Management of Osteoporosis in The Community: How Well Can Family Physicians Do? A Retrospective Case Series in a Local Public Clinic in Hong Kong.

With the ageing population, osteoporosis is increasingly becoming a global health concern. Previous research showed that management of osteoporosis in Hong Kong had been suboptimal. In view of the need to provide quality care to osteoporotic patients, a Multidisciplinary Osteoporosis Clinic (MOC) organized by primary care physicians was set up in a public primary care clinic. This study aimed to explore the clinical effectiveness of MOC. We performed a retrospective case series study. All osteoporotic patients who had attended MOC during the period 1 January 2015 to 31 December 2018 were included. Changes in Dual X-ray absorptiometry (DEXA) T-score of recruited patients after two years of management at MOC were analysed. Subgroup analyses of the 2-year interval DEXA T-score changes among patient with or without history of fragility fracture, and among patients with or without pharmacological treatment , were performed. Serial interval DEXA T-score changes of recruited patients who had completed 5 years of bisphosphonate treatment were also analysed. Paired Student’s t test was used to analyze the interval DEXA T-score changes.


Abstract Background
With the ageing population, osteoporosis is increasingly becoming a global health concern. Previous research showed that management of osteoporosis in Hong Kong had been suboptimal. In view of the need to provide quality care to osteoporotic patients, a Multidisciplinary Osteoporosis Clinic (MOC) organized by primary care physicians was set up in a public primary care clinic. This study aimed to explore the clinical effectiveness of MOC.

Methods
We performed a retrospective case series study. All osteoporotic patients who had attended MOC during the period 1 January 2015 to 31 December 2018 were included. Changes in Dual X-ray absorptiometry (DEXA) T-score of recruited patients after two years of management at MOC were analysed. Subgroup analyses of the 2-year interval DEXA T-score changes among patient with or without history of fragility fracture, and among patients with or without pharmacological treatment , were performed. Serial interval DEXA T-score changes of recruited patients who had completed 5 years of bisphosphonate treatment were also analysed. Paired Student's t test was used to analyze the interval DEXA T-score changes.

Results
Totally 186 osteoporotic patients were recruited. After two years of management at MOC, the T-score at lumbar spine and femoral neck were improved, from -2.71±0.76 to -2.35±0.83 and -2.40±0.75 to -2.10±0.76 respectively, P<0.001. For subgroup analysis on patients with or without history of fragility fracture and patients with or without bisphosphonate treatment, improvement in T-scores of both lumbar and femoral neck were all signi cant at two years, P<0.05.
Among those who had completed 5 years of bisphosphonate treatment, progressive improvement in Tscores of both lumbar and femoral neck were observed, with P <0.001.

Conclusion
MOC run by family physicians in public primary care setting could effectively treat osteoporotic patients with signi cant improvement in DEXA T-scores.

Background
With the ever-ageing population and longer life expectancy, osteoporosis is increasingly becoming a global health concern [1]. Currently it is estimated that more than 200 million people suffer from the disease worldwide [2]. Approximately 30% of all postmenopausal women have osteoporosis in the United State and in Europe. [3] According to a meta-analysis published in 2016, there has been an obvious increase in the prevalence of osteoporosis in China over the past 12 years (prevalence of 14.94% before Page 3/15 2008 and 27.9% during the period spanning 2012-2015). The pooled prevalence of osteoporosis in people aged 50 years and older was 34.7%, meaning around one third of people over 50 year-old are suffering from the disease [4]. In another local study, the prevalence of osteoporosis in Hong Kong Chinese female increases with age dramatically. In women aged 70 or above, over half have osteoporosis at the hip. [5] Despite the enormous social and economic impact of osteoporosis and its related complication, namely fracture, osteoporosis remains underdiagnosed and undertreated worldwide. [6] A recent study conducted in Asia, including data from Hong Kong, highlighted the poor performance in osteoporosis diagnosis and management, even among high risk groups with prior fragility hip fractures. In the study, bone mineral density (BMD) measurement was only performed in 28.2% of patients prior to hip fractures; 51.5% were informed that they had osteoporosis after hip fractures, and 33.0% received medications for osteoporosis 6 months after discharge. [7] According to Osteoporosis Society of Hong Kong, much of the responsibility for delivering effective treatment to osteoporotic patients falls on endocrinologists and orthopaedic surgeons in Hong Kong. With the long waiting time in specialty clinics, many patients miss the golden time for treatment, i.e. before the happening of fragility fracture and its associated morbidities and mortality. In view of this

Study design
A retrospective case series study conducted in a public primary care clinic of KCC of HAHK.

