Quality improvement programme

Quality improvement initiative: implementing routine vertebral fracture assessments into an Australian Fracture Liaison Service

Abstract

Osteoporosis is a global health concern and a major contributor to worldwide mortality rates. Vertebral fractures due to osteoporosis are common and often undetected. Since vertebral fractures are often missed, evidence and guidelines suggest that vertebral fracture assessment (VFA) may optimise current approaches to bone density tests. We aimed to integrate VFA into an Australian Fracture Liaison Service (FLS) and measure the impact it had on osteoporosis treatment initiation.

A retrospective case note review was undertaken to determine the number of clinic patients receiving VFA before the change in practice. Proctor’s seven domains of implementation strategy were used to facilitate quality improvement outcomes.

The percentage of eligible patients receiving a routine VFA at the FLS imaging centre increased from 0% to 90%. The remaining 10% of patients did not receive a scan due to the patient not being able to assume the correct position, skilled staff being unavailable to perform the scan, or the patient declining. Post implementation, almost half (41%) of patients who underwent a VFA displayed abnormalities and 16 (4%) of these recorded a normal bone measure density score but abnormal VFA. Despite the successful adoption of the new screening protocol, there was no statistically significant increase in treatment initiation rates for patients with normal bone mass density scores.

The FLS successfully integrated routine VFA into the osteoporosis care pathway. However, the introduction of VFA did not significantly increase treatment initiation. It may be more effective to offer VFAs to a proportion of patients based on a tailored approach rather than offering them routinely to all patients who access the FLS.

Problem

Osteoporosis is a serious global health concern1 and is a major contributor to worldwide mortality rates.2 It is estimated that cases of osteoporosis will continue to rise as the population grows and increases in age.3 Traditionally, osteoporosis diagnosis has been based on a minimal trauma fracture and/or a low bone mineral density (BMD) score, assessed via a dual-energy X-ray absorptiometry (DXA), in a person older than 50 years of age.4 Vertebral fractures are common5 6 and are high-risk factors for future vertebral and non-vertebral fractures.7 8 Often these fractures are asymptomatic, and with conventional DXA scans not detecting vertebral fractures, estimates suggest up to 70% of cases may go undetected.9 These figures are alarming as vertebral fractures increase the risk of future fractures more than fourfold,9 lead to other serious health conditions,10 and significantly raise mortality rates.11 12 Since vertebral fractures are often missed, evidence13 14 and guidelines15 suggest that vertebral fracture assessment (VFA) may optimise current approaches to bone density assessments. Developments in technology now allow for VFA during routine DXA scans with low additional radiation exposure to the patient and minimal cost to the healthcare system.16 Evidence suggests that identifying vertebral fractures using VFA can have a significant impact on secondary fracture probability, signifying that VFA may be an effective approach to improving fracture risk calculations.17 Furthermore, treatment initiation and medication continuation may be aided by the additional information provided by a VFA.8

The primary aim of this study was to evaluate the success of an Australian Fracture Liaison Service (FLS) in integrating VFA into their osteoporosis care pathway. The FLS where the evaluation took place is based at Flinders Medical Centre within the Southern Adelaide Local Health Network (SALHN). The service offers assessment for people aged 50 years or over who have attended the regional hospital and sustained a minimal trauma fracture. On referral, patients receive fracture risk assessment, fall risk assessment, dietetic-focused assessment, self-management education, investigations (pathology, bone density testing), and medication review and/or the initiation of osteoporosis treatments (predominately antiresorptive therapy). There is a paucity of data reporting the number of patients receiving routine VFA in FLS’, making it challenging to benchmark against other services. Schousboe and colleagues explored the relationship between a positive VFA and consequent treatment initiation within clinical practice.18 Using a prospective observational cohort design the researchers examined the Canadian Manitoba Bone Density Programme database between 24 February 2010 (when VFA testing was introduced) and 30 March 2016. In total, 47 944 patients underwent a DXA scan, of which 10 053 (21%) also received a VFA at the same time. A Dutch study evaluated the implementation of VFA in an FLS, reporting that VFA increased from 4.6% to 97.1%. Accounting for differences in international osteoporosis assessment frameworks and healthcare systems (e.g., differences in healthcare insurance), we aimed to increase VFAs from 0% to 50%. We were also interested in whether the VFA positively impacted early identification and treatment initiation for secondary osteoporotic fractures.

