Osteoporosis Foundation (IOF) and American Society for Bone and Mineral Research (ASBMR) initiatives , the term FLS is used in this chapter. FLS would generally include Type A or Type B models according to the classification system developed by Ganda and colleagues . FLS has been established in a growing number of countries, including Australia , Canada , France , Ireland , Northern Ireland , Singapore , Spain , Switzerland , The Netherlands , the UK . and the USA . Several case studies of well-established FLS are discussed subsequently.


Australia: Concord Repatriation Hospital, Sydney


The Minimal Trauma Fracture Liaison (MTFL) service was established in 2005 and is a Type A model . The MTFL is a physician-led service that provides care for non-frail patients with fragility fractures. Frail patients, including the majority of hip fracture patients, receive secondary preventive care from the Orthogeriatrics Service at the same hospital. The MTFL is delivered by an advanced trainee (i.e., a physician in his/her fourth–sixth year of post-graduate training) who requires a 0.4–0.5 full-time equivalent (FTE) appointment.


The MTFL was evaluated in a prospective controlled study. The control group was comprised of patients who chose to decline the free consultation offered by the service in favour of arranging follow-up with their PCP. After 4 years, refracture incidence in the MTFL group was 80% lower than that of the control group. A subsequent formal cost-effectiveness analysis reported favourable results, including an incremental cost-effectiveness ratio (ICER) well below the Australian maximum accepted willingness to pay (AU$ 17,291 versus AU$ 50,000) and very modest costs for delivery of the MTFL at AU$ 146 per patient per year .


Canada: St. Michael’s Hospital, Toronto


The Osteoporosis Exemplary Care Program (OECP) was established in 2002 and is a Type A model . The Division of Orthopaedic Surgery and the Metabolic Bone Disease Clinic (MBDC) collaborated to design and implement the OECP, which was facilitated by an osteoporosis coordinator. The OECP provided care to patients managed in both the inpatient and outpatient settings. During the first year, appropriate care was provided to 96% of patients:




  • Thirty-six percent of outpatients were treated for osteoporosis prior to assessment by the OECP.



  • Fifty-six percent of outpatients were referred to the MBDC or their PCP for osteoporosis treatment.



  • Thirty-one percent of inpatients were treated for osteoporosis prior to assessment by the OECP.



  • Treatment was initiated for a further 24% of inpatients and another 34% were referred to the MBDC or their PCP for an osteoporosis consultation.



A subsequent formal cost-effectiveness analysis of the OECP concluded:




  • A 9% reduction in the incidence of secondary hip fractures could be achieved in the first year, by a hospital that manages 500 fragility fracture patients annually.



  • Savings of CN$ 48,950 would be achieved in the first year from the prevention of hip fractures alone, which would increase in subsequent years as rehabilitation and dependency costs rise.



  • Hiring an osteoporosis coordinator would be a cost-saving measure for a hospital that receives ≥350 fragility fracture patients per year.



UK: The Glasgow University Teaching Hospitals, Scotland


The Glasgow FLS was established in 1999 and is a Type A model for inpatients and Type B model for outpatients . The Glasgow FLS is a ‘doctor light’ model which is primarily delivered by clinical nurse specialists (CNSs) who work to locally agreed protocols for case-finding and management of fragility fracture patients. During the first 18 months of operations:




  • More than 4600 patients with fragility fractures were seen by the CNSs.



  • Almost 75% were considered for BMD testing and 20% were recommended treatment without BMD testing.



  • Eighty-two percent of patients who underwent BMD testing were found to have low bone mass (T-Score of ≤−1 standard deviation (SD) compared to the young adult normal).



During the previous decade, the Glasgow FLS evaluated an excess of 50,000 fragility fracture patients (Personal communication, A.R. McLellan and S.J. Gallacher). A national audit in Scotland compared the assessment rates for patients managed by the Glasgow FLS with several other service configurations in place in other Scottish hospitals . The Glasgow FLS dramatically outperformed other models with 97% of hip fractures and 95% of wrist fractures receiving assessment compared to less than 30% at other sites. A subsequent formal cost-effectiveness analysis of the Glasgow FLS concluded that 18 fractures were prevented, including 11 hip fractures, and GBP 21,000 saved for every 1000 patients managed by the service .


