Room for improvement in osteoporosis diagnosis
Monash Centre for Health Research and Implementation clinician researchers, Endocrinologist A/Prof Amanda Vincent and Endocrinologist and PhD student Dr Alicia Jones, have found significant room for improvement in the diagnosis of osteoporosis in a study recently published in The Journal of Clinical Densiometry.
Osteoporosis affects one in three 3 women and one in five men aged over 50 years, and is responsible for over nine million bone fractures per year in women globally. Bone loss may be slowed or prevented by medication, so many of these fractures are avoidable. They cause untold physical pain, prevent people engaging in their normal daily activities, can lead to serious long-term physical and mental health complications, and cost health systems millions of dollars. With an ageing global population, the impact is only set to worsen.
Diagnostic technology is available in Australia and other developed nations in the form of dual X-ray absorptiometry (DXA) scans, which measure bone mineral density. They are used for diagnosis, risk stratification and monitoring treatment response. DXA scans are covered by guidelines issued by the International Society for Clinical Densitometry (ISCD), which include criteria for clinical reports. However, the only two prior studies assessing DXA scan reports and adherence to ISCD guidelines – one in Canada, the other in the United States - demonstrated errors in as many as 80 per cent of scan reports.
The utility of DXA relies on the accuracy of the scan and its interpretation, as well as the adequacy of clinical reports. Errors in any aspect can lead to incorrect diagnosis, unnecessary investigations and inappropriate or missed treatment, causing patient distress and wasting critical healthcare resources.
A/Prof Vincent and Dr Jones aimed to determine whether DXA reports for patients receiving care at a large academic teaching hospital in Melbourne are adhering to required ISCD reporting standards, and determine whether this differs depending on patient factors and the imaging service used. In the process their study became the largest study in the world to review this.
They manually compared 459 DXA reports with ISCD guidelines. The reports were from patients aged 18 years and over, receiving care at specialist outpatient clinics between January 1, 2018 and December 31, 2019. Two hundred and fourteen of the scans were conducted internally within the hospital, while 245 were performed at 23 different external imaging services.
Each adherence to an ISCD criterion was scored as one point, giving a total score of 14 for baseline DXA scans and 18 for repeat DXA scans. The score was then converted to a percentage.
The overall median report score was just 57.1%. The two ISCD criteria with the lowest adherence scores were around recommendations and timing of future DXA scans - included in just 1.1% of reports - and recommendations for investigation for secondary causes of osteoporosis - included in 1.2% of reports. An earlier Canadian survey of primary care physicians’ preferences for DXA reports, found that the majority consider these specific components of the DXA report to be “extremely helpful” or “very helpful”.
The researchers found significantly higher compliance with reports performed within their own hospital service, compared to external services, after adjusting for age, sex, indication, and type of scan. Baseline DXA reports had slightly higher scores than repeat DXA scans, and, among external imaging services, rural services had higher scores than metropolitan services.
So where does this leave us? A/Prof Vincent says, “Avoidable fractures arising from osteoporosis are a massive hidden burden on Australian women. The fact that we have the technology for accurate diagnosis, but are generally failing to apply best practice when using it, is potentially a huge missed opportunity. We need to understand why this is happening to drive improvements, which may be a fairly simple matter of audit and feedback.
“We were surprised by the differences between rural and metropolitan services, but this warrants further investigation – the number of scans done by rural providers in our study was relatively low, and more robust data should ideally be secured.”