Background

Osteoporosis is a skeletal disorder characterised by diminished BMD and bone microarchitecture, leading to an increased risk of fractures.19 Globally, it is estimated that up to 18.3% of the population is living with osteoporosis,1 with this figure likely to rise with increasing life expectancy.3 20 The cost of the disease has significant social and economic burdens,21 22 yet, despite this, there are widening treatment gaps.23 24 The proportion of individuals with osteopenia or osteoporosis who require intervention but do not receive one is increasing for several reasons, including concerns about medication side effects, lack of awareness, and differences in national and international healthcare policies.25 26

The most common osteoporotic fractures are vertebral fractures5 6 yet there is a worldwide underdiagnosis.27 A large European study reported that up to 12% of females aged between 50 and 79 will experience a vertebral fracture, the majority being osteoporotic in nature.6 This figure is estimated to rise to 20% in females over 80 years of age.6 Currently, osteoporosis is diagnosed through a DXA scan that typically measures the density of the bones at the hip and lower spine, meaning that vertebral fractures are often missed. This oversight is further exacerbated by the asymptomatic nature of vertebral fractures, with prediction rates suggesting that between 50% and 70% of cases will not receive clinical attention.9 These statistics are concerning because spinal fractures can elevate the risk of subsequent fractures by 9over fourfold, and can increase other morbidities and even mortality.25 26 The identification of vertebral fractures offers healthcare systems a prime opportunity to reduce the frequency of secondary fractures. To achieve this, an additional scan, known as a VFA, can be undertaken, allowing clinicians to acquire a lateral view of the thoracic and lumbar spine to determine if there are any fractures.

In Australia, VFAs are not routinely administered and, as such, many vertebral fractures are undetected, leaving individuals untreated and at high risk of subsequent fractures.4 An illustration of this comes from a large Australian community-based study (n=482) of postmenopausal women aged 70 and over who reported no previous knowledge of vertebral fractures.28 VFA identified that 18.3% of the sample had experienced a vertebral fracture. Solely using BMD scores to diagnose osteoporosis resulted in a diagnosis range between 40% and 74%, suggesting that between 26% and 60% of osteoporotic women may not have received appropriate treatment. Knowledge of vertebral fractures, as well as BMD scores, significantly improves secondary fracture risk prediction15 and the evidence base is growing supporting the effectiveness of VFA as an additional clinical marker for osteoporosis.28–30 Routine VFAs for all patients who access an FLS may modify the patient’s risk category and thus impact on the initiation and type of intervention offered.8 Furthermore, collecting a baseline for vertebral fractures allows clinicians to discriminate between old and new fractures and to monitor treatment plans.8

This study aimed to evaluate the success of an Australian FLS in introducing routine VFAs into their osteoporosis care pathway. We also wanted to assess whether VFA positively impacted early identification and treatment initiation for the detection of future osteoporotic fractures for patients with normal BMD. Research evaluating the use and effectiveness of VFA within public FLS is imperative for quantifying the diagnostic gap,31 influencing clinical guidelines, and informing funding decisions around osteoporosis care pathways.

Measurement

This study used routinely collected clinical data. A retrospective electronic medical record review was undertaken, and baseline indicators were collected from clinical data during the pre-implementation (i.e., before VFA was introduced) period from 1 September 2017 to 29 February 2020 and compared with post implementation data from 1 March 2020 to 30 October 2022. We aimed to measure how many patients received a VFA before and after the service had implemented the change in practice and whether this had an impact on treatment initiation. See online supplemental figure 1 for an illustration of the routine FLS care pathway before and after VFA was introduced.