USA: Kaiser Permanente Health Maintenance Organisation


In 1998, several Kaiser Permanente medical centres in Southern California (SCAL) began establishing osteoporosis disease management programmes. By 2002, all Kaiser SCAL centres had implemented a fully integrated Healthy Bones Program which was subsequently extended across the Kaiser system throughout the USA . The Kaiser Healthy Bones Program is a Type A model delivered primarily by care managers and nurse practitioners and is underpinned by the state-of-the-art HealthConnect ® electronic medical record. In addition to providing comprehensive secondary preventive care for fracture patients managed within the Kaiser system, Healthy Bones has established a systematic approach to primary fracture prevention, with particular focus on individuals aged ≥75 years who are at a high risk of suffering their first fragility fracture.


The Healthy Bones Program has been subjected to actuarial analysis to model outcomes. In 2006, 2510 hip fractures were predicted for Kaiser SCAL but just 1575 were actually observed. This 37% reduction in the expected hip fracture rate was associated with an avoidance of hip fracture expenditure of US$ 31 million.


Effective systems for patients who have fractured in the past


The studies reviewed earlier illustrate that FLS can eliminate the secondary fracture prevention care gap for patients presenting to urgent care services with new fragility fractures in a highly cost-effective manner. However, what is to be done to reduce fracture risk for the substantial pool of patients who have suffered fragility fractures before an FLS was implemented in a locality, when ‘usual care’ would likely result in less than 20% receiving appropriate care? If fracture prevention efforts are to be optimised, then patients who have fractured in recent years present the next most obvious high-risk group for fracture risk assessment. Retrospective case-finding initiatives are required.


A good example of such an approach was developed by Brankin and colleagues in the UK . All community-dwelling women aged ≥65 years, who were registered with a PCP in Coatbridge, Scotland ( n = 4045), were mailed a questionnaire regarding osteoporosis, which focussed on fracture history, treatment and risk factors for fracture. Thirty-seven percent ( n = 852) of the 2286 women who returned a completed questionnaire reported a history of fracture after the age 50 years; 5% ( n = 43) had previously undergone BMD testing and 9.4% ( n = 80) were currently receiving treatment for osteoporosis.


In order to close the care gap identified, a system that could be described as a primary care-based FLS was established, which comprised of a nurse specialist working under the clinical direction of a PCP with a specialist interest in osteoporosis and ready access to BMD testing. Low bone mass (T-Score ≤ −1 SD) was reported for 86% of fracture patients and 52% had osteoporosis (T-Score ≤ −2.5 SD). Osteoporosis treatment was prescribed according to the local clinical practice guidelines, resulting in 64% of patients receiving treatment.


It is noteworthy that since April 2012, secondary fracture prevention was included as a domain of UK PCPs’ Quality and Outcomes Framework (QOF), a financial incentive scheme for all UK PCPs which rewards particular activities in specific therapeutic domains . The osteoporosis indicators are:




  • OST1: The practice can produce a register of patients:



    • 1.

      Aged 50–74 years with a record of a fragility fracture after 1 April 2012 and a diagnosis of osteoporosis confirmed on dual-energy X-ray absorptiometry (DXA) scan and


    • 2.

      Aged 75 years and over with a record of a fragility fracture after 1 April 2012.




  • OST2: The percentage of patients aged between 50 and 74 years, with a fragility fracture, in whom osteoporosis is confirmed on DXA scan, who are currently treated with an appropriate bone-sparing agent.



  • OST3: The percentage of patients aged 75 years and over with a fragility fracture, who are currently treated with an appropriate bone-sparing agent.



The impact of introduction of the osteoporosis indicators in the QOF needs to be evaluated over the coming years. However, the inclusion of fragility fracture registries in QOF sends a clear message to UK PCPs, and policymakers elsewhere, that secondary fracture prevention is considered to be clinically important by decision makers in the National Health Service (NHS). This creates a positive environment for implementation of long-term care plans for fragility fracture sufferers in the UK.


Systems for case-finding vertebral fractures


Ironically, the most common fragility fracture is the most frequently overlooked. Rates of recognition and diagnosis of patients with vertebral fractures are low throughout the world . Furthermore, vertebral fractures constitute a small proportion of fractures that will be identified by an FLS. Of the 4600 or more patients assessed by the Glasgow FLS during the first 18 months of operations, only 2% were patients with vertebral fractures. In response to this care gap, a number of strategies have been reported to improve case-finding:




  • Chest radiographs: Establishing protocols to consistently report vertebral fractures identified opportunistically upon chest radiographs is the most obvious way to improve vertebral fracture ascertainment .



  • Vertebral Fracture Assessment (VFA): VFA technology is available on most modern DXA scanners. Adding vertebral fracture assessment (VFA) to routine BMD testing has been shown to significantly increase identification of vertebral fractures and enhance subsequent fracture risk prediction .