Setting of the Multidisciplinary Osteoporosis Clinic (MOC)
Patients with con rmed osteoporosis or osteoporosis related conditions such as history of fragility fractures were referred to MOC for further management. The referring source was mainly from different general outpatient clinics (GOPCs) of KCC and other primary care providers in Hong Kong. This multidisciplinary clinic was specially designed to cater the needs of osteoporotic patients.
Patients referred to the clinic were invited to a Community Based Specialty Nursing Session which included a comprehensive osteoporosis management health talk and an individual counselling session.
Fracture risk assessment tool (FRAX), a scoring system assessing one's 10-year osteoporotic fracture risk based on individual's clinical risk factors as well as BMD at the femoral neck [8], was used and FRAX score was calculated during the session. Advice was given accordingly.
After the nursing educational session, a doctor consultation session was arranged in the following weeks.
Doctors conducting the consultation acquired at least fellowship quali cation of both Hong Kong College of Family Physicians (HKCFP) and Royal Australian College of General Practitioners RACGP (RACGP).
They were also equipped with updated knowledge regarding osteoporosis management. The consultation time was on average 15 minutes per patient, which was longer than the 7-minute consultation time in GOPC. This was to ensure su cient time had been given to attending doctors to convey necessary pharmacological and non-pharmacological advice, and to provide quality medical assessment and management. [9] Bisphosphonates are recommended as rst line treatment in most guidelines including National Osteoporosis Foundation (NOF) guidelines and the American Association of Clinical Endocrinologists (AACE) guidelines [10,11]. Oral alendronate is the only pharmacological agent that can be reimbursed in HAHK for patients with history of osteoporotic fracture. Therefore, guideline directed medical therapy, namely alendronate was regularly prescribed in our clinic.
DEXA scan was performed upon joining MOC, i.e. baseline DEXA scan. It was repeated after 1-2 years of initiation of pharmacological treatment and every 2 years thereafter if BMD had been stabilized or improved, compatible with most guidelines' recommendation including NOF guidelines, AACE guidelines and the International Society for Clinical Densitometry (ISCD) guideline [10,11,12]. To ensure uniform data comparisons, patients were urged to have their DEXA scans repeated in the same diagnostic center, assuming the same DEXA machine was used. Doctor follow up appointment was arranged depending on clinical needs.
Allied health services such as dietitian counselling, physiotherapy and occupational therapy were also available upon doctor's referral.  [13]. The difference between the patient's BMD and mean BMD of young female adult, divided by the standard deviation (SD) of the reference population, yields the T-score. A DEXA T-score of -1.0 or above was regarded as normal. While a DEXA T-score between -0.1 to -2.5 was regarded as osteopenia. Osteoporosis was de ned as DEXA T-score of ≤-2.5, according to World Health Organization [14].  From previous studies, the mean BMD difference between treatment and non-treatment group was 0.03 g/cm 2 , with a SD of 0.13 [15,16]. At 95% con dence level and a power of 0.8, with the use of paired sample T test for sample size calculation, the sample size required is 147. To allow the room for data exclusion (~20%), totally 186 patients were included into the data analysis.

Statistical analysis
All data were entered and analyzed using computer software (SPSS version 23, Chicago, IL, US).
Categorical variables were presented as frequencies and percentages. Continuous variables were presented as mean, plus standard deviation. Paired Student's t test was used to analyze continuous variables. All statistical tests were two-sided, and a P value of <0.05 is considered signi cant.

Results
A total of 507 patients with ICPC coding L95 (osteoporosis) attended MOC during the period 1/1/2015 to 31/12/2018. 321 patients were excluded while 186 patients were included in the nal data analysis. Among the excluded, majority of them were excluded because they had been treated with osteoporosis medication by other doctors before joining MOC (N=160, 50%) or they had no interval DEXA scan performed during the study period (N=102, 32%). The owchart of case recruitment is summarized in Majority of them were nonsmoker, nondrinker, with a mean BMI 21.5 +/-3.1 kg/m 2 . The demographics of the included patients are summarized in Table 1a.   During the follow up period, 46 patients received and completed 5 years of bisphosphonate treatment. Progressive improvement in T-scores of both lumbar and femoral neck were observed, with the mean Tscore of lumbar spine improving from -2.79±0.73to -2.22±0.74 (P<0.001) and that of femoral neck improving from -2.48±0.60 to -1.93±0.74 (P<0.001). Results of the interval T-scores of lumbar spine and femoral neck during 5 years of drug treatment are shown in Figure 5.
Regarding the secondary outcome, 1 out of the 186 osteoporotic patients suffered from new osteoporotic fractures among the 2-year follow-up period. The incidence rate was 2.7 cases per 1000 person-years. This patient suffered from left hip fragility fracture after two years of oral alendronate.