Design

Before the implementation project proceeded, VFA in the FLS was only performed if a clinician specifically requested the scan. After implementation, VFAs were routinely offered to patients as part of their standardised osteoporosis assessment. The pre-implementation period occurred over 30 months and the study used a retrospective cohort design. Case notes for all patients assessed during this time were audited by a primary reviewer (DS—FLS nurse) and checked by two secondary reviewers (BB—FLS non-physician clinician and LG—researcher). The primary outcome was the number of eligible patients who received a VFA after the change in practice was introduced. In terms of eligibility for the VFA, it should be noted that the publicly funded healthcare insurance scheme in Australia, Medicare, only allows an individual to be rebated for one bone density test every 24 months. Therefore, if a patient has received imaging within another service, for example, prior to having surgery after a hip fracture, then they will not be eligible to receive a second scan with the FLS. The FLS will request scan results from the other service. We also wanted to explore the number of cases where the patient had normal BMD, demonstrated through a DXA scan, but the VFA identified vertebral fractures and treatment initiation was recommended.

An FLS quality improvement group was formed to implement the VFA into the osteoporosis care pathway. There was no hierarchical structure within the group, and we found that an inclusive team approach was effective for decision-making and promoted the onboarding of FLS staff. All FLS staff were consulted during the quality improvement process and were amenable to the changes. This positive engagement was primarily centred around the staffs’ understanding that the service was aiming for a ‘gold standard’ of care, and this was the driving incentive for the team. To be successful, we had to gain the approval of the bone densitometry team who administer the FLS DXA scans. The bone densitometry team do not work exclusively with the FLS and accommodate other healthcare services. The motivators for the bone densitometry team implementing the change were less explicit. To address this, we emphasised that the change was linked to a quality improvement which would offer patients a gold service of care. There was hesitation from the bone densitometry staff regarding the additional time needed to undertake the scan and interpret the results. A primary concern was the feasibility of administering a VFA with the FLS’ patient population who are usually older, can be frail, and often present with painful fractures. Within our service, there is typically one unaided bone densitometry staff member on duty so there was apprehension about the time needed to position patients and manual handling for individuals who may need assistance. To mitigate this, the FLS staff, who are located next to the imaging facility, provided reassurance that they would be available to help with manual handling.

Strategy

We used the Expert Recommendations for Implementing Change32 discrete implementation strategies and Proctor’s framework33 to identify and measure outcomes related to the implementation of the VFA into the osteoporosis care pathway (table 1). The implementation required an integrated approach across the entire FLS including the diagnostic bone densitometry team and consulting endocrinologists. We also used the Standards for Quality Improvement Reporting Excellence 2.0.34

Table 1
|
Implementation strategies used in the current evaluation per the (ERIC)32 discrete implementation strategies and Proctor’s framework33

Results

Table 2 outlines the pre-implementation and post implementation descriptive statistics. Before the implementation, 0% of patients were referred for a VFA. Post implementation, out of the total number of referrals to the FLS, 80% of patients were eligible to receive a VFA and 90% of these underwent a VFA at the FLS imaging centre. There were 10% of patients who underwent their DXA scan at the FLS imaging centre but did not receive a VFA. There were several reasons for this including the patient being unable to assume the position needed for the VFA, the bone densitometry staff being unable to perform the scan (i.e., lack of training or confidence), or the patient declining. It should be noted that out of 34 participants who underwent BMD assessments at another centre, only two of these received a VFA.

Table 2
|
Data for number of referrals, tests and treatment initiation

Post implementation, almost half (41%) of patients who underwent a VFA displayed abnormalities and 4% of these recorded a normal BMD score but abnormal VFA. Four per cent of patients received a normal BMD score but abnormal VFA. Inferential statistics found there was no significant difference in treatment initiation between the pre-VFA and post-VFA implementation periods χ2(1, 134)=3.41, p= .065.

Lessons and limitations

We proactively initiated a change in practice to try and optimise our approach to bone density assessments as part of our osteoporosis care pathway. Our FLS introduced routine VFAs for all eligible patients accessing the osteoporosis care pathway. The implementation has been successful with 90% of eligible patients receiving the additional VFA during their routine osteoporosis assessment. When compared with similar VFA implementation studies, for example, van der Velde and colleagues35 who reported 97.1% of patients receiving VFA, our imaging rates are marginally lower, possibly relating to variations in patient populations or differences in staff capacity and training.