  • Reformatting Computed Tomography scans: Reformatting of computed tomography (CT) examinations of the chest or abdomen in a hospital in New Zealand identified a prevalence of vertebral fractures of 13%, the majority of which were previously unknown .



Putting in place systems with local radiology/imaging departments that ensure reliable reporting of vertebral fractures has the potential to dramatically improve case-finding of vertebral fractures for a negligible investment of new resources.




What is the best system to prevent a further fracture?


What an effective secondary fracture prevention system does


The key elements in effective provision of secondary preventive care have been described as the 3i’s: identification, investigation and initiation (of intervention): .




  • Identification: A robust system must be established to ensure that when patients present to urgent care services with a fragility fracture, it is recognised as such, documented and the patient is scheduled for investigation and, if appropriate, intervention.



  • Investigation: An assessment of future fracture risk is likely to include BMD testing for a significant proportion of fracture patients, particularly those under the age of 75 years. For patients who have suffered non-vertebral fractures, spinal imaging can identify previously undiagnosed/unrecognised vertebral fractures, which have implications on future fracture risk . Investigations for secondary causes of low bone mass should also be undertaken.



  • Initiation: Fracture patients with osteoporosis/low bone mass should be treated with osteoporosis medicines in accordance with national/regional/local management guidelines. The need for non-pharmacological interventions and identification of lifestyle risk factors should be considered. Individuals who may be at an increased risk of falling should be counselled and/or referred to falls prevention services for mitigation of falls risk factors.



An additional key step is delivery of care on a long-term basis. As with many other chronic diseases, adherence with osteoporosis treatment has been shown to diminish rapidly for around half of patients who start taking anti-osteoporosis medicines . In this regard, ascertainment that a previously undiagnosed individual has osteoporosis on account of presentation with a fragility fracture provides a potentially life-altering opportunity; agreement on and implementation of a long-term management plan with the patient could prevent them from suffering a hip fracture in the future. Tailoring education about osteoporosis and future fracture risk to the individual patient’s needs is likely to influence adherence with the management plan . In health-care systems that rely upon primary care physicians (PCPs) to manage long-term conditions, effective communication between the system that delivers the 3i’s of secondary preventive care and the PCP is essential.


Effective systems for new fracture patients: Fracture Liaison Services


Several systematic reviews have analysed studies from systems that have been established to deliver secondary fracture prevention . The most recent of these sought to classify systems based upon the intensity of the service provided: .




  • Type A: A 3i model that assumes direct responsibility for identification, investigation and initiation of osteoporosis treatment for fragility fracture patients.



  • Type B: A 2i model that identifies and investigates fracture patients, but leaves initiation of the osteoporosis treatment to the PCP.



  • Type C: A 1i model that identifies fracture patients and sends an alert to the PCP to suggest that further assessment is needed. These models do not schedule BMD testing.



  • Type D: This could be termed ‘Zero i’ model because the service provides educational material to the patient, but neither informs nor educates the PCP on the need for fracture risk assessment or osteoporosis treatment.



The proportion of patients who underwent BMD testing and received osteoporosis treatment for each of the classifications is shown in Table 1 . Clearly, Type A and Type B models deliver considerably higher treatment rates than the less intensive Type C and Type D models do. The proportion of fragility fracture patients who receive treatment will be dependent upon intervention thresholds in national/regional/local clinical practice guidelines. The Technology Appraisal Guidance on osteoporosis treatments for post-menopausal women published by the National Institute for Health and Clinical Excellence (NICE) in the UK in 2005 provides an illustration . NICE assumed that 50% of fragility fractures occur in women aged over 75 years and 25% in women aged 65–74 years and 50–64 years. Adherence to the guideline would allow for up to 100% of the over 75-year-olds to receive treatment, 60% of the 65–74-year-olds and 20% of the 50–64-year-olds. Based on these assumptions, up to 70% of post-menopausal women who suffer fragility fractures would be eligible for osteoporosis treatment. Given that it is unlikely that all women over 75 years of age with fractures would receive treatment, on account of comorbidities and clinical appropriateness for individuals with a very short expected survival, the optimal evidence-based osteoporosis treatment rate is likely to be in the range of 50–70%. However, there is currently no consensus on a precise figure or range.


Nov 10, 2017 | Posted by in RHEUMATOLOGY | Comments Off on Secondary prevention and estimation of fracture risk

Full access? Get Clinical Tree

Get Clinical Tree app for offline access