Fracture risk assessment in long-term care:a survey of long-term care physicians

Osteoporosis is a common disease characterized by low bone mass and increased risk for fractures [1]. Osteoporosis-related fractures are strongly associated with recurrent fractures, and are responsible for lasting disability, morbidity and excess mortality [2‐6]. In Canada, the incidence of osteoporosis-related fractures is 7.2 and 15. 3/1000 person-years for men and for women over the age of 50 years, and the cost to the healthcare system has been estimated to be up to $3.9 billion annually [7].

In recent years, the focus of osteoporosis management has shifted from the treatment of low bone mineral density (BMD) to an integrated approach of fracture risk reduction through recognition of important clinical risk factors. In 2010, Osteoporosis Canada (OC) published evidence-based clinical practice guidelines for the diagnosis and management of osteoporosis that reflect the new approach to risk assessment. The guidelines strongly recommend the use of the validated fracture risk assessment tools FRAX or CAROC (Canadian Association of Radiologists and Osteoporosis Canada). These tools evaluate the risk of osteoporosis-related fractures based on individual risk factors and are appropriate for use in clinical practice [8]. Both tools are country- and sex- specific, and based on sets of risk factors that include age, BMD of the hip (FRAX, but not CAROC, can also provide a score in the absence of BMD measurement), prevalent fragility fractures, use of glucocorticoids and others, and predict the 10-year probability of major osteoporotic fractures (hip, spine, distal forearm and humerus). Results obtained from the FRAX or CAROC predictor tools are used to guide treatment so that pharmacotherapy is recommended for patients with an elevated 10-year probability of major fractures. Many countries have also recently reviewed and updated their clinical practices guidelines using a similar framework [9]. Knowledge may not benefit patients unless it is disseminated and translated to healthcare providers, patients and policy makers in a form they can use. Attention must be devoted to adapt this knowledge to the context and its stakeholders [10].

Elderly residents of long-term care (LTC) facilities have a higher prevalence of osteoporosis and higher rates of fracture morbidity and mortality following fractures than their age- and sex-matched community-dwelling counterparts [11]. Eighty-five percent of LTC residents are reported to have osteoporosis and 40% of all hip fractures occur in this population [12, 13]. Nevertheless, the majority of these frail individuals are not treated for osteoporosis [14]. To better understand this care gap, and to inform the development of practice guidelines specific to the LTC setting, we conducted a survey to collect information and opinions on fracture risk assessment from frontline physicians in the LTC environment. Our objectives were to explore which elements of the OC guidelines’ recommendations are perceived to be applicable in the LTC setting, to inquire about current practices for fracture risk assessment by LTC physicians and to describe barriers to applying the OC guidelines for fracture assessment and prevention in LTC.

Three hundred and forty seven questionnaires were distributed and 87 (25%) participants submitted completed questionnaires.

Table 1 shows participants’ demographics; mean (SD) time in practice was 32 (11) years and most had over 50 LTC residents under their care. Eighty-seven percent of physicians considered prevention of osteoporosis-related fractures as important, and over 90% felt confident in their ability to assess fracture risk in their LTC residents.

In this survey, physicians with a clinical practice that includes LTC residents report that the prevention of osteoporosis-related fractures is an important aspect of their practice; however the majority do not use OC recommendations for evaluation of fracture risk or osteoporosis management. LTC physicians report that they are confident in their ability to assess fracture risk, but almost half of them are not using the fracture risk assessment protocols recommended in the guidelines. Many respondents are aware of the OC guidelines’ recommendations and properly recognize patients at high risk of fractures, as demonstrated by their answers but they acknowledge a number of barriers to applying the recommendations in the LTC setting. There seem to be more uncertainty in identifying those who are at moderate risk who may not require pharmacotherapy. This inability to discern which resident actually might benefit from therapy may lead to under-treatment of this population. Furthermore, because of perceived barriers associated with using FRAX or CAROC, many LTC physicians may be adapting their own strategies for fracture risk assessment in LTC leading to suboptimal bone health management. Therefore, the responses collected in this survey support the need to establish the effectiveness of OC current guideline recommendations in the LTC population (for example the fracture prediction tools that require BMD measurements), to take into consideration the particular LTC setting and to enlist effective partnerships with clinicians, leaders and patient groups for future development and implementation of relevant best practice guidelines for this population.