The proximity of the FLS offices to the imaging centre has been pivotal with the high number of VFAs performed. Informal conversations between the bone densitometry team and FLS have been invaluable and have had an implicit impact on the implementation, for example, informal debriefs after VFAs. Building rapport throughout the process has been important and we feel the face-to-face element has assisted compared with if we had taken a virtual approach. The bone densitometry staff can also access rapid support from FLS staff if they require assistance positioning a patient. Positioning frail, older patients for the VFA can be challanging and having two staff members on hand is often necessary. If this support was not available, the number of VFA performed would have likely been lower.

The effectiveness of the VFA was mixed. On one hand, the assessment highlighted the high percentage of patients who displayed vertebral abnormalities (41%). Interestingly, 4% of patients received a normal BMD score but abnormal VFA, suggesting that some patients may not have been offered osteoporosis treatment if clinicians based their initiation decisions solely on BMD scores. However, it should be noted that our service offers a tailored model of care and clinicians in the FLS base their treatment initiation decisions on multiple variables and would not use BMD scores in isolation.

Overall, there was no significant difference in treatment initiation between the pre-implementation and post implementation periods. It may be the case that treatment initiation did not increase post implementation because of our patient population. As our FLS is a multidisciplinary programme designed to prevent secondary fractures in individuals who have already sustained a fracture, it is likely that treatment initiation would be offered based on this information alone. As well as fracture history, clinicians also use risk assessments which consider the demographics, medical history, and DXA scores of patients. Therefore, it may be the case that VFA provides little additional information that would impact treatment management for the population we serve. However, VFA may be beneficial when an individual without a known low trauma fracture has a DXA scan that indicates osteopenia and may not otherwise be started on treatment.10 13 14 Rather than implementing a routine VFA protocol for all FLS patients, it may be more appropriate to offer the additional scan to a proportion of patients based on a tailored approach that takes into consideration local guidance, available evidence, and the individual patient’s needs.13 14 An innovative improvement would be for VFA to be offered to all patients who are referred for a DXA who do not have a fracture history.13 This remodelling may allow us to capture individuals who have not knowingly sustained a minimal trauma fracture yet but who have unidentified vertebral breaks and are at high risk of secondary fractures.

One of the limitations of this study was that we relied on case note data and did not collect data prospectively. For future evaluations, we will incorporate a short reflective process to capture this information in real-time. This lesson highlights the complementary nature of quantitative and qualitative evaluations and the necessity of both to be able to see the ‘full picture’. On reflection, we acknowledge that the absence of a significant effect of VFA on treatment initiation may have been linked to inadequate investigation during the planning phase of the project. It is possible that we made an error by assuming that implementing guidelines would automatically enhance diagnosis or treatment initiation rates of the service we provide. As a competent and versatile multidisciplinary team offering a personalised approach to osteoporosis assessment, we may not have encountered any significant issues in the area that the guidelines are intended to address. It could have been advantageous for us to carry out a thorough baseline audit to gain a complete understanding of the number of patients who could potentially derive additional benefits from the implemented assessment. As a clinical team dedicated to enhancing our research skills, we now recognise the importance of choosing an appropriate conceptual framework during quality improvement projects. Looking back, a framework such as a Plan-Do-Study-Act cycle, which offers a more comprehensive approach to designing and evaluating quality improvements in health interventions, may have been more suitable. To optimise our future efforts, we aim to collaborate with an implementation researcher for any upcoming projects. This individual will provide us with the necessary guidance and support, including assisting us during the planning stage, to develop interventions that are both effective and sustainable in the long term. Future research could formally investigate whether the implementation of VFA had an impact on other aspects of clinical practice or patient experience, such as the nuanced medication choices made by clinicians, or the treatment preferences expressed by patients.

Conclusion

Vertebral fractures are frequently undiagnosed as they can be asymptomatic, and routine osteoporosis assessment does not include spinal imaging.9 Current evidence13 14 and guidelines15 suggest that VFA may optimise current approaches to bone density assessments. This project aimed to integrate routine VFA into an Australian FLS and to measure their effect on treatment initiation. There is a dearth of research evaluating the success of implementing routine VFA into an osteoporosis assessment pathway, so this article aims to fill this knowledge gap. We achieved our objective as the percentage of eligible patients receiving a routine VFA increased from 0% to 90%. The introduction of VFA did not have a significant quantitative difference to treatment initiation suggesting that it may be more effective to offer the additional scan to a proportion of patients based on a tailored approach.