Discussion
To our knowledge, no similar team based management model in primary care for osteoporotic patients has been described. In this retrospective case series analysis, we demonstrated that a dedicated team led by primary care physicians could signi cantly improve the bone mineral density of osteoporotic patients.
Our patient demographics are in line with that of other studies on osteoporotic patients conducted in china and Asia. [17,18,19], with a mean age of 66.1 and a majority of them being post-menopausal women with menopause at a mean age of 48.9. The results of our study were very promising that DEXA T-scores of the recruited patients improved signi cantly in both lumbar and femoral neck, P<0.001. Over 80% of osteoporotic patients joining the MOC were started on osteoporosis drug treatment which was signi cantly higher than the treatment rate of previous studies [7]. For patients with pharmacological treatment, predominately treated with alendronate, their 2-year interval T-score improved 0.38SD (from -2.76 to -2.38) at lumbar spine and 0.33 SD (from -2.45 to -2.12) at femoral neck, which were comparable with other published clinical studies [20,21,22]. Furthermore, subgroup analysis of patients who were not on drug treatment also showed signi cant improvement in T-scores of both lumbar and femoral neck.
This suggests that the T-score changes and BMD improvements could not be solely explained by the pharmacological effects of bisphosphonates but also by the comprehensive services provided by the MOC.
The reasons why this MOC is highly effective is multi-factorial. First of all, the MOC is a well-designed clinic tailored to manage osteoporotic patients. There is a protocol for staff to refer to and follow once the patients attend the clinic. Secondly, our nurse would meet every patient during the individual counselling session when they would collect information on patients' demographics and calculate the FRAX scores and give individualised advice accordingly. Thirdly, Doctors who have attended the MOC are relatively experienced family physicians who have achieved intermediate family Medicine training quali cations. In addition, a longer consultation time (around 15 minutes per case) would also allow the doctor to discuss the management plan with the patients more thoroughly compared with GOPC settings (around 6-7 minutes per case). Indeed, previous research on the doctors' barriers on osteoporosis treatment have revealed that a lack of knowledge on osteoporosis drug treatment and speci cally its adverse effects may deter many doctors from treating the disease [23,24]. By centralizing this group of osteoporotic patients to MOC with dedicated doctors and nurses, many of these obstacles of care could be overcome, resulting in improvements in outcomes.
This study has several strengths. This is the rst study ever in primary care of Hong Kong to evaluate the effectiveness of a community based osteoporosis clinic, in terms of DEXA T-score changes in osteoporotic patients. Previous studies published mainly focused on the detection and treatment rate of osteoporosis [6,7,25], or tried to explore factors preventing the delivery of effective osteoporosis treatment to osteoporotic patients [7,26]. Our study is comprehensive and impactful in terms of the outcome measurements. It not only assessed the treatment rate of osteoporotic patients but also the changes in their BMD and the occurrence of fall during the follow up period, which are importance parameters directly in uencing the mortality and morbidity brought about by osteoporosis. Another strength of the study is that the clinical data were precisely documented and complete. In MOC, nurses would document patients' epidemiological data, using a preset template, in the nursing notes of CMS of HAHK. Doctors would also document patients' clinical condition, DEXA scan and blood investigation reports in consultation notes of CMS with a similar format. These measures minimized the possibility of missing data or recall bias.
There are some limitations in this study. Firstly, a large proportion of patients were excluded from the nal data analysis, which might result in selection bias. Majority of the patients excluded had been treated with bone resorptive agents before joining MOC (N=160) or they had no interval DEXA scan within the follow up period (N=102). Bone resorptive agents could have lasting effects on residual BMD and fracture prevention despite stopping the drug [27,28]. Patient's DEXA T-score change after joining MOC could be related to previous treatment instead of the true effect of MOC, therefore were excluded. we believe this treatment naive group of patient could provide a fair re ection of the effectiveness of MOC. Secondly, this study was carried out in a single public primary care clinic in HAHK and therefore, the results of this study may not be generalized to all primary care clinics in Hong Kong or in the private setting. In addition, most patients followed up in MOC had to pay for their own drugs and DEXA scans as they were not readily available from HAHK for free, except for osteoporotic patients with past history of fragility fracture. Therefore, those who continued to follow-up in MOC could well represent a group of patients who were more health conscious and nancially capable. Thirdly, since this is a retrospective study without control group, we should be cautious in attributing all the positive outcome to the effect of MOC. However, available literature has reported that age-related T-score of Chinese women who are treatment naive deteriorates with age [29]. Hence, despite the lack of a control group, in view of the signi cant improvements in T-score of patients who joined the MOC, especially in the non-pharmacological group, it is justi able to credit improved BMD of osteoporotic patients to the effective management at MOC.

Conclusion
With its growing incidence and increasingly signi cant complications, osteoporosis represents a major burden to healthcare. This study provided invaluable evidence that a multidisciplinary osteoporosis clinic run by dedicated family physicians in a public primary care setting can effectively treat osteoporosis patients with signi cant improvement in T-scores. We believe that, based on the ndings of this study, an integrated, comprehensive and multi-disciplinary service model could be developed in primary care to meet the service demand in managing osteoporosis patients in the community.  Interval T-scores of lumbar spine and femoral neck of patients who completed 5 years of bisphosphonate treatment