Other work has also shown that a minority of physicians use recommended validated fracture prediction tools to assess fracture risk in their LTC residents. Difficulty accessing BMD and other evaluation modalities, unknown cost-benefit effectiveness of interventions in this population and lack of resources were the most often cited barriers to optimal fracture risk assessment. Our results mirror those obtained by McKercher et al. a decade ago in a survey of the members of the medical directors and advisors to LTC facilities in Ontario following the publication of the initial OC clinical guidelines in 1996 [15, 20]. In their survey, 46% percent of respondents did not routinely assess fracture risk, 52% based their assessment on clinical factors only and 23% on BMD or spine radiographs. Perceived barriers to initiating treatment for osteoporosis included lack of access to BMD, unproven effectiveness of interventions in the LTC population, possible side effects of pharmacological treatments, time and cost of diagnosis and treatment, and patient reluctance. This situation is similar in many countries around the world [21, 22]. In the United States, Colon-Emeric et al. surveyed LTC home administrators, physicians and nurses on the use of clinical practice guidelines in LTC [23]. The most frequently cited barriers to their implementation were provider concerns that guidelines were “checklists” to replace clinical judgment, limited facility resources, conflict with family representatives and facility policies that conflict with guidelines’ recommendations. A more recent survey also demonstrates the need for education and adaptation of osteoporosis guidelines of front-line staff in LTC in management of osteoporosis [24]. These results underscore the fact that guidelines targeted at community-dwelling men and women cannot be readily applied to those living in residential care, even if these guidelines are updated and broadly disseminated, as OC has done in 2010 [8]. Moving from evidence to practice in the clinical world requires integrated knowledge translation that include taking into consideration more than the knowledge to be transmitted but also the context or setting where the this will take place, the target audience and the facilitators (human resources and others) that will ensure changes in clinical practice can take place [10, 25].

Nevertheless, there have been efforts to provide guidance for fracture prevention in residential care. A scoping review of strategies for the prevention of hip fractures in elderly nursing home residents documented that vitamin D supplementation and, in some cases, alendronate and hip protectors were associated with reduced fracture risk [26]. Consensus recommendations for fracture prevention in LTC have been published [27, 28]; however, most recommendations are based on data obtained in clinical trials that excluded LTC residents, which may reduce physicians’ confidence in applying them [29]. Recently Rondondi et al. demonstrated that in an older population of LTC residents the 10-year fracture probability appeared to be mainly determined by age and clinical risk factors obtained by medical history, rather than by BMD or the presence of vertebral fractures on radiography [30], thereby supporting the concept that prediction rules in LTC may not necessitate evaluation of BMD or imaging. Additional barriers we have identified such as lack of resources to administer fracture assessment tools, resistance from family members, costs, polypharmacy, may help explain why LTC physicians do not initiate therapies in patients at high risk for fractures including those who have recently sustained a hip fracture [14, 31, 32]. There have been successful multifaceted interventions in LTC that have demonstrated reduced fracture rates associated with vitamin D supplementation and anti-osteoporosis therapy use in LTC [33, 34]. Therefore, if barriers associated with applicability of treatment could be addressed or knowledge regarding interventions effective in LTC could be effectively translated into practice, the prevention of fractures in LTC could be optimized.

Our study is limited mostly by a modest response rate and the geographic limitation of the survey distribution (Ontario, Canada). Nevertheless, we feel that our questionnaire resulted in data consistent with those of previous surveys conducted in study populations in other jurisdictions, suggesting that our findings are translatable. It is possible that those that responded are physicians with an interest in osteoporosis, or who are more confident in their knowledge and practice around osteoporosis management. We are not able to confirm whether physician perceptions are consistent with practice. Like others, we have highlighted the urgent need to address barriers to fracture risk assessment in LTC with the development and dissemination of setting-specific best-evidence guidance with particular attention to the culture and nature of the environment and the identification of local champions.

Osteoporosis-related fractures cause significant morbidity and loss of autonomy in the general population and even more strongly affect the frail elderly who live in LTC. To ensure optimal management of osteoporosis and prevent fractures in this population, evidence-based guidance applicable in the LTC setting must be developed and disseminated to physicians and front-line staff. We have shown that perceived barriers to fracture risk assessment and osteoporosis management in LTC have not changed over the last 10 years, contributing in part to the ongoing care gap in osteoporosis management. The results of this survey and of the published literature can greatly assist efforts by expert and community stakeholders to adapt consensus guidelines and bone health research priorities for use in the LTC setting in Canada. Such efforts are currently in